Part C - Kernel Extension Specifications
Preface node
heading:part-c-kernel-extension-specifications:53362
What this page is
This is generated FPF reference text from the specification preface or supporting sections. It helps interpret FPF; it is not FPF Reference product documentation.
Methodology
Use it to understand how the specification wants to be read, then return to a route, pattern, or work packet for active work. Cite generated IDs only when the wording changes the task decision.
Content
Epistemic holon composition (KD-CAL)
Status: Stable
Use this when you need to combine epistemes for a claim and intended use, or determine what an epistemic change or proposed substitution preserves.
First useful result. Start with the conclusion the recipient needs for the intended use. For a synthesis, work back to the premises and connections that could warrant that conclusion (B.1.3:4.4). A separate claim about each part can need different grounds from a claim about what those parts can do together. When choosing the parts or models is itself at issue, use C.39.SA:4 to decide which distinctions to retain. For a proposed change, recover its source and proposed result. Establish the scope, support dependencies and preservation conditions needed for the conclusion. Return the justified synthesis or change assessment, including any unresolved condition. An adequate synthesis need not have a common numerical R. Apply the F–G–R assessment or CL substitution conditions when the receiving operation uses them. If the question is only which episteme is meant, C.2.1 can suffice; identification or publication alone does not require a KD-CAL assessment.
Scope & exports. A substrate-neutral calculus for composing epistemic holons (U.Episteme) and reasoning about their change and equivalence.
Exports:
- Three distinct assessment aspects—Formality F, ClaimScope G, Reliability R—for one claim-bearing episteme and stated use.
- A pairwise ladder of Congruence Levels (CL 0…3).
- Four Δ-moves (Formalise, Generalise/Specialise, Calibrate/Validate, Congrue).
- Composition rules (Γ_epist) for aggregates.
- Propagation laws for CL through mappings and notation relations.
C.2.1 identifies an episteme by its claim content, EntityOfConcern, and effective reference scheme under EpistemeConstitutionRelation.
Empirical grounding and edition are separate C.2.1 relations. Viewpoint selection and U.View conformance use E.17.0; expression construction and scheme changes use A.6.3.RT.OE/A.6.3.RT; use C.29 when a question about choosing, transferring, or relying on a mathematical representation remains unresolved; publication uses E.17/E.24.PUB; a carrier remains a distinct entity.
Every F–G–R computation names the claim and its U.ClaimScope. If a path changes scope, notation, kind, reference plane, source-local meaning, model-use basis, or evidence basis, it names the actual relation traversed and applies only the loss that relation declares.
Formality F is the assessment of expression properties governed by C.2.3. KD-CAL uses the properties required by the receiving operation, with any local ordered summary explicitly declared. It supplies no universal rank of languages or logics.
Problem Frame
U.System (physically embodied and acting) and U.Episteme (knowledge holon) are two admitted holon kinds. KD-CAL is the primary composition pattern for U.Episteme, giving engineers a way to say (a) which expression properties an episteme supplies for the receiving use (F), (b) where its claim is asserted to apply (G), (c) how well that claim is warranted for the receiving use (R), and (d) how closely two epistemes coincide (CL). C.2.1 supplies the constitution test: claim content, one EntityOfConcern, and one effective reference scheme. Grounding, edition, viewpoint, view, representation, publication, form, and carrier remain neighboring objects or relations governed by their respective patterns.
Problem
A proof is treated as evidence that an actual system meets its assumptions, or a program as if it entailed a theorem. Warrant becomes opaque when support and its applicable currentness conditions are not explicit. Attributing performed Work to a standard conceals the system that performs it. Aggregation can hide a missing necessary premise, erase complementary support, or conceal contrary evidence behind an unjustified score. KD-CAL keeps constitution, input meanings, any scales used, and support dependencies explicit.
Forces
- Universality vs domain idioms. One calculus must cover physics theories, legal codes, safety specs, algorithms, and formal proofs without flattening their differences.
- Meaning vs materiality. The same claims can appear on different carriers. When a historical use depends on a particular source, its provenance must remain recoverable.
- Deductive vs empirical. Axiomatic certainty and empirical trust have different evidence-continuity profiles; both must compose.
- Abstraction vs enactment. Epistemes constrain action; systems act. The calculus must keep the roles distinct.
Solution
Coordinates, constitution, and neighboring relations
KD-CAL assessments for one episteme and use.
- Formality F. The relevant properties of the expressed claim or argument: terms, predicates, semantics and inferences. Litmus: does this expression supply what the receiving interpretation, execution or check needs? An expressed derivation and evidence that it passed a checker remain distinguishable.
- Claim scope (G). Set-valued applicability over
U.ContextSlice, with ∩/SpanUnion/translate algebra. Litmus: where is the claim asserted to apply, and under which assumptions? An applicable C.2.2 warrant model routes congruence-loss penalties to R; formality and scope changes use their own operations. - Reliability R. Warrant for this claim and receiving use. Litmus: what supports this conclusion, under which assumptions, and what limits it? R-claims MUST identify the grounds that support the claim for the receiving use. A numerical R requires the B.3/C.2.2 meaning, scale, and model; otherwise retain separate support and a bounded reasoned conclusion. A proof under axioms needs no empirical score, and F cannot be substituted for R. Relevance windows and B.3.4 currentness rules apply where the relied-on support consumes them.
Congruence Level (CL), pairwise ladder.
CL‑0Opposed/Disjoint (contrastive; no substitution).CL‑1Comparable / Naming‑only (label similarity; no substitution).CL‑2Translatable (structure‑preserving correspondence in a declared fragment with stated loss).CL‑3Near‑identity (the declared invariants match).
CL is a characteristic of a relation between two epistemes; it is not a fourth member of the F–G–R assurance tuple and it is not a characteristic space of its own.
KD-CAL substitution constraint: plane preservation and CL ≥ 2 are necessary for a substitution under this calculus; substituting type‑structure additionally requires CL = 3. These conditions do not establish suitability for the proposed use.
State the substitution, direction, correspondence rule, and tolerated loss under the pattern governing the proposed use, and establish reliance separately.
When local meanings cross semantic contexts, first test the F.9 Bridge predicate on the F.17 cells for those local meanings; keep the obtaining Bridge, bounded-use claim, and reliance result distinct. A CL value establishes no local system-role-kind membership or assignment.
Constitution and neighboring relations. For a KD-CAL F–G–R assessment, recover F, G, and R for one claim and use of one C.2.1 episteme. The episteme's claim content, EntityOfConcern, and effective reference scheme identify the episteme through EpistemeConstitutionRelation. F characterizes the claim's form; G is the separate U.ClaimScope; R relies on evaluation, evidence-use, and assurance relations. Empirical grounding and edition remain separate C.2.1 relations. Viewpoint selection and view conformance remain under E.17.0; expression construction and scheme changes use A.6.3.RT.OE/A.6.3.RT. Use C.29 when a question about choosing, transferring, or relying on a mathematical representation remains unresolved. Publication occurrence, form, and carrier remain under E.17/E.24.PUB. When reasoning between notations, establish their correspondence and the preservation or loss that the inference depends on. Apply C.2.2's R-only penalty rule when its warrant model applies; changes to formality or scope use their own governing operations.
Four Δ‑moves (epistemic motion)
- ΔF — Formalise. Make the expression property needed by the receiving use explicit or stricter: resolve terms, supply a predicate or semantics, or express a required derivation. Name the changed property; a different calculus alone establishes no increase.
- ΔG — Generalise / Specialise. Widen or narrow the claim scope (assumptions & scope). Changes to decomposition granularity are an orthogonal view and do not change G unless they alter the envelope.
- ΔR — Calibrate / Validate. Revise warrant through support that bears on the claim: proof or reasoning, calibration, severe tests, or monitoring as applicable. State what the contribution changes. A formalization alone is ΔF, not an R increase; choosing new inquiry is a separate decision.
- ΔCL — Congrue. Establish or revise the congruence relation between two epistemes and state its warranted level (CL 0…3) under the declared correspondence rule. Moves compose into paths. A CL chain minimum retains only the ordered congruence meaning justified by the relation family; it is not a numerical reliability loss.
Composition (Γ_epist) and propagation
Let Γ_epist compose epistemes {Eᵢ} for one declared claim and use. B.1.3 supplies the synthesis/compilation Method; B.3 and C.2.2 govern warrant and scale discipline.
- R (Reliability). First distinguish indispensable premises, alternative sufficient arguments, complementary evidence, different scope slices, and counterevidence. Identify duplicated data and shared assumptions or bias. A numerical fold requires warranted input meanings, compatible scales, dependencies, and a receiving model. Neither series nor parallel syntax supplies a default minimum or maximum, and there is no universal cap at the best support line. Where no common model is justified, retain separate contributions and limitations in a bounded reasoned synthesis.
- F (Formality). Assess the expression properties needed from the essential constituents and their connecting inferences under C.2.3:15.1. The composition must preserve the claimed properties. A minimum is only a conditional bound on a shared local ordered scale with an established composition rule, not a universal value for the whole. Formalization changes its stated properties; neither an axiomatic mode nor a
line=formaltag converts them into empirical warrant. - G (ClaimScope). Required premises compose only on their overlapping scope. Distinct supported slices may form
SpanUnion({G_path})under A.2.6 and C.2.2's type-before-scope rule; retain their support separately and drop unsupported regions. A new source does not by itself generalise the claim. Scope change remains an explicit ΔG± move. - CL (Congruence). Keep each traversed mapping and the ordered meaning of its declared CL visible. A chain minimum is usable where that relation's congruence rule justifies it. A notation, scope-translation, kind, plane, source-local, model-use, or evidence-reuse relation contributes only its own warranted loss. A numerical Φ needs its receiving model; a monotone table or clipped output does not supply one.
For example, two necessary independent conditions with probabilities 0.9 each have conjunction probability 0.81, not minimum 0.9. Conversely, a limited complementary source need not reduce the support already available. A credible contrary result changes the affected conclusion. A theorem A ⇒ P remains valid as a formal result while evidence violating A can defeat its use as assurance of an actual system.
Γ remains defined on holons and respects the core's identity and boundary discipline. Its support account establishes neither a new action permission nor the worth of acquiring further evidence.
What must not be conflated (normative guards)
- Representation structure ≠ carrier. Files, PDFs, or repositories are carriers outside the episteme; they never count as parts of
U.Episteme(see C.2.1:4.1–4.2, :4.6, and :11 item 11). - Work and guidance. Systems perform Work. Epistemes can guide or constrain that Work through their claim content; keep the guiding content distinct from its performer.
- CL is a qualitative ladder of preservation classes. Do not average its values.
Archetypal Grounding
Composition rule. Compose knowledge by Γ_epist with explicit support roles, dependence, scope, input meanings and any scales used. Use the receiving model for any R calculation; where no common model is justified, retain a bounded synthesis with its separate support and limitations. Identify the episteme by claim content, EntityOfConcern, and effective reference scheme; keep empirical grounding, edition, viewpoint selection, view conformance, representation, publication form, publication occurrence, and carrier in their own direct relations.
System case. Consider a battery-pack thermal subsystem integrating a physics model of heat flow and an operating envelope for fast-charge. As a system, it composes pumps, sensors, and controllers through the system, composition, boundary, state, and dynamics guidance in A.1, A.14, A.22, and A.3.4, with conservation constraints made explicit; B.1.6 and C.16 govern resource and measurement claims as applicable. The assurance story depends on epistemes about the model and envelope; the system acts, epistemes constrain.
Episteme case. Consider a climate projection episteme: its claim content covers PDEs and parameterisations; its EntityOfConcern identifies what the projection claims concern; and its effective reference scheme supplies the interpretation rules. A separate U.ClaimScope names historical forcings, resolution, and assumptions. Any empirical-grounding occurrence names the grounding holon and covered claim subgraph separately. Its representation may include domain equations and a tabular schema; reasoning between them uses their correspondence and any preservation limits. When composing radiation, cloud, and ocean-mixing contributions, identify which assumptions the projection requires and what each hindcast tests. Shared models or data do not establish independent confirmation. An aggregate R needs the domain quantity and dependency model; otherwise retain the separate tests, assumptions, scope limits, and any disagreements. F describes the required expression properties of the essential constituents and their connections, including the interpretation of the coupled equations.
Bias‑Annotation
- Metric worship. Treating
[F,G,R]as ends rather than means; mitigation: state the supporting grounds and limits for the claim and its use. - Category slip. Equating a notation, view, publication form, or carrier with claim content, EntityOfConcern, effective reference scheme, or an empirical-grounding participant; mitigation: apply C.2.1 constitution and then the pattern governing the neighboring relation.
- Analogy inflation. Presenting naming similarity as identity; mitigation: establish the correspondence and preserved invariants. A claimed CL retains its rung and the substitution limits in §4.1.
Conformance Checklist
- C2-1 (Episteme constitution and neighbors). Every
U.EpistemeMUST satisfy C.2.1 constitution through claim content, one EntityOfConcern, and one effective reference scheme. Empirical grounding and edition are stated through their separate C.2.1 relations. Viewpoint selection andU.Viewconformance use E.17.0; expression construction and scheme changes use A.6.3.RT.OE/A.6.3.RT, with C.29 for an unresolved question about choosing, transferring, or relying on a mathematical representation; publication occurrence, form, and carrier use E.17/E.24.PUB. None is treated as an episteme slot or identity component merely because a record or notation places it beside the constitution values. - C2-2 (Assessment meaning). When a KD-CAL F–G–R assessment is stated or used,
[F,G,R]SHALL be recoverable for one claim and use, with a brief rationale. Identifying or publishing an episteme does not by itself require this tuple. F names the relevant expression-property assessment under C.2.3; a local rank requires its declared scale and mapping and is not a global F value. Where R has no justified numerical model, retain its unquantified support and bounded conclusion. Formal validity needs no empirical score. CL is declared for pairs only. - C2‑3 (Composition). Authors SHALL identify support roles and dependencies under B.1.3/C.2.2 before combining inputs. Any numerical R or loss MUST have justified meanings, scales, assumptions, and a receiving model under B.3; no universal min/max or F-to-R conversion applies. Otherwise return separate support and a bounded synthesis. F requires the properties of the essential constituents and their connections to support the claimed whole under C.2.3; G uses applicable path intersections and supported SpanUnion under A.2.6. For reuse that depends on a changed value or mapping, the actual operation or relation and its warranted limitations MUST remain recoverable; do not hide contrary evidence or unsupported scope.
- C2‑4 (Cross-notation use). Cross-notation reasoning MUST recover the source and target expressions, their correspondence, and the preservation needed by the inference. A
NotationBridgeedge may record these relations and their limitations. State CL where the receiving rule uses it, including the KD-CAL substitution constraint in §4.1. - C2‑5 (Work performer). The performer of
U.WorkMUST be a System. Keep any guiding episteme distinct from that performer.
Consequences
Benefits. A compact map for knowledge epistemes; visible support dependencies and limitations; useful formal and qualitative results alongside justified calculations; disciplined reuse across domains with stated correspondence and preservation limits; consistent separation of meaning from material carriers. Trade-offs. Authors must identify the support relation and any calculation model; multi-notation work keeps its relation-specific basis. Some useful syntheses have no common scalar. Mitigation: the C.2.1 constitution test and the patterns governing neighboring objects and relations keep the ordinary entry brief while preserving recoverable precision.
Rationale
KD-CAL composes C.2.1 epistemes through the relations among their claims and support. Claim content, EntityOfConcern, and effective reference scheme keep episteme identity stable; assessment of expression properties (F), claim scope (G), warrant for the claim and use (R), and pairwise congruence (CL) remain distinct in composition. Direct grounding, edition, view, representation, publication, and carrier patterns prevent category collapse. G concerns the claim’s applicability; changing decomposition detail changes G only if it changes that applicability.
Relations
- Depends on:
C.2.1 U.Episteme: Constitution, Empirical Grounding, and Edition Relationsfor episteme identity, the constitution relation, and the separate grounding and edition relations;E.17.0for viewpoint selection andU.Viewconformance;A.6.3.RT.OE/A.6.3.RTfor expression construction and scheme changes;C.29for an unresolved question about choosing, transferring, or relying on a mathematical representation; andE.17/E.24.PUBfor publication occurrence, form, and carrier. - System-related claims: system composition, boundary, state, conservation, resource, and measurement claims use
A.1,A.14,A.22,A.3.4,B.1.6, andC.16as applicable. KD-CAL composes epistemes and feeds assurance lenses in Part B.
Worked mini‑examples
- Formalization (ΔF). Recasting a variational free-energy narrative into a typed calculus can make its variables, assumptions and inferential relations explicit. State those changes and any corrected scope. A correspondence between biological and ML formulations still needs its own mapping and preservation basis; formalization alone establishes neither that correspondence nor an empirical gain (R).
- Complementary hindcast evidence. Two hindcast lines supporting a climate projection may address different errors, reuse data, or cover different conditions. Identify those relations before combining them; a maximum may select one attested argument only under that declared meaning, not measure their combined corroboration. Keep disagreement and unsupported scope visible. This illustration supplies no climate-specific reliability model.
- Preservation across notations. Suppose a rendering of a type-theoretic specification in a semi-formal DSL omits a declared source invariant. State the correspondence, omitted invariant and resulting limit on inference. On the KD-CAL ladder this preservation loss is below CL 3 and cannot justify type-structure substitution under §4.1. A transported theorem requires preservation of the invariants its proof uses.
C.2:End
U.Episteme: Constitution, Empirical Grounding, and Edition Relations
Type: Pattern Status: Stable Normativity: Normative except where a section is explicitly marked informative
Plain name. Episteme constitution.
Use this pattern when. You are developing, using or changing an account and need to recover what it says, how its claims can guide work, or what a change affects. Use the common method in :4.0, then the identity, grounding or edition branch needed for the question. A pump specification with a changed pressure threshold carries different claims. Publishing the same claims about the same pump under the same interpretation in another file preserves the episteme.
First useful move. Establish what the account must help its recipient understand or do. Recover the relevant claims, their subject and interpretation; write or draw what is understood when the account is still developing. A corrected sentence, an interpretable table or a justified answer about a changed dependency can finish the task. Section :4.0 connects these moves; :4.1.1 gives the identity comparison when needed.
One-line summary. A U.Episteme is a claim-bearing holon identified by exact claim content, one exact EntityOfConcern, and the effective ReferenceScheme that makes the claims interpretable about that entity. Changing any of these three discriminators identifies another episteme. The representation language can be shared by records whose effective rules differ. A document-specific convention for an empty cell can alter the claims a reader may use even where both readings are admitted by the language (E.5.2.L:4.1).
Primary working reader and viewpoint. The engineer or researcher comparing, revising, teaching, grounding or publishing that knowledge object. The working concern is to reidentify it through those uses and locate a change when it occurs.
Primary governed object and architecture. One U.Episteme and its EpistemeConstitutionRelation, EpistemeEmpiricalGroundingRelation and EpistemeEditionRelation. The EntityOfConcern of the episteme is the entity its claims concern; for a pump specification, it is the pump, while C.2.1 governs the specification's identity.
What goes wrong if missed. The reader confuses a whole account with selected claims, its subject or its expression; a shared filename can also hide changed claims or interpretation. What this buys. The practitioner can recover the needed meaning and grounds, locate a consequential change and repair the dependent account, expression or use.
A theory, model, specification, proof or diagnosis can be an episteme when the selected object is that claim-bearing whole. A diagram can carry such claims too. When the task concerns its layout or the calculations available through its notation, use the publication or representation branch in :4.0.
Not this pattern when. Use the subject's direct method to inspect or change the pump, perform work or apply a calculation when its account and grounds already suffice. Open C.2.1 when interpretation, episteme identity, grounding or a change in the account leaves the receiving use unclear. A separately inspected classification assertion has its own claim content, subject and scheme; its criterion remains under A.1 or C.3.2, with E.24.UK used for public U-kind admission.
Problem Frame
FPF treats an episteme as a holon, not as a document class or a filled record. It is a claim-bearing whole about an EntityOfConcern under an effective reference scheme. Carriers, notations, and admissible operations can differ; the identity question is what makes this one episteme and what changes it.
The core constructive question is whether a ClaimGraph, U.Entity, and effective ReferenceScheme stand in the relation that makes the claim content interpretable and evaluable as claims about that entity. When they do, their selected organization yields a whole-level epistemic characteristic: the resulting holon can be used as one defeasible or deductive body of knowledge. That characteristic is not supplied by any one participant alone.
Any U.Entity can participate as the EntityOfConcern. An episteme can therefore concern another episteme or itself without changing the constitution relation.
Different representation regimes can support different operations. Make the representation and its admitted operations explicit when the current use needs them. Identify the episteme by its claim content, EntityOfConcern, and effective ReferenceScheme.
Problem
Without one direct episteme ontology, several practical failures recur.
- Carrier and episteme collapse. A PDF, database row, proof script, dashboard, or neural-model file is treated as the knowledge holon. File replacement is then reported as epistemic change even when claim content, EntityOfConcern, and interpretation are unchanged.
- Subject drift. A specification or model keeps one label while the entity it concerns changes. Comparison and evidence use then combine claims about different entities.
- Interpretation drift. The same tokens or graph are read under different designation, measurement, or evaluation rules while users assume one unchanged episteme.
- Neighboring-relation collapse. Grounding holon, viewpoint, view, claim scope, model-use structure, evidence, edition, and publication become optional fields of one omnibus record. Their different obtaining and identity rules disappear.
- Representation-first ontology. Tuple components, graph nodes, schema fields, and database keys are treated as actual relation participants or subject identity discriminators merely because a tool exposes them.
- Agency leakage. An episteme or its claim graph is said to perform Work. Systems perform Work; the claim content can guide or constrain it.
- Dependent-kind identity fork. A method description or view is assigned another identity merely because the pattern for that kind supplies a membership condition. The same episteme can then appear twice, and a viewpoint or method-description use can be mistaken for a change of knowledge object.
The familiar Symbol-Concept-Object triangle can still introduce the difference among expression, meaning, and subject. It cannot serve as the ontology because it suppresses reference scheme, grounding, viewpoint, evidence, and the distinction between a relation and a representation of that relation.
Forces
Solution
Begin with the work the account must support and use :4.0 to develop, interpret or revise it. When comparison requires episteme identity, EpistemeConstitutionRelation identifies what it says, its exact subject and its effective interpretation. The later branches supply grounding, edition and dependent-use judgments at the precision the receiving work needs.
Local episteme mantra. Recover what the recipient needs to understand or do. Connect the claims, their subject and interpretation with the grounds and expression that let the recipient use them. Follow a change to the results that depended on it. Compare exact episteme identity through claim content, EntityOfConcern and effective reference scheme when that comparison matters; otherwise repair the affected expression, relation or use and finish at the useful result.
The mantra is a recall aid. Section 4.0 explains the common method; sections 4.1–4.9 supply constructions and change rules. Section 4.10 is a later reference for relation and neighboring-object distinctions.
First-use completeness questions
When this helps
Use this guidance when you are preparing, using or changing an account and need to know what its content says, how another person or system can use it, or what a change affects. A useful first result can be a corrected sentence, an interpretable table or an answer about which earlier conclusion needs reconsideration.
Begin with the work the account must support. Someone may need to understand a mechanism, compare two options, check a requirement or perform a calculation. Establish enough of that use to judge the account's adequacy. A familiar task can already supply its purpose; developing an account can also expose a question that was previously unclear.
The general approach is to connect the claims about the subject with the grounds for checking or using them, the expression through which the recipient recovers their meaning, and the operation the recipient needs to perform. When one changes, follow the affected connections. The following sections explain those connections and give the Methods for working with them.
If the account and its use are already clear, perform the subject work directly. Open the additional distinctions when they resolve an uncertainty about meaning, preservation, expression or availability.
Knowledge work: claims, grounds and what recipients can do
When a team develops knowledge about a subject, it can obtain several connected results. It articulates claims that can be examined, establishes grounds for relying on them, and learns to use the relevant distinctions in its work. Developing an account, supporting its claims and acquiring that ability can require different actions.
An episteme, E of kind U.Episteme, is an organized claim-bearing whole about an identified subject. A condition can qualify an assertion, a premise can support a conclusion, and an exception can limit a rule. These connections make it possible to examine and use the claims together. In FPF, the exact claims G, their subject T and their effective interpretation R identify E. The next section explains those constituents and participants.
A kind classifies independently identified entities by a membership criterion. U.Episteme is one such kind; E is a member. A.1 and C.3 govern that distinction. A specialized kind such as U.WorkPlan adds its own membership condition while retaining the identity of the same episteme.
In a request to “revise this knowledge,” establish which result needs revision. If the team means its account of the pump's operating range, identify E and the claims that need correction. If it means the pressure-limit premise used by a calculation, identify that selected claim content. If the team instead needs technicians to recognize when the limit applies, the intended change concerns what those people can do. The appropriate action follows from the subject of the request.
Knowledge engineering can retain hypotheses, disputed assertions, known errors and requirements as identifiable epistemes. They remain available for testing, comparison and correction. When a practice calls a claim established knowledge, recover the claim and the grounds and conditions under which the practice relies on it. Episteme identity alone supplies no judgement about that reliance.
For a requirement, distinguish the assertion that a requirement has been imposed from the requirement it expresses. “The drawing requires a clearance of at least 0.4 mm” reports something about the drawing. “The clearance shall be at least 0.4 mm” states a condition on the intended assembly. The first can be checked against the drawing; relying on the second involves the applicable requirement and checking practice. A plan similarly preserves its intended coordination claims from the planning perspective, including after the planned time.
When the receiving question is whether a performer can apply the knowledge, name the work they must perform and the conditions of that performance. A.2.2 supplies the ability-and-fit comparison. For example, a technician may be able to explain a pressure limit while still needing help to select the correct limit for a changed operating condition. Qualify each claim by its actual evidence.
The common work is to maintain the connections: observation and reasoning inform the claims; an expression makes those claims recoverable; a person or system uses them in a particular operation; the operation and its consequences can expose a need to revise the account or its use. The following sections explain how to follow those returns.
Recover what the account says
Read enough connected material to recover the claims and the relations that change their meaning. Identify the subject of those claims and how names, quantities and modal expressions are interpreted. The same numerical string can state an observed value, a proposed value or a required limit. Recover that force from the account.
When developing an account, start with the work the account must support. Its recipient may need to carry out an operation, reconstruct what happened, decide between alternatives or explain a result. If its content is yours to choose, work back from that use to the claims and connections it requires. When the subject has relevant constituents, ask which of them could change the answer. Which premises or exceptions belong in the account about them? For a conclusion about simultaneous operation, check that the claims about the parts hold under the conditions of that operation. For a comparison across configurations, establish a common basis for the features at issue. A decision can also be justified under rival accounts if it follows under each. An account can omit a distinction when that preserves the answer and the grounds its recipient needs. C.39.SA:4 helps choose the composition scales and model distinctions for this purpose.
Write or draw what is already understood when doing so can expose a missing condition or connection. Obtain grounds for a missing fact, revisit an unsupported inference and improve the expression where the intended claim cannot be recovered. These repairs can be needed together. C.2.4 explains progressive articulation. A.6.3.RT.OE explains how to construct an expression that supports the next operation.
Once the claims are determinate, compare an account through the C.2.1 constitution:
- Claim content: what is asserted, including consequential conditions, qualifications and relations among claims.
- EntityOfConcern: the particular subject those claims concern.
- Effective reference scheme: the designation and interpretation rules actually used to understand the claims about that subject.
These are semantic discriminators. For this whole, the subject T is an independently identified U.Entity serving as EntityOfConcern. Claim content G and effective rules R are abstract U.Entity individuals under C.2.1:4.1.1. G is the semantic constituent containing the interpreted claims; it is not a second name for the whole E. For reusable typed uses, C.2.1:4.1.2 supplies the ranges NonemptyClaimContent and EpistemeInterpretationScheme. Those declared ranges are not public U-kinds named U.ClaimGraph and U.ReferenceScheme.
A file boundary is useful for finding material, while the selected claims and their organization determine the account being compared. Several files can express one account; a file can contain several independently identified accounts. When only a claim or a subgraph is needed, select that constituent explicitly. When a new episteme is constituted about it, the selected content becomes the EntityOfConcern of that new episteme.
If a word, sign or reference still permits materially different readings, retain that uncertainty and obtain the missing interpretation before relying on the disputed conclusion. Where exact constituent identification matters, C.2.1 supplies NonemptyClaimContent, EpistemeInterpretationScheme, ClaimContentDesignation and SchemeDesignation. Ordinary writing needs only the explanations and references sufficient for its actual use.
Using an episteme as a model. For a claim-bearing informational model, identify this episteme once by its claim content, EntityOfConcern and effective interpretation. State which reading or operation on it can answer the inquiry and what warrants that use. A room number can designate a room, a timetable can state its scheduled occupancy, and that same timetable can be used to predict when the room will be free. Only the prediction relies on the relevant correspondence between scheduled and actual occupancy. If the inference concerns the episteme's own EntityOfConcern, use that same subject. If the account is used by analogy for another entity, keep its original subject and give the grounds for transfer. A.1.1 governs its named, bounded uses with a holon.
The thing used as a model can also be a physical specimen or a directly observed performance. A report about it is a separately identifiable episteme. The tested specimen can be the EntityOfConcern of one report and the GroundingHolon for selected claims about a different system, when the empirical relation in §4.3 obtains. For observations from which a repeatable method is to be recovered, use A.3.1.MR.
Connect the claims with their subject and grounds
Aboutness identifies what a claim concerns. To decide how that claim can guide work, recover the connection required by that use. Establishing empirical support requires an observation or test bearing on what the assertion says; a formal conclusion needs its derivation under the stated rules; applying a requirement needs the grounds of its force and a way to check compliance.
When empirical grounding is needed, identify what would be observed, measured or affected in a test. Recover the actual observation, measurement, intervention or test and explain how its outcome bears on the claim under the account's interpretation. Name the whole participating in that empirical connection. C.2.1 calls that participant the GroundingHolon: H of kind U.Holon, involved in the mapped empirical relations. GroundingHolon names its participation in this relation; GroundingHolonSlot is the corresponding declaration position.
Select the claims this grounding covers. A test of one pressure range can ground the specified claims about that range; a statement about fatigue life needs the corresponding test or other applicable grounds. A description of a general Method can concern the Method while a particular system participating in its trial supplies the grounding holon. The subject described and the holon participating in its test therefore have separately recoverable jobs.
An evidence account is an episteme whose claims report a result, observation or argument offered for a target claim. State which claim it supports, challenges or constrains, and why that inference is warranted. A.2.4 governs this evidence-use relation; A.10 recovers the source, production and reliance conditions that the present use needs. The observation or test and the report about it can change independently. A report can be corrected after discovering that it identifies the wrong specimen, or remain intact while a later application falls outside the tested conditions.
Grounding makes a designated empirical claim inspectable through its actual connection with the world. The empirical result can support or refute that claim. Evidence use explains the direction and scope of support; an assurance judgement answers the particular reliance question that requires it. C.2.6 connects these contributions with the receiving work. Keep an ordinary adequate explanation when it supplies the needed connection.
For a formal derivation, recover the interpreted premises and inference that establish the conclusion. When that conclusion is applied to a real system, also recover the model-to-subject correspondence and empirical premises of that application. For a normative account, recover the rule or decision that gives the requirement its force; empirical inspection can then establish whether the subject meets it. These connections allow formal and normative content to inform practice while retaining their actual grounds.
When several accounts contribute to a conclusion, give each its contribution: for example, a necessary premise, an independently sufficient argument or evidence that limits applicability. Preserve a material conflict and distinguish repeated reports of one observation from additional observations. B.1.3 supplies the corresponding synthesis and target-scheme compilation methods. Use C.2.1 to identify the resulting account and the original accounts on which it relies.
Recover what description names in the situation
“Description” is ordinary wording whose referent depends on the passage. In “revise the description of the pump's operating range,” it can name the episteme E carrying those claims about the pump T. In this use, description and episteme designate the same individual. The word gives no additional membership test.
The technical term U.Episteme gives the claim-bearing whole one reusable constitution and identity rule across descriptions, plans, theories and other accounts. That rule lets a practitioner compare the same account across files and expressions, distinguish changed claims from changed presentation, and retain the original plan for comparison after performance. Where ordinary “description” already designates that whole, there is one individual with two contextual designations. An additional description kind would need a positive membership condition that changes a real classification or use; the different word supplies none.
To understand what this description says, recover G: for example, the operating range, its conditions and the stated relation between pressure and flow. To understand what it concerns, recover T, the EntityOfConcern. To check an empirical claim, recover the appropriate claim-to-world mapping and H. This separates three questions: which episteme is being revised, what its claims say, and what grounds their use.
The intended action can select another object. “Shorten this description to fit the card” may concern an expression and the meaning that it must preserve. “The description on page 2 is unreadable” can concern the printed signs or the wording. Name that object when the difference affects the repair. Treating every occurrence of the noun as the same technical kind would conceal the actual work.
A WorkPlan can describe intended work. Its plan character comes from its intended coordination claims, horizon and plan-item organization under A.15.2. The intentions keep their own temporal perspective after execution or cancellation. A plan for Monday's pump inspection can remain the same episteme on Tuesday and supply the baseline for a later comparison. A report that the inspection occurred has other claims and its own grounds. Both can be called descriptions in ordinary language.
Similarly, a MethodDescription states claims about an identified Method as a way of doing. Its specialized membership comes from A.3.2's subject and claim criterion. The word description supplies no additional general superclass between these specialized kinds and U.Episteme. Their different names alone establish no mutual exclusion.
In describing something, identify the work performed: observing relevant features, composing claims, choosing an expression, or communicating an already available account. The performed work and the resulting episteme or expression have their respective identities. The action may use an existing episteme without changing it.
When the same account is used to choose a pump, compare a requirement or teach a maintenance procedure, name that operation and the claims it relies on. There is no need to put these overlapping uses into exhaustive “describing” and “non-describing” classes. If the task is specifically to provide a characterization for a concern, state what the recipient needs to learn about the subject and use that requirement to choose and judge the supplied claims.
A viewpoint can provide shared requirements for that account. Specification use adds another question: which stated conditions are to be checked, by what means, for what use and with what authority when authority matters. Recover the tolerances, comparison rules and relevant viewpoint that affect that judgement. An account can describe a required condition, explain why it matters and supply a criterion for checking compliance. Each contribution is understood through its claims and use.
Choose the concern and check viewpoint conformance
A viewpoint states requirements for an account: which subjects and concerns it covers, what information is required and what forms, omissions and consistency conditions are allowed. A reader can use those requirements to ask whether a particular account is suitable for that concern.
In FPF, the viewpoint P is itself an episteme containing these fixed requirements; E.17.0 classifies that same individual as U.Viewpoint. Distinguish P from the convention or rule collection it defines when a receiving operation uses the rules as its argument. An ordinary self-contained P can concern the target kind to which its requirements apply. A separate structured organization of conventions is needed only when independently maintained components and their relations are part of the actual viewpoint problem. E.17.0 explains both cases.
A view is that same episteme E, classified as U.View when its conformance to a viewpoint P obtains under E.17.0. To establish this, recover the subject of E, the applicable requirements in P and the corresponding claims or other admissible evidence for each requirement. Determine whether the coverage, interpretation and permitted omissions meet P. The same E can satisfy several viewpoints.
An author can write a suitable view directly or derive an account from an earlier source. These paths answer different questions. E.17.0 establishes conformance; A.6.3 establishes the claimed relationship between source and derived content. Use both when both relationships matter.
Select a viewpoint because the work needs its requirements. A question such as “What is the pump's maximum pressure?” can often be answered directly. A team repeatedly exchanging safety accounts may need a maintained viewpoint so that every recipient can depend on the same coverage.
Construct an expression and choose its organization
Here a particular expression X is the identified sign construction, an entity in its selected notation domain, that can be read or processed under applicable conventions. A filled table has selected entries and relationships between them. Its reusable empty layout specifies where entries can go. To recover the table's assertions, the recipient needs the entries together with the interpretation of the headings, units and relationships. When exact expression identity matters, select the construction being compared and its rules: for example, a character sequence or a typed table with defined entries. Copies can supply the same construction through different carriers. A different wording can preserve the claims while changing that expression.
Start from the operation the recipient must perform. Choose an expression that makes the required distinctions available. Then recover its meaning and compare that meaning with the source or the developing account. Inspect consequential omissions and added connections. Try the intended operation with the expression. A.6.3.RT.OE develops these construction steps; A.6.3.RT governs a change between expression schemes and its preserved and changed content.
In a representation use, an identified entity stands for an identified subject or selected content under a correspondence. The entity keeps the kind and identity supplied by its own domain. Establish what represents what, the correspondence between relevant elements and relationships, and the interpretation under which the intended operation is warranted. A route graph can represent passages in a building; the graph's edge and path rules determine which route conclusions can be returned to the building. Use C.29 for additional grounds when a question about choosing, transferring, or relying on a mathematical representation remains unresolved.
An expression can represent a subject and express claims about it. Those are different relations of the same construction. For example, a table's rows can correspond to passages, while its entries assert their widths. The correspondence explains what the rows stand for; the interpreted entries explain what the table claims.
When presentation means an arrangement for a recipient's encounter, identify the actual order, grouping, emphasis and access provided in the expression. When it means the presenting activity, identify the performed work. These meanings do not require an additional entity between the expression and its carrier. Choose that arrangement from what the recipient must understand or do. Show a qualification where it constrains the claim the recipient will use. A presentation can be prepared for a discussion before public release. E.17 supplies the publication-facing application, including the condition of preserving the accepted source claims when that is the task.
A publication form F supplies reusable organization rules selected for expressing material in a publication. F is the rule collection; an episteme defining those rules and an expression following them are independently recoverable objects. A card layout can require a subject, a value, its unit and a qualification. A particular filled card supplies the actual selections. When the form is reusable, distinguish the rule from its particular application: two cards with values 0.3 and 0.4 can follow the same layout and express different claims.
Use the form where it helps the recipient recover the account. An ordinary sentence can already be a sufficient expression. Machine processing may require a particular encoding and interpretation, with no separate human presentation task.
Repair the conventions when an available expression is insufficient
A familiar notation can already support the operation. When its conventions leave an important distinction unexpressible, a reference ambiguous or a reading operation undefined, design the missing contribution.
Start from the recipient's subject question and identify the distinction, referent or reading operation that the available conventions leave unresolved. When a distinction is missing, choose two cases that the recipient must treat differently, state their difference in subject terms and try to express both under the candidate rules. If the same expression results while different answers are needed, supply the missing distinction or retain an accessible source that provides it. This derives a notation requirement from work the expression must support.
Next define how the expression is formed. Identify the relevant constructors and their operands, the rule for grouping them and the referent of each consequential name. Give the interpretation of their combination and show how the recipient obtains the needed consequence. A legend that explains each sign can be sufficient for a familiar construction; a new composition also needs its reading operation.
Test a change that can affect the interpretation. Reusing one measured value twice differs from performing two successive measurements. A shortened expression must preserve whichever operation the account actually specifies. A correct formal transformation needs the applicable interpretation and its conditions; a reported observation retains its own history.
For translation, construct the target from interpreted source parts and relationships. Perform the target operation and return its answer to the original subject question. Check whether the translation merges source cases that the receiving operation must distinguish. Preserve a lost distinction when the receiving operation depends on it, using a richer target or a defined return to the source. A translation can remain sufficient for a narrower question.
When complementary representations stay in use, establish their shared referents and the rules connecting their contributions. Before propagating an edit, determine whether it changes a source fact, a requirement, a model parameter or only a display. Recompute a derived prediction when its parameter changes; preserve an independently recorded observation so that the difference can inform the subject inquiry. Resolve conflicting change demands as a subject decision.
The construction above supplies the common operation for recovering missing expression conventions. For a larger notation-design problem, FPF Library's Notational Engineering DPF develops these methods further: NOT.1 derives the needed distinctions; NOT.2 fixes formation and reference; NOT.3 gives operative interpretation; NOT.4 develops transformations; NOT.5 handles translation and loss; NOT.6 maintains complementary representations. Bring the resulting expression, rule or located limitation back to the account being developed. A.6.3.RT and A.6.3.RT.OE provide the direct construction when available conventions suffice.
Provide an expression and identify what carries it
A carrier is the physical or digital entity from which the expression can be recovered under the relevant access and interpretation conditions. Select the concrete entity needed for the question: a particular sheet, a continuing file object or a fixed digital snapshot.
State which identity is in use when continuity matters. Overwriting a continuing file can preserve that file object while changing the recoverable expression. A fixed byte snapshot changes when its bytes change. Printing a second sheet supplies another physical carrier. A pathname or URL locates something through an access mechanism; establish what it denotes when the comparison depends on exact identity.
For a precise comparison, identify the carrier separately from its relevant state. Recover the expression from that state using the selected decoding or reading conventions. A screen can remain the same screen while displaying successive expressions. If the state needed to recover an earlier expression is unavailable, that earlier availability needs another source.
Publication, in the availability-reading used here, is the continuing availability of a selected account through a particular expression and carrier to a declared audience for a stated use. Establish that the recipient can obtain the expression and recover the required content under those conditions. Where a chosen publication form matters, check how that expression follows the form.
The activity of publishing includes actions such as placing a file or granting access. Its success contributes to availability. Actual access, interpretation and reliance by a recipient are further occurrences. Keep the specific claim clear when reporting a publication or its use.
An exact publication can identify one expression even when several expressions convey the same episteme. This makes it possible to distinguish an HTML text from a spreadsheet supplied for calculation and to establish which one the recipient obtained. A collected release can include several such publications and several epistemes. Its packaging alone does not determine their semantic unity.
Follow changes to the results that depend on them
Locate the change at the object or relation actually changed. Then follow the reasoning or operation that used it.
A change to the interpreted claims, their subject or their effective interpretation changes episteme identity. Reconsider a derived conclusion when its derivation used the changed claim or interpretation. A preserved passage can remain usable on its established grounds if the changed material is outside those dependencies. Uncertain dependence requires investigation rather than an assumption of preservation.
New external evidence can change the warrant for existing claims. If the author instead changes the grounds asserted within the account, the claim content also changes. C.2.1 and C.2.6 explain these different returns.
A changed expression can preserve the episteme while altering usability or availability. First distinguish rules that determine the account's meaning from rules that express already determined values. Replacing prose with a labeled table can preserve both meaning and interpretation; changing an operative rule from “blank means unknown” to “blank means zero” changes the interpretation. Compare the complete claims and effective rules, then test the affected operation. Removing a qualification can strengthen a claim; changing a font can be harmless when font has no relevant meaning or operational role. The reason for the judgement comes from the expression conventions and intended operation.
A changed viewpoint reopens conformance to that viewpoint. Selecting another already defined viewpoint changes the current comparison. A changed carrier or access condition reopens the relevant bearing or availability claim. The account's subject and claims may remain unchanged.
An episteme E2 is a later edition of E1 when it has a different identity, its revision actually uses E1 as a source, and its preserved and changed features satisfy the applicable continuity rule. That rule states what must remain, what may change and which departures count as a fork, translation, retargeting or independent reconstruction. C.2.1:4.5 defines this continuation as EpistemeEditionRelation(E1, E2). E2 is the edition; the relation connects E1 and E2.
When selecting an edition for publication or identifying claims from it, identify that particular episteme. A reference to E2 selects E2 with its claims, subject and effective interpretation. Establishing how E2 continues E1 is a separate judgement, needed when that continuity matters to the use.
For an ordinary correction, finish when the needed meaning is clear and the affected use has been reconsidered. Preserve enough explanation of a consequential change for its recipients to understand what they can now rely on. The extent of that explanation follows the change and the receiving work.
Connected case and changes in its conditions
A warehouse team maintains an account of passages used by a robot. A dispatcher who needs one passable route can stop with a connected route whose passages are wide enough in the same warehouse arrangement. To promise the shortest admissible route, the dispatcher also needs grounds to rule out a shorter one. Lengths or bounds can supply those grounds. For the bounded comparison below, a passage is admissible when its clear width is at least 1.0 m. The other operating conditions have already been resolved, and the account covers all the passages under consideration.
The account's subject is the selected warehouse arrangement. Its claims include each attributed length and width. The interpretation fixes the endpoints, measurement meanings and units. The viewpoint chosen for this route table calls for a length and width for each passage under consideration and a recoverable admissibility condition. A request for one admissible route can already be answered from the passages that establish it.
The table is an expression of the relevant claims. A graph representing the same passages allows a shortest-path calculation after inadmissible edges have been removed. Under the stated correspondence and route assumptions, A–B–D has length 7 m and A–C–D has length 10 m. A robot-readable record and a diagram for a meeting can express the same route result while supporting different recipient operations. A publication provides a particular expression through an identified carrier under the applicable access conditions.
A check of the survey identifies a transcription error: the observation for B–D recorded 0.8 m, while the table copied it as 1.1 m. Recover the passage, measurement meaning and relevant conditions of that observation. Here the warehouse system participating in the inspection is the grounding holon; the survey report is an evidence account for the width claim. First repair the warehouse claim and its expression. Then return to the route operation: B–D is excluded, so the earlier 7 m route conclusion loses its premise and the remaining shown route is A–C–D. The route viewpoint can remain unchanged. The table layout can remain unchanged. A previous calculation that used only the lengths for a different, explicitly width-independent question may remain valid.
If the passage was instead narrowed after the earlier survey, retain the earlier observation as an account of its stated time and create the new current claim from the later inspection. That is a different change from correcting a transcription. The claim-to-world mapping, observation history and evidence-use question determine which treatment is warranted.
The general operation demonstrated here is to recover the changed claim and its grounds, follow the exact dependency and revise the affected result. The graph calculation supplies the subject-specific inference. The epistemic method supplies the connection between that inference, its source account, expression and later use.
Several changes require other branches of the same explanation:
- Dynamic source. A dashboard can replace the displayed width while the screen and source service continue. Recover the state and interpreted claim used by a calculation. A new redraw with the same relevant expression and claims leaves that calculation's input unchanged.
- Several materials in a release. A maintenance report can accompany the route account in one download. Identify the claims of each account and the expressions that make them available. Their common carrier does not supply a relation between their conclusions.
- New audience and shortening. A short guide that says “A–B–D is available” but removes the condition on clear width presents a stronger claim. Restore the condition or establish a different qualified claim; then revise the dependent route advice.
- Machine use. A parser needs an agreed relation between fields, units and passage identities. Reformatting is acceptable only when the required interpretation and operation are preserved.
- Normative account. A drawing can require B–D to have width at least 1.0 m while an inspection reports 0.8 m. The requirement, observation and report about the drawing have different claim content and checks. Use them together to determine the discrepancy.
- Small correction. Correcting a misspelling in “clear width” can finish after confirming that the heading still has its established meaning and that the consuming parser, if any, accepts it. The full publication analysis is needed only if the correction affects that use.
Read the relationships
The arrows below name relations between selected objects. They show how to interpret the connections, rather than an order of stages.
Constitution connects the claims, subject and effective interpretation into E. View membership classifies that same E by established conformance to P. Expression relates X to the claims recovered from it. Form conformance relates X to reusable organization rules. Bearing relates a carrier's state to the recoverable expression. Edition continuity relates two epistemes through actual revision and its continuity rule. The grounding branch connects designated empirical claims with observations or tests involving H; the evidence branch states what an account contributes to a claim's support. The selected claims in the drawing are part of E's content, not an extra constituent of its identity.
Publication combines the availability conditions involving E, X, the carrier, the audience and the intended use. To find the next needed Method, start from the unresolved relationship: C.2.1 for constitution or edition, E.17.0 for viewpoint conformance, A.6.3.RT and A.6.3.RT.OE for expression, and E.24.PUB for publication. The general path is complete when the reader has the account and the assurance needed for the current work.
An ontic, of kind U.Ontic under E.24, is a connected ontology unit. It coordinates a primary subject kind, the identity or recognition rule for entities of that kind, and the minimal relation set needed by named dependent uses. EpistemeOntic joins U.Episteme with constitution, grounding, edition, viewpoint-conformance and publication relations. It supplies their common basis while the pattern for each relation retains its own predicate. An episteme describing that ontic is another claim-bearing object, with the ontic as its subject.
Identify the next necessary relation
When the receiving question requires exact episteme identity, answer the following three questions. They are identity questions, not fields to fill. When the answers are still being developed, :4.1.1 connects the writing work in C.2.4:4.6 to the resulting constitution judgment.
If the task needs only the episteme's identity, stop after the three answers. Otherwise state the concrete receiving use. Ask for an unresolved uncertainty or choice only when that use is a real inquiry or decision.
Open a row below only when its first column names the reader's next sentence or action. Each positive answer adds an independently governed object or direct relation; none adds another slot or identity discriminator to the episteme.
Stop when no row describes the next sentence or action. A readable sentence naming the claims, EntityOfConcern, and effective reference scheme is then enough. Do not complete the table as a record. Use section 4.10 only when a later sentence or action actually needs the full relation and neighboring-object reference.
Identify the episteme by its constitution
The shared C.2.1 identity of one U.Episteme is:
claim content is the identity-bearing ClaimGraph carried as the episteme's constitutive claim structure. Each episteme selects one U.Entity as the EntityOfConcern of the current claim-bearing whole. Its ClaimGraph may also designate other independently governed entities as participants in relational, comparative, negative, counterfactual, or modal claims.
When the subject is an admitted relation kind, an already individuated obtaining relation occurrence, an admitted collection-as-whole, or another independently identified joint entity, that object may be the EntityOfConcern. Several participant designations do not by themselves constitute such an object. A negative or counterfactual relation claim can designate the relation kind and its participants in the ClaimGraph without requiring an obtaining world-side occurrence.
Split the ClaimGraph only when it combines independent subjects in a way that makes the selected EntityOfConcern untruthful or breaks the intended identity or comparison use; do not split merely because one predicate connects several participants. The reusable predicate-definition boundary is stricter: before publication, name one truthful EntityOfConcern and state what the definition claims about it. If no such single concern can be selected, remain at local compound-claim level.
The effective ReferenceScheme supplies the designation and interpretation rules needed to read this ClaimGraph as claims about this EntityOfConcern. Add measurement, comparison, or evaluation rules only when the claims' meaning uses them. The corresponding measurement or evaluation entities and relation occurrences remain under their direct patterns.
A theorem read under a formal vocabulary and calculus needs designation and interpretation rules; it needs measurement or evaluation rules only if its meaning uses them. By comparison, a pump-tolerance episteme uses the applicable units, measurement procedure, and pass/fail criterion to make its claim meaningful; the actual measurement and evaluation occurrences remain neighboring objects rather than episteme constituents.
Changing any identity discriminator yields another episteme. Changing a carrier, layout, rendering, publication occurrence, evidence item, viewpoint assignment, or model-use setting does not by itself yield another episteme. The pattern for a dependent episteme kind may recognize the same individual as a member through its stable membership condition, but it does not add another identity discriminator. A.3.2 governs U.MethodDescription membership; E.17.0 governs U.Viewpoint and U.View membership through fixed predicates over already identified epistemes. A.6.3 governs an optional viewing construction between source and receiving epistemes, not view membership. If any work or construction changes claim content, EntityOfConcern, or effective reference scheme, those changed C.2.1 discriminators identify the resulting episteme, not the dependent-kind label.
This identity is constructive. The claim graph and reference scheme are epistemic constituents; the EntityOfConcern remains an independently governed entity related through aboutness and reference. When EpistemeConstitutionRelation obtains, their organization yields the whole-level characteristic of being one interpretable claim-bearing whole. The relation occurrence and the resulting episteme are distinct but reidentified from the same three discriminators.
Obtain the content, recover its interpretation, and compare a change
Writing can help obtain the content before it is a finished knowledge object. Following C.2.4:4.6, put the cue or earlier expression beside a proposed claim. Recover what it is about, what is asserted, and the conditions that make that assertion meaningful. Compare the proposal with the observations and reasons that prompted it; use the subject's Method when a missing distinction requires further investigation. Keep unresolved alternatives visible when they affect the next use. A note or partial card can support this work before its content satisfies the episteme-constitution predicate.
Recover the whole asserted content. ClaimGraph names the semantic claim content organized as a whole: the claims, their participants and predicates, polarity, conditions, and the asserted relations among claims. Include quantification, modal force, and grounds asserted as part of that account where they change what it says. A field named claim may carry only part of this content. A graph representation can express it, but graph nodes and edges have the identity of that representation. An episteme describing this content is another object, with the described content as its EntityOfConcern. A separately obtaining justification or evidence relation remains separate unless the account also asserts something about it.
Recover the effective rules. ReferenceScheme names the rules that actually govern designation and interpretation of this content about its EntityOfConcern. Recover what the expressions designate and how their combinations, qualifications, and inferential uses are read. Include measurement, comparison, or evaluation rules only where the claim's meaning uses them. A document describing those rules is an expression of the scheme; a rule elsewhere in that document enters the effective scheme only when it governs this account. A citation can identify the rules when its selected content is unambiguous; otherwise state the missing rule. An unresolved choice between materially different readings leaves that part of the constitution unresolved.
Individuate content and rules before assigning a kind. A body of interpreted claim content and a collection of operative rules are abstract U.Entity individuals under A.1:4.1. For the content, the discriminator is the complete selected claims with their interpreted meanings and asserted organization; for the rule collection, it is the operative rules with their application domains and interpreted dependencies. These discriminators identify the individual independently of the text, graph, document or episteme through which a reader obtains it. Giving those claims or rules under an unambiguous interpretation can designate the individual directly; :4.2.6 also permits designation through an episteme edition and selection. Different such designations can return the same individual. Changing any of these identity discriminators identifies another content or rule collection; an expression change that preserves them does not.
This individuation and reference basis permits the content or rules to be an actual relation participant or the EntityOfConcern of an episteme about them. Neither a slot assignment nor a classification judgment supplies that basis. Entity status alone establishes no holonhood, episteme membership or obtaining relation. When the needed interpretation or identity remains unresolved, preserve the partial written result and recover the missing distinction before using it as a participant. That limit on the use does not prevent writing from helping obtain the content.
C.2.1 identifies R by its operative semantic rules, their application domains and interpreted dependencies. A notation's conventions can supply those rules, the means of expressing values under them, or both. The word convention therefore does not decide which identity changes.
The boundary. A convention belongs to the effective R when it supplies a rule needed to determine what the account's designations refer to, what its predicates and combinations mean, or how its qualifications, modalities or inferential commitments are interpreted. Its semantic rule, domain and dependencies enter the rule-collection comparison. A convention implements an expression of an already determined account when it maps signs or arrangements to the same already identified semantic values and leaves those rules, domains and dependencies intact. The corresponding mapping must be recoverable; calling a change a translation or rendering establishes none of its preservation conditions.
Apply this distinction to the actual meaning-bearing dependency. A layout rule that places one already identified value beside its label can implement expression only. A spatial rule that makes adjacency mean causation supplies semantics and belongs to R when the account uses that meaning. A fixed parser can recover an already defined semantic value from a serialization; a parser rule that changes the value, relation or interpretation cannot be omitted from the comparison simply by placing it in software. The same notation can have both kinds of convention.
Establish the interpreted rules before comparing the change. Do not redefine the source R after seeing the target so that equality becomes automatic. An alternative spelling of a rule is compared as an expression of that rule. An added semantic rule, a changed application domain or a changed interpreted dependency changes R. Two different rule collections can agree on a selected calculation, or even have a proved equivalence for a bounded use; that result does not identify the complete rule collections. When the semantic correspondence cannot be established, retain the useful partial comparison and leave E identity unresolved.
Why T and R still contribute when G is interpreted. G supplies what this account asserts, including the meanings of its predicates and participants. It does not by itself select which participant is the EntityOfConcern of the whole. The same relational content about pump P and valve V can support an account about P or an account about V when each selection is truthful; C.2.1 distinguishes those whole accounts by T.
Likewise, one interpreted assertion need not determine the full operative rule and domain under which it is stated. Rules that agree on the values used in one assertion can differ on another admissible input. R retains the actual semantic rules and dependencies needed for interpretation and reuse. It is compared independently of the assertions selected into G. Neither difference turns T into a constituent or makes R another description episteme.
The cases below fix the relevant assumptions before judging identity:
Full identity uses semantic equality of the actual constituents and T. It neither uses byte identity nor identifies distinct schemes merely because they answer one question alike. A source-preserving expression can be constructed under A.6.3.RT or A.6.3.RT.OE without constituting another E. A B.1.3 compilation whose target scheme S differs from the source R instead returns an episteme identified with S and the resulting claim content. The preserved-content and reliance judgements still have their own scope.
Choose the precision the receiving work needs. Express the claim and rules well enough to perform the named comparison, interpretation, or revision. Ordinary prose, a diagram with a clear interpretation, or a partial written comparison can be sufficient. Add a C.3.2 classification only when the receiver needs a membership, substitution, or quantified-range judgment; identify the candidates and criterion for that particular classification. Classifying whole epistemes does not establish a kind for each of their constituents. Section :4.2.1 governs a receiving use that specifically needs typed constitution participants.
Judge identity separately from admissibility and use. For full episteme identity, compare the complete claim content and its semantic organization, the EntityOfConcern, and the effective interpretation rules. Supply the correspondence that makes the preserved claims and their conditions recoverable on both sides. In a formal subject, the declared semantics and a proof or exhaustive comparison may supply it. In ordinary language, use the subject's distinctions and stated bounds to compare what each account commits to; expose an ambiguity that prevents the comparison. Agreement on selected questions establishes only that bounded agreement. A conservative derivation may omit content and therefore does not by itself establish full identity.
A change to asserted meaning or grounds is a content change even if the principal sentence remains. Choosing a different interpretation rule changes the effective scheme even if all signs remain. Making an already determined meaning easier to recover can preserve both; adding a previously unsettled condition or commitment requires a new identity judgment. When only part of the account has been compared, preserve that partial result without asserting equality of the whole. There is no general procedure here for deciding equality of arbitrary natural-language meanings.
Reconsider what depended on the changed value. Use :4.8 to locate the change. A :4.2.5 ClaimAddress continues to designate its claim in the original episteme while the edition reference resolves and that ClaimGraph declares exactly one claim with the intrinsic identity. To use corresponding content in another episteme, establish the claim-to-claim correspondence and resolve an address in that episteme; the old address remains an address in the original episteme. Changed claims or interpretation can defeat a proposed transfer of empirical coverage, an A.6.6 claim selection, or a predicate evaluation. Retest the affected viewpoint-conformance, edition-continuity, or receiving-use conclusion under its own rule. Reuse an unaffected part only with the correspondence and conditions that support that reuse. A different expression or carrier with the same complete content and scheme leaves episteme identity intact, while its suitability for the receiving action still needs that action's criterion.
Declare the range needed by a receiving interface
Use an exact value directly when the receiving question concerns that value alone. A fixed-scheme operation can put its rules in the operation declaration, as in B.1.3. A reusable typed interface with a variable argument or replaceable participant needs a range: the receiver must be able to distinguish an eligible replacement from an expression that merely looks similar. State that distinction before choosing a type name.
For the recurring interpretation and claim-content uses below, the following declarations identify two C.3 kind individuals over the abstract entities individuated under :4.1.1. The candidate domains retain the distinction between those entities and their documents, graph representations or describing epistemes. Each row supplies the candidate interpretation and membership rule of its C.3.2 KindSignature; no context-slice-dependent extension is needed for these content-fixed individuals.
Under these candidate interpretations, EpistemeInterpretationScheme is a subkind of U.Entity, and NonemptyClaimContent is a subkind of U.Entity. C.3.1's criterion-entailment branch applies: every member must have the independent semantic identity and reference basis in :4.1.1 and therefore satisfies A.1:4.1. A.6.5 can use either range as an entity ValueKind; actual participation still requires the receiving relation's predicate to obtain.
Recover the candidate independently of the classification. C.3.2 first checks that the proposed candidate is a value in the stated domain; then it judges membership as true, false or unknown. Missing knowledge of a needed rule or content relation leaves that judgment unknown. It does not turn a known document into a rule collection, nor turn an unidentified expression into a fully identified value.
For scheme identity, compare the actual rules and domains, including the dependencies that determine their meaning. A restatement can preserve the scheme when an explicit alignment preserves those rules; agreement on a few outputs cannot establish that identity. A changed interpretation rule or domain identifies another scheme even if its name and document remain. For content identity, apply :4.1.1 to the whole selected content. Two designations from different source editions may return the same content under the same scheme. Source, acceptance, currentness and witness qualifications attached to a designation do not enter the designated semantic base merely because they accompany it, as in A.6.6. When the selected content itself asserts such a fact, that assertion belongs to the content being compared. A change of graph encoding preserves content only when the interpretation supplies the complete correspondence; graph shape alone supplies none.
The kinds themselves continue when their candidate domains and operative membership distinctions remain the same under C.3.1's comparison. Rewording the declaration or citing it from another pattern need not create another kind. Replacing the semantic rule/content criterion with the presence of a schema label would create another classification distinction. A change of one member's rules or claims instead identifies another value; it need not change the kind.
These declarations supply a shared range for the receiving interfaces that cite them. A.1.1 uses EpistemeInterpretationScheme for its coherence participant R; A.2.6, A.19.CPM and A.19.SelectorMechanism use it for a bound scheme argument. A.6.0's R7/R8 predicates use NonemptyClaimContent for exact dependent content and semantic bases. Reusing one criterion preserves the same meaning of admissible input across those interfaces. A receiver adds any narrower language, semantic-domain or representation condition its operation actually needs; that condition does not become another kind merely because another pattern states it. A recurring independently different classification requires its own C.3 declaration.
Test replacement in the receiving use. Resolve the proposed value, check the declared range and any narrower input conditions, then apply the receiver's own predicate or operation law. An eligible scheme need not interpret every actual input. Two eligible schemes need not be equal or admit a meaning-preserving translation. An eligible content value need not be a sufficient basis or an actually used premise.
Missing interpretation stops the dependent operation; a false subject predicate remains a different result. For a cross-scheme use, establish the required correspondence, F.9 Bridge and bounded-use assertion where the receiving rule requires them. To rely on that Bridge-use assertion, recover the A.10 evidence-provenance path and RelianceDisposition for the same bounded use; add B.3 only when a named assurance claim is at issue. Membership supplies none of these results.
Retest dependent bindings and conclusions when the resolved value or a used rule changes. Reuse unchanged content only with its established correspondence and applicability. Keep the old value and its qualified conclusions available instead of rewriting them as statements about the replacement. These bounded C.3 declarations do not admit public U.ReferenceScheme or U.ClaimGraph; E.24.UK governs any separately needed durable U-kind proposal.
Govern the core direct relation
Tech name: EpistemeConstitutionRelation.
Plain reading: these claims, under this reference scheme, are claims about this entity and together constitute one episteme.
Participants and the shared reusable-declaration rule
The constitution predicate uses three actual participants, each a U.Entity. The content and rule collection have the independent semantic identities given in :4.1.1; the EntityOfConcern has its own subject-governed identity. Recover each before judging whether the relation obtains:
This is sufficient for a direct constitution judgment with values. A reusable typed interface has an additional question: which candidates may replace a participant under its declared range? For that use, A.6.5 requires a complete SlotSpec with an independently governed ValueKind and designation mode. State the candidate domain, membership criterion, identity rule, and applicability needed by the receiver; use C.3.2 for a local classification and E.24.UK for a proposed public U-kind. Carrying the designation ByValue does not remove the membership requirement.
For reusable designation of the EntityOfConcern, EpistemeConstitutionRelationSignature declares EpistemeConstitutionRelation as its subject and U.Entity as the ranged kind for that participant. Its vocabulary includes the relation designator and the SlotSpec below; its laws are the obtaining and participant-determined identity rules in :4.2.2. Its applicability is the reusable designation and typing of the EntityOfConcern in a constitution assertion or description whose claim content and effective rules are recovered under :4.1.1. This declaration's own claim content, relation-kind EntityOfConcern, and effective FPF interpretation identify one declaration episteme; these subject, range, vocabulary, laws, and applicability contents satisfy A.6.0's U.Signature predicate.
Use this SlotSpec for the EntityOfConcern designation in A.6.REL:5.5 or B.2:5.1. An eligible designation must resolve to an independently identified entity; that typing alone does not establish that the claims concern it or that the constitution predicate obtains. The other two participants remain necessary constituents. Section :4.1.2 supplies bounded C.3 ranges for a receiving typed use; this EOC declaration does not thereby acquire further SlotSpecs. A use that additionally needs ClaimGraphSlot or ReferenceSchemeSlot must select the range appropriate to its participants and complete its declaration under A.6.5. If the available range does not cover that use, establish the missing criterion before claiming the stronger interface. E.24.UK:4.2c separately retains the public U-kind questions. A card field or tuple position acquires none of this declaration meaning by its label alone.
The separately needed EpistemeEmpiricalGroundingRelationSignature and EpistemeEditionRelationSignature remain declarations over the governed episteme and holon kinds stated in :4.3 and :4.5. Each declaration is first a C.2.1 episteme: its own claim content declares the relation, its EntityOfConcern is that direct relation kind, and its effective scheme gives those declarations meaning. The A.6.0 predicate judges that same individual's U.Signature membership; RelationSignature is its relation-facing use, with no second identity.
A complete declaration states its relation-kind designator, the SlotSpecs required by its typed use, direct predicate, occurrence identity, applicability, and any actual dependencies or provided names. A manifest may expose those dependencies or names when a consumer needs them. A readable one-off assertion needs neither a signature nor a manifest; assertion wording does not make its relation obtain.
Obtaining and occurrence identity
EpistemeConstitutionRelation obtains exactly when the effective reference scheme supplies a coherent designation and interpretation of the claim graph as claims about the EntityOfConcern, and the three participants are constitutively organized as one claim-bearing whole whose claims can in principle be evaluated under that scheme. Merely placing three designations in a card does not make the relation obtain.
The relation occurrence is participant-determined by the <ClaimGraph, EntityOfConcern, ReferenceScheme> triple. The same triple cannot constitute two distinct U.Episteme instances under the shared C.2.1 identity rule. Recognition of that individual as a dependent episteme kind adds a membership judgment under the dependent kind's subject pattern, not another constitution occurrence or discriminator. A tuple may represent the triple, but tuple order and storage keys contribute nothing to identity.
The episteme and the relation occurrence are not identical. The relation is the obtaining organization among the three participants. The episteme is the knowledge holon constructively identified through that organization and its whole-level claim-bearing characteristic.
Ordinary assertion, classification assertion, and explicit occurrence use
An ordinary assertion can state that claim content concerns an entity under a scheme without explicitly naming a relation occurrence. For every direct predicate, keep four jobs separate:
- The pattern for that predicate defines participant meanings, the obtaining predicate, applicability, and the occurrence-identity rule.
- The current case supplies the facts that satisfy or fail that predicate.
- The assertion carries affirmative or negative polarity.
- A separately governed evaluation or evidence-use judgment assesses support for the assertion and the limits of reliance when the receiving use needs it.
When a receiving relation or claim needs the exact constitution occurrence, inspect the current ClaimGraph, EntityOfConcern, and ReferenceScheme facts against C.2.1's predicate. Only after those facts satisfy the predicate may the participant-determined identity rule individuate the occurrence for designation. The assertion, designation, occurrence, case facts, and reliance judgment remain different objects.
Each classification judgment has one pattern governing its criterion. A.1 governs constructive recognition of a candidate as an instance of an already admitted holon kind. C.3.2 governs a local-kind membership judgment. E.24.UK governs the ontology-level decision that admits a public U-kind; it does not classify a project candidate. None of these judgments is a direct admission relation created by C.2.1.
When project work needs a classification judgment as a separately reviewable claim, identify one claim-bearing episteme whose exact EntityOfConcern is the candidate entity. For an admitted holon kind, its claim content states affirmative or negative polarity for the exact classification predicate, names the kind, cites the A.1 constructive criterion and any kind-specific criterion, designates the direct part-relation occurrences used in the assessment, and cites any evidence-use relations needed to make support for the classification and the limits of reliance inspectable for the declared use. For a local kind, its claim content states the same polarity distinction for the candidate, local kind, selected KindSignature edition, context slice, and judgment governed by C.3.2; its reliance posture remains separate. A value classification inside another claim can remain claim content of that episteme instead of fabricating a value-shaped EntityOfConcern.
The assertion does not create the candidate, admit a U-kind, or make the candidate change kind when an FPF host is renamed or republished. For example, the assertion that Pump #37 satisfies the constructive U.System criterion may be revised when evidence changes, while Pump #37 and the criterion it satisfies retain their independently governed identities.
A card that calls a listed collection a holon is still only a classification assertion episteme. Its assertion polarity is affirmative, but the card alone leaves reliance unresolved for any use that requires A.1 to recover the exact constituents and grounded part relations, their constructive assembly, the whole's reidentification rule, actual compatibility with a governed larger-assembly construction, a composition-grounded whole-level characteristic, and the already admitted holon kind with its kind-specific criterion. The card form supplies none of those facts and does not make the classification predicate true.
The same constitution rule applies when a reader proposes to use an obtaining F.9 Bridge. Say first in ordinary words what the reader proposes to compare, substitute, translate, publish, or otherwise do; name the direction d, use-specific correspondence rule r, tolerated semantic loss t, and affirmative or negative polarity for named use u. Identify that statement as one ordinary C.2.1 assertion episteme: the exact Bridge b is its EntityOfConcern, its ClaimGraph designates <u,d,r,t> and polarity, and its effective ReferenceScheme makes those designations, the rule, and the tolerance interpretable. The exact <ClaimGraph, b, effective ReferenceScheme> triple identifies the assertion. Changing u, d, r, t, or polarity changes the claim content and therefore the assertion episteme, not fixed Bridge b. Keep this local claim form in ordinary wording: it introduces no public U-kind, universal use relation, or durable CamelCase claim name. Reopen F.18 only if an independent later use actually needs a reusable name.
An affirmative bounded-use assertion is one premise for that use; it is neither permission nor proof that the use occurred. A negative assertion leaves an otherwise obtaining Bridge in place. Use A.10 to classify ordinary bounded reliance on the exact evidence-provenance path: pass supports only the named use, degrade supports only its named narrower use, and another disposition supplies no support for the attempted use.
Use B.3 only when a named assurance claim about this bounded use is at issue. That assurance result remains separate from the Bridge-use assertion and adds assessment Work, System, Method, assignment, bindings, witnesses, or a reusable note only when the assurance use depends on those identities. A domain rule may require an assurance claim for a consequential use, but neither consequence nor a display creates the claim. Neither the A.10 nor B.3 branch authorizes the use. If the use occurred, recover the Work under A.15.1, assertion episteme under C.2.1, publication occurrence under E.17, direct relation under its domain predicate, operation application under A.6.1, or another result under its own pattern.
State rule-content and subject assertions
In ordinary prose, cite the PatternID and state the concrete contribution: what the cited content defines, constrains, tests, distinguishes, or helps the practitioner do. This readable branch is normally sufficient. For example, “use A.1 to test constructive holon recognition” states the needed contribution.
Open an exact defining or constraining episteme edition or ClaimGraph only when its identity changes interpretation, migration, conflict analysis, publication, dependency repair, or reuse. Then identify the subject, predicate or constraint, polarity, exact defining content, case facts, and only the scope, time, scheme, or bounded-use qualifications that change the assertion. Do not fabricate an assertion episteme merely to avoid an ordinary pattern citation.
Definition or constraint is not actual rule-content use. State derivedUsingRuleContent(dependentContent, baseContent) only when one identified derivation claim used the exact nonempty base subgraph as a formal premise under a declared inference rule. State evaluatedAgainstRuleContent(dependentContent, baseContent) only when one identified criterion-selection claim selected that base for one exact bounded evaluation. Consultation, influence, quotation, provenance, evidence, evaluation Work, and later sufficiency establish neither predicate.
An E.4.PFR row is optional and opens only for a named framework-maintenance, edition-impact, comparison, publication/dependency-repair, or refresh receiver. It represents an already identified assertion; it creates neither the assertion nor a pattern-owner fact.
Address one claim inside an exact episteme edition
U.EpistemeRef is the admitted RefKind for designating one already identified U.Episteme. Under the effective reference scheme of the receiving assertion or description, its resolution method returns exactly one episteme satisfying the C.2.1 identity rule. A value that resolves to none or more than one is unresolved. The reference, its token or serialization, the resolution act, and the episteme remain different objects. Retargeting the reference designates another already identified episteme; it does not revise either episteme.
Use the reusable C.2.1 value ClaimAddress only when a receiving claim or work item needs one exact claim inside a larger ClaimGraph:
The second component is not a printed node label interpreted by the episteme's general ReferenceScheme. It is a claim identity that the exact ClaimGraph itself declares and preserves across its admissible representations. Resolve the edition first, then require that its ClaimGraph contains exactly one claim with that intrinsic identity. Resolution fails when the edition is unresolved, the identity is absent or non-unique, or the token belongs only to one rendering or serialization.
Two ClaimAddress values are equal only when they resolve the same exact episteme edition and the same intrinsic claim identity in that edition. Reusing the same visible token in another edition does not preserve the address. An EpistemeEditionRelation also does not preserve it by itself; a receiving migration rule must state any claim-to-claim correspondence it uses.
When a ClaimGraph declares no stable intrinsic identity for the needed claim, cite the whole episteme or constitute the claim as its own C.2.1 episteme. Do not invent an address from a heading, row number, file location, or display token.
C.2.1 ClaimAddress designates claim content carried by the exact edition. It is neither a U-kind nor a RefKind, turns no claim content into a U.Entity or another U.Episteme, and carries none of the claim content itself. Use U.EpistemeRef for the whole episteme and the admitted reference kind for an independently identified entity or relation occurrence.
Resolve a constituent designation to its exact value
Use a constituent designation when the receiving work needs to recover content or rules from an addressable episteme instead of carrying them by value. Ordinary writing can continue with the expressions and rules available in :4.1.1; it need not create a reference or a finished episteme merely to fit these forms.
Resolve the exact episteme first under :4.2.5. For ClaimContentDesignation, recover its whole claim content or the uniquely declared selection, retaining the selected claims' semantic qualifications and selected interclaim relations. A single intrinsic claim can use a :4.2.5 ClaimAddress as the selection. A subgraph selection must likewise be declared by the exact ClaimGraph and resolve uniquely across its admissible representations. A row number, heading or diagram position alone supplies no such identity. When that selection is unavailable, carry the exact content by value, cite the whole episteme, or constitute the needed content as its own episteme.
For SchemeDesignation, effectiveScheme selects the rules governing the source episteme itself. schemeDefinedBy selects content that defines one exact rule collection, with its domain and semantic dependencies; interpret that defining content under the source episteme's own effective scheme. A document can describe a different scheme from the one under which the document is read. A mention, name or example of a scheme is insufficient to select all of its rules. Follow an imported rule only through its exact source and interpretation; an unresolved or circular definition with no independently recoverable rules does not resolve the scheme.
Resolution succeeds only when the exact edition, selection and interpretation determine one value. Preserve an unresolved result when the source is absent, selection is ambiguous or a required rule is missing; do not silently choose the latest edition or a nearby document. The receiving interface then checks its declared range and use conditions under :4.1.2. Resolution alone establishes neither membership nor successful use.
These are composite designation values, not new U-kinds or RefKinds. Use an explicit designation field, such as claimContentDesignation or schemeDesignation, with the applicable C.2.1 form; an A.6.1 ByRef argument can specify this resolution method while keeping the resolved value as its argument range. The admitted U.EpistemeRef inside the form retains its own referent range. Two designations are the same designation only when they select the same constituent through the same exact episteme edition and selection. Different designations can resolve the same semantic value. Retargeting changes the designation and requires resolution again; it changes neither the old episteme nor the old value.
Apply the distinctions together. A declaration D describes a rule set S in ordinary mathematical prose, while S declares a propositional language with material implication and the inference rule that permits q from p and p => q. Selecting D's effectiveScheme returns the rules for reading D; selecting the exact content that defines S returns S. A formatting-only stylesheet is an identified rule collection but fails EpistemeInterpretationScheme membership. An exact partial S can pass membership and still leave unless uninterpreted in a proposed input; a consuming operation requiring that interpretation stops.
For an A.1.1 comparison, let R be the complete S and let M and E have the same effective S. Here the actual participant R is S, the abstract rule collection described by D. R passes the declared range; the coherence claim still requires its independently admitted predicate P to hold for the exact M and E content. Changing the interpretation of => while retaining the signs yields another R and requires a new coherence judgment, including the bridged branch if its conditions now apply. Reprinting D while preserving its complete defining content changes no R.
For an A.6.0 basis use, let the exact dependent content be q and the selected base be the two claims p and p => q, under the stated inference rule. Both are NonemptyClaimContent. The base can be sufficient for deriving q, but R7 actual use additionally needs the identified derivation claim that used those premises. Replacing the base with the single claim p preserves membership and loses that sufficiency under the stated rule. Adding an asserted new reason to the dependent account changes its content even if the field still displays q. Two exact editions that designate the same unchanged base under S provide two source qualifications of that semantic base, not two bases merely because the addresses differ. An episteme asserting that this two-claim base is sufficient for deriving q under S can take that exact abstract content entity as its EntityOfConcern. That assertion has its own claim content and effective scheme; it does not replace the base it concerns.
Add empirical grounding through its own relation
Tech name: EpistemeEmpiricalGroundingRelation.
Plain reading: these designated empirical claims of this episteme are inspectable through exact observation, intervention, measurement, or test relations involving this grounding holon.
C.2.1 uses explicitly designated partial coverage. For one candidate grounding occurrence, select one exact claim subgraph C from the episteme's already constitutive ClaimGraph. C may include formal claims but must contain at least one empirical claim. The occurrence says that every empirical claim in C is grounded; it says nothing about empirical claims outside C. To claim full empirical grounding for the episteme, C must contain every empirical claim in that episteme. Purely formal epistemes or selected claim subgraphs fall outside this relation's applicability; their formal derivation and any assurance remain under their applicable rules.
For every empirical claim in C, state a concrete claim-to-world mapping under the episteme's effective ReferenceScheme to independently governed direct observation, intervention, measurement, or test relation occurrences involving the exact grounding holon. The mapping names what observation, intervention outcome, measured characteristic, or test result bears on that claim. One measurement involving the same holon cannot ground an unrelated claim.
The selected claim subgraph is by-value predicate content drawn from the episteme's ClaimGraph; it is not a third world-side relation participant, another episteme constituent, or a new U-kind. An assertion about grounding designates that subgraph and the claim-to-world mappings but creates none of the mapped occurrences.
Applying the declaration rule in 4.2.1, EpistemeEmpiricalGroundingRelationSignature is one declaration episteme whose exact EntityOfConcern is EpistemeEmpiricalGroundingRelation; the same individual has U.Signature membership and relation-facing RelationSignature use only under A.6.0. Its complete declaration includes the covered-claim-subgraph rule, obtaining predicate, maximal-continuous-interval identity rule, applicability, actual dependencies and provided names, and these participant SlotSpecs:
EpistemeEmpiricalGroundingRelation over participants (E,H), with covered=C, obtains exactly while exact covered claim subgraph C contains at least one empirical claim and every empirical claim in C has a current claim-to-world mapping to the required independently governed direct observation, intervention, measurement, or test relation structure involving H under E's effective ReferenceScheme. Every mapped relation required by that coverage must obtain. An exact direct evaluation relation counts as part of the empirical test only when the mapping states its concrete use in that test; otherwise evaluation and evidence can support or challenge an assertion about grounding but are not its world-side base.
One occurrence is identified by <episteme, exact covered claim subgraph, grounding holon, maximal continuous interval during which the complete coverage predicate is true>. Closing the open end of that interval refines the description of the same occurrence. Demonstrated failure of any required mapping followed by restored complete coverage yields another occurrence.
Evidence or evaluation availability alone establishes neither obtaining nor nonobtaining and proves no temporal gap.
- If the complete coverage predicate is known to obtain, grounding continues without a stored report or work log.
- If it is known not to obtain, the relation does not obtain.
- If its truth is unknown, an affirmative grounding assertion has unresolved reliance for the declared use; that posture is not a third world-side grounding state.
The grounding holon need not be identical to the EntityOfConcern. One method-description episteme may have one grounding occurrence for a claim subgraph mapped to exact enactment work and another for a different claim subgraph mapped to the system whose behavior was observed. Each occurrence names its own C, H, and mappings. Sharing one grounding holon makes comparison inspectable but proves neither the same subject, the same claim content, nor coverage of any unlisted claim.
Keep neighboring uses under their direct relations
Names ending in Slot are admissible here only as SlotKinds inside the exact RelationSignature governed by the neighboring direct relation pattern. A card or other episteme form carries participant designations in ordinary fields; it does not acquire SlotKinds by using similar field labels. None of those neighboring SlotSpecs belongs to EpistemeConstitutionRelationSignature.
Relate distinct episteme editions explicitly
Tech name: EpistemeEditionRelation.
Plain reading: this later episteme continues this earlier episteme as an edition under one applicable continuity rule.
EpistemeEditionRelation has exactly two direct participants. Applying the declaration rule in 4.2.1, EpistemeEditionRelationSignature is one declaration episteme whose exact EntityOfConcern is EpistemeEditionRelation; the same individual has U.Signature membership and relation-facing RelationSignature use only under A.6.0. Its complete declaration includes the direct predicate, participant-determined identity, applicability, actual dependencies and provided names, and these SlotSpecs:
The relation obtains only when all of these conditions hold:
- the two epistemes have different C.2.1 identities;
- the later episteme actually uses the earlier episteme as the source for the claimed revision;
- one applicable edition-continuity policy or rule states which claim, EntityOfConcern, and effective-reference-scheme features must be preserved, which may deliberately change, and what counts as continuation for this episteme family;
- the exact preserved and deliberately changed features satisfy that rule;
- no failure condition in that rule classifies the case as a fork, translation, retargeting, or independent reconstruction instead.
Work, an enacted Method, provenance, and change results supply case facts. Their labels do not make continuity true. C.2.P may recover the source expression and source-to-revision use; the direct change patterns supply exact changed features. If the continuity claim separately consumes a first-existence fact, apply the shared boundary in 4.9. A missing required rule or fact blocks only that positive edition claim.
One occurrence is identified by the exact <earlier episteme, later episteme> pair. Two revision Work occurrences do not create two edition occurrences for the same pair. The relation is acyclic in its earlier-to-later direction. A renamed file, later publication, shared title, bare provenance edge, or Method named “revision” establishes no occurrence.
When comparing these epistemes as models, use B.5.MR to establish refinement or abstraction for the receiving inquiry. That comparison uses their interpreted assertions and admitted answers. It can hold between independently produced models; conversely, a later edition can correct an earlier assertion without preserving it.
Several edition occurrences form a lineage structure only when a receiving use depends on their organization. A separately identified edition collection remains under A.14; collection membership does not establish continuity. PhaseOf may describe one unchanged episteme over a proper interval but does not connect two different C.2.1 identities.
When claim content, EntityOfConcern, or effective reference scheme changes, the later object is another episteme. Apply A.6.4 separately when the current claim is effect-free retargeting between epistemes with different EntitiesOfConcern; retargeting alone does not establish edition continuity. A changed publication form alone identifies neither another episteme nor an edition relation.
Keep descriptions, cards, publications, and representations downstream
The organized claims expressed by a filled card can constitute an episteme when G, T and R are recoverable. The particular signs and arrangement supply expression X; reusable arrangement rules can supply F; a selected graphical or tabular construction can represent its subject through an identified correspondence. Use C.29 when a mathematical-lens choice, transfer or reliance question remains. Identify the object selected by the actual question.
When a card or other form designates the participants of one direct relation, its field labels may correspond to SlotKinds in that relation's RelationSignature, and its field values may be by-value designations or references of the declared refModes. The form is not a filled direct relation occurrence; supplying fields does not make the predicate obtain or provide occurrence identity.
In a relational assertion, the claim graph designates the actual participants and states affirmative or negative polarity for the direct predicate. The direct pattern defines that predicate and the occurrence-identity rule; current case facts determine whether the predicate is satisfied or failed. For a forecast, scenario, counterfactual, permission, or another claim family, name its governing pattern rather than using one common catch-all field. When explicit evaluation is needed, judge support for or refutation of the assertion under its direct rule or receiving evaluation. Use A.10 when the declared use needs a replayable source-to-use account and bounded reliance judgment. An affirmative assertion may designate an occurrence only after the case facts satisfy the predicate and the direct identity rule individuates that occurrence; a negative assertion creates no failed world-side occurrence. In a relation-occurrence description episteme, the EntityOfConcern is that exact already individuated occurrence. The assertion and description retain their own C.2.1 identities. Their forms do not by themselves supply case facts or the occurrence-identity rule, or establish obtaining; evidence supporting the case conclusion and any constitutive contribution are governed separately by the relevant evidence and direct-relation rules.
A designator designates an already recoverable referent. A governed reference resolves to that referent under an effective reference scheme.
For publication, E.17 and E.24.PUB govern the occurrence that makes a selected episteme available through one exact expression and carrier to a declared audience for a bounded use. A reusable form constrains the expression when selected. Plain published episteme names the episteme's contingent participation in that occurrence. Its C.2.1 identity stays fixed before, during and after that availability relation.
One completed inspection card shows why the distinctions matter. Its filled claims can identify one episteme; its reusable layout can be a publication form; its paper sheet or file can be a presentation carrier; and a publication occurrence can make the selected card episteme edition available to the maintenance team. None of those uses makes the others identical.
Republishing the same claims about the same subject under the same effective scheme with another form or carrier preserves the episteme. Changed claim content, subject or effective scheme identifies another episteme. If the account also needs the rendering Work, use the conditional Work branch in :4.0.
A tuple can represent the identity triple and a graph or hypergraph can represent claim, justification, dependency, or relation structure. Use C.29 for an unresolved mathematical-lens choice, transfer or reliance question. ClaimGraph names the semantic content being represented; JustificationGraph names the separately governed inferential organization. Nodes and edges in a selected graph remain representation elements. An explicit correspondence can relate one selected representation element to an independently recovered object, but it neither identifies the two nor makes the representation element a participant of the represented direct relation.
Preserve description and meta-description recursion
If episteme E1 describes pump P, P is the EntityOfConcern participant in the constitution relation that identifies E1. If review episteme E2 describes E1, then E1 is the EntityOfConcern participant for E2. The two relations have different triples and therefore identify different epistemes.
An episteme may describe itself when its own identity remains recoverable. Self-reference never closes an assurance argument by itself. Each justification or evaluation path terminates in independently governed evidence, observation, or formal derivation rather than in a cycle of claims that cite one another.
An episteme used as a description is admitted for specification use only when the :6 conditions are satisfied: checkable claims and a named harness or validation relation. If the relying use selects a viewpoint, name that describing use and preserve or update its exact selection only when the selection affects reliance. Formal notation alone does not grant specification use or change the episteme's kind.
Locate the change before updating episteme identity
Hand episteme transformations to their subject patterns
Shared identity-inception boundary. Work or transformation can explain how an entity came about, but C.2.1 by itself establishes neither when that entity first existed nor that the work caused its inception. Open this boundary only when a current receiving claim asks whether a new entity began. Then use the subject's direct inception rule: its pattern defines the predicate and identity rule, and current case facts must satisfy them. If no such rule is recoverable, return one missing-governor blocker naming the entity or candidate basis, available case facts, required inception predicate, and receiving use. When the current question is only changed episteme identity, form, representation, view, or publication, do not open this boundary.
When inception is attributed to Work, apply A.15.PROD:4.5. Begin with the subject's exact productIdentitySpecification and its named applicability predicate or filled local claim for the candidate basis, subject context, and boundary. Then recover actual Work effects and the declared Work-to-change and change-to-identity bases, locate the earliest satisfying inceptionBoundary, and identify exact identityClosingWork. Designate the resulting entity only on the after-side of that boundary and retain the exact specification and applicability basis used for the claim. A missing required basis stops the positive claim; missing recoverable Work granularity for identityClosingWork returns that section's work-granularity blocker.
A.6.2-A.6.4 define episteme-to-episteme morphing, source-to-receiving viewing construction, and retargeting. Identify every source and receiving episteme independently under C.2.1 before testing the exact transformation relation. Each transformation pattern states which identity discriminator is preserved or changed and names the exact correspondence, reinterpretation, or retargeting relation on which it relies. When a possible Bridge between exact F.17 cells is at issue, use F.9 to test whether that relation obtains. If the morphism relies on that Bridge for a proposed use, state a separate C.2.1 assertion with the Bridge as EntityOfConcern and <u,d,r,t> plus polarity in its ClaimGraph, then recover the A.10 evidence-provenance path and RelianceDisposition for that same bounded use, adding B.3 only when a named assurance claim is at issue; none of those facts makes the morphism application occur. Categorical function, mapping, or tuple notation creates no direct relation occurrence.
For an A.6.3 source-to-receiving viewing construction, the two identified epistemes have the same EntityOfConcern, and the receiving claims are licensed under the viewing rule by the source together with any additional named sources and admitted facts that rule requires. Claim content or the effective scheme may be restricted without changing that subject. E.17.0 alone judges whether the receiving episteme conforms to an exact viewpoint and therefore has dependent U.View membership. Direct authoring or query generation can yield a candidate episteme without an A.6.3 construction, and neither route creates a multi-view family. For retargeting, the EntityOfConcern changes and the case names the exact domain correspondence, retargeting rule, or relation on which it relies. An F.9 Bridge is additional only when the case separately asserts a semantic relation between exact F.17 local senses from different semantic contexts and the F.9 predicate obtains. For a same-EntityOfConcern representation transition, use A.6.3.RT; the represented episteme may remain unchanged while the way it is represented and the available operations change.
Relation and neighboring-object reference
This reference table keeps the neighboring objects and relations visible after the application method. Ordinary prose names only the current object and its direct relation. A sentence such as “Model M concerns Pump P under Scheme S” is sufficient until another use needs explicit empirical grounding, a classification assertion, occurrence identity, edition continuity, publication, or representation correspondence.
Semantic triangle as a didactic projection (informative)
The Symbol-Concept-Object triangle is a teaching projection, not the episteme ontology.
A triangle diagram may be used to introduce expression, meaning, and subject if its caption says that it compresses the C.2.1 episteme ontology. Its corners and arrows are representation elements. They supply no SlotKinds, direct relation occurrences, or identity rules.
This limitation matters in practice. A proof-assistant term, wiring diagram, clinical chart, learned embedding, or verbal explanation can all occupy the Symbol corner while supporting different operations and different losses. Those questions belong to the patterns for choosing and transforming representations (C.37 and A.6.3.RT); one geometric picture does not answer them.
Description and specification-use boundary (normative)
In a claim-bearing-result use, “description” designates episteme E, a U.Episteme identified by its claim content G, EntityOfConcern T and effective ReferenceScheme R. The word supplies no additional subtype criterion or relation. For a comparison of the claims alone, identify G; for a correction to wording or layout, identify the expression and the meaning it must preserve. Section :4.0.5 explains how to recover the intended object from the actual use.
When these values affect the current description use, keep them recoverable:
For specification use, recover all applicable conditions:
- the exact description episteme and its C.2.1 constitution;
- checkable claims, invariants, criteria, or acceptance conditions in its ClaimGraph;
- a named harness, validation, conformance, measurement, or evaluation relation capable of checking those claims for the stated use;
- when viewpoint selection affects reliance, the named describing use and its exact selected viewpoint are preserved or explicitly updated.
Declared formality, notation discipline, comparators, tolerances, and measurement rules are named when the claims depend on them. They do not substitute for the checkable claims or the harness. If the conditions are absent, call the episteme a description and present proposed criteria as proposals; a Spec suffix, schema, signature, approval, or publication does not supply specification force.
Specification use does not create another episteme identity. A revision that changes ClaimGraph, EntityOfConcern, or effective ReferenceScheme identifies another episteme under C.2.1; a changed harness, evaluation result, publication, or relying use changes its own neighboring object or relation.
E.10.D2:4.1.3 applies these conditions to a proposed ...Spec name.
Self-description uses the same rule. If an episteme describes itself, its EntityOfConcern designation resolves to that episteme. If a review episteme describes it, the review episteme has the first episteme as EntityOfConcern and its own claim content and reference scheme.
Episteme morphing, viewing, and retargeting (normative)
C.2.1 governs episteme identity discriminators and neighboring relations. A.6.2-A.6.4 govern transformations between epistemes.
Effect-free episteme morphing
For a morphism from episteme X to episteme Y, state by value:
- which of claim content, EntityOfConcern, and effective reference scheme are preserved, restricted, bridged, or changed;
- any exact viewpoint selected for the named use, and any empirical-grounding, claim-scope, model-use, evidence, or representation occurrences that the arrow rule reads or compares; name only those dependencies and the endpoint facts it uses; the arrow changes none of those occurrences and makes none obtain or cease;
- how claims in
Ycorrespond to or are licensed byXand any additional sources or admitted facts required by the named morphism rule, the exact correspondence or retargeting relation governed by that pattern, and anyF.9Bridge that governs cross-context sense use when current.
The morphism declaration and its mathematical arrows are different objects. The declaration is a C.2.1 episteme, normally an A.6.0 FormalSubstrate signature, whose EntityOfConcern is the local mathematical family and whose claim content declares vocabulary, laws, and applicability. One arrow f : X -> Y is a local mathematical object identified inside that substrate by its exact endpoints, arrow rule or designator, and declared formal equivalence.
The mathematical statement f : X -> Y names no execution. When an exact operation application is current, A.6.1 separately identifies its argument and result bindings. For any precise performed-Work claim, use A.13 to identify the actual performer and A.15.1 to admit the dated Work independently. If that claim must also identify the assignment under which the Work was performed, check that relation separately through F.6. Identify the affected or newly constituted episteme, its C.2.1 discriminators, and any production or change relation under their direct governors. The same arrow may be used in several applications, and an arrow may relate already existing epistemes. No bare result, generic Work result, or universal production relation follows from an arrow or declaration.
A claim that one arrow is suitable for one exact use is another C.2.1 assertion. Its complete claim content names the arrow, use, conditions, and polarity. Evidence and reliance qualify that assertion; they do not identify the arrow or operation application.
Epistemic viewing
A.6.3 governs a viewing construction between independently identified source and receiving epistemes with the same exact EntityOfConcern. Under its declared rule, the receiving claims are licensed by the source together with any additional named sources and admitted facts that rule requires; claim content or the effective reference scheme may be restricted. The relation can hold between already existing epistemes and does not establish their production history. It neither grants U.View membership nor performs work. Direct authoring and query generation can identify receiving epistemes without a viewing claim; any claim that their production used an earlier source needs its own historical grounds.
E.17.0 independently asks whether one fixed receiving episteme E conforms to one fixed viewpoint episteme P; formally, whether EpistemeViewpointConformanceRelation(E,P) obtains. Only an obtaining relation gives the same E dependent U.View membership. A system may perform viewing, query, authoring, or rendering work, but neither that work nor an A.6.1 result position grants U.View membership or supplies C.2.1 identity.
Empirical grounding continues only while every mapped direct relation required by the receiving episteme's exact covered claim subgraph obtains. Changing publication, current use, evaluator, or evidence alone changes neither the conformance of fixed E to fixed P nor grounding. Several source or receiving epistemes do not automatically form a multi-view family; identify any current collection under C.13 and any selected organization under A.22.
Epistemic retargeting
A.6.4 governs a local class of effect-free arrows r : X -> Y whose exact endpoint epistemes concern different exact entities. It does not admit a durable U.EpistemicRetargeting kind. The A.6.0 FormalSubstrate signature that declares the class, one arrow r, a bounded-use assertion q, a current-case judgement, and any actual operation application remain different objects.
For one receiving use, q states one proposition with affirmative or negative polarity: whether r preserves the named invariant and makes the stated loss acceptable under named conditions. A separate current-case judgement compares exact current facts with that proposition and reports satisfies, fails, or cannot decide. The same r can have different q assertions and case results for different uses without changing arrow identity. A missing deciding fact yields cannot decide, names that fact, and states what would reopen the question; it does not create unresolved assertion polarity or require an assurance record.
Use A.20 when the proposition is an internal constraint, A.10 when the evidence source-to-use account must remain replayable, and B.3 when a named assurance claim is at issue. The named predicate and applicable direct facts supply the case judgement. If the case also asserts a semantic relation between exact F.17 local senses, F.9 defines that separate Bridge and its bounded-use claim.
A system may perform retargeting work. Identify its enacted method, any exact A.6.1 operation application and binding, affected or newly constituted episteme, and actual change facts separately. The arrow itself performs no work, and the mathematical statement r : X -> Y infers no bare result or universal production relation.
Examples include retargeting from an episteme about a physical cabinet to one about a selected functional U.Structure, or from observations to a learned model, when the independently identified source and receiving entities differ. For Cab-7 as the cabinet and Route-A as the selected functional structure, an independently stated behaviour-test result can supply q's current-case basis; the judgement is satisfies only when that result meets q's named conditions. The source expression Realises(Cab-7, Route-A) has no current direct relation governor here and therefore stops at A.6.RCD missing-governor; it contributes nothing to that judgement. For a Fourier representation change, identify the endpoint subjects under C.2.1. If the same signal remains the EntityOfConcern, use A.6.3.RT or the applicable representation-transition pattern. Use C.29 when a mathematical-lens choice, transfer or reliance question remains.
Multi-view description and publication (normative)
C.2.1 identifies each candidate episteme E and each viewpoint episteme P separately. E.17.0 then asks whether E conforms to P; its formal name for that relation is EpistemeViewpointConformanceRelation(E,P). Only when the relation obtains is the same E a U.View under P. For one named describing use, say that the use selects P. Keep E, its EntityOfConcern, the use, and P distinct. Selection changes neither episteme identity nor conformance and does not make E a view. If another receiving use selects an already identified view, name that use and the pattern that defines or constrains it separately.
Several conforming views remain a plurality. Recover an exact C.13 collection only when a receiving use depends on that plurality as a collection. Recover an exact A.22 U.Structure only when the use additionally depends on organization among those views, and state the exact direct organizing relations. A shared EntityOfConcern, package, table, heading set, diagram, or carrier creates neither a view family, a collection, nor that structure.
Use A.6.3 when the account needs the same-EntityOfConcern relation licensing receiving claims from named sources. Establish actual source-use history separately when that history matters. A system performs any viewing, authoring, query, comparison, or repair work; direct authoring or query generation can identify an episteme without a viewing claim. Neither that work route nor an A.6.1 result position grants U.View membership or identifies a multi-view family.
E.17 governs multi-view publication forms and uses, while E.24.PUB governs publication occurrences, forms, and carriers. The same recognized view can participate as the selected episteme in several publication occurrences without changing identity or conformance. Publication establishes no view membership and no cross-view subject relation.
When cross-view correspondence, consistency, realization, trace, or change impact matters, name the exact participants and apply the pattern governing that subject relation. A C.2.1 assertion or description episteme may carry that claim; it does not make the relation obtain. If that governing pattern cannot be recovered, return a missing-relation blocker naming the participants, required predicate and use, and missing governor; do not invent a relation. E.17, a matching heading, graph edge, diagram position, or shared carrier invents no correspondence relation.
Archetypal Grounding — Worked Cases
These cases ground the pattern in practice; only a case that names an obtaining EpistemeEmpiricalGroundingRelation asserts empirical grounding.
Physical engineering
A pump-maintenance specification has a claim graph about exact pump P under a reference scheme that resolves part names, states, units, and measurement procedures. Those three participants identify the episteme. For test bench B, exact covered claim subgraph C_B contains the discharge-pressure-tolerance and leakage claims. Its claim-to-world mapping names the direct pressure-measurement and leakage-inspection relations involving B; a maintenance-interval claim outside C_B is not grounded by those measurements.
The grounding relation over (E,B), with covered=C_B, continues for the maximal interval during which every required mapping obtains. If that coverage continues while an evidence archive or inspection-work log becomes unavailable, only a separately governed support, warrant, confidence, or evidence-use assertion may change.
A publication occurrence makes the episteme available through a rendered checklist form borne by an exact carrier. When checklist marks assert particular maintenance or inspection Work, use :4.0's Work branch; assignment attribution is added only when the checklist account requires it.
When a receiving use needs a separate assertion that P satisfies the constructive U.System criterion, identify it as another episteme about P; renaming or republishing the governing FPF pattern does not change P or create its systemhood.
The classification assertion changes only when its own claim content or reference scheme changes. Pump continuity is judged instead under the A.1 reidentification rule; a changed or unchanged assertion does not establish that continuity.
Medicine
A diagnostic model concerns one patient-state entity or one admitted patient cohort under a scheme that defines observations, measurements, and diagnostic interpretations. Each EpistemeEmpiricalGroundingRelation identifies the exact covered diagnostic-claim subgraph, grounding holon, mapping to the required current observation, measurement, or test relations, and maximal continuous interval of complete coverage; it grounds no unlisted claim. If a threshold revision changes claim content or the effective reference scheme, that changed discriminator identifies another episteme; moving the unchanged model to another screen changes only the exact publication or representation object that actually changed.
Competence claims, teaching Work, and learner-facing views
A curriculum model concerns an exact competence structure under a scheme that relates assessment or performance evidence to competence claims. If taught or learned wording leaves the changed subject or result unclear, use E.10.LRN to recover the claim first: receiving an intervention and acquiring a capability call for different evidence.
A learner-facing episteme is a U.View when it conforms to an exact learner-facing viewpoint under E.17.0. If the account needs to establish how the source curriculum model licenses its claims, A.6.3 tests that viewing relation; an assertion that authoring actually used the source requires its own historical grounds. Changing the selected competence claims changes the receiving episteme; changing only its display preserves it.
If empirical grounding is needed, name the covered competence claims and their mappings to the relevant assessment or performance relations involving an exact admitted course-cohort or learning-environment holon, as required by :4.3. A later capability, transfer or retention claim needs the applicable grounds connecting that result to the claimed use; use A.10 when this source-to-use account must remain replayable. If particular teaching, assessment or construction Work must be asserted, open :4.0's Work branch. The identity-only use finishes with the model's claims, exact subject and effective scheme.
Episteme about an episteme
Simulation model M is one episteme. Review R concerns M, so the EntityOfConcern in R is the episteme M, not the physical system modeled by M. Claims in either episteme may cite separately governed evidence-use relations concerning the simulated or physical system. Publishing R does not revise M.
A theory episteme is recognized through its claim-bearing constitution and whole-level inferential characteristics. A textbook publication can make one edition of that theory available, but the publication occurrence, form, and carrier are not constituents of the theory and do not establish its holonhood.
Edition succession
Episteme E1 is the exact source used to produce candidate edition E2. The applicable continuity policy says that the EntityOfConcern remains the same, specified core claims must remain traceable, listed claims may be corrected, and a translation into another reference scheme counts as a derivative rather than an edition. The current case identifies the preserved core claims, deliberately corrected claims, unchanged EntityOfConcern, and source-to-revision use. Those facts satisfy the policy, so the positive EpistemeEditionRelation(E1,E2) assertion is available.
If the same Work instead retargets the claims to another EntityOfConcern, translates them under a rule that the policy classifies as a derivative, or reconstructs similar content without using E1 as source, the edition predicate fails even when the Method is named “revision.” Work, Method, provenance, change facts, evaluation, and evidence remain outside the two-participant relation. Later repackaging or publication establishes neither another episteme nor edition continuity.
Grounded identity across two observations
A morning-observation episteme concerns an object designated M; an evening-observation episteme concerns an object designated E under another reference scheme. A.1.RI supplies the Method for deciding whether these observations concern the same continuing object: recover its continuation criterion, construct connections under the subject and observation rules, and compare the alternatives. Carry the identification with its supporting premises; retain ambiguity when the observations leave materially different connections possible.
The two observation epistemes can retain different identities after the object is identified: they can make different time-indexed claims or use different reference schemes. C.2.1:4.2.3 permits the identifying claim in ordinary language. If a later use needs a separately designated relation occurrence, apply A.6.REL with the direct relation pattern's obtaining predicate and occurrence-identity rule. A missing relation definition then blocks that occurrence-dependent use. It does not erase a conditional identification already established under stated premises; an unspecified object-continuation criterion, by contrast, leaves the identification itself unresolved.
Construct a bounded identifying result. A.1.RI:5.1 develops the complete two-object case: independent persistent point objects at 0 and 10 cm are observed one second later at 1 and 9 cm, with exact positions in one frame and crossing permitted. With one observation of each object at each end of the interval and a 2 cm/s speed bound, only the association requiring 1 cm per object is admissible. Raising the bound to 10 cm/s also permits the crossed association requiring 9 cm per object. The observation claims retain their different occasions after an identification; the changed premise instead leaves the object correspondence unresolved. The receiving use determines whether a discriminator is worth obtaining.
In Rodin's astronomical reconstruction, the connecting trajectory is supplied through theory and observations. In the HoTT reconstruction, MS and ES are represented as terms of a point-object type Pt; p : MS =_Pt ES expresses the identifying construction as a term of their identity type.
Readable wiring diagram as a proxy
Wiring-model episteme E1 concerns exact harness H under reference scheme S1, which resolves connector designators, pin identities, and connection predicates. If only layout changes in a wiring-diagram representation, identify the exact representation transition and preserved connector, pin, and connection correspondence; E1 remains the same. If instead only an exact publication form, carrier, or rendering changes, identify that E.17/E.24.PUB object and relation; E1 again remains the same. If a connection claim is removed from E1's ClaimGraph or the legend changes the effective reference scheme, the changed claim content or scheme identifies episteme E2.
The wiring-diagram representation under its stated diagram scheme represents the connectivity of H; its mapping resolves connector marks and pin marks to the independently identified connectors and pins. A layout-only transition preserves connector identity, pin identity, and connection predicates. If the diagram omits a connection that remains in E1's ClaimGraph, that representation loses one connection predicate. If a revised legend no longer resolves an earlier mark to its connector, that mark-to-connector reference is lost. The diagram remains admissible for maintenance diagnosis only while the connections on which that diagnosis depends are preserved and recoverable; stop that use or return to the source relation structure when they are not.
For a concrete typed receiving use, let local kind K_H admit already identified epistemes about H whose complete content consists of affirmative or negative physical-connectivity claims between identified pins under S1. The criterion excludes other subjects, proximity readings, and additional claim families; it classifies whole epistemes, not claim-content or scheme individuals. C.3.2 judges candidates against that criterion. Let E1 assert that pin P1 is connected to pin P2.
A readability score can improve while the diagram makes the connections needed for diagnosis harder to recover. When that score is used as the practical value, apply E.13: name the intended diagnostic value, the readability proxy, and what became worse. Use the representation branch in :4.0 for the transition and its preserved or lost structure.
Trained or probe-derived representation and tool-using inference
A distributed activation pattern observed during inference is a system-side phenomenon. A trained probe or decoded rendering can represent it for a declared use through an identified correspondence. A.6.3.RT governs a same-subject representation transition; C.29 supplies the additional grounds when a mathematical-lens choice, transfer or reliance question remains. To identify a probe report as an episteme, recover what it claims about that exact phenomenon and the interpretation scheme that gives those claims meaning. Decodability or a readable label alone leaves that identity question unanswered. Use E.10.LRN when learned representation still leaves the subject or result unclear.
For a claim about particular inference or tool-call Work, open :4.0's Work branch. If a trace participates at an exact A.6.1 result position, state that binding. When the trace carries claims about the call, identify its episteme through those claims, that Work as subject, and the effective scheme. A question about when the trace first existed separately opens :4.9.
An answer may carry a proposed solution, while an evaluation report states whether it passes a named test. To rely on the solution, inspect the support relevant to its claim; to assert empirical grounding, test :4.3's exact covered-claim mappings. A successful call or score does not supply those mappings.
When a changed tool regime changes inference behavior, locate the changed method, Work or representation and the evidence used to assess it. Update the grounding occurrence if its coverage predicate changes. Reidentify an episteme only when claim content, EntityOfConcern or effective reference scheme changes.
Develop a naming claim in writing
A naming note starts with a concern that the proposed name hides a distinction. Following C.2.4:4.6, the author writes competing interpretations, identifies what the name would designate, and states the reason for preferring one reading. The note helps obtain the claim; a completed NameCard is useful only when the naming work needs its arrangement and comparison. Identify the resulting claim-bearing whole under :4.1.1 once its content, subject, and interpretation are recoverable.
Suppose the designation claim in that card stays the same, but SelectionRationale now asserts a different reason for the choice. The complete asserted content changed, so the card's episteme changed. Merely reordering fields or making the same reason easier to read can preserve it. A newly asserted reason needs its own support; preserving the designation does not transfer the earlier justification. If only an external evidence-use relation changes while every assertion in the card stays fixed, reconsider that relation and reliance without automatically changing the card's episteme.
Bias-Annotation (informative)
C.2.1 deliberately favors explicit aboutness and interpretation because claims without an exact EntityOfConcern and effective reference scheme are difficult to compare or test. The mitigation is the A.6.REL minimum-current-object rule: ordinary use adds another object only when the reader's next sentence or action requires it, and states that object's direct relation to an already recoverable object.
The pattern also resists representation bias. Formal calculi, diagrams, learned representations, and interactive tools can materially change available reasoning operations, but their convenience or geometry cannot establish subject identity. State those differences under the selected representation correspondence or transition predicate. Use the corresponding pattern for any explanation needed to apply it, and add C.29 only when a mathematical-lens choice, transfer or reliance question remains.
Finally, the pattern has a claim-bearing-holon bias. Decodability alone does not make the decoded entity an episteme. The decoded entity is admitted as an episteme only when claim content, an exact EntityOfConcern, and an effective reference scheme are recoverable and together satisfy the constitution relation.
Conformance Checklist (normative)
- Episteme identity. Complete claim content, exact EntityOfConcern, and effective
ReferenceSchemeare recoverable under :4.1.1. Distinguish obtaining that content in writing, candidate admissibility, full identity, and suitability for the receiving action. A dependent episteme kind such asU.MethodDescriptionorU.Viewadds a governed membership judgment for the same individual, not another identity discriminator. - Direct constitution and case judgment.
EpistemeConstitutionRelationhas its three identified participants, obtaining predicate, and participant-determined occurrence-identity rule. C.2.1 defines that rule; current case facts satisfy or fail the predicate; an assertion states affirmative or negative polarity; and evaluation or evidence use states reliance only when needed. Designate an occurrence only after positive case facts and the identity rule individuate it. - Declaration identity and Slot discipline. A complete reusable declaration is first a C.2.1 episteme; A.6.0 independently judges its
U.Signaturemembership. Grounding and edition declarations use the ranges in :4.3 and :4.5. The constitution declaration supplies the EntityOfConcern SlotSpec; any additional typed participant use needs its independently governed range and complete SlotSpec; a direct exact-value constitution judgment alone supplies no general ClaimGraph or ReferenceScheme ValueKind. Preserve predicate, occurrence identity, applicability, and actual dependencies. Signature-local SlotKinds never become participants, and a one-off assertion needs no signature or manifest. - Classification discipline.
A.1governs recognition under an admitted holon kind,C.3.2governs local-kind membership, andE.24.UKgoverns public U-kind admission. A classification assertion needed as a separate claim-bearing object is a C.2.1 episteme about the exact candidate; it states affirmative or negative polarity for the exact classification predicate and keeps the assessment of that assertion and the limits of reliance separately governed. It neither creates the candidate nor changes the kind's admission. - Empirical-grounding discipline.
GroundingHolonSlotoccurs only insideEpistemeEmpiricalGroundingRelationSignature. Each occurrence names one exact covered claim subgraph containing at least one empirical claim and maps every empirical claim in it to the required current direct observation, intervention, measurement, or test relations involving the grounding holon. Unlisted claims receive no grounding from that occurrence. One occurrence is reidentified from the episteme, covered claim subgraph, grounding holon, and maximal continuous interval during which the complete coverage predicate is true; demonstrated coverage failure followed by restoration yields another occurrence. Evaluation counts in the empirical base only when its exact direct relation and use in the test are stated; otherwise evaluation and evidence support or challenge an assertion. Availability or loss of a report, store, or Work log alone neither makes nor unmakes grounding. - Edition discipline.
EpistemeEditionRelationhas exactly the earlier and later epistemes as participants and is acyclic in that direction. Positive continuity requires exact source use, an applicable edition policy or rule, and preserved and deliberately changed claim, EntityOfConcern, and scheme features satisfying that rule. Test the failure conditions of that rule, including any distinction it makes between an edition and a fork, translation, retargeting or independent reconstruction. Work, Method, provenance, and change facts supply case facts but no label makes continuity true. - View and neighboring-relation discipline. C.2.1 identifies epistemes; E.17.0 alone tests the conformance of fixed E to fixed P and the resulting same-individual
U.Viewmembership. One named describing use may select one exact viewpoint P, but that selection creates no context value, selects no view, and remains separate from A.6.3 source-to-receiving construction. Several views remain a plurality. Recover a C.13 collection only when the use depends on that plurality as a collection, and an A.22 structure only when it depends on their organization. Cross-view claims use the pattern for their direct subject relation or return a blocker naming the participants, required predicate, use, and missing defining or constraining pattern. Use E.17 for view or publication form and E.24.PUB for publication occurrence, form, and carrier, not for view membership or correspondence. - Description boundary. Recover the EntityOfConcern and description episteme under their C.2.1 identities. Different constitution triples identify different epistemes. In permitted self-description, the EntityOfConcern designation resolves to that same episteme; self-reference supplies no assurance by itself.
- Specification use. Specification force is admitted only when the E.10.D2 conditions obtain: checkable claims and a named harness or validation relation. A selected viewpoint is preserved or updated only for the named describing use whose reliance depends on it. Naming and appearance do not grant specification force.
- Agency, work-result, and identity-inception boundary. Only systems perform Work.
A.6.1declares typed argument and result positions; neither a position nor its binding says when the bound entity first existed. When a current claim asks that question, the subject's direct inception pattern must define the predicate and identity rule, and current case facts must satisfy them. When inception is attributed to Work, apply A.15.PROD:4.5 and retain the exact specification and applicability basis, actual effects, earliest satisfying boundary, andidentityClosingWork; a missing required basis stops that positive claim. If no subject inception governor exists, return onemissing-governorblocker naming the entity or candidate basis, facts, required predicate, and receiving use. When no first-existence claim is current, do not open the inception boundary. No morphism, heading, representation, form, bare A.6.1result, generic work result, or universal production relation supplies that fact. - Publication boundary. Distinguish the episteme from its publication occurrence, particular expression, reusable form and carrier. A
U.Viewis that same episteme classified under E.17.0's conformance condition. Plainpublished epistemestates publication participation; a receiving classification need follows C.3.2 and public-kind admission follows E.24.UK. - Representation boundary. Tuple components, graph elements, schema fields, and notation tokens remain representation elements. An explicit correspondence may relate one to an independently recovered object without identifying the two or changing the represented direct relation's participants.
- Transformation and Bridge-use boundary. A morphing, viewing, or retargeting declaration states which C.2.1 identity discriminators are preserved or changed and names the exact correspondence or retargeting relation used. For cross-context sense use, F.9 separately establishes the exact Bridge; one C.2.1 assertion about that Bridge carries
<u,d,r,t>and polarity; A.10 handles ordinary evidence reliance; B.3 adds a result only when a named assurance claim is at issue; and the direct receiver defines or tests any actual Work, assertion, publication, relation, or operation application. The mathematical morphism performs no work, and none of these objects authorizes another. - Recursive assurance. Self-reference and meta-description do not form a minimal justification cycle; assurance terminates in independently governed evidence, observation, or formal derivation.
- Minimum current object. Readable prose adds no object beyond the current use's dependency and states the direct relation to an already recoverable object.
Common Anti-Patterns and How to Avoid Them
Consequences (informative)
Benefits. Episteme identity becomes stable across carrier and publication changes. Aboutness, empirical grounding, viewing, edition continuity and representation questions can be repaired locally through their respective relations. Self-description and multi-view use need no second ontology. The same pattern works for physical engineering, medicine, learning, formal work, and computational modeling.
Costs. A load-bearing episteme use has a recoverable exact EntityOfConcern and effective reference scheme rather than relying on a title or file. Empirical-grounding and edition claims depend on their own obtaining and identity evidence. Existing record-shaped schemas sometimes need to distinguish their fields from actual relation participants.
Limits. C.2.1 does not decide whether an epistemic claim is true, sufficient, current, or authoritative. It does not prescribe a file format, graph database, proof calculus, or publication layout. Those questions remain with evidence, evaluation, temporal, representation, and publication patterns.
Rationale
The three constitution discriminators identify the knowledge holon; empirical grounding, viewpoint, scope, edition, and publication remain separately governed. A change in one of those neighboring relations can therefore be examined without reidentifying an unchanged episteme.
The constitution relation explains how the identity-bearing participants are organized. The episteme is the resulting holon with a whole-level capacity to carry interpretable claims. A relation declaration is first its own C.2.1 episteme; A.6.0 judges whether its complete content satisfies U.Signature membership, including the declared range, and RelationSignature names its relation-facing use. A complete declaration carries the direct relation for its stated typed reuse; an assertion claims that the relation obtains, and a publication occurrence makes a selected episteme edition available. None replaces another.
SoTA-Echoing
When does grounding cease? For the pump case in :9.1, a compact record with one evidence-status flag is useful for finding the latest available report. If that flag also defines grounding, an unavailable archive would end the grounding occurrence even while the required measurements continue; an available pressure report could also be mistaken for coverage of the unrelated maintenance-interval claim. The questions about current empirical coverage and its continuity require more information.
Sections :4.3 and :4.8 therefore select explicit covered claims, their mappings and the maximal continuous interval of complete coverage. This costs claim-to-world mapping and continuity evidence when that relation is asserted. It allows the practitioner to keep grounding continuous through an archive outage, withhold reliance when coverage is unknown, and recognize another occurrence after a demonstrated break and restoration. The cheaper evidence-status record remains sufficient for the question of report availability. Reopen the grounding account when a required mapping fails, its covered claims change, or the purported evidence no longer establishes continuity.
When is a revision an edition of this knowledge object? PROV-O, 2013 Recommendation, wasRevisionOf, provides a usable provenance assertion: derivation as a revised version retaining substantial original content. It is economical for exchanging a producer's revision assertion. C.2.1's :9.5 asks the additional domain question of which changes count as continuation of this episteme family. A translation can retain substantial content while that family's policy classifies it as a derivative.
Section :4.5 therefore uses provenance as a case fact and tests an explicit continuity policy over the two episteme identities, source use, preserved features, permitted changes and failure conditions. The cost is recovering the policy and the changed features; the gain is a decision that distinguishes the accepted correction from translation, retargeting or independent reconstruction under the same policy. A trusted revision assertion can be consumed for a use that accepts its producer's criterion. A use requiring the local edition judgment must test the local criterion. Reopen that judgment when the source-use facts, policy or required preserved features change.
When may observations be identified as concerning the same object? Rodin's reconstruction, §§2–7, makes a theoretically supported connection between observations part of the identifying construction. Section :9.6 retains the episteme/object distinction and directs the identifying construction to A.1.RI:5.1. The two association candidates require displacements of either 1 cm or 9 cm. The 2 cm/s premise selects one; the 10 cm/s premise admits both.
For this ordinary conditional identification, :4.2.3/:9.6 retain an ordinary identifying claim supported by A.1.RI's domain criterion, motion premises and comparison of associations. Requiring a separately governed relation occurrence first would add a declaration dependency without changing that calculation. The occurrence account is worth its additional cost when a later assertion must designate that exact occurrence; then A.6.REL supplies its predicate and identity requirements. A changed speed bound reopens the identification itself, whereas a missing occurrence rule blocks only the occurrence-dependent use.
These comparisons select different information for different questions. The three identity answers suffice for identifying the episteme; empirical coverage, edition continuity and occurrence designation are added for the respective uses above.
Material interaction and grounding. Malafouris, People Are STRANGE (2026), develops the material-engagement account of how experienced self-boundaries change with tools, surroundings and ways of engaging with them. This motivates a question for :4.3 and :9.3: has a change of material interaction changed what can be observed or inferred about the subject? Reconsider the covered claims and grounding holon where that change affects their empirical support.
Internal representation and reasoning. Anthropic's A global workspace in language models (2026) reports interventions in which changing an internally represented intermediate concept changes a later answer in the studied reasoning tasks. This supports a causal contribution of those representations in those tasks. Use :9.8 to distinguish the changed activation, the probe's rendering and the answer when interpreting the result.
Tool availability and inference quality. Cheng et al. (2026) report degraded reasoning under tool-enabled evaluation even when the studied models made almost no tool calls, and develop training methods that preserve reasoning while adding tool use. For an inference-quality claim, use evidence from the applicable tool regime and distinguish availability, actual calls and evaluated performance through :9.8.
Relations (overview)
-
A.1.RI constructs observation-based reidentification of the subject while C.2.1 retains the identities and claims of the observation epistemes.
-
Builds on:
A.1for holon recognition,A.6.RELfor direct relation occurrences,A.6.0for independent same-individualU.Signaturemembership and relation-facingRelationSignatureuse,A.6.5for declaration-local SlotSpecs and participant designations,A.7for entity-description distinction, andC.29for remaining mathematical-lens choice, transfer or reliance questions. -
Coordinates with:
A.3.2forU.MethodDescriptionmembership without a second episteme identity.C.3.2for local-kind membership judgments.E.24.UKfor ontology-level U-kind admission.E.10.D2for contextual description and specification-use naming, including selection that creates neither conformance nor membership.A.6.1for typed operation positions and exact current application bindings.A.6.2,A.6.3, andA.6.4for morphing, source-to-receiving viewing construction, and retargeting.A.6.3.RTfor representation transitions.E.17.0for conformance of fixed E to fixed P andU.Viewmembership.C.13andA.22for separately current multi-view collections and structures.- The pattern for the exact direct subject relation, or a missing-relation blocker naming the participants, required predicate, use, and missing defining or constraining pattern.
F.9for exact Bridge semantics when a claim concerns a bounded cross-context use.E.13when a visible representation-quality proxy is used as practical epistemic value.A.2.6for claim scope.A.1.1for bounded model-use structure.A.10andB.3for evidence and assurance.A.14only when a phase or separately selected edition collection is current.C.2.P,A.3.1, andA.3.4when source use, revision method, or actual change is current in edition-continuity evaluation.- C.2.1:4.9 only when a local claim separately asks when a new entity began.
E.17for multi-view publication forms and uses.E.24.PUBfor publication occurrences, forms, and carriers.G.11for currentness.
-
Used by: every pattern that identifies, describes, classifies through an explicit assertion, compares, grounds, transforms, views, publishes, or refers to a
U.Episteme.
C.2.1:End
Clarify Episteme, Publication, and Source Wording (Epistemic Precision Restoration)
Type: C.2 precision-restoration pattern for episteme, publication, source wording, and source-relation wording Status: Stable Normativity: Normative unless a section is explicitly informative
Use this when
Use C.2.P only when one unresolved distinction prevents the reader from selecting or safely using the direct pattern. The unresolved point must concern a source expression, claim-bearing episteme, publication, publication unit, view, carrier relation, source-to-use relation, or use disposition.
Recognizable situation. A sentence is readable, but it leaves the reader unable to tell whether it refers to claim content, a publication or bounded unit, a carrier or display, a source-bearing relation, or a project-side use of one of those objects.
First useful move.
- Say what the sentence is doing: defining, claiming, instructing, comparing, locating a source, describing a publication, or supporting a project use.
- Recover the one unresolved episteme, publication, source-to-use, carrier, or use-disposition distinction.
- Rewrite the sentence, or hand the remaining claim to the pattern that defines, constrains, or tests it. Then stop.
Not this pattern when. If the receiving pattern and its field at issue are already recoverable, apply that pattern directly. A clear ordinary phrase needs no C.2.P record. Use A.6.P for a relation problem whose publication or source-expression side is already clear, F.18 for a stable reusable name, A.7 for EntityOfConcern-description-carrier separation, E.17 for reader-facing publication of an accepted account, and E.24.PUB for publication-occurrence, form, or carrier distinctions.
Smallest outputs.
- Direct repair: the repaired sentence. Add a reason or non-use boundary only when it changes interpretation or action. No record.
- Compact row: the exact sentence, its function, one recovered kind or relation, the selected wording or disposition, and the reader's remaining use.
- Full check: only when several recovery fields interact, a source-to-FPF use is contested, or a reusable ontological or naming decision is being made.
- Other outcomes: reduced-use cue, blocked use, incomplete rewrite, or not triggered.
Changing an FPF pattern is not by itself a reason to use the full check. Use the cheapest product that preserves the needed distinction.
Source-expression boundary. External or ordinary prose may be clarified without forcing all of its vocabulary into FPF. If no FPF-governed use is recoverable, keep the phrase as source-local wording, a source-finding cue, or a blocked claim-bearing use.
Source-to-use continuity. Do not close a repair merely by replacing the word source. Name the source expression or selected source episteme, the publication occurrence only when availability matters, the relation or path that carries it into the current use, the permitted use, and the condition that requires return to the source. Keep physical raw material with its constituent, resource-use, supply, transfer, or transformation relation.
Ordinary-language survival. Words such as source, view, support, route, and display may stay ordinary when they make no FPF kind, relation, authority, evidence, gate, work, decision, or reliance claim. Repair by sentence function and consequence, not by trigger word.
What this buys. The practitioner recovers the one distinction that matters without building a second ontology or a second review procedure. The final wording still tells a cold reader what to do and where to stop.
What goes wrong if missed
Episteme-publication-heavy text starts to build a parallel ontology. A generic publication face becomes a U.View, a file becomes an episteme, a dashboard tile becomes evidence, a pattern name becomes a procedure, or a slash list becomes a group kind. A broad word such as source can hide several different recovery fields: an FPF pattern or DRR; a publication field; a document named for source, evidence, architecture, or review use; a reviewed publication, review packet, review record, or review state; a project-side FPF kind and reference named by value; or a relation.
The immediate cost is that engineers and FPF authors start making action, evidence, gate, decision, or engineering-justification claims from the wrong entity, publication, record, relation, or carrier.
What this buys
C.2.P gives one small epistemic precision-restoration move: recover the FPF kind and relation set first, then write wording that preserves the needed distinction without adding another claim. It prevents string-replacement cleanup, keeps FPF-side and project-side episteme and publication work separate, and blocks unclear wording from having FPF-governed use by guesswork.
Successful repair condition. Type-correct wording is not enough. A repair must satisfy the E.2 Pillars, especially P-2 Didactic Primacy, together with E.12 and the register rule in E.10:6.2. It closes when the repaired passage, together with sources the reader can use, supports the intended action, judgement or explanation. When the passage supplies a premise for a larger conclusion, show how the reader can use it and which other grounds the full conclusion needs.
When Tech and Plain registers are both current, the Tech interpretation remains recoverable and the Plain or didactic line maps back to it. An ordinary or metaphorical Plain line may stay light when it carries no FPF-governed use; if it carries an ontological, evidence, causal, assurance, bridge, gate, work, decision, or use-boundary claim, that claim must be recoverable through the Tech fields, named FPF kind, recovered relation, project-side reference, or disposition.
A repair to a Problem frame, recognition text, example, or worked slice is incomplete if it improves typing but hides the working situation, why it matters, or the first useful action; name the applicable FPF pattern when that pattern carries the claim. Preserve the remaining action guidance as well as the recovered meaning.
Recovery focus in plain terms. The use being made is one episteme-publication-heavy wording use inside conformant text: the word or phrase, the sentence function it carries, the FPF kind or relation it must recover, and the remaining declared use boundary after recovery.
Primary working user. The first user is a practitioner maintaining conformant FPF-style or project text: an author, reviewer, or engineer-manager who must repair wording without losing ontology. The downstream user is the practitioner who will rely on the repaired pattern or project text in a working situation.
Practical payoff. State what the repair lets the reader distinguish or do, and any remaining question that requires another named FPF pattern. Explain a rejected reading only when F.19 establishes an independent ground, a plausible intended-reader interpretation, and a consequence for truth or use. If the repair blocks a false claim, retain the useful remaining statement or state the blocked use.
Problem frame
FPF already has episteme, publication, view, carrier, presentation, relation, naming, and pattern-application concepts. FPF-governed project, review, draft, pattern, and architecture prose, plus source prose being unpacked for possible FPF use, can still introduce convenient intermediate words that survive into final guidance without their kind and relation set recovered.
The recurring situation is simple: a sentence is understandable enough to feel worth keeping, but its head kind is not recovered. If it is repaired by replacing one broad word with another broad word, the ontology gets worse while the text looks cleaner.
Purpose and Scope
This pattern gives the current glossary and rewrite rules for terms around epistemes, publications, views, publication forms, generic publication faces, MVPK faces under E.17 constraints, carriers, records, and bounded publication units.
It exists because episteme-publication-heavy texts can use locally convenient heads that collapse EntityOfConcern, publication unit, publication face, carrier, record, source relation, and project-side FPF kind and reference. Those words may be useful recognition handles, but they are not safe FPF heads when they carry ontology, authority, or authority-changing meaning.
The rewrite discipline here is ontological and use-facing, not lexical; in this pattern the repair is bounded to episteme, publication, source wording, and source-relation precision:
- do not replace one broad token with one new broad token by string substitution;
- first recover the FPF kind and relation set, whether the wording carries a claim, the publication, view, carrier, or relation construction, and any work, action, or authority crossing;
- then choose the smallest wording that preserves the FPF-governed distinction without creating a second ontology.
For FPF-governed wording, use the E.10 cues and routes within the connected F.19 reading. C.2.P works in the C.2.1, E.17, and E.24.PUB epistemic-publication ontology rather than in a lexical registry.
Problem
Without an epistemic precision-restoration discipline for episteme-publication-heavy wording:
- broad publication words hide which field family is current: episteme or view, publication form or face, bounded publication unit, carrier relation, named document use, review-state use, or project-side FPF reference;
- FPF pattern-application claims and project-side fields for work occurrence, work plan, decision, action invitation, method, record, carrier relation, or front-end relation get mixed in one sentence;
- slash lists and heterogeneous rows become false group kinds;
- unclear source meaning is guessed into FPF-governed wording rather than blocked or assigned to an accepted FPF extension;
- authors copy the same loose wording into
DRRs, patterns, source-relation notes, source-ref target notes, or project texts.
Forces
Solution
Repair episteme-publication-heavy wording by epistemic precision restoration, not by dictionary replacement.
A successful rewrite satisfies these field-validity constraints:
-
the head kind and sentence function are recoverable under
E.10; -
a stable reusable name has an
F.18naming result; -
a relation, comparison, dependency, support, sameness, grounding, mapping, or endpoint claim has
A.6.Prelation precision, with use-boundary and project-side reliance questions split into their own fields; -
a claim-bearing episteme has a recoverable
C.2.1identity; dependent-kind membership, such asU.View, uses its defining pattern. A project-side object or relation keeps its own kind, applicable reference, and declared use; -
publication, view, face, and carrier distinctions satisfy the applicable
E.17.0view-conformance,E.17/MVPK publication-face, andE.24.PUBpublication-occurrence, form, and carrier rules; -
the repaired text satisfies
E.2Pillars, especiallyP-2 Didactic Primacy, by preserving or restoring the action, judgement or explanation this reader needs. A recognition reason or a reference to a pattern can supply the part of that work it actually supports.When both Tech and Plain registers are current, the Plain or didactic line maps back to the recovered Tech kind, relation, or FPF pattern application under
E.10:6.2.Ordinary Plain wording and intentional didactic metaphor stay light when they carry no FPF-governed use, but ontological, evidence, causal, assurance, bridge, gate, work, decision, or use-boundary claim in a more expressive Plain line must be recoverable through the repaired Tech fields.
FPF-governed Problem frames, Problem sections, recognition texts, examples, and worked slices must still show the broad working situation and first useful move, or the rewrite is incomplete.
-
the final phrase preserves the distinction without adding another claim;
-
unrecoverable meaning, kind, register mapping, or remaining reader use fails closed.
Progressive recovery products
An ordinary application ends with the repaired sentence. Add a reason or non-use boundary only when it changes interpretation or action; no record is required.
Use a compact row only when another reader must later inspect the recovery:
Add a source identity, publication occurrence, carrier relation, declared-use boundary, project-side reference, naming decision, evidence relation, or assurance reference only when that claim is live. Optional fields do not become a common schema.
Use a full check only when several unresolved fields interact, the source-to-FPF use is contested, or a reusable ontological or naming decision is being made. Record the original sentence, its function, the selected kinds and relations, the neighboring-pattern contributions, selected wording, remaining reader use, and reopen condition. Record competing interpretations or rejected overreads only when they are independently grounded, plausible for the intended reader, and consequential under F.19. Omit unused fields.
When the wording exposes a field defined elsewhere, name the contributingPattern and the concrete definition, constraint, or test it supplies.
Carrier-specific recovery. Words such as carrier, file, dashboard, screen, front-end, and rendering are recognition cues. First say what the carrier is being used for. If the next pattern is not already clear, use the compact category-to-contribution route in §4.1.3 for publication, evidence or currentness, generated results, framework packages, Work or reliance, architecture or structure, and base or support questions. Do not close on the word carrier alone.
General Recovery Check
Run this check only while one C.2.P distinction remains unresolved.
- Name the sentence function. State what the sentence would let a reader claim or do.
- Recover the blocking distinction. Read the expression using its governing conventions. In a filled table, use the headings, units and applicable key to recover what one row asserts. A missing convention matters if two readings would classify the item differently or change the action. Obtain that convention or the deciding fact, or retain both readings. A.6.P can resolve a remaining question about the relation. Stop here if the recovered assertion answers the reader’s question. Use C.2.P.DR when someone infers a further classification, operation or result from how the expression looks.
- Select the next result. Rewrite directly, write the compact row, perform the full check under its material conditions, or return a non-use result. If another pattern now supplies the needed definition, constraint, or test, name that contribution and stop C.2.P.
Fail closed when the kind, relation, use, or remaining reader action cannot be recovered. A type-correct sentence that no longer shows the working situation or useful action is an incomplete rewrite.
Slash Discipline
In conventional abbreviations, source titles, mathematical notations, standards, URLs, file paths, and ordinary notations, a slash can be part of an accepted designation rather than a hidden FPF kind. In FPF-facing episteme and publication ontology, a slash is still a recovery trigger before it is a synonym marker unless the mark is part of such accepted notation, carrier syntax, or conventional designation.
Before leaving a slash expression in current prose, classify the expression as one of these cases:
- accepted notation or conventional designation: keep established names, mathematical and unit notation, file paths, URLs, and quoted source tokens whose slash belongs to the designation or carrier syntax. Examples include
ISO/IEC,ISO/IEC/IEEE, and1/2; - a plain-language synonym pair with no ontology, authority, evidence, or use-boundary claim;
- an unresolved
and/or-style join that must be split or recovered before FPF-governed use; - a proposed composite kind: establish its meaning under the kind’s subject pattern; use
F.18only for a durable reusable name andA.6.Pfor an unresolved relation; - a relation claim: recover the direct predicate, its participants, and its defining pattern; use
A.6.Ponly while one of these remains unclear, then use the direct subject pattern. Add declaration, occurrence-identity, qualification, or representation detail only when the receiving use needs it; - a tuple-like record: state the record’s purpose and component meanings; use
A.6.5SlotSpecs only for a reusable typed declaration that needs them; - a failed ontology signal where the sentence lists unlike values because the FPF kind under repair, relation record, relation phrase, tuple-like record, alternative-case disposition, or not-triggered disposition has not yet been recovered.
If the expression is not one of the safe notation, conventional-designation, carrier-syntax, quoted-source, or plain-language cases, do not keep the slash as final wording. Do not repair it by replacing the slash with one equally vague grouped word. Write the recovered FPF kind, relation record, relation phrase, tuple-like record, alternative-case disposition, or not-triggered disposition by value.
Unclear Source Meaning and FPF Extension Candidates
Sometimes the problem is not a bad word but one of two different cases:
- the intended claim cannot be determined from the surrounding source, current
FPFkinds, or current FPF episteme and publication ontology; - the claim is understandable, but current
FPFdoes not yet contain the kind, pattern, relation record, or method guidance needed to carry it.
Do not merge those cases.
An unclear claim is not current architecture truth merely because deleting it feels risky, and it must not be rewritten by guessing a likely author intention.
An understandable uncovered claim may be retained as a candidate FPF extension when the problem situation, current FPF gap, and first user action that would improve are stated. State an overread or rejected use only when it has the independent ground and practical consequence required by F.19.
Classify the case explicitly:
- recovered by value: the sentence names the object or relation, its defining or constraining pattern, and the declared use. When it uses
projectSideFPFRef, recover the exact project-side kind and reference described in :4.2.1. Stop at the selected value and the relation needed for that use; - understandable FPF extension candidate: the thought is clear enough to state as a candidate new or amended FPF kind, pattern, relation record, method guidance, accepted
DRRcontent decision, or campaign-scoped content question, but it does not carry current authority, evidence, or use-boundary claim until an accepted architecture decision, acceptedDRR, or accepted FPF pattern supplies that authority; - source wording without FPF-governed use: the phrase has no current authority, evidence, or use-boundary claim;
- reduced-use cue: the phrase is kept only as a recognition cue or anti-case, not as a claim-bearing architecture decision;
- blocked use: the phrase is blocked for claim-bearing architecture, pattern, or project text while the needed meaning, kind, or relation is missing.
- rewrite incomplete: the repaired wording may be kind-correct, but it does not yet state a remaining reader use, recognition reason, Tech-to-Plain mapping when both registers are current, or FPF pattern application, or a Plain or didactic line carries ontological, evidence, causal, assurance, bridge, gate, work, decision, or use-boundary claim that cannot be recovered from the Tech interpretation; continue repair or demote to a non-use disposition before the text has FPF-governed use.
These dispositions are recovery results, not a meta-governance authority over all of FPF.
When recovery names another FPF kind, use the pattern that defines or constrains that kind, its declared use boundary, and its conformance checks.
After that identification, C.2.P no longer defines or constrains the recovered kind.
C.2.P only makes the kind under repair, relation, and use boundary explicit enough to apply the appropriate pattern.
No other disposition is closed. In particular, "seems to mean", "probably about", a cleaner paraphrase, or a broad umbrella replacement is not a successful recovery.
Keep source, return, work, and next-pattern questions separate
When source data or source material is epistemic, identify the source expression and selected source episteme. Add a publication occurrence only when availability matters. Then name the relation or path that carries the source into the current use.
Treat source-use as a cue, not as a relation name. Do not call an endpoint value unless the direct pattern declares a value slot.
If the wording also makes a claim about a Method, Work, transformation, evaluation, transfer, or receiving use, use its direct pattern. Use A.6.P.WMR only while the work-side relation or what is being claimed about it remains hidden. Physical raw material stays with its constituent, resource-use, supply, transfer, or transformation relation.
Use source-return only for a reverse or escalation move from a derivative, coarsened, extracted, compressed, rendered, or reused object to a named source. Such a move is current when a stronger use, dispute, freshness change, hidden loss, or missing distinction requires the source again. For ordinary movement from a source into current use, say source-to-use path or name the relation. Name a rule-bearing ClaimGraph only when later comparison or reuse depends on the identity of that rule.
C.2.P does not decide whether Work may proceed. Use A.15.4 only after the wording repair shows that a publication, display, or other appearance is being relied on as a reason for intended Work and the exact project-side object or relation is still unresolved. If the direct pattern is already known, use it without an intermediate reliance-repair branch.
Once the remaining question is clear, stop C.2.P and choose the direct branch:
A.6.Pfor relation precision;F.18for a reusable name;C.30.Pfor a hidden architecture or structure distinction;C.16.PorC.16.Qfor a hidden characteristic, scale, or evaluative-quality distinction;- the applicable episteme or publication pattern for an already identified episteme or publication question.
These branches are alternatives. C.2.P may expose a naming or authority question, but it neither renames an accepted FPF pattern nor admits a reusable head; use F.18 and the applicable accepted decision source for that change.
Carrier-like words are only recognition cues. Once the category is clear, use one row below and stop C.2.P as soon as that contribution closes the question.
Core Glossary
Cross-Side Fields That Must Stay Split
These fields are current episteme-publication precision vocabulary for DRR, architecture, and pattern-drafting work.
They exist to prevent one sentence from mixing FPF-side use-boundary, project-side records, actual work or action, method selection, carrier access, and authority records.
They are local recovery aids, not FPF kinds, not record kinds, and not a universal record ontology.
Close each field only by naming the FPF kind by value, relation record, relation phrase, project-side FPF kind and reference by value, or explicit non-use disposition at issue in the sentence.
The same local-aid rule applies to neighboring field names such as sourceRelationClass, explanationSourceRelationClass, comparativeRelationClass, representationValidityUseBoundaryValue, allowedUse, misuseRisk, and worldContactPolicy: they help record a local recovery or reader-use boundary, but they do not become kinds. These local fields do not instantiate evidence, gate, assurance, work, commitment, speech act, decision, release, authority, representation kind, world-contact kind, or policy kind. Read allowedUse as a local reader-fit field under declaredUseBoundary, not as permission, evidence relation, or authority.
Episteme, Publication, and Carrier Distinctions
Trigger Boundary
For FPF-governed wording, use E.10:0.2, E.10:0.2a, E.10:0.2b, E.10:0.2c, and E.10:0.2d for lexical cues within the connected F.19 reading and for selection of an already known applicable pattern.
This pattern is applicable only when one C.2.P-specific source-expression, episteme, publication, carrier, source-to-use, or use-disposition distinction is still needed to select or safely use the direct receiving pattern. When the pattern and field at issue are already recoverable, apply it directly.
When this pattern is applicable, do not restart from word taste. Keep any E.10 cue or routing result as input and recover source-expression clarification, FPF-governed use, current episteme-publication relation set, use disposition, and remaining reader use.
Current Preferred Vocabulary
Use PublicationUnit when the reader inspects a bounded unit inside a publication. A person or an AI agent can perform that inspection.
Do not use it for UI behavior, carrier behavior, front-end behavior, file identity, dashboard behavior, or export behavior; use A.7, specific carrier or front-end wording, or the applicable named FPF pattern instead.
Use the current cluster names directly: PublicationUnit Stability Discipline, Local Head Restoration, and PublicationUnit Primary-Subject Discipline.
When the current entity is a bounded unit inside a publication, use PublicationUnit; when the current entity is authoring or editing work, name that work directly.
Use EntityOfConcern and its applicable EntityOfConcernRef when local wording names the participant of a claim-bearing episteme or episteme-lane view. publicationUnitPrimaryEntityOfConcern has the union meaning defined in C.2.P:4.2.1 and used by E.17.AUD: the carried EntityOfConcern or the non-claim-bearing kind, topic, or subject value. For E.17.AUD.OOTD's separate primary-subject question, use its local publicationUnitPrimarySubject; the two fields are not aliases.
For describedEntity, DescribedEntityRef, primary described entity, EntityOfInterest, or EoIClass, use the exact EntityOfConcern participant, its applicable entityOfConcernRef or EntityOfConcernRef, EntityOfConcernChangeMode, EntityOfConcernClass, publicationUnitPrimaryEntityOfConcern, or the local FPF kind named by value.
Use EntityOfConcernSlot only while inspecting the exact reusable C.2.1 constitution RelationSignature; it is not an episteme field. If no claim-bearing episteme or same-individual U.View is current, use a non-claim-bearing kind named by value or plain topic or subject instead of inventing an EntityOfConcernRef.
For a publication unit’s primary subject, use publicationUnitPrimarySubject under E.17.AUD.OOTD. A section about scheduling maintenance may quote a report about a pump’s pressure. The section concerns the schedule, and the cited report concerns the pump. For this technical identification, establish that the unit carries the claims of the identified report and that the unit’s primary subject is the report’s EntityOfConcern. An index can point to the report under the pump’s name without carrying its claims.
Use ordinary topic, subject, or local referent when those words preserve the distinction needed by the receiving use; use the defined technical designation when episteme constitution or another direct subject rule depends on it. Preserve the applicable register requirements of E.10.
Do not mint any other reusable FPF name from this pattern alone. The E.17.AUD cluster PublicationUnit Stability Discipline defines and constrains PublicationUnit; this pattern only recovers bounded-publication-unit wording into that head and points to the cluster for its tests. FPF-governed uses keep the nearby definition or explicit publication relation set.
PublicationUnit use and non-use boundary
PublicationUnit means one named, bounded unit inside a publication that a reader inspects as one unit: for example, one pattern body, section, table, note, card, sheet, or screen block inside that publication.
Use it when the unit boundary matters to authoring, review, navigation, or a claim about what that bounded unit exposes. Do not use it for the underlying episteme, a view, publication form, whole publication, carrier, file, interface behavior, dashboard behavior, authoring Work, or review process. Name those objects and relations directly.
The bounded unit may carry or expose claim-bearing content, but it is not identical with that content. PublicationUnit keeps the inspected publication-side boundary visible without mixing the unit with authoring action or reader action.
Epistemic Precision Restoration After E.10
Keep any E.10 cue or routing result as input. Apply C.2.P only while a source-expression, episteme, publication, carrier, or use-disposition distinction is still needed to select or safely use the direct receiving pattern. If the pattern and field are already known, apply them directly.
Rewrite execution modes
Direct repair
Use this mode for the ordinary case. Rewrite the sentence; add a reason or non-use boundary only when it changes interpretation or action. No row or note is required.
When the intended function is navigation, “The note helps the reader find section 4.2” is a sufficient result. Stop.
Compact row
Use the five-field row in 4.0a when the recovered distinction must remain inspectable. Add only the optional fact that changes the current claim.
Full check
Use a full check only when:
- several unresolved fields interact;
- the source-to-FPF use is contested; or
- a reusable ontological or naming decision is being made.
The full check records the exact span and function, selected kinds and relations, direct pattern contributions, final wording, remaining reader use, and reopen condition. Add competing readings or rejected overreads only when they pass F.19’s independent-ground, plausible-reader, and contribution tests. Pattern authorship or a source citation alone does not require this mode.
Publishing the compact row as a local note
When the five-field compact row must remain beside the repaired text, publish that same row as a local note. This changes only its presentation.
Completion and reopen boundary
The application is complete when the smallest selected product:
- preserves any E.10 cue or routing result rather than restarting from word taste;
- names the recovered kind, relation, or non-use disposition;
- hands any remaining relation, naming, publication, evidence, work, decision, or assurance claim to the applicable pattern;
- leaves the reader able to take the intended action or understand why the grounds stop short of it. When a sentence gives a result for one claim, extract or publication, retain the connection needed to use it for the original question (C.39.QR:4.1).
Reopen when a replacement head hides another umbrella, a later use adds a material field, the direct receiving pattern changes, or the repair becomes technically exact but hard to use.
Also reopen when an entry cue, summary, dashboard, retrieval snippet, or source note still carries the pre-repair reading. Do not reopen merely because a larger form could be filled.
Archetypal Grounding
Cheap case: ordinary reader help
Starting sentence: “The note supports the reader.”
Questions: What does the note let the reader do? Does the sentence claim evidence, authority, gate passage, work permission, or assurance?
Result: It only helps navigation. Rewrite: “The note helps the reader find section 4.2.” The recovered function is ordinary reader help. No FPF kind, relation, compact row, or full check is needed. Stop after the sentence.
Mixed case: display and decision
Starting sentence: “The dashboard approves launch.”
Questions: Is the dashboard the decision, or does it display one? Which exact project object carries the approval claim?
Result when a passing gate decision is established: “The dashboard shows GateDecisionResult GD-17 for release candidate R; the decision, not the display, records that the gate passed.” E.24.PUB separates publication form, carrier, and occurrence; use E.17 when publishing a reader-facing account of the accepted decision. A.21 supplies the gate-decision meaning. The reader may find and cite GD-17, but must use the direct release or permission rule for launch. Stop.
Result when no decision resolves: “The dashboard is only a cue; launch approval is unresolved.” Block approval-bearing use until the required decision is recoverable under the direct approval rule.
Boundary and Anti-Cases
Transfer Coverage
C.2.P is intentionally narrow but must transfer across three recurrent publication situations:
- FPF-side drafting: pattern text, DRR text, source-relation notes, review-use notes, and pattern draft prose;
- project-side publication: dashboards, explanations, cards, documents, front-ends, rendered files, and generated summaries used around evidence, work, gates, decisions, assurance, or methods;
- external source-expression clarification: seminar fragments, papers, reviews, standards, and tool outputs being clarified before possible FPF use.
In all three situations the same invariant holds: before accepting the wording as current FPF text, recover the distinction between source wording and current FPF wording, claim-bearing episteme, publication construction, carrier-relation construction, relation-like slice, the applicable neighboring pattern for any non-C.2.P field, and remaining reader use.
Bias-Annotation
Conformance Checklist
Current Scan Boundary
For FPF-governed wording, use E.10:0.2, E.10:0.2a, E.10:0.2b, E.10:0.2c, and E.10:0.2d for lexical cues within the connected F.19 reading.
C.2.P conformance begins only when one C.2.P distinction remains unresolved under §0's entry condition. This same condition covers non-FPF source text being unpacked before possible FPF transfer; source status creates no separate entry branch.
Do not copy the E.10 trigger list into this pattern as a second registry. Keep any E.10 cue or routing result as input and recover source-expression unpacking mode, FPF-governed use mode, current episteme-publication relation set, use disposition, and remaining reader use. Use A.6.P only while the remaining direct predicate or participant question is unresolved.
Common Anti-Patterns and How to Avoid Them
Consequences
Operating consequence
For each repair:
- start from the sentence's function and practical consequence, not from a word list;
- preserve any E.10 cue or routing result and recover only the one still-hidden episteme, publication, carrier, source-to-use, or use-disposition distinction;
- use the direct pattern as soon as it is known;
- choose direct repair before a compact row, and a compact row before a full check;
- keep the final sentence readable to a cold practitioner and state what action or non-use remains;
- reopen only when a material field or use changes.
Do not perform global lexical replacement. Do not treat every FPF pattern edit as a full epistemic-restoration case. Do not retain a compatibility field whose only function is to remember a previous internal schema.
Rationale
FPF already contains the relevant ontology. The recurring defect was not lack of concepts but ad hoc wording that bypassed them: source, target, display, object, host, route, supported use, and similar trigger terms packed several FPF kinds and relations into one convenient phrase; they are examples of wording to unpack, not current replacement vocabulary.
For FPF-governed wording, use F.19 with the E.10 cues, then use only the pattern whose field remains unresolved: F.18 for a reusable name, A.6.P for relation precision, and A.7, C.2.1, E.17.0, E.17, MVPK, or E.24.PUB for the relevant EntityOfConcern, episteme, view, publication, or carrier distinction. C.2.P stops as soon as the direct pattern and field are known.
Because every normative FPF pattern must satisfy E.2, epistemic precision is not a value apart from P-2 Didactic Primacy. A stricter repair is incomplete if it turns reader-facing problem text into a kind inventory with no working situation or first useful move. The remedy is recognition wording whose claim remains recoverable through the Tech interpretation or a named FPF pattern application, with a declared-use boundary when that boundary matters.
SoTA-Echoing
The question is when ordinary contextual interpretation is sufficient to repair an episteme-publication claim, and when its intended use requires further recovery of the actual participants, relation or governing condition. Sections :4.0a–:4.1.3 select progressive recovery: understand the sentence in its context, recover only the unresolved distinction needed for the receiving use, and stop when that use is clear under its direct pattern. This is an FPF synthesis whose benefit and cost depend on what remains unresolved.
An alternative is a competent reader’s contextual explanation. It can recover the whole intended claim with less effort than an explicit record. In :5.0.1, understanding “supports” as helping the reader find a section is enough to write the direct sentence and finish. In :5.0.2, the meaning of “approves” can be understood while the particular decision and its scope still need to be recovered. The useful extra work is locating that decision under A.21; repeating a definition of approval would leave the receiving launch decision unsupported. These cases support the conditional choice in :4.1.2, not a claim that contextual readers generally fail.
The comparisons below retain designation, contextual explanation and explicit-constraint contributions for these different questions. They do not supply the local FPF relations themselves.
The selected line accepts the cost of consulting a direct pattern or recovering a source fact when that contribution is needed to settle the receiving claim. The ISO row constrains designation work; the SHACL row makes the explicit-constraint analogy and its limit inspectable; the two contextual-recovery comparisons explain why a full record is conditional. This is a reasoned choice for the cases and conditions above, not a measured advantage over every capable reader. Reopen the comparison when a cheaper explanation preserves the same receiving use, an actual repair loses a necessary distinction, or the governing claim or source changes.
Relevant FPF Patterns
The current FPF corpus already has patterns that contribute to this discipline:
E.10supplies word/head, morphology, register, and naming cues and routes within the connectedF.19reading.E.10.D2gives the "thing vs words vs rules" discipline and the carrier humility rule.F.18gives the local-first naming protocol: governed value and kind; the pattern and contribution that define or constrain them; by-value scheme; local-sense claim; intended use; candidate comparison; and any separately obtaining relation.A.6.Prestores an ordinary sentence with the direct predicate, participants, and defining pattern. It adds a candidate note, reusable declaration, occurrence identity, or representation only when unresolved meaning or the receiving use needs that detail.C.2.1gives episteme identity through exact claim content, EntityOfConcern, and effective ReferenceScheme and keeps empirical grounding and edition continuity as distinct direct relations.A.7keeps EntityOfConcern, Description episteme, and publication carrier distinct.E.17.0tests view conformance to a viewpoint;E.17governs reader-facing publication forms for an accepted account;E.24.PUBdistinguishes publication occurrence, form, and carrier.A.15.4shows how to keep an encountered publication, display, or low-articulation cue distinct from the project-side FPF kind and reference used for work or reliance.A.16coordinates language-state moves;A.16.0supplies an optional trajectory account when their history matters;A.19defines the CharacteristicSpace and its reusable predicates;B.2.5governs supervisor-subholon feedback;C.27distinguishes state or rate readings from intervention-sensitive temporal claims;A.3.3defines the reusable state-space and transition-law episteme.E.19compares a materially revised pattern edition with its earlier supported uses and tests the changed wording.A.6.Acarries action-invitation discipline when a publication, representation, or cue invites an action without itself becoming authority, evidence, gate passage, or work completion.C.11carries decision-making and decision-record discipline when the question under repair is a decision rather than generic action.A.15andA.15.4split system-role kind and assignment, Method, WorkPlan, and actual-Work alignment from appearance-based reliance repair; do not useA.15as a universal repair for episteme-publication wording.E.9is the campaignDRRpattern for campaign-level content decisions;E.11is only for entry-discoverability situations and must not organize an episteme and publication repair by default.
These internal FPF patterns remain primary:
Use the internal patterns above for the particular claim left unresolved. Under :4.6, a repair closes when the intended kind, relation and relevant use condition are recoverable; an ordinary sentence can supply that result. A remaining unrecovered distinction keeps the affected use unresolved under its direct rule, while an understood claim outside the existing FPF account can require a separate framework decision.
E.19 Review Profile Carry-Through
Use this section for an E.19 judgment of a pattern edition. Select only the profiles whose triggers hold for that edition and use. For an ordinary sentence repair, :4.6 supplies the completion test.
Run PCP-TERM when the repair changes episteme-publication-heavy naming, umbrella words, slash compounds, trigger-word replacements, or relation wording.
Run PCP-BRIDGE when the repair imports claims, terms, or norms across contexts, disciplines, or reference planes under E.19.
Run PCP-ENTRY only when the repair changes which FPF pattern a working reader should apply in a problem situation. Do not use PCP-ENTRY as a substitute for the epistemic precision restoration itself.
Run PCP-PRAG when the repair changes reader action, practical payoff, or what the working reader can safely do next. A type-correct but inert repair fails this line even when every recovered kind is technically right.
When the repair changes evidence, proof, witness, grounding, explanation, gate, release, or engineering-justification claims, apply the relevant FPF pattern (A.10, E.17.EFP, A.20, A.21, B.3, or another applicable pattern) and select the current E.19 profile by the changed claim. Do not mint a local evidence-review profile inside C.2.P.
Relations
- Builds on:
E.2Pillars, especiallyP-2 Didactic Primacy;F.19,E.10,E.10.ARCH,A.7,F.18,A.6.P,C.2.1,E.17.0,E.17, MVPK,E.24.PUB, andA.6.A. - Coordinates with:
E.6,E.7,E.8,E.9,E.12,E.19,A.10,A.15,A.15.4,B.3,A.20,A.21,A.6.F,A.6.P.WMR,A.6.3.CSC,A.6.3.CR,A.6.3.RT,C.30.P,C.16.P,C.16.Q,E.17.EFP, andE.17.ID.CR. - Does not replace:
E.10general lexical rules,F.18naming protocol,A.6.Prelation precision, or local episteme and publication patterns. It says when those patterns must be applied to episteme-publication-heavy wording.
C.2.P:End
Reliability R in the F–G–R triad
Reliability (R) concerns the warrant for one typed claim under an explicit claim scope (G). Keep the support useful and its limitations visible. A numerical R needs a receiving model for its meaning, scale, inputs, and dependencies; no universal fold is supplied. On reuse, name the actual relation and apply only its warranted loss to R, not to F or G.
Type: Architectural (A) Status: Stable
Problem frame
KD‑CAL asks: “What does the available support warrant for this claim and use?” FPF separates expression, applicability and warrant in the F–G–R assessment. Relations traversed by a justification path are named separately:
- F (Formality) describes how the claim is expressed and how strongly it supports verification workflows (C.2.3).
- G (Claim scope) describes where the claim is asserted to apply as a set-like object (A.2.6).
- R (Reliability) describes how strongly the claim is warranted by linked evidence under that scope.
- CL / CL^k / CL^plane (Congruence Levels) describe fit or loss for the relation families that define them—for example, a semantic relation, kind relation, or reference-plane relation (B.3, C.3, F.9). A CL value belongs to the declared relation or traversal used by the path, not to the claim as a fourth coordinate. Shared wording about a "context" creates no relation and no loss value.
In practice, the triad is frequently used before it is made explicit:
- Authors implicitly “average” disparate evidence and report a single confidence.
- Teams treat higher formality (F) as if it automatically implies higher warrant (R).
- The claim's scope grows implicitly through phrasing instead of explicit scope operators (A.2.6).
- A claim or its evidence is reused after a change of scope, kind, plane, notation, source-local meaning, model-use basis, or evidence basis without naming the actual relation and routing its declared loss into R.
This pattern makes R explicit in KD‑CAL and fixes the triad discipline required by Kind‑CAL (C.3) and the Trust & Assurance calculus (B.3).
Problem
FPF needs a reliability coordinate that is:
- Auditable. A reader can recover the grounds for the supported conclusion, the argument connecting them to this claim and use, and the effect of actual reuse limitations.
- Composable. Support can be combined under its warranted meanings and dependencies without illegal scale arithmetic; where no numerical aggregate is justified, the separate contributions remain usable.
- Orthogonal. R is not conflated with F (expression) or G (scope).
- Relation-aware. Any loss declared by an actual scope-translation, kind, plane, notation, source-local, model-use, or evidence-reuse relation is explicit and affects R only.
Forces
Solution
Canonical triad relation
Definition DEF‑C2.2‑1 (Epistemic location).
An epistemic location for a claim c is the tuple:
Loc(c) = ⟨F(c), G(c), R_eff(c)⟩
where:
F(c)is the assessment of the expression properties needed for this use under C.2.3. It may use a declared profile or a justified local ordered summary; it is not a universal ordinal value.G(c)is Claim scope (A.2.6), treated as a set-like scope object.R_eff(c)names the effective warrant for this claim and use. Its meaning and scale come from the B.3 receiving model, not from the letter R. A probability-like or ratio-scale value in[0,1]needs that model; an ordinal proxy keeps its declared ordinal meaning. If no common quantitative model is justified, report R as unquantified with the separate support and bounded conclusion, rather than substitute zero or invent a score. When a guard consumes a path-specific value, identify the actual PathId and its model (§4.8.A / G.6); a declared policy name alone does not warrant collapsing paths into one scalar.
A location always concerns one exact claim. G carries its U.ClaimScope; any stance, reference plane, effective scheme, model-use basis, working situation, evidence basis, or validity window is stated separately when it changes interpretation or use:
- No generic
Kor Context value is part of epistemic-location identity; the exact subject-specific values above remain independently governed. S ∈ {design, run}is the claim’s stance value; keep design-time and run-time assurance separate.ReferencePlaneis declared where applicable; plane crossings applyCL^planeand penalize R only.- When a Working-Model publication elects B.3.5, or another applicable assurance requirement calls for it, declare the required
validationModeand grounding under that profile (E.14).
Mode and assurance contribution. A declared validationMode states the justification posture under its selected profile. B.3.5 defines the allowed modes and grounds for its covered claims. Use B.3.3 to judge what the actual support establishes for this claim and receiving use; the mode alone does not select a fixed mix of assurance contributions.
In all modes, R remains warrant, not ontological truth; “proof ⇒ R=1 in the world” is a category error.
Formal-input rule. Empirical R may be N/A for a strictly axiomatic claim. Preserve the proof and its conclusion under the stated axioms; do not set R_proxy := F for an R fold. The tag line=formal, a postulative mode, or rescaling F into [0,1] supplies no conversion model. A declared F-derived ordinal proxy is valid only for its own ordinal meaning.
Any F-derived value used as another quantity needs a receiving model establishing its meaning, scale, conversion, assumptions, and warranted application; in particular, checkability alone does not establish a probability about a real system.
⟨F,G,R⟩ is an assurance tuple, not a U.CharacteristicSpace; G remains a U.ClaimScope. Successive assessments may be compared through their actual expression properties, scope changes and warrant grounds. A separate characteristic-space chart requires its own Characteristics and Scales under A.19; the tuple alone supplies no metric or dynamics.
What Reliability R means in KD‑CAL
Definition DEF‑C2.2‑2 (Reliability as warrant).
R is a conservative, evidence-bound indicator of how strongly the claim "holds as stated" under its declared U.ClaimScope and the separately named evidence and use conditions. It is interpreted as warrant strength, not as truth.
Interpreting the assessments.
- A higher
Rmeans stronger warrant only within the same declared quantity, scale, claim, and receiving model. A number from another model is not automatically comparable. - A change in F identifies changed expression properties or a justified change on a declared local scale. It can enable a particular check or reuse without establishing the claim's warrant or widening its scope.
- A larger
Gmeans “the claim applies to more cases,” but does not itself imply the claim is warranted in those cases.
Dependence-sensitive support composition
Definition DEF‑C2.2‑3 (Support composition).
For claim c, identify the support needed by its receiving use. Distinguish indispensable premises, alternative sufficient arguments, complementary evidence for the same question, support for different scope slices, and counterevidence. A source's presence in the graph does not make it an indispensable premise. B.1.3 supplies the synthesis Method and its guidance and controller examples.
Let SpineClaims(P) name premises and lemmas required by argument P; SpineRelations(P) names the scope, kind, plane, notation, source-local, model-use, and evidence-reuse relations it traverses. Satellite citations are not required premises. Retain each contribution's source, assumptions, scope, and limitations, including shared datasets, duplicated evidence, and common bias.
Choose the operation from the model. B.3 requires the target quantity, compatible scales, dependency assumptions, and warranted operation before an aggregate is calculated. An indispensable weak premise can limit an inference, but min is not a universal probability or warrant fold. Two necessary independent conditions with probabilities 0.9 each give 0.81 for their conjunction; minimum 0.9 overstates it. With only those marginal probabilities, the joint probability lies between 0.8 and 0.9. This settles a required lower bound of 0.8. A requirement of 0.85 needs more information about dependence. Monotonicity and boundedness of a proposed rule are insufficient.
Alternative and complementary support. A sufficient argument may be usable without the others. A maximum can select the best attested argument value under a model whose result has that meaning; it does not measure combined corroboration. Complementary evidence may strengthen or qualify a conclusion by addressing different rival explanations or limitations, even when neither source is sufficient alone. Count neither publications nor method names as independent confirmation. A shared bias may leave apparent agreement uninformative. There is no universal “never exceed the best source” cap and no entangled-source fallback to minimum.
Scope and conflict. Retain different G_path slices under A.2.6; do not use maximum to hide unsupported regions. For overlapping-scope p and ¬p, preserve credible contrary evidence. Separate claims only by distinctions established by the sources; otherwise narrow, qualify, or withhold the affected conclusion. An uninformative study, a lack of decisive support, and evidence against the claim are not interchangeable.
Useful synthesis without a common numerical R. If there is no warranted common model, retain separate support and limitations and give a bounded reasoned synthesis. This may finish the receiving question without a score, penalty table or extra study. The feasibility and worth of further inquiry are separate C.11/C.19.2 questions; their cost does not alter what the current evidence supports.
Relation-specific congruence penalties route to R only
A reused claim may traverse more than one independently governed relation. Before calculating R_eff, state what changed and use the rule for that change. A.2.6 owns claim-scope operations; C.3/C.3.3 owns kind relations; F.9 owns a semantic Bridge between exact local-sense cells; notation, reference-plane, model-use, and evidence-reuse relations keep their own definitions. None is a universal crossing relation.
Invariant INV-C2.2-1 (R-only penalty routing). For each traversed relation r whose rule declares a congruence loss:
F_out = F_in.G_out = translate(r, G_in)only whenris an applicable A.2.6 scope translation; otherwiseG_out = G_in.R_out ≤ R_inon the named ordered warrant scale for a loss-only transformation, with any numerical penalty justified by that relation's receiving model.
A scope translation may narrow or re-express G; it never widens the claim silently. Assess changes in expression under C.2.3. C.2.1 determines whether a changed expression still expresses the same episteme. A correction to a previous F assessment can leave both the expression and the episteme unchanged. A semantic Bridge changes neither kind nor scope by itself. A kind or plane relation supplies no semantic correspondence unless that separate relation also obtains. Evidence reuse changes warrant only through its own evidence-use or reliance claim.
There is no implicit crossing. If a reuse depends on a changed value and its required relation or operation is absent, unresolved, or outside its applicability, the reuse is non-conformant. Each path records the relations it traverses and routes their declared losses to R, while every other coordinate changes only under its own rule.
Worked micro-example: scope revision and evidence reuse
A materials-lab claim says:
c_lab:"Adhesive X on Al6061 retains at least 85% of its initial tensile strength throughout exposure at 120–150 °C for up to two hours."
The claim covers Al6061, actual adhesive temperatures from 120 to 150 °C and exposures of up to two hours. Call this scope G_lab. The supporting tests used rig Calib-v3. The lab accepts results from this rig for one year after testing. A plant engineer can restrict the operating conditions covered by the claim and then judge whether the lab tests support the restricted claim at Plant B.
- State the plant claim and its scope. At Plant B, assume calibration establishes
|T_actual − T_reported| ≤ 2 °Cthroughout the intended operation (C.16). Restrict the claim to Plant B operating on Al6061 for up to two hours with readings in[122,148]°C. Every such exposure satisfies the original temperature condition under the stated bound. A reported 149 °C could correspond to an actual 151 °C, so this plant rule excludes that reading. A.2.6 governs this restriction. - Judge reuse of the lab evidence. Identify the tests run on
Calib-v3, establish whether their results are still current for the intended use, and check whether the material, loading and measurement conditions support the claim for the plant exposures. The temperature bound establishes which exposures the rule covers. It supplies no tensile-strength measurement. A declared fitCL=2records the relation's fit, not a probability decrement. State the actual reuse limitation. Calculate a numericalR_effonly if a receiving model establishes the R quantity and this loss; otherwise keep the separate support and qualified plant conclusion. This judgement does not perform the scope edit.
If lab and plant use distinct local meanings for a material term, F.9 separately tests a Bridge between their exact F.17 cells. Its semantic loss is not the calibration correction or the evidence-reuse result. A further safety narrowing of that reported-temperature interval to [125,145]°C is another explicit A.2.6 ΔG− decision.
The example therefore preserves one simple rule: name each changed value and relation once, change G only through the scope rule, and reduce R only through the loss rule that actually applies.
Effective reliability under reuse: a justified loss model
Definition DEF‑C2.2‑4 (Effective reliability under reuse).
A relied-on relation may introduce loss in the support for the receiving claim. Name that relation and its scope, semantic, notation, model-use, evidence-reuse, kind, or reference-plane rule. The corresponding CL, CL^k, or CL^plane is an ordinal summary belonging to that relation family; the ranks are not amounts to subtract from R.
A quantitative loss model names the receiving quantity and scale, input meanings, dependencies, loss interpretation, derivation or calibration, and applicability assumptions under B.3. If the model uses functions Φ, Ψ, Φ_plane, and a combining operation Π, cite the definitions and versions used by the model. A policy identifier, table, monotonicity, boundedness, or clipping to [0,1] does not by itself justify any of them.
For a loss-only interpretation on an ordered scale, worsening fit cannot by itself count as an improvement in warrant. The model must justify any pathwise CL minimum, repeated-loss treatment, or neutral term for an absent relation. Preserve separately justified ordinal chain-congruence operations in C.3.3; they do not provide a numerical R penalty.
Quantitative illustration: a lower bound for two required events. Suppose a receiving claim requires events A and B. An applicable model and evidence establish P(A) ≥ 0.82 and P(¬B) ≤ 0.15 for the same use. The probability bound P(A ∩ B) ≥ max(0, 0.82 − 0.15) = 0.67 follows without an independence assumption. Here 0.67 is a lower bound, not a point estimate or a generic confidence score. The 0.15 term comes from the stated bound on failure of B, not a CL rank. If those event meanings or bounds are unavailable, the calculation is unavailable.
Reuse conditions. Apply a relation's justified admissibility or protection condition to the named use before relying on it; neither this pattern nor a bare CL rung creates a universal waiver obligation. If the condition is unsupported, limit or stop that reliance while retaining any independently supported source conclusion.
Formality and scale discipline
- A local ordinal formality summary, CL, and ordinal R proxies permit only operations justified for their ordered meanings, not arithmetic pretending they are ratio-scale measurements. A formality profile is not one number to rescale.
- A numerical R requires a justified receiving model even at high formality. A complete formal proof remains useful with empirical R marked N/A; a missing empirical score does not demote the theorem.
- When support has no common numerical model, state its separate contributions, limitations, and the bounded conclusion. Use validity windows, empirical reproducibility information, and B.3.4 decay only where the claim actually consumes them.
Assurance lanes
Use B.3.3 to select the assurance contributions needed by the claim and receiving use. In this KD-CAL account, these labels distinguish contributions when reporting them is useful:
- TA (Typing assurance): establish the types and semantic correspondences needed for this transport or composition. Use B.3.3’s Concept-Bridge Assurance when the conclusion depends on whether meanings, referents or conditions correspond across descriptions. When the argument also relies on a tool, establish separately whether that tool is qualified for the operation it performed.
- VA (Verification assurance): logical/algorithmic checking, proof, model checking, static guarantees.
- LA (Validation assurance): empirical adequacy under declared conditions, tests, benchmarks, telemetry.
Lane reporting distinguishes logical soundness, empirical adequacy and typing or semantic alignment. Keep the applicable contributions separately interpretable. SCR, notes or lane tags may express them when the receiving use needs that form. Any combined numerical warrant still requires the model in §4.3.
Scope operations are kind-safe (and use the ClaimScope algebra)
Establish the kinds of the claim subjects before combining their scopes.
Well-formedness constraint WFC‑C2.2‑1 (Type before scope).
Let G1 and G2 be claim scopes for claims about entities of kinds K1 and K2. A scope operation that combines them—such as G1 ∩ G2 for serial intersection or SpanUnion({G_i}) for parallel coverage—is defined only if:
K1 = K2; or- an exact C.3/C.3.3 kind relation or cast makes the operation well typed for these participants and this direction.
An A.2.6 scope translation changes G only under its own rule. A kind relation does not translate scope. If distinct source-local meanings also matter, an actual F.9 Bridge and its bounded-use claim are separate; neither repairs an ill-typed scope operation.
This constraint prevents “type-by-scope” anti-patterns where scope manipulation is used to hide type mismatch.
Minimal authoring recipe
A minimal, conforming KD‑CAL authoring flow for reliability is:
- Fix the typed claim. State the claim, its EntityOfConcern and the type distinctions needed to interpret it (C.3).
- Declare claim scope. Identify the exact
Gand state the conditions that determine membership under A.2.6. Use the applicable scope operator when combining, translating or revising scopes. - Declare interpretation conditions. State design or run stance,
ReferencePlane, effective scheme, model-use basis and working situation where each changes the claim or its use. DeclarevalidationModewhen an elected B.3.5 profile or another applicable assurance requirement calls for it.Galready carries claim scope; do not add a generic Context identifier. - Connect the supporting grounds. Make the relied-on grounds and their bearing on the claim inspectable under B.3.3. Use evidence references, stubs or lane tags when the receiving account requires them. Retain validity windows and currentness rules where they affect the use (B.3.4).
- Identify support roles and dependence. Distinguish required premises, sufficient alternatives, complementary support, scope slices, and counterevidence. Identify duplicated data, shared assumptions, and plausible common biases.
- Choose a justified calculation or a synthesis without a common numerical R. Name the receiving quantity, compatible scales, assumptions, and model before any numerical fold. Otherwise retain separate support and a reasoned bounded conclusion; do not substitute a universal min or max.
- Name actual relations on reuse. Use A.2.6 for an applicable scope translation, C.3/C.3.3 for a kind relation, F.9 for a semantic relation between exact local-sense cells, and the pattern governing notation, plane, model-use, or evidence reuse. Record the fit or loss declared by each traversed relation. If a required relation is absent or unresolved, stop that reuse; a generic cross-context Bridge cannot substitute for it.
- Return the usable result. State F, G, and the supported conclusion with its warrant and limitations. State R numerically only under the justified receiving model, with the actual calculation and relied-on loss definitions. A formal conclusion does not require an empirical score. Inquiry or action choice, if needed, remains separate.
Authoring template: claim-local support summary
When publishing a path-specific R for a guard or decision, include enough of the support summary to identify its quantity, model, inputs, and use. G.6 PathId references can carry this information; no new Core type or mandatory table is introduced.
Retain actual CL summaries where their relation uses them; any chain minimum needs that relation's ordinal meaning. Empirical time limits and the fixed theory version of a proof remain different conditions. If several paths are consumed, retain their distinct scopes and models and cite their PathId(s). A synthesis without a common numerical R may instead give its separate contributions and limitations in ordinary prose.
Archetypal Grounding
Informative; non-binding.
System illustration
System. A brake controller S has a claim:
c1:“For road friction μ ∈ [0.2, 0.9] and vehicle mass m ∈ [900, 2200] kg, wheel slip stays in [0.05, 0.25] under ABS control.”
-
The formality assessment of
c1identifies the declared model variables, constraint predicates and execution semantics of the test harness under C.2.3. These properties enable the specified checks and tests; their results and the model's fit to the physical plant require the separate evidence below. -
G(c1)has the stated μ/m bounds, but this illustration leaves the speed domain and admissible tire set unspecified. Under A.2.6, those domains and any coupled restrictions are needed to decide membership of a slice satisfying the stated bounds; that membership remains unresolved here. A product set in(μ, m, speed, tire)space is justified only if its scope predicate admits every combination of the selected domains. -
Evidence:
- VA: model-checking of a simplified plant/controller model (strong, but only for the simplified plant).
- LA: HIL simulation + track tests under sampled conditions, with the observation windows and any currentness qualification required by the receiving claim.
- TA: typed alignment between “μ” in simulations, “μ” in the estimation pipeline, and “μ” inferred from real-world sensors.
If track telemetry is used as evidence for the road claim, establish the grounds for that reuse, including the road claim, telemetry edition, operating scope, intended use, and any validity window on which that use depends. Apply only the fit or loss declared for that evidence reuse; G(c1) changes only through a separate A.2.6 scope revision.
Episteme illustration
Episteme. A paper asserts two claims about an algorithm A:
c2:“A terminates for all inputs in domain D.” (axiomatic / proof-carrying)c3:“A achieves ≥ 0.92 F1 on dataset family F under deployment preprocessing P.” (empirical)
c2 can achieve high VA through its proof under the stated axioms and domain; an empirical LA contribution may be N/A. If the conclusion is also used for an implementation, establish the correspondence between “domain D,” the proved algorithm and that implementation’s input and execution model.
c3 needs empirical support for its stated dataset and preprocessing conditions. For production use, establish the dataset/preprocessing applicability and the grounds for reusing that evidence, then apply the warranted loss of that reuse to R_eff; change G separately if the production claim has another scope. Use B.3.4 when changed conditions or an applicable review rule can alter that support; retain any required validity window or shift condition. An unchanged bounded use with still-applicable support needs no new freshness policy.
Bias-Annotation
Informative; non-binding.
- Ontological and epistemic bias: High formality is often mistaken for high warrant (“proof therefore true in the world”). Report the applicable assurance lanes and route declared reuse loss into R. A formal proof keeps its conclusion under its axioms; empirical adequacy needs its own support.
- Pragmatic bias: Teams may Goodhart R by narrowing scope or selecting easy tests. Explicit scope and its revisions under A.2.6 expose the cases covered by the claim.
- Governance bias: Overconfident reuse after a changed scope, scheme, model use, evidence basis, kind, or plane is a recurring failure. This pattern requires the actual relation and its declared loss instead of one generic crossing label.
- Didactic bias: A single scalar can hide what kind of warrant exists. Separate lane reporting keeps the kinds of support visible.
Conformance Checklist
Normative.
Common Anti-Patterns and How to Avoid Them
Informative; non-binding.
Consequences
Informative; non-binding.
Rationale
The triad separates expression, applicability and warrant so that improving one is not mistaken for improving the others.
- G states applicability. It states where the claim is asserted to apply. If G is implicit, teams argue about “what was meant” instead of updating scope.
- F describes expression properties. It identifies what the claim's terms, semantics and inferential relations make available for interpretation, scrutiny and reuse. If these properties are conflated with R, formalisation becomes a rhetorical substitute for support.
- R carries warrant. It describes support for this exact claim and use under a stated meaning and any scale used. Its inputs must not erase the distinction between a necessary premise, complementary evidence, and a credible contrary result.
Each changed value has its own rule. Route a traversed relation's declared congruence loss into R; change the claim's expression or scope only through the applicable expression or scope operation. Re-establish warrant for the receiving use instead of carrying it forward without a basis.
No universal fold is conservative for every support model. Minimum can overstate a conjunctive probability and can also suppress useful complementary evidence. The small common rule is to establish input meanings and dependencies, calculate only what they warrant, and otherwise preserve a useful bounded synthesis.
SoTA-Echoing
Normative.
SoTA pack use. When the current warrant relies on a result from a G.2 SoTA Synthesis Pack, cite the relevant ClaimSheet IDs and CorpusLedger entries. Cite a BridgeMatrix row only when the current path uses an F.9 cross-local semantic Bridge represented by that row. Directly identified sources and support can suffice where no pack result is used.
The current methodological comparison is Gutierrez, Glymour and Davey Smith, Evidence triangulation in health research (2025): compare target questions, design assumptions and possible shared biases; use qualitative synthesis where quantitative combination is unwarranted. Adapt those distinctions for heterogeneous support. The paper supplies neither a universal R scale nor a requirement to commission another study for every useful conclusion.
Relations
Builds on:
- C.2 (KD-CAL overview).
- A.2.6 (claim scope and scope revision).
- C.2.3 (Formality F).
- B.3 and B.3.3/B.3.4 (assurance, evidence lanes, and refresh).
- B.1.3 (Γ-fold patterns).
- C.3.3 (cross-kind use).
- G.6 (EvidenceGraph PathId discipline).
- A.6.3.RT.OE/A.6.3.RT (expression construction and scheme changes).
- C.29 (an unresolved question about choosing, transferring or relying on a mathematical representation).
- A.1.1 (selected model-use structure).
- A.10 for source-to-use and bounded-reliance accounts when they must remain replayable.
- B.3 when a named assurance claim is being judged.
F.9 is used only for an actual semantic correspondence between distinct local meanings. Reference-scheme or reference-plane changes keep their own rules.
Coordinates with: C.16 for measurement claims, E.14 for working-model assertions, F.17 for optional local-meaning addresses, and E.18/E.17/A.21 when the work involves their transfer, publication, or gate objects. G.2 supplies relevant source-pack entries; G.7 remains the conditional calibration path for its declared cross-Tradition/F.9 Bridge use, not a universal calibration owner.
Used by: C.3.3 for cross-kind reuse discipline, guard macro bundles in C.3.A and C.21, and acceptance or gating logic that consumes R_eff while preserving F and G.
Clarifies: the KD-CAL meaning of reliability implicit in C.2:4.1 and the relation-specific reuse claims referenced across B.3 and C.3; it does not create a universal transport relation.
C.2.2:End
Language-State Chart over U.CharacteristicSpace
Type: Architectural (A) Status: Stable Normativity: Normative unless marked informative
Plain-name. Language-state chart.
Builds on.
A.19, E.10, F.18.
Used by.
C.2.LS, C.2.3, C.2.4, C.2.5, C.2.6, C.2.7, A.16.0, A.16, A.16.1, A.16.2, B.4.1, B.5.2.0, F.9.1, A.6.P, C.16.Q, A.6.A.
Use this pattern when. Routing, comparison or publication needs a language-state position, a comparison on declared Characteristics and Scales, or an evaluation of a slot predicate. Recover the separate readings and conditions hidden by labels such as early, ready or settled.
First question. Which expression, articulation, closure or representation readings can change the next action, and which source or use conditions must be stated beside them?
What goes wrong if missed. Teams flatten the relevant Characteristics, source and use conditions, grounds for assessment and publication lane into one maturity adjective. A polished document can then look ready although the claim is still open, or a useful early cue can be rejected because it is not yet polished.
What this buys. A recoverable position claim that shows the distinctions needed by the next action while keeping the episteme publication separate from its grounds, publication form and carrier.
First useful chart
When a chart position is needed, make the following content recoverable; a row is one possible form:
- the episteme publication being positioned;
- only the facet values or supported sets of possible values relevant to the next action; use intervals where the Scale admits them and mark a decision-relevant unknown as
unknown; - the grounds for those readings, with any source, evidence, access or action conditions on which the next use depends;
- any local threshold used by the named next action and whether the readings meet it, fail it, or leave the result undecided;
- publication form, face, or carrier only when that distinction changes interpretation, preservation, or the next action.
Filled example. Pump-alert-17 is the alerting episteme publication being positioned. Its affected asset and discrepant reading are visible, but the claim structure is not yet complete; rival diagnostic routes remain open; the alert enters the operator's acknowledgement-and-check procedure; and its representation mixes text with telemetry. Formality is omitted because the present routing decision does not depend on it. The grounds are trace T-17 and operator note N-4. The local rule permits routing to the diagnostic-question seam once the affected asset and discrepant reading are explicit. Form, face, and carrier are omitted because none changes that decision.
Sufficient result and non-use. Use an ordinary account, a direct facet reading or an existing chart directly when it supplies the distinctions, grounds and conditions required by the receiving use. Where that use needs a chart position, comparison or predicate evaluation not yet supplied, declare or reuse the needed chart basis and state the missing readings or result.
Practical result. The position account helps the practitioner decide whether to continue, stop, or route the publication to the pattern that governs the next question; it does not itself authorize that action.
Problem frame
In engineering, inquiry, operator, and management practice, teams often need to say where a U.Episteme publication stands before it has reached an endpoint subject pattern. That publication may appear through several cue-bearing, route-bearing, or endpoint-bound publication forms. The chart claim remains about the U.Episteme publication rather than about a local alias or a carrier lane.
Reusable arrangements or notations for cue packs, routed cue sets, abductive prompts, typed route-bounded projection publications, partial normal forms, and endpoint records are publication forms. A completed cue or record can express the episteme being assessed; its arrangement does not determine its chart position. An MVPK face is a selected publication form under E.17; it remains separate from the episteme, publication occurrence and carrier under E.24.PUB.
By contrast, a service disturbance, a model-vs-observation discrepancy, a bodily tension, a telemetry trace, a model output, or a carrier document may trigger, witness, or carry that episteme, but none of those is itself a coordinate in the space.
Practitioners still have to determine the relevant readings, any threshold used, the publication form needed and the claims those readings support. Labels such as raw, early, settled, or ready can hide these separate values and decision conditions.
Problem
When relevant facets and use conditions remain implicit:
- teams collapse several facets into one maturity story;
Fis silently misused as a surrogate for articulation, closure, source and use conditions, and representation factors;- thresholds are published as vague readiness statements instead of explicit facet conditions;
- source phenomena, epistemes, publication forms, publication faces, and carriers are conflated;
- bridge and endpoint work inherit under-described upstream states.
Forces
Solution
A language-state chart is a declared U.CharacteristicSpace whose basis slots use selected language-state Characteristics and Scales under A.19. A profile containing several Characteristics expands into their separate slots; it is not one atomic Coordinate.
Chart, position claims and publication
The space declaration identifies the chart and makes its basis, optional overlays and typing hooks recoverable under A.19. A position claim states facet readings about the selected published U.Episteme edition; the E.24.PUB publication occurrence remains separate. A score table or local route map can express the declaration and position claims. Its reusable form and carrier remain distinct from those contents.
Changing a slot's Characteristic, Scale or position requires a changed space declaration under A.19. Revising a reading changes the position claim and can use the same space. Changing a threshold changes the predicate under A.19:5.1.8. Changing only the form or carrier of these unchanged contents changes neither the space, reading nor predicate.
Core use
Use the declared chart to answer five questions:
- which basis slots define where the episteme stands;
- what a position claim in that chart means;
- which thresholds are locally declared over those slots;
- what comparisons are admissible without cross-facet collapse;
- and how the same position claim stays distinct from the publication form expressing it.
Position reading under A.19
A language-state position is a partial, slot-explicit coordinate claim in the declared language-state U.CharacteristicSpace.
ValueSet(Scale) in A.19 is the full Scale domain. Publish each selected slot's reading as one value or as a supported set of remaining possible values within that domain. This set describes incomplete information about the slot's value. Use an interval only where the Scale admits it; state a decision-relevant unknown separately when the available information does not determine the needed reading. For closure, {CD1, CD2} can report two remaining possible values while the assessment is incomplete. Separate source and use claims follow C.2.6:4.3: two unresolved possible sources differ from an established contribution by both.
When a decision depends on several slots together, retain the dependencies that can change it. A constraint or bound can already decide the question while some joint readings remain unknown. Suppose the grounds leave only (AE4, CD1) and (AE2, CD3) possible. A rule requiring both AE4 and CD3 fails in every remaining case, although each required reading is individually possible. If all the remaining combinations give the same decision, the exact position can remain open.
position here abbreviates these slot-explicit A.19 coordinate claims.
Facet basis
The language-state chart is coordinated by explicit facet subject patterns rather than by an informal master progression. Select the Characteristics and Scales needed for the next action from:
C.2.3for the selected expression properties in a formality assessment, or an explicitly justified local summary;C.2.4for articulation explicitness;C.2.5for language-state closure degree;C.2.7for the language-state representation-factor bundle.
The selected basis slots together define the chart under A.19. Each Characteristic has its own Scale and slot. In particular, neither a formality profile nor a representation-factor bundle is silently treated as one Coordinate. C.2.2a does not govern the internal meanings or Scale semantics of those Characteristics.
C.2.6 supplies the source, evidence, access and action-condition account needed beside the coordinates. That account contributes no slot by itself. A local summary contributes a slot only when its one aspect and Scale are defined under C.2.6:15.1 and A.18/A.19; changing that basis requires an explicit chart change.
Position, grounds and publication
Keep the positioned episteme publication separate from the grounds for its assessment and from its publication form and carrier:
- positioned episteme publication - the
U.Epistemepublication whose current position is being claimed; - grounds / witnesses - disturbances, discrepancies, traces, model outputs, bodily tensions, exemplars, or contrasts that justify the current reading;
- publication forms - arrangements or notations used for publications such as cue packs, prompts, typed projections, partial normal forms and endpoint records; an MVPK face is a selected publication form, identified when that distinction matters;
- carriers - physical or digital media, such as a paper sheet, file or console display, that bear the selected form.
C.2.2a governs the coordinate reading of the position claim. Keep that claim distinct from its grounds, publication form and carrier. C.2.6 also distinguishes a reading's derivation from evidence for its target claim and from the conditions under which someone uses it.
Position publication rule
A published position identifies the episteme publication, states the facet readings relevant to the present use, and gives their grounds. A decision-relevant unknown is stated as unknown; unrelated facets need not be filled merely to complete a form.
For a movement, routing, comparison or endpoint use, name the receiving action and any language-state threshold it uses. If that action also requires a source, evidence or interaction condition, state or cite its predicate beside the facet readings; the condition does not become a Coordinate by being included in the same account. Add publication form, MVPK face, carrier, or SCR/RSCR lane only when that distinction affects interpretation, preservation, distribution, or the next action.
The position is reviewable when the named or inherited episteme publication, relevant readings, grounds, and any relied-on threshold are recoverable, and when any form, face, or carrier mentioned remains distinct from the positioned episteme publication. A polished note or better carrier does not by itself prove a new chart position.
Non-substitution of F
The formality assessment supplies the selected expression Characteristics, each with its own basis slot. A justified local summary may occupy one slot; a multi-property profile does not. These readings describe only the expression properties selected under C.2.3.
A conforming account shall not infer:
- closure from formality alone;
- source or use relations from publication-face format alone;
- representation factors from articulation alone;
- or routing admissibility from a lone
Fstatement.
Where operationally meaningful thresholds exist, they must be published on the relevant slots rather than being disguised as informal F sublevels.
Position versus publication form
Identify the expression or component whose properties supplied each reading, and the expression in which the assessor reports it. Reformatting the assessment table can preserve every reading about the earlier alert. Rewriting the alert can change its articulation even if the assessor uses the same table. Check the affected facet on the changed expression. The table can follow a reusable publication form, including an MVPK face, and its carrier can be identified under E.24.PUB. A subsequent publication has to satisfy the conditions of its receiving pattern.
Threshold publication discipline
If a threshold is used to justify a move or endpoint entry, that threshold shall be stated on explicit basis slots in the chart. Statements such as this is now ready, this has matured, or this is still too early are non-conformant when they substitute for undeclared slot conditions.
Comparison and bridge note
Positions may be compared directly only under one shared chart meaning and the applicable local comparison rule. When comparison or reuse depends on distinct local meanings, recover the two exact F.17 local senses and test the F.9 predicate. Cite a Bridge only when that predicate obtains. State the proposed bounded use and any reliance separately; a loss note is optional and belongs to that use, not to the fact that the contexts differ.
If no Bridge obtains, preserve both local positions. Name the comparison or translation relation needed by the receiving use, or return the applicable missing-governor result instead of inventing correspondence.
Use the recovered position
Use the position facts for the named receiving question. If the next action uses a local threshold, compare the relevant readings with it and state whether they meet it, fail it or leave the result undecided. Use the supported result to decide whether to keep the publication where it is, route it through a seam or prompt pattern, or open a facet or endpoint question. An unknown reading prevents only the action or conclusion that requires it. The receiving pattern governs the route, comparison or publication decision.
Archetypal Grounding
Tell. One note can be used in an operator's acknowledgement procedure yet still have low closure. Another can report an architecture document's assertions in symbolic form while its route choice remains open. Both notes have positions in one language-state chart, but not on one maturity progression.
Show (System). A service disturbance is a system-side phenomenon. The positioned U.Episteme publication is the alerting episteme published from that disturbance. In Pump-alert-21 (:13.2), its position combines declared observation terms and unit-bearing values, CD1, and local/sparse/symbolic readings for the defined indicator code. The account separately states the operator acknowledgement and diagnostic-check condition.
Show (Episteme). The research note in :13.1 names a discrepancy and two possible causes. Its reader can compare the causes but cannot yet attribute the discrepancy to either one. To assess the internal encoding of the readings, the reader would need evidence about the meters. The typography of the note supplies no such evidence.
Bias-Annotation
The pattern favors separate facet readings over one stage adjective when a decision depends on them. Ordinary descriptions remain sufficient under the non-use boundary in :0.
Conformance Checklist
CC-C.2.2a-1The language-state chart SHALL be a declaredU.CharacteristicSpaceunder A.19, not a disguisedFprogression.CC-C.2.2a-2Published positions SHALL cite explicit facet subject patterns when those positions matter for movement, routing, or endpoint entry.CC-C.2.2a-3Position claims SHALL use slot-explicit values, supported sets of remaining possible values, or intervals admitted by the slot's Scale; uncertainty SHALL NOT be hidden inside stage words such asready,early, ormature.CC-C.2.2a-4A position claim in the chart MUST NOT be conflated with the current ground, witness, publication form (including an MVPK face), or carrier.CC-C.2.2a-5Comparison or reuse that depends on distinct local meanings SHALL recover the two exact F.17 local senses and test the F.9 predicate. Cite a Bridge only when it obtains; keep the bounded-use claim, reliance, and any optional loss note separate. If it does not obtain, keep the local positions distinct and name the missing comparison or translation governor.CC-C.2.2a-6Corridor and navigation notes MUST NOT be read as relocation of facet, seam, bridge, or downstream subject-pattern semantics into the chart subject-pattern set.CC-C.2.2a-7If a position claim is used for routing, endpoint entry, or gate-adjacent reasoning, its relied-on threshold SHALL remain explicit or recoverable from the cited applicable upstream publication. The positioned episteme publication SHALL remain identifiable. Include form, MVPK face or carrier distinctions when they affect interpretation or the next action, as in :4.5.
Common Anti-Patterns and How to Avoid Them
- Maturity monism. Replace the needed distinctions with one stage word. Publish the relevant slot values and state any required source or use condition separately.
- Formality capture. Use
Fto stand in for articulation, closure, or source and use conditions. Repair by naming the facet subject pattern. - Carrier collapse. Treat a document, cue pack, or routed note as if it were the position itself. Repair by distinguishing the position claim, publication form (including an MVPK face) and carrier.
- Threshold folklore. Hide a relied-on facet threshold behind a readiness label. Repair by making its slot conditions explicit or citing the applicable declaration.
- Bridge by vibe. Similar stage language is treated as equivalence. Recover the two exact F.17 local senses and test F.9; cite a Bridge only when its predicate obtains. Otherwise keep both positions local and route the comparison or translation question.
- Corridor inflation. Treat the navigation cluster or corridor map as if it were the subject-pattern set for all downstream semantics. Repair by naming the pattern that governs the current chart, seam, or downstream claim; identify any seam publication form separately.
Consequences
Practitioners can state the language-state readings of a U.Episteme and their grounds without hiding them inside one maturity adjective. Where the receiving use needs a chart position, the publication must make its relevant slots, grounds and any relied-on threshold recoverable.
Rationale
Language-state work needs one explicit statement of what this chart is before individual facet, move, and endpoint patterns start using it. Its declared basis and each facet's rule let readers interpret a position and test any threshold used by the next action.
SoTA-Echoing
The external comparisons below concern publication distinctions and scoped profiles. The chart basis remains grounded in A.19 and its facet patterns.
FPF architecture-description use. Keep the position claim and the expression, selected form and carrier through which it is published distinguishable when their identity changes interpretation or use, under C.2.1, A.19 and E.24.PUB.
Characteristic-space and profile discipline. The multi-facet chart is grounded in FPF A.19 and the Characteristics defined in C.2.3–C.2.5 and C.2.7. The cited NIST profile method supplies the narrower comparison of tailoring an account to a specific use.
Local stance. When the receiving use needs a chart position, use A.19 with the selected facets, any threshold used, and the publication distinctions that affect that use. The ISO and NIST contributions above are bounded comparisons; A.19 and the facet patterns govern this chart.
Relations
- Builds on:
A.19,E.10,F.18. - Coordinates with:
C.2.LS,C.2.3,C.2.4,C.2.5,C.2.6,C.2.7,A.16.0,A.16,F.9,F.9.1,E.17.1. - Constrains: threshold publication, positional claims, and anti-collapse discipline across the language-state cluster.
Worked Examples
Inquiry cue before endpoint capture
An inquiry note says that two meters gave different readings under the same recorded operating conditions. It identifies the meters and their units. It names calibration drift and a sampling-time difference as possible causes, with no observation yet to distinguish them. A reader can recover a diagnostic question and two live explanations. The display uses names and numbers. If the comparison concerns the code inside either meter, its distribution across internal units is still unknown. Use the established readings for the diagnostic inquiry. Entry to a pattern that needs a specific relation or causal conclusion depends on that pattern’s own conditions.
Routed operator alert note
Use Pump-alert-21 from C.2.LS:13.1 to compare the three stated diagnostic alternatives before selecting a check. The alert names Pump P7, its indicated pressure of 70 kPa, the alarm condition p_indicated(P7,t21) < 90 kPa, and the contrasting implications of sensor bias, low supply, and an open bypass. Its chart readings and separate use condition are:
F: the pressure variables and units are defined, the alarm predicate is explicit, and the question and diagnostic contrasts are expressed in prose;AE = AE4: the diagnostic question has the participants, conditions, and bounds needed for this comparison;CD = CD1: the three alternatives have explicit contrasts, but none has yet been excluded and no diagnostic route has been selected;- use condition under C.2.6: the console connects the alert to the operator's acknowledgement and diagnostic-check selection; this is stated beside the coordinates;
RepresentationFactorBundle = {local, sparse, symbolic}for the alert's indicator-code component only: one of eight named units is active, the fixed decoder gives its report label, and the active fraction1/8meets the example's local sparse criterion. C.2.LS:13.1 supplies the unit order, decoder, and criterion.
If an accepted reference-gauge reading of 72 kPa for the same sample t21 excludes sensor bias under the stated case rule while the other two alternatives remain, update CD to CD2. The question remains fully stated and the code, expression properties, and operator-loop connection are unchanged.
Keep the live responses explicit in the route-bearing seam publication. Before publishing an endpoint-subject-qualified work record or reliance record, test the applicable downstream pattern's conditions; these chart readings do not decide that result.
Viewpoint-bound adequacy note
A review note asks whether the architecture description specifies who acknowledges a storage fault. The cited edition names two possible responders and leaves their assignment open. A reviewer can identify the unresolved choice and the evidence still needed. Before assigning a closure level, check whether either responder has been ruled out or chosen and what would allow that choice to be reopened. Compare the note with an observation report on the same chart only for facets assessed on the same basis. Keep the source of each assessment available.
Polished prose is not closure
A prose rewrite may look cleaner, more compact, or more manager-readable than the source cue and still remain low in closure or articulation explicitness. If the underlying slot values, uncertainty, and route plurality remain unchanged, then publication polish changes the rendering or carrier lane, not the chart position by itself.
Position Publication Package Discipline
An account of a chart position identifies the episteme publication, the decision-relevant facet readings or intervals, their grounds, and any local threshold used by the named next action. Add form, face, carrier, freshness, or inherited pins only when the receiving use relies on them.
Minimum self-check:
- Does the account name the positioned episteme publication rather than a ground, form, face, or carrier?
- Does it state every facet that can change the named next action and mark a relevant unknown as unknown? Does it preserve the combinations on which the action depends?
- Are the grounds and any relied-on local threshold recoverable?
- Does the account supply the position facts needed by the receiving question and, if a local threshold is used, say whether the readings meet it, fail it or leave the result undecided, while the receiving pattern governs the action?
- If the use depends on distinct local meanings, was F.9 tested on the two exact local senses rather than inferred from similar wording?
Extended corridor map
After the first position account is understood, the wider Language-State & Semantic Routing Corridor can be read as a distributed overlay over C.2.2a, C.2.LS, C.2.4–C.2.7, A.16, A.16.0–A.16.2, B.4.1, and B.5.2.0.
A.16.1 / PreArticulationCuePack is the earliest durable seam publication form in that corridor. B.4.1 is the explicit route-bearing seam after cue preservation, and B.5.2.0 is typed prompt entry. C.16.Q, A.6.A, A.6.P, B.5.2, A.15, and C.25 are downstream subject patterns, not members of the language-state chart. The map explains navigation only; it relocates none of their semantics into C.2.2a.
C.2.2a:End
Formality of an Episteme
Type: Definitional (D) Status: Stable Normativity: Normative unless marked informative
Plain-name. Formality assessment.
One-line summary. Formality concerns how explicitly and precisely an episteme's content is expressed. Assess the properties needed for its use, including the interpretation of terms, constraints, operations and inferences. These properties form a declared profile when several matter together; they do not supply a universal rank of languages, logics or tools.
Problem frame
Use this when a reader, engineer or researcher must decide whether an expression is precise enough to interpret, compute with, combine or check, or what formalization would make that use possible. A research hypothesis, interface specification, controller model, decision policy and formal proof can all need this judgement, but they need different properties of expression.
The first useful move is to name the expression and the operation it must support. A reviewer may need unambiguous acceptance conditions; an interpreter needs an execution rule; a proof checker needs a statement and derivation in its admitted calculus. This assessment identifies the needed repair without requiring a more elaborate formalism for every task. Use an already sufficient expression directly. Readability alone is a different question; truth, empirical support, organizational approval and permission to act retain their own grounds.
Problem
“More formal” can hide several different changes. A document can acquire a clear structure without fixing its ambiguous terms. A precise specification can be non-executable. An executable program can omit the property its user needs established. A checked proof in one logic need not become stricter when translated into another logic. Ranking these changes on one ladder can reject adequate work and reward an irrelevant change of notation or tool.
Composition adds a further difficulty. Several precisely expressed claims can be connected by an informal inference. Their presence does not make that whole argument a formal derivation. Conversely, an informal observation used as empirical evidence does not make the separately stated mathematical claim less precisely expressed. Without an explicit assessed expression and support relation, “the least formal part” can refer to the wrong object.
The practical problem is therefore to establish which expression properties the receiving operation needs, which are present, which must change and how those properties survive combination or translation.
Forces
Solution - Assess the required properties of expression
Formality is assessed on a named expression of an episteme's content. Establish what its next use requires, recover the relevant interpretation and inferential relations, and inspect the expression against those requirements. Keep independent properties distinct. A shared profile makes their combination visible; a local summary is useful only when its declared criteria preserve the distinctions the receiving use needs.
Subject and representation of the assessment
F names the formality assessment in an account that also uses G and R. It is not a globally ordered atomic U.Characteristic, and this pattern supplies no universal U.Formality value or F0...F9 scale. The assessed subject is an exact claim, declaration, model or argument expressed by an episteme. State which parts and connecting inferences belong to that assessment.
An ordinary assessment can be a sentence: “The budget condition has declared units, a typed variable and an explicit predicate; no derivation of its adequacy for the project is supplied.” That is enough when it answers the receiving question. For comparison, automation or repeated use, describe the selected expression properties and their values as a profile. The description remains an episteme; a reference such as formalityRef points to that assessment and its basis.
If a CharacteristicSpace is used, each assessed Characteristic has its own declared Scale and slot under A.18 and A.19. A profile is not one Coordinate containing several unnamed characteristics. Record missing or unknown assessments separately from values; unknown precision is not evidence of low precision. Keep the relevant calculus, assumptions and interpretation recoverable alongside the property assessments.
Place beside G and R
A named assurance argument may consume F alongside G or R under B.3. The notation [F,G,R] is a view of distinct assessments, not automatically a point in one numeric space.
- F concerns expression properties. G concerns applicability or claim scope.
- A derivation expressed in a calculus is part of the assessed content. Evidence that a particular checker accepted it supports that checking claim. Confidence in the checker, its assumptions and the resulting conclusion belongs to the relevant assurance argument.
- A fully expressed empirical claim can have weak or missing empirical support. Formal validity under stated premises and applicability to the world answer different questions.
- Mapping loss is assessed for the receiving use. Any effect on R needs its B.3/C.2.2 basis; an ordinal CL label supplies no automatic penalty.
- A notation, carrier or rendering change preserves the assessment only when it preserves the expression properties on which that assessment rests.
Comparison and local ordered summaries
Compare profiles only after aligning the assessed properties, their meanings and relevant conditions. One expression can provide a formal predicate while another provides an executable procedure; neither fact alone makes one more formal for every use. A changed use may need both, or may need only one.
When an order is useful, declare its domain and comparison rule. For a fixed compatible basis of independently established requirements, an expression that satisfies all requirements satisfied by another and some additional ones can dominate it under that declared rule. This does not rank the entire languages or establish a preference for every use. Preserve incomparability when different requirements are satisfied; unresolved assessment is a separate result. Use A.19:5.1.3 for an order overlay when a CharacteristicSpace is involved.
A local ordinal summary is possible when its levels have nested criteria. For example, one team's specification-expression scale can require:
- named objects and stable terms;
- those conditions plus explicit acceptance predicates and their interpretation;
- those conditions plus expressed derivations for a specified set of obligations.
The last level includes the earlier criteria by declaration. Its proof-checking evidence is still assessed separately. State the exact scale, mapping from the profile, domain and conditions before using its levels or a threshold. Another team’s levels are comparable only to the extent that their criteria support the same comparison. When the expressions are used as models, B.5.MR asks a different question. For a fixed inquiry and interpretation, which answers does each model admit? Rewriting a ceiling of thirty minutes as an equivalent predicate can make it executable while admitting exactly the same durations. Lowering the ceiling to twenty minutes can exclude a previously admissible duration in the same notation.
Arithmetic, averages and numerical differences are not licensed by an ordinal scale. Where a change is described as an increase in formality, identify the strengthened property or the established order that makes that claim meaningful. A new logic, extra subsystem or larger toolchain is not such a property by itself.
Requirements for a receiving use
Express a formality requirement through what the recipient needs to do. “The invariant has explicit types and semantics, and every inference needed for this proof is expressed in the selected calculus” is a recoverable condition. If checked validity is also required, name that separate evidence requirement. A local ordinal threshold can abbreviate the expression condition only when its named scale captures the required criteria.
Normative content must make the expression conditions needed for its intended use recoverable. A suitable ordinary explanation can do this without a separate score, profile document or formal proof. A richer assessment is required when a comparison, assurance argument or automated consumer relies on it. Neither a satisfactory formality assessment nor a threshold result grants permission to act.
Improving an expression can change its assessed properties while preserving the episteme's claim content. Correcting an erroneous assessment can leave the expression unchanged. When the receiving requirement changes, return to the property, formal apparatus, derivation or external support it needs; do not automatically raise every part of the expression.
Assess and improve an expression
- Identify the exact claim or argument and the next operation: interpretation, execution, combination or checking. Name the required result and any applicable binding requirement.
- Recover the terms, domains, variable bindings, constraints, operation meanings and inferential connections needed for that operation. Section 13 supplies common distinctions, not a mandatory assessment form.
- Inspect the expression. State which needed properties it provides and where interpretation, an operation or an inference remains unspecified. Keep a proposed derivation separate from the result of checking it.
- Test the receiving condition. A sufficient expression ends this inquiry. If a missing property prevents the use, repair that part or obtain a suitable expression. Use C.19.2 to choose and configure formal apparatus when that is the question; use C.39 when a way to obtain the needed result still has to be developed.
- After the change, repeat the affected use, including any connecting inference. Preserve unchanged scope and support assessments unless their own grounds changed. If the repaired expression reveals a different claim or assumption, assess that change explicitly rather than calling it a purely notational improvement.
Archetypal Grounding
Tell. Ask what the expression enables its recipient to interpret or do, and which property enables it.
Show (System). A controller requirement in controlled language can state a clear bound. Typed invariants make its variables and condition explicit. An executable model enables simulation. A formal derivation can establish an invariant under declared plant assumptions. These contributions can be connected, but execution is not a required intermediate stage for every proof, and a proof does not establish the plant assumptions empirically.
Show (Episteme). A research conjecture can gain stable terminology, explicit predicates and an expressed argument. Each change removes a particular obstacle to examination. A mathematical conjecture may remain non-executable and unproved while being fully precise about what is conjectured.
Bias-Annotation
The pattern favors recoverable expression properties over a universal prestige ranking. This makes a comparison more explicit but can be unnecessary for an already adequate ordinary use. Keep the assessment as small as that use permits. Do not replace one compulsory score with a compulsory inventory of every possible property.
Conformance Checklist
CC-F-1An assessment relied on in use SHALL identify the assessed expression, selected properties and their grounds. Normative content SHALL make its necessary expression conditions recoverable; a universal F value is not required.CC-F-2A local ordinal summary SHALL declare its scale and mapping. Arithmetic over its levels is invalid unless a separately justified quantitative scale is established.CC-F-3A claim of greater formality SHALL name the strengthened property or declared order. It SHALL NOT imply greater truth, trust or scope by itself.CC-F-4Status, maturity or tool labels MUST NOT substitute for the assessment of expression properties.CC-F-5A subdivision of a local formality scale SHALL retain the criteria for the level it subdivides. Comparing levels on different scales requires a basis that preserves the meaning of the properties being compared.CC-F-6A composite assessment SHALL cover the essential content and inferences of the named whole. It SHALL NOT inherit a property from constituents without the composition preserving that property.CC-F-7Implementations MUST NOT average unrelated formality properties or local ranks into an undeclared score.CC-F-8Changes in G, R or CL SHALL NOT change an expression assessment unless its own grounds change.CC-F-9Transport SHALL preserve a formality assessment only when its relied-on properties and interpretation are preserved. A receiving rewrite that changes claim content, EntityOfConcern or the effective reference scheme has its own episteme identity under C.2.1.CC-F-10Translation and bridge limitations SHALL remain explicit for the receiving use. Assess assurance effects under their governing models.CC-F-11Assigned properties SHALL be supported by the terms, predicates, semantics, derivations or other assessed content. A claimed checking result needs its own evidence.CC-F-12A named notation, prover, calculus or execution harness SHALL NOT by itself establish the required expression property or a successful check.CC-F-13Publication and approval status MUST NOT substitute for formality.CC-F-14A policy consuming formality SHALL state the required properties, or a justified threshold on a named local scale, and keep separate assurance and authority conditions explicit.CC-F-15Language-state facets MUST NOT be hidden inside a formality rank. Each Characteristic used in a space retains its own Scale and slot.
Common Anti-Patterns and How to Avoid Them
Consequences
Rationale
Formality matters because an expression can make interpretations, operations and inferential commitments available for shared examination. The properties that enable these uses are related but not identical. A formal specification can define what is required without giving a procedure; a procedure can run without expressing a proof of the required property. A richer logic can express additional distinctions without improving an already sufficient statement of an ordinary invariant.
A profile preserves these relations without pretending that every pair has one overall order. Declared local chains remain economical when their criteria are nested. Comparison then rests on the receiving operation and preserved meaning. Composition follows the same reasoning: a whole must supply the connections its result depends on, while evidence about the world retains its separate role.
SoTA-Echoing
These sources support distinctions among capabilities and formal bases. They do not establish a total order from prose through execution to successive schools of logic. Apply the assessment method above to the expression and use at hand; consult the sources when their particular language or apparatus is in question.
Relations
- Uses: A.18 and A.19 for declared Characteristics, Scales, profiles represented in spaces and any order overlay.
- Coordinates with: C.2 and C.2.2 for epistemic composition and assurance; B.3 when an assurance argument consumes the assessment.
- Coordinates with: C.2.LS, A.16 and C.2.4–C.2.7 for language-state properties that retain their own meanings.
- Uses: C.19.2 when selecting or configuring a formal apparatus is the missing practical contribution; C.39 when a way to obtain the result must be developed.
- Coordinates with: F.9 for a cross-context transfer whose mapping and limitations affect the receiving expression.
Distinctions for describing expression
The following descriptions identify common contributions. They are neither mutually exclusive classes nor consecutive levels. Select only distinctions needed by the receiving operation.
Unstructured prose
Free natural language may leave vocabulary, assumptions and connections unstable. It can still preserve an observation or start an inquiry. Identify the ambiguity that obstructs the next use before requiring a more elaborate form.
Scoped notes
A note can identify its topic and use terms consistently while leaving criteria informal. Naming the subject helps interpretation; it does not by itself formalize a predicate or narrow claim scope. Assess scope separately.
Structured outline
A coherent section structure helps readers find claims and their grounds. Placeholder criteria can remain inside a complete outline. Editorial organization is useful, but inspect what the filled content defines.
Controlled narrative
Constrained prose can stabilize interpretation and state acceptance or refusal conditions. This may be sufficient for a competent reader's decision. Automatic interpretation needs an additional mapping or semantics when the prose itself does not supply one.
Explicit predicates and constraints
Typed entities, bound variables, units where applicable, domains and an interpretation make a predicate recoverable. The claim's consistency or consequence can then be examined under the declared semantics. A written formula with undefined symbols does not supply this contribution; an explicit predicate need not be executable or proved.
Executable semantics
An expression can define how an interpreter or procedure produces results. Declare the relevant input meanings, operation and output interpretation. This enables a run under those conditions; it does not establish a separate invariant or substitute for its specification. A non-executable formal statement can remain sufficient for a mathematical argument.
Coupled formalisms
A hybrid controller model can relate discrete transitions to continuous evolution; another account can connect several formal subsystems. State their shared meanings and interconnection obligations. Additional formalisms introduce relations to establish, not an automatic increase in rigor. Assess whether the whole preserves the properties its intended use needs.
Formally expressed derivation
A derivation states how its conclusion follows from declared premises under an identified calculus. Inspect the essential inferential steps, including their connection to the stated conclusion. A checked derivation in higher-order logic can satisfy this need without translation into dependent type theory.
Properties expressed in types and proof terms
A dependent type can express a property of a program or mathematical object; a term can carry a derivation of that property under the calculus. Recover the exact proposition, assumptions and term. Distinguish that expression from evidence that the required checker accepted it. Its usefulness may include preventing a particular invalid construction or connecting a program to its specification.
Higher equalities and univalent foundations
When a question needs relationships among equalities or a particular treatment of equivalence, name those structures and the foundation used to express them. This can change the available construction or proof. It does not strengthen every unrelated invariant or make simpler formal foundations inadequate.
Keeping the distinctions together
A useful assessment explains which of these contributions cooperate. A controlled requirement can be related to a formal invariant and an executable model, with a derivation connecting the model to the invariant. The links carry claims of their own. Check the required link when its meaning or assumptions change; do not infer the whole connection from the presence of all three documents.
Assessing formality in practice
First-pass questions
- Which claim, declaration or argument must be understood, executed, combined or checked?
- Are its terms, domains and interpretation sufficient for that operation?
- Are the required predicates, operations and inference steps expressed?
- Which calculus and assumptions matter, and what checking result is available when required?
- Would changing the expression property change the receiving result, or would it only add unused apparatus?
Short assessment
A short assessment names the subject, required properties, observed basis and consequential gap. For example: “The interface condition gives typed inputs and outputs and an explicit postcondition. The procedure has executable semantics. The preservation of this postcondition across the retry branch is not expressed.” The next useful change is the missing retry argument, not the acquisition of a more elaborate specification language.
For repeated comparisons, use the same property meanings and conditions for every candidate. A compact profile can record them. Add a local rank only when its declared mapping preserves the distinctions used by the comparison.
Typical improvements
- Replace unstable terms with explicit definitions and consistent references when interpretation is blocked.
- State a necessary acceptance condition as a predicate with its types, applicable units and meaning when it must be evaluated or reasoned about.
- Supply operational semantics when the result must be computed.
- State the relations and obligations joining formal subsystems when the whole must be analysed.
- Express the missing derivation when the result depends on a conclusion from premises, then obtain the checking evidence required for that use.
These changes can occur independently or together. Describe the property changed; use a local rank difference only when the named scale gives it meaning.
Composition and interaction
Essential-content and connection rule
Identify the assessed episteme's expressed claim or argument and its compositional support relation. A property of the whole requires the relevant properties of its essential content and of the inferences that connect it. A machine-checked annex does not make an informal connecting inference machine-checked. Listing common constituent properties is also insufficient when the composition can fail to preserve them.
External evidence has a different role. A typed claim supported by an informal observation can retain its explicit predicate and interpretation while its empirical warrant remains weak or unresolved under B.3. If the assessed subject is instead the whole argument, include its essential observation-to-conclusion inference in that assessment. Do not take a minimum over every external source in an evidence chain.
A minimum can abbreviate an upper bound on a local ordered scale only when all assessed parts use that scale and the composition rule establishes the bound, including the connection. A minimum alone does not prove that the whole meets that level; additional joining obligations may remain. Without that basis, inspect the needed properties directly.
Relation to G
Formality concerns expression; G concerns applicability or claim scope. Making an assumption explicit can reveal that the claim applies less widely than expected. Record that scope correction separately from the improvement in expression. A precise claim can still have broad scope.
Relation to R
An explicit expression can make a claim easier to test or its derivation easier to check. It does not create empirical support. A checked formal argument establishes its conclusion under the admitted premises and calculus; applying that conclusion to the world additionally needs the relevant interpretation and evidence. Assess freshness, corroboration and mapping limitations under their assurance models.
Relation to CL and Bridges
A Bridge may expose loss or mismatch across contexts. Preserve that limitation for the receiving use. Reuse a formality assessment only if the properties and interpretation on which it relies survive the transfer. A receiving rewrite that changes claim content, EntityOfConcern or the effective reference scheme identifies a new episteme under C.2.1; assess the expression of that episteme. A changed carrier alone does not require a new assessment.
Worked examples
Research hypothesis
A short note proposes a scaling law with stable terms and acceptance conditions in prose. That can be enough to decide which measurement to attempt. Writing the relation with explicit variables, units and parameter conditions enables a quantitative comparison. The expression becomes more determinate in those respects; the fit to observations remains a separate question. A poor fit is not repaired by making the same formula typographically more formal.
Interface specification
An interface specification declares preconditions, postconditions and invariants with their types and interpretation. It can be used to compare designs before an implementation exists. A faithful executable reference model adds the ability to run scenarios. A running implementation that omits the postcondition does not dominate that specification: the two expressions supply different contributions. To claim conformance, connect the implementation behavior to the required postcondition and supply the needed argument or test.
Safety controller
A controller and plant model express discrete transitions, continuous evolution and the obligations connecting them. A proof of an invariant in higher-order logic can satisfy a requirement for an explicitly derived invariant under those assumptions. A checked dependent-type development can satisfy the same requirement. Choose between adequate options by compatibility, available proofs and application cost through C.19.2; there is no general requirement to replace HOL with a supposedly higher foundation. If the plant-assumption evidence is weak, retain that limitation even when the derivation is fully checked.
Decision policy
A policy says in controlled prose that a request exceeding a thirty-minute planned-duration ceiling fails the budget condition. To evaluate that condition automatically, introduce a Plan variable p, define plannedMinutes(p) in nonnegative real minutes and state meetsBudget(p) iff plannedMinutes(p) <= 30. The expression now gives the needed domain, unit and predicate. Its evaluation can answer the budget question; it supplies neither the other eligibility conditions nor permission to proceed.
If “duration” later means elapsed duration rather than planned duration, retaining the formula's printed shape is insufficient. The input interpretation and therefore the condition have changed. Return to the affected policy meaning and uses.
Proof-bearing algorithm
An algorithm's dependent type states a length-preservation property, and its term expresses a derivation under the chosen type theory. Inspect the property and term, then recover the required checking result. A separate HOL proof of the same stated property may be equally sufficient for the recipient. If a new task instead requires reasoning about higher equalities or a specific univalent construction, it changes the apparatus requirement. That new requirement does not retrospectively make the earlier sufficient derivation less rigorous.
Executable ML recipe
A training-and-evaluation recipe declares the inputs, parameters, operations and output interpretation needed to repeat a run. This supplies executable meaning. It does not express a proof of generalization merely because the harness is sophisticated. If the receiving use needs an empirical performance estimate, obtain the corresponding observations and inference. If it needs a formal property of the procedure, state and derive that property under a suitable model. The two continuations repair different gaps.
Guidance for developing and examining expressions
For authors
Make the expression sufficient for its intended use. State a remaining ambiguity, unspecified operation or missing derivation when it changes what a reader can do. An early informal expression can be the appropriate result. Improve the property that blocks the next use, and preserve its connection to the original question.
For reviewers
Read the assessed claim or argument and its expression. Check its terms, predicates, semantics and required inferences, including essential connections. Recover the basis for any claimed checking result. Keep missing expression, missing proof, weak empirical support and inadequate readability distinct, because they call for different repairs. A complete template or named tool answers none of those questions alone.
For integrators and assurance leads
State the expression properties a receiving operation consumes. For a composition or transfer, establish their preservation through its connections. Reuse a declared local scale when it captures those needs; retain the underlying meaning when several participants use different scales. Apply separate scope, evidence, authority and language-state conditions where they matter.
Glossary and notation
- Formality assessment (F). An assessment of selected properties of an episteme's expression, with its subject, interpretation and basis recoverable.
- Formality profile. A structured account of those selected properties; when represented in a CharacteristicSpace, each Characteristic has its own declared slot.
- Local ordered summary. An explicit mapping from such assessments to a named ordinal scale whose levels have justified nested criteria.
- Change in formality. A change in an assessed property or in a justified order, not a numerical difference between unrelated labels.
- Essential content and inference. The content and connecting reasoning required by the assessed whole under its named composition relation; external warrant is assessed separately.
- Example notation.
F: explicit typed invariant; derivation not supplied, or a threshold on a declared local specification-expression scale.
Reusing an earlier assessment
Recover the assessed meaning
An earlier label such as F4 or F8 is a locator for an assessment, not enough information to use it. Recover the expression, criteria and basis. Preserve useful properties already established, such as explicit predicates, executable semantics or a checked derivation. Do not infer a new profile or admissibility result from the number alone.
Reuse a sufficient result
When the same expression properties and conditions still answer the receiving question, reuse the assessment. If an earlier requirement was written only as a rank, recover what operation and guarantee it was meant to protect before replacing it with explicit criteria or a justified local threshold. A binding requirement is changed through its governing authority; a revised vocabulary does not silently waive it. Reassess only the properties or dependent conclusions whose basis has changed.
Boundary to language-state facets
C.2.4 governs articulation explicitness, C.2.5 governs closure degree, C.2.6 recovers source and use claims, and C.2.7 governs the representation-factor readings. C.2.LS brings selected results together in an optional profile. A formality assessment can refer to these results when relevant, but must not relabel them as steps of one rigor scale.
C.2.3:End
Language-State Facet Profile: Publish Readings and Use Conditions Together
Type: Definitional (D) Status: Stable Normativity: Normative unless marked informative
Plain-name. Language-state facet profile.
Technical alias. LanguageStateFacetProfile names the optional record form below. A filled profile is an assessment account governed by C.2.1.
Use this pattern when. Use C.2.LS when the next use of a published U.Episteme needs selected language-state readings to be interpreted, compared or reused together. Select the required formality, articulation, closure and representation readings, together with any source or use conditions needed to interpret or apply them.
Sufficient result and non-use. The profile can point to existing readings and include only the facets needed together. If ordinary wording, one direct facet reading or an applicable existing profile already serves the receiving use, use it without constructing another bundle.
What goes wrong if missed. Teams replace the facet profile with a maturity adjective such as ready, raw, or stable, then use that label to choose routes, decide whether to reopen, interpret Bridges, or make publication decisions. The label hides the facet values.
What this buys. A decomposable profile bundle: each facet keeps the definition and tests supplied by its own pattern, while the profile gives authors, assurance readers, and integrators one place to publish selected readings and the conditions on their joint use.
Problem frame
When several language-state readings are needed together, collect their references, assessment scopes, grounds and use conditions in one recoverable account. This makes their separate contributions to the receiving judgement inspectable. If that judgement needs a language-state chart position, its declared U.CharacteristicSpace supplies the basis.
Problem
When authors need several facet readings together, a vague maturity label can conceal their different contributions, or F can become their surrogate. That blocks admissible threshold publication, undercuts A.16 move guards, and makes school-to-school bridge work harder than it needs to be.
Forces
Solution
A language-state facet profile records selected readings and use conditions about one identified episteme edition. Use C.2.2a:0 and :4.2–4.5 to select the needed facets, obtain readings and grounds, and identify any conditions used by the receiving action. The profile brings those results together and can refer to existing readings or to the source and use account supplied by C.2.6. Each Characteristic reading retains its meaning, Scale, value, assessment scope and grounds. A source or use claim retains its predicate, participants and grounds. Use a declared language-state chart over U.CharacteristicSpace when the receiving use needs a position or an operation in that chart. The optional fields below can refer to the readings and accounts that supply these contributions:
formalityRef-> the expression-property assessment governed byC.2.3, including its selected basis and any justified local ordered summaryarticulationExplicitnessRef-> the articulation reading governed byC.2.4languageStateClosureDegreeRef-> the closure reading governed byC.2.5languageStateAnchoringModeRef-> the source, evidence, access or action-condition claims selected underC.2.6; cite their account or an explicitly defined local summary with its criterionlanguageStateRepresentationFactorBundleRef-> the selected representation-factor readings governed byC.2.7thresholdRefs?-> context-local threshold declarations over the named facetsrouteNotes?-> informative notes that help interpret routing or reopening decisions
C.2.LS therefore defines only the profile bundle; it defines neither an individual characteristic nor a trajectory. A formality assessment or representation-factor bundle can contain several Characteristics. If placed in a CharacteristicSpace, each of those Characteristics receives its own declared Scale and slot; the reference to the assessment is not itself a Coordinate. The C.2.6 account is likewise separate from those slots; a local summary needs its own declared Characteristic and Scale before it can supply a Coordinate. A.18/A.19 supply characteristic semantics, A.16 defines admissible moves, and E.18 governs an independently selected TransformationFlowStructure.
Kind and profile-bundle boundary
The optional LanguageStateFacetProfile form expresses an ordinary C.2.1 assessment account. The episteme edition being assessed is its EntityOfConcern. Its claims give the selected readings and their grounds within the assessment scope, using a declared chart when the receiving use needs one, together with any required source or use conditions and any thresholds used in the assessment. C.2.1 governs the account's identity; E.24.PUB governs its expression and publication. No additional profile individual or dependent U-kind is needed for this use.
Distinguish the selection of facets from the filled assessment. A selection identifies the Characteristics and Scales needed by the receiving use and any separate C.2.6 questions. If a chart is needed, bind those Characteristics and Scales in its A.19 basis. The same selection can be used for several assessments, just as the same record form can express them.
A filled assessment states readings about its identified episteme edition and scope; changing a reading or a stated relation changes that claim content even if every definition reference stays the same.
An A.19 CharacteristicSpacePredicate has its own by-value identity under A.19:5.1.8. Reusing readings with a different predicate requires a new evaluation of that predicate, while the readings can remain unchanged.
For reuse, recover the assessed edition, facet meanings, any chart basis needed by the use, scope, readings and relied-on grounds from the profile or its applicable cited source. Recover a threshold or a separate source or action condition and its receiving use when the conclusion relies on it. Saying which of the selection, readings, threshold or whole assessment is being reused prevents an unchanged label from concealing changed content. Rearranging the same claims in a dashboard or table leaves their meanings available under the same account-identity rule.
Choose direct readings or a shared profile.
Use direct facet readings when they already provide the inputs for the receiving judgement. A shared profile is useful when several readings and their conditions must be interpreted, compared or reused together: recover their joint scope and how each contributes to that judgement.
When readings concern different expressions or components, keep that distinction visible. Identify which readings or relation claims each receiving condition uses, including any local threshold, and report that condition as met, not met or unresolved as its grounds warrant. This saves reassembling dispersed references. The profile adds no stronger warrant to the readings. Keep it partial when the task needs only part of the facet family, and reuse an existing applicable assessment rather than writing another.
Contribution boundary
C.2.LS defines only profile composition and requires the language-state facets to remain explicit and non-collapsed. It does not:
- redefine
F; - invent a universal formality progression or an undeclared local summary;
- redefine the scale semantics of
AE,CDor the representation factors inC.2.7, or replace the direct predicates recovered through C.2.6; - define reopen/backoff moves;
- define endpoint classification or bridge kinds.
Threshold publication discipline
When the receiving use relies on a threshold, make its condition and input readings recoverable from the profile or an applicable cited source. Do not describe hidden sub-levels of F when the issue is articulation, closure, a source or use condition, or the representation-factor bundle. Test a required relation under its direct predicate. An A.19 CharacteristicSpacePredicate retains its declared space, Characteristic and Scale bindings, domain and test under A.19:5.1.8; only declared Characteristic readings supply its slot values.
Local profile-reading witness
For this pattern, a published facet profile is reviewable when:
- the assessed episteme edition, any chart used, scope, selected refs, readings and any relied-on source or use claims are explicit or recoverable from a cited applicable upstream publication;
- any threshold-bearing use identifies the threshold and the readings it consumes;
- route notes or local overlays remain informative and visibly docked to the explicit facet bundle;
- and any claimed move, Bridge, gate state or downstream result retains its own governing definition and test.
A polished label, one strong facet, or one memorable route note does not by itself yield an admissible profile reading. The profile remains conformant only when the named facets stay explicit and decomposable.
Composite readings
For a diagnostic note, ask what the wording lets a reader identify and which explanations the observations still permit. A note can name two sensors, their readings and the time window precisely while leaving both a calibration fault and a sampling delay open. Cite the trace from which the readings were derived. If the decision also depends on how the sensor encodes them, inspect that encoding. A table listing the two readings can supply the user with symbolic labels while leaving the internal code unknown.
Corridor map note
For a map of the related language-state patterns, see C.2.2a:15.
Archetypal Grounding
A team may say a draft is "still forming" for different reasons. The profile lets the team identify whether the issue is low articulation, low candidate-space closure, an unavailable observation or use condition, or an unresolved representation-factor bundle.
Operator alert. An operator alert note can have stable alert terms and structure and low closure, while participating in an operator acknowledgement-and-check procedure.
Inquiry note. A station note records the service identifier, date and actual arrival time for each of three trains at the same station. The reader can identify all three events. To compare their delays, obtain each scheduled arrival from the applicable timetable and compare it with the matching actual arrival on the same time basis.
Bias-Annotation
Making selected facet readings explicit costs more than using one stage adjective. It makes their contribution to the receiving judgement inspectable, including any threshold that judgement uses; retain the ordinary wording or existing reading when that already serves the use.
Conformance Checklist
CC-C.2.LS-1A language-state facet profile SHALL reference the patterns that define its facets rather than invent local unnamed factors.CC-C.2.LS-2C.2.LSMUST NOT redefine the C.2.3 assessment or conceal several expression properties in one undeclared rank.CC-C.2.LS-3A threshold relied on for routing, reopening, or lexical repair SHALL be stated or cited with the readings it consumes. Required relation conditions retain their direct predicates.CC-C.2.LS-4Trajectory accounts that rely on facet profiles SHOULD reuse A.16 move kinds. Apply E.18 only to an independently selected TransformationFlowStructure; use E.18.2 when its mathematical description is needed.CC-C.2.LS-5Composite labels such asearly,settled, orreadySHALL NOT stand in for the explicit facet bundle when those states matter operationally.CC-C.2.LS-6Composite readings, overlays, and route notes SHALL remain decomposable into named facets and MUST NOT behave as hidden master factors.CC-C.2.LS-7Profile references or notes MUST NOT replace the governing move, Bridge, gate or downstream rule. Cite the applicable rule when a proposed consequence depends on it.
Common Anti-Patterns and How to Avoid Them
- Shadow progression. Treating
early/lateas a master scale. Split the judgement into the named facets. - Formality capture. Letting
Fstand in for closure or articulation. Publish those facets explicitly. - Bundle inflation. Turning the profile into a second
A.19. Keep it thin and referential. - Opaque readiness. Using words such as
readyormaturewithout naming which facet justifies the claim. - Route-note capture. Letting an informative route note act as a move rule, gate state, or endpoint rule. Keep route notes informative. Use
A.16for admissible moves, the applicable pattern for a downstream definition or test, the applicable gate or Work pattern for those claims, and anauthoritySourceRefonly when an external authority supplies the rule.
Consequences
The benefit is clearer source and rule references: early cue work, bridge annotations, and reopen moves can all refer to one explicit facet profile. The trade-off is making the selected readings and their use conditions recoverable, including any threshold used.
Rationale
The profile brings together the references needed to interpret facet readings together and test any local threshold used by the receiving judgement. Each referenced pattern supplies the corresponding definition and test; the profile preserves their different contributions to the receiving judgement.
SoTA-Echoing
Source boundary. The facet meanings and profile rules come from A.18/A.19 and the neighboring FPF patterns. The external comparisons below contribute scoped conditions and publication distinctions.
Local stance. The useful result is a decomposable profile of the readings and conditions needed for the receiving use.
Relations
- Builds on:
A.18,A.19,C.2.1,C.2.2a,C.2.3,E.24.PUB. - Coordinates with:
C.2.4,C.2.5,C.2.6,C.2.7,A.16.0,A.16,A.16.1,A.16.2,B.4.1,B.5.2.0,E.18,F.9for any Bridge and bounded-use claim, andF.9.1only for an optional stance note about that claim. - Constrains: language-state threshold publication and profile composition.
Worked Examples and Composition Notes
Operator-facing early alert
In this constructed case, the console publishes Pump-alert-21 for an operator comparing diagnostic alternatives before selecting a check:
At sample t21, Pump P7 indicates 70 kPa, below its alarm threshold of 90 kPa:
p_indicated(P7,t21) < 90 kPa. Which of sensor bias, low supply, or an open bypass explains this indication? Keep these three alternatives live until an observation distinguishes them.
For this comparison, the three candidates have these stated implications:
Initially no reference-gauge reading, upstream-pressure reading, or bypass-state observation is available. The alert makes the contrasts explicit without excluding a candidate or selecting a diagnostic route. Its receiving use is comparison within this stated candidate field. The console connects the alert to the operator's acknowledgement and selection of the next diagnostic check.
The alert's code component uses eight indicator units in this order: NORMAL, LOW_PRESSURE, HIGH_PRESSURE, LOW_FLOW, HIGH_TEMPERATURE, VIBRATION, SENSOR_FAULT, LINK_LOSS. Exactly one unit is active at a time. The fixed decoder assigns the active unit its corresponding report label; LOW_PRESSURE denotes the alarm predicate above. Thus [0,1,0,0,0,0,0,0] denotes this low-pressure alert. For this code comparison, the local criterion calls an active-unit fraction at most 1/4 sparse and a greater fraction dense.
The profile states:
- its formality assessment covers the displayed alert condition and question: the pressure variables and units are defined, the threshold predicate is explicit, and the question and diagnostic contrasts are expressed in prose;
AE = AE4for the initial diagnostic question, because its subject, indicated condition, alternatives, contrasting implications, and bounded receiving use give a complete branch-appropriate expression (C.2.4:4.2);CD = CD1, because the contrasts weakly stabilize the stated candidate field while all three alternatives remain live and no route is selected (C.2.5:4.2);- the C.2.6 account states that the console connects the alert to the operator's acknowledgement and diagnostic-check selection; this action condition is separate from the chart readings;
RepresentationFactorBundle = {local, sparse, symbolic}for the indicator-code component: one unit identifies the report, the active fraction1/8satisfies the stated local criterion, and the named units have an explicit decoding rule (C.2.7:4.1). These readings concern the code; assess the surrounding prose separately if its representation factors matter.
Changed condition. Suppose an accepted reference-gauge reading for the same sample t21 becomes available: 72 kPa. Under the stated case implications it excludes sensor bias, while low supply and an open bypass remain live. The meaningful reduction from three candidates to two supports CD2; it selects neither remaining route. The updated diagnostic question still has its complete expression, and the pressure vocabulary, alarm predicate, indicator code, and operator-loop connection are unchanged. Apply the receiving pattern's own conditions when choosing the next question, check, or endpoint use.
The alert has a complete question and still-open diagnostic alternatives. Its code factors and operator-loop connection answer different questions; no one facet replaces the others.
Research cue before lexical repair
An observer reports a vibration just after startup. The note identifies the machine, the observer and the occasion. It leaves open what changed and which part vibrated. The next investigator can locate the observation and seek a comparison or a trace. If a later summary proposes a cause, identify which evidence supports that claim. A.16.1 can preserve the initial cue. Use A.6.P for an unclear relation or B.5.2 when the material supplies the conditions for an abductive prompt.
Architecture-description case
A note evaluates whether a named edition of an architecture description assigns a responder to each failure mode. For one mode the description names two possible responders and gives no rule for choosing between them. The note can express this finding precisely. The assignment remains open in the description. To determine who responded in service, inspect the record of that incident.
Same F, different profile
Two notes may supply the same required expression properties and still differ sharply in articulation, closure, source and use conditions, and representation factors. One may be used in an operator's check procedure with diagnostic alternatives still open; another may assess a document's assertions with comparatively closed alternatives. The profile bundle keeps that difference visible instead of letting F behave like a master factor.
Reuse the selection while readings or thresholds change
In :13.1, the facet definitions and record form remain available after the reference-gauge reading arrives. The closure assessment changes from CD1 to CD2; retain the assessed alert edition and the new grounds in the updated assessment. This is reuse of the facet selection with changed claim content, not reuse of the earlier closure reading.
Now consider two stipulated receiving rules for that same assessment. A diagnostic-comparison meeting requires an explicit question and contrasting alternatives; it can use the available AE4 reading. An action-selection rule additionally requires a selected diagnostic route. CD2 still leaves two alternatives live, so that rule's condition is not met.
The same readings support these different conclusions because the receiving conditions differ. If both assessments publish their threshold conditions, their full contents differ; the common readings may be cited once. Neither rule changes what AE4 or CD2 means, and each applies only to the stated receiving use.
Authoring and Review Guidance
For authors
When publishing a language-state facet profile:
- identify the decision or operation for which the recipient will use the profile;
- choose the facet readings and any conditions that could change that result, and cite the assessments that supply them;
- add threshold refs only when a threshold changes routing, repair, or another operative decision;
- when a label such as "mature", "raw", or "ready" carries an operational claim, make the relevant facet readings and conditions visible.
For assurance readers
Use :18.1 to check the profile's meanings, relied-on conditions and claimed consequences.
For integrators
Integrators should preserve the profile references. A local alias is acceptable when the underlying facet references remain recoverable and unambiguous.
Extension and Migration Notes
Local extension rule
Contexts may extend the profile with local threshold refs, route notes, or additional descriptive aids, but they shall not add a new master facet that collapses the named facet set into one summary factor.
Migration from surrogate prose
When a stage label carries a claim used in routing, comparison or another decision, recover the readings and conditions it abbreviates. Examples include:
- "the episteme is still early",
- "the issue is not mature enough",
- "the note is ready",
- "the cue is still raw".
For that use, state the relevant readings or relation claims and their grounds, whether any relied-on threshold is met, and the governing rule for the resulting move. An ordinary description with no such operational claim needs no profile expansion.
Boundary reminder
Use the profile to refer to facet readings and applicable thresholds. Definitions and tests for moves, Bridges, Scales and bundles remain with their governing patterns.
Use a facet selection in an assessment
A record containing only references to facet definitions supplies a selection of Characteristics and any separate source or use questions to be answered. To use it for a position assessment, add or cite the required readings and relation claims, identify the episteme edition being assessed and state the assessment scope. Retain the definition references so the meaning of each reading remains recoverable.
Profile Publication Package Discipline
Minimal publishable profile package
A publishable LanguageStateFacetProfile carries what its receiving use needs:
- the selected refs from :4, with the relevant readings or relation claims, stated here or cited;
- any threshold refs that substantively affect routing, repair, bridge interpretation, or review load;
- the relation to
Fwhen readers might otherwise treatFas a surrogate; - any missing or unknown reading or unresolved source or use condition that could change the receiving judgement.
One-line publication is admissible only if the applicable facet definitions and tests, and the grounds of any relation claims, remain recoverable.
Partial-profile rule
A partial profile identifies its selected facets and any missing reading that could change the receiving judgement. Publishing AE and CD can suffice when a source or interaction condition is irrelevant to that use. When such a condition matters, state the relation and its grounds or the decision-relevant unknown under C.2.6:4.3. For an independently defined local Characteristic, use C.2.2a:4.2 for its reading and uncertainty.
One facet can be sufficient for the named use. Refer to other readings only when that use needs them; no note about irrelevant omissions is required.
Overlay discipline
Local overlays such as "explicit-but-open", "derived from trace T", or "requires operator acknowledgement" are admissible only when their meanings are recoverable from the readings or relation claims stated or cited in the profile. Overlays remain secondary to the declared profile and must not replace the facet bundle.
Cross-Facet Reading Rule
No master-facet reading
Do not infer the whole language-state profile from one facet. High AE does not entail high CD; an operator acknowledgement condition does not fix AE or CD; symbolic representation does not establish the expression properties needed by the receiving use; low CD does not imply low operational consequence.
Threshold interaction rule
For a threshold expressed over one facet, identify any other facet readings or relation conditions that also constrain the receiving use. A local rule may require sufficient AE for an explicit open explanatory question for B.5.2.0 and cap CD so rival explanations remain live. Another may require AE3+ and an open-interpretation CD condition for an exploratory A.6.P use of a relation that needs clarification. Apply these thresholds to the named semantic branch under the receiving pattern's entry conditions. If the readings are uncertain, use C.2.2a:4.2 to check what their joint constraints establish. A bound can settle the condition. If the answer still differs between possible combinations, identify the difference the receiving use needs. Any binding consequence retains its own governing rule and authority.
Transition reading rule
Read changes in the profile's readings and source or use conditions separately. A note may become more explicit without becoming more closed. Replacing direct access to an observation with access to its written report can leave closure unchanged while changing which check the recipient can perform. A representation may become more symbolic without adding the expression properties needed by its receiving use. Use the applicable move rule to determine which current readings and source, use or route conditions must hold. State any changes to them that the transition makes.
Review Matrix and Migration Tests
Review matrix
An assurance reader should ask:
- does each published facet keep the definition and test supplied by its own pattern rather than by surrogate prose;
- does any overlay introduce a hidden scalar or gate decision;
- are threshold claims tied to the facet that bears them;
- do cited moves in
A.16,A.16.1,A.16.2,B.4.1, orB.5.2.0match the facet bundle; - if reuse depends on distinct local meanings, were their exact senses recovered and the F.9 predicate tested, preserving a false or unresolved result; if a Bridge obtains, does its separate bounded-use claim support the proposed reuse under F.9:4.2, with any optional loss account or F.9.1 stance note kept separate from the facets;
- if viewpoint conformance is claimed, was it tested under
E.17.0?
Migration test for source prose
When a receiving use depends on a source phrase such as "still immature" or "already stable enough", recover the claimed reading or relation, its grounds, any threshold or route consequence, and the cited rule that warrants that consequence. Add an external authoritySourceRef only when an external authority supplies the rule.
Comparative profile use
Compare profiles on the named facets. A missing reading leaves that facet's comparison unresolved. A reporting aggregate needs a declared local model and operations admissible for the Scales under A.18/A.19; publishing an aggregation rule alone does not establish that basis. The receiving rule determines whether the aggregate supplies a sufficient input. Keep component readings or an accessible return to them whenever that rule, interpretation or reliance needs their distinctions or grounds.
C.2.LS:End
Articulation Explicitness — How Explicit Is an Episteme's Meaning for Its Use?
Type: Definitional (D) Status: Stable Normativity: Normative unless marked informative
Plain-name. Articulation explicitness.
Use this pattern when. Use C.2.4 when the next move depends on how explicitly an episteme's meaning is articulated for its semantic branch and receiving use.
What goes wrong if missed. A formal-looking sentence is treated as semantically ready, a real early cue is discarded as too vague, or a partly explicit note is forced through relation repair even when its receiving question concerns another kind of content, such as a plan or ordinary domain statement.
What this buys. One branch-neutral ordinal characteristic: a stable cue can acquire recoverable governed structure and a complete form for its actual semantic branch, with meaning that remains stable in a receiving use, without confusing articulation with formality, closure, truth, trust, or endpoint authority.
Problem frame
A U.Episteme may matter before its meaning is fully explicit. In the language-state chart over U.CharacteristicSpace, AE assesses how explicit that meaning is for its receiving use. The content may be a relation claim, a plan, a question or another domain statement; :17.1 gives the applicable branches.
Problem
When articulation explicitness stays implicit, authors either overstate readiness for a receiving pattern or hide useful early structure. Reusing F for this judgement creates a category error: formality assesses the expression properties needed by a use, while articulation concerns whether the meaning needed by the receiving use is explicit. A relation-shaped ladder creates a second error by scoring other well-formed claim kinds as immature merely because they are not relations.
Forces
Solution
Articulation explicitness (AE) characterizes how explicitly an episteme's semantic structure is expressed for its semantic branch and receiving use. Its ordinal Scale orders progression from a cue to a complete branch-appropriate expression whose meaning remains stable in that use. The common direction is independent of the semantic branch.
Characteristic, reading and assessment
Articulation explicitness is the Characteristic; AE0–AE5 name values on its ordinal Scale (:4.2). An AE reading attributes a supported value, or a set of remaining possible values, to an identified episteme edition within a stated assessment scope. Identify a selected fragment when only that fragment is assessed. The branch and receiving use determine which semantic structure must be recoverable. For a cue-bearing account, assess the meaning needed for the intended next use; the underlying observation or felt tension is part of its grounds.
A.18 distinguishes the Characteristic from its Scale and values. A chart slot binds that Characteristic and Scale under A.19; an assessment states the reading and its grounds. A note, table or C.2.LS profile can express the assessment. Its form does not determine the reading.
Choose a sufficient articulation claim
Start with the intended question or action and the meaning it requires. Read the available expression to recover its subject, semantic branch, and the participants, conditions and bounds that matter to that use. State what is already explicit and what is still missing. A direct condition can be sufficient for the next action; in that case, use it without assigning an AE level.
Use an AE reading when the receiving work needs comparison on this Scale, a position in a declared chart, or a test against a declared AE threshold. Match the available content to the anchors in :4.2 and state the grounds for the reading. Reuse an applicable assessment when its edition, scope, branch and receiving conditions still match. If the grounds do not determine the needed value, retain the supported possibilities or a decision-relevant unknown. The applicable rule for the receiving action supplies its condition; an AE value supplies only the articulation input that condition uses.
Develop and revise the content in writing
Writing can help obtain the meaning that will later be assessed. Put the cue or earlier wording beside a proposed statement, question or plan. Try to make the missing participant, distinction or condition explicit, then compare the expressions with the observations and reasons that motivated them. Retain what supports the next use and expose a gap that still prevents it. Use the relevant subject pattern when developing the missing content requires its method.
Assess closure separately under C.2.5 when the use depends on which alternatives remain live or which route is fixed. Making a claim explicit can raise AE while leaving its warrant weak; restoring a lost alternative can improve the articulation while reducing closure. Preserve the observations, qualifications and rivals needed to reconsider the conclusion. A.16.0:14 shows this comparison across three written accounts; use a separate history account only when that history changes the next use.
When the content, scope, branch or receiving use changes, reconsider the affected assessment and any decision that relied on it. A changed receiving requirement can make the earlier reading inapplicable even when the assessed episteme is unchanged. C.2.1 governs any claim-content or episteme-identity change; A.16 and A.16.2 govern a selected publication move or retreat.
Characteristic specification
- Assessed subject: the identified episteme edition, with assessment scope and receiving use stated.
- Scale discipline: ordinal.
- What rises: semantic shape becomes more explicit.
- What does not follow automatically: truth, trust, closure, admissibility, or formality.
AE is therefore independent from F, from closure degree (CD, C.2.5), and from endpoint authority.
Use C.2.8 to compare how much selected structure a reader can correctly extract under stated preparation, access and budget, including when both expressions already meet the same AE threshold. Reader recovery supports an AE judgment about explicit articulation; it does not turn the AE0–AE5 progression into a scale of extractable structural amount.
Starter anchor set
The anchors are a starter set. A local use may refine them, but it shall keep the common direction and the distinction from F intact. A refined anchor may be branch-relative; it may not make relation structure the universal measure of articulation.
Use discipline
- An entry threshold may be stated in
AEfor the direct semantic branch named by the current use. A.6.Pis only the relational branch. A note does not enter it merely because a table, arrow, or sentence looks relation-shaped.AEmay justify why an episteme remains inA.16.1orB.4.1while its branch or required structure is still unresolved.- A local threshold shall name both the intended receiving pattern and the branch-appropriate structure it requires.
AEshall not stand in for closure, confidence, truth, trust, or authorization. Judge articulation and the C.2.3 expression properties by their own criteria; neither assessment label substitutes for the other.
Change discipline
Raising AE requires additional recoverable anchors, participants, actions, conditions, bounds, or stable receiving-use structure appropriate to the current branch. Lowering AE is admissible under A.16.2 when a prior articulation proves over-committed, misleading, or routed to the wrong branch.
Archetypal Grounding
Tell. “Something is off” may preserve a real cue while the meaning needed by the next question remains unclear. If the cue says only “role,” route it through E.10.ROLE; do not presume a work-facing kind or assignment.
Show (System). An operator alert cue grounded in a disturbance trace may be stabilized as a candidate intervention cue before the intended intervention and its conditions are fully stated.
Show (Episteme). A research note may name a contrast and exemplars before it has a fully articulated claim.
Bias-Annotation
The pattern legitimizes early cues. The counter-bias is explicit: low AE never licenses hidden semantics or unreviewable leaps.
Conformance Checklist
CC-C.2.4-1AESHALL NOT be treated as a synonym forF.CC-C.2.4-2A route change SHOULD require the local articulation threshold declared for that receiving branch;A.6.Papplies only to a current relational branch.CC-C.2.4-3AEjudgements that drive routing or use SHALL cite the anchors, participants, actions, conditions, bounds, or other branch-appropriate structure that justifies the chosen level.CC-C.2.4-4RaisingAESHALL NOT be described as if it automatically settled closure, truth, trust, authority, or admissibility.CC-C.2.4-5Relation-shaped notation SHALL NOT raise the level or selectA.6.Pwhen the claim belongs to another branch.
Common Anti-Patterns and How to Avoid Them
- Precise syntax, unclear meaning. A template constrains the sentence form while needed meanings remain unresolved. State the expression properties and the articulation gap separately.
- Mystical cue immunity. Low
AEis presented as exempt from authoring discipline. Preserve it, but state what remains unresolved. - Ready-by-tone. A sentence sounds precise, so authors assume
AE3+. Publish the recoverable structure. - Relation-shaped shortcut. Arrows, columns, or grammatical subjects and objects are treated as proof that the claim is relational. Recover the claim kind and route its branch before assigning the higher level.
Consequences
The benefit is admissible publication of early cues and branch-aware threshold setting. The trade-off is that authors must distinguish not yet explicit from already formal and must name the receiving branch rather than treating relation repair as the universal destination.
Rationale
AE supplies a shared ordinal interpretation for explicit articulation in the language-state chart. A.16.0, A.16.1 and B.4.1 can cite these anchors when an articulation threshold helps state their entry or completion condition. A directly stated sufficient condition needs no extra level merely to repeat it.
SoTA-Echoing
Gendlin's Introduction to Thinking at the Edge (2004), pp.3–5, is a historical comparison for :4.6. It describes trying several words against an initially unclear felt meaning, then developing sentences that express what each attempted word left unsaid. The resulting distinctions can change what is being articulated; writing and reading back can further develop the meaning. The ordinal AE anchors are FPF's construction for assessing that content against the stated receiving use.
Relations
- Builds on:
A.18,C.2.2a,C.2.LS. - Coordinates with:
C.2.1,C.2.5,C.2.8,A.16.0,A.16,A.16.1,A.16.2,A.3.1,A.3.2,A.6.P,A.6.3.RT,A.13,A.15,A.15.1,C.29,F.6,B.4.1,B.5.2.0,C.2.P.DR, andC.16.P. - Constrains: articulation thresholds for routing and repair.
Worked Examples and Edge Cases
Precise syntax, unclear meaning
A template may enforce precise syntax while leaving the meaning needed by the receiving use unclear. For example, a plan template can name a field without identifying the actor or the completion condition. A local system-role kind, classification, or assignment is selected only after evidence supports that branch. Formal-looking language does not make its semantic route recoverable.
Plain plan, high articulation
The note At 14:00, Maintenance Team 2 will isolate Pump P-17, replace Seal S-4, and restore service only after Leak Test LT-9 passes uses ordinary language and little formal notation. It can still reach AE4: the planned actor, affected entity, ordered actions, time, and completion condition are explicit enough for the planning branch.
It routes to A.15; the note describes planned actions, while A.15.1 governs any admitted Work occurrence. The note needs no relation repair merely to count as explicit.
For a planning check whose required content is exactly those items, inspect them directly. Add the AE4 reading only when a comparison, chart or declared threshold needs that Scale. The plan's explicitness still leaves the planned action and any authorization for it to their own tests.
Relation-looking but wrongly routed
A row Maintenance Team 2 | Pump P-17 | isolate | 14:00 looks slot-shaped. If it records intended action, however, it is a compact plan row rather than proof of a durable relation. Its layout does not raise AE until the planning meaning and conditions are recoverable, and it does not select A.6.P.
Threshold edge case
A cue that names a stable trigger and two candidate interpretations, but leaves which one is intended open, can support AE2 for that question. If the assessor has only an excerpt and the missing passage could settle the branch, the assessor cannot yet establish whether the complete expression reaches the AE3 threshold. The excerpt alone supplies no upper bound on the complete expression’s articulation. Recover the missing passage when the difference changes the receiving decision. If an assistant supplied the interpretation, identify what the original expression allowed the reader to recover and what the assistant added. Use the receiving branch’s own conditions for entry.
Recovering what a clearer summary lost
In the constructed case in A.16.0:14, an initial note preserves three timeouts just after a restart, a later successful request, and two rival explanations. A summary says that the system is permanently undersized and recommends adding replicas. That diagnosis and action are more explicit, but comparison with the original note reveals the lost observation window, success and rivals.
The revised account returns to the original observations, withdraws the permanent-capacity conclusion and formulates a question about an observation that could distinguish the rivals. Under the case's stated readings, the summary raised AE from AE2 to AE3 and CD from CD2 to CD3; the revision keeps AE3 while reopening to CD2. Comparing the written contents supplies the reason for revision. The levels report different aspects of that change; neither a clearer summary nor its higher closure establishes the capacity claim. The full case distinguishes the supported edition-continuation claim from the rejected one and identifies the permitted next use.
Authoring and Review Guidance
Author prompt
To assign AE, ask:
- is a stable cue, question, or contrast nameable;
- which anchors, participants, actions, fields, or meanings are visible;
- is the direct semantic branch recoverable;
- is a complete form for that branch publishable;
- can the named receiving use interpret it without reconstructing the meaning from the source?
Review prompt
A reviewer should reject an AE claim based only on rhetorical confidence or relation-shaped presentation. The claimed level should be supported by the branch-appropriate anchors, participants, actions, fields, meanings, conditions, bounds, or stable receiving-use structure.
Threshold publication reminder
If AE determines whether an episteme stays in A.16.1, passes through B.4.1, or enters a direct receiving pattern, publish the local threshold and name that branch. Use A.6.P only for the relational branch; do not borrow its threshold for plans, methods, Work, representations, prompts, characteristics, or ordinary domain claims.
Extension and Migration Notes
Local anchor refinement
Contexts may refine the starter anchor set with subanchors, but the refinement must preserve the ordinal direction and the distinction from F and CD.
Migration from vague articulation prose
When routing or use depends on words such as "still vague", "more explicit now", or "ready for formalization", recover the meaning already explicit, the missing content and the condition for the next use. Use that direct condition when it is sufficient. Add an AE reading only when the receiving work needs its Scale, with the grounds and any declared threshold. A claim about the resulting move or route still follows its own governing pattern.
Boundary reminder
AE does not govern closure, confidence, or warrant. If authors want those meanings, they must publish them through their own governing patterns.
Articulation Publication Package Discipline
Minimal articulation package
An AE claim that changes routing or use should be published with more than a level token. State:
- the episteme publication being judged;
- the stable cue, question, or contrast;
- the candidate anchors, participants, actions, fields, or meanings relevant to the current branch;
- the branch-appropriate structure already recoverable and any decision-relevant gap;
- the grounds for the level;
- the named next use and its local threshold when the level is used to change route.
Unresolved claim-bearing role wording goes first to E.10.ROLE. A bare AE3 label is insufficient when the supporting structure and intended receiving branch are absent.
Threshold package for route change
When AE controls a route change, publish the intended receiving pattern, its semantic branch, the minimum recoverable structure it requires, and the current evidence that the publication meets or misses that threshold. A relational route may name A.6.P; another branch names its own receiving pattern.
Evidence-limited rise rule
AE may rise only as far as the published branch-appropriate structure—including any relevant anchors, slots, or contrasts—warrants. Stylistic polish, templates, or rhetorical confidence do not raise AE on their own.
Threshold Crossing and Split Handling
Branch-aware high-articulation exits
At AE3+, use the local threshold of the direct branch rather than one universal destination:
- for a relation claim, restore precision through
A.6.Pwhen needed, then use the pattern that governs that relation; - for a plan or intended activity, use
A.15; - for a Method, use
A.3.1; useA.3.2to test aC.2.1episteme's membership as a MethodDescription; - for a dated Work-occurrence claim, recover the performer under
A.13and independently test Work admission underA.15.1; useF.6only when precise assignment-bound attribution is needed; - for a representation claim, use the represented subject's rule and the representation's correspondence; use
C.29when choosing or transferring a mathematical representation or relying on it beyond its supported correspondence leaves a separate question. UseC.2.P.DRwhen declarative form is being overread, addingA.6.3.RTonly when a representation transition is at issue; - for an abductive prompt or explicit open question, use
B.5.2.0or the pattern that governs the question; - for a Characteristic or Scale claim, use
A.17andA.18, withC.16.Pwhen its scalar wording hides the construction; - for an ordinary domain claim, use the domain pattern that governs its subject; use
C.2.1when episteme identity is needed.
If the branch or threshold is unresolved, preserve the episteme through B.4.1 or A.16.1 and state what is missing. Use an applicable recognition or clarification step to resolve that question before relying on the receiving result. AE reports recoverable articulation; it does not itself choose the branch or authorize the receiving use.
High-articulation, low-closure cases
A note may reach AE4+ while remaining low or mid in CD. In such cases state that articulation is sufficient for precise handling while closure still leaves rival routes or frames live.
Articulation claims for separate fragments
Scope each fragment's articulation claim to its receiving use; do not assign one averaged level that hides the route structure. The fragments may remain in the same note or publication when their scopes are clear. If the recipient must use the fragments together, read the connection as part of that use. A test result and a restoration instruction can each be explicit while leaving it unclear whether this result releases this pump. One sentence naming the pump and the applicable test can supply the missing connection.
Review Matrix and Endpoint Boundary Tests
Review matrix
A reviewer should ask:
- are the named anchors present in the published expression rather than presupposed;
- does the level rest on recoverable meaning rather than tone, formatting, arrows, or columns;
- is the direct semantic branch explicit, or is a relation route being inferred from presentation alone;
- are the participants, actions, fields, meanings, conditions, bounds, or ordinary wording required by that branch still hidden;
- has an intended activity or plan been mislabeled as Tech
Workbefore an occurrence is admitted; - if
AEjustifies a route change, does the publication meet the named receiving pattern's entry conditions?
Endpoint-boundary test
High AE does not by itself authorize endpoint claims, gate claims, or quality ascriptions. If such consequences appear, show which downstream governing pattern takes over.
Migration note for false precision
Rigid templates, capitalized labels, or tidy sentence rhythm can simulate articulation. Test whether the claimed branch-appropriate structure, including any required anchors or slots, is recoverable. If it is not, revise the assessment to the level supported by the available structure, or leave the level unresolved when its grounds cannot be established.
C.2.4:End
Language-State Closure Degree — How Fixed Is the Current Candidate Space?
Type: Definitional (D) Status: Stable Normativity: Normative unless marked informative
Plain-name. Language-state closure degree.
Use this pattern when. Use C.2.5 when a U.Episteme publication must say how fixed its candidate space, route space, or frame space has become before endpoint use, reopening, or retreat.
What goes wrong if missed. A confident tone is mistaken for closure, closure is mistaken for truth or gate authority, or a closure drop leaves readers unable to tell which endpoint uses and route commitments remain supported.
What this buys. A separate ordinal characteristic for closure degree, so teams can distinguish exploration, stabilization, selected route, guarded fixation, and admissible retreat without collapsing closure into formality, articulation, warrant, or obligation.
Problem frame
A U.Episteme may already be explicit enough for publication while its declared position claim remains intentionally open to rival routes or frames. The declared language-state chart over U.CharacteristicSpace therefore needs a separate pattern defining the closure-degree basis slot: how fixed or closed the current candidate space has become.
Problem
Closure is often hidden inside vague words such as "ready", "settled", or "open". When closure is not explicit, teams cannot reason cleanly about reopen, sketch-backoff, or the admissibility of endpoint docking.
Forces
Solution
Language-state closure degree (CD) characterizes how firmly a specified candidate space is fixed in the position expressed by an episteme for a named use. The candidates may be answers, frames or possible next moves. Identify which choice is being assessed and which alternatives remain live. The ordinal Scale in :4.2 describes the progression from open exploration to stronger fixation.
Characteristic, reading and assessment
Closure degree is the Characteristic; CD0–CD5 name values on its ordinal Scale. A reading attributes a supported value, or a set of remaining possible values, to an identified episteme edition within a stated assessment scope and receiving use. A language-state chart binds the Characteristic and Scale in a slot under A.19. The Characteristic, its Scale values, a reading and the chart slot remain distinct under A.18/A.19.
A closure assessment states the fixity and its grounds. Claims about a binding route commitment, gate or authority relation follow their own rules and case facts. A note or C.2.LS profile can express the closure assessment; its reusable form and the filled assessment remain distinct.
Recover the choice before assigning a level
Name the question or choice whose candidate space matters. Recover the live alternatives, any selected route or frame, and the reason that limits further choice. State the scope of that fixation and what would justify reopening it. The different sources of narrowing in :16.2 determine what must be reconsidered when conditions change. If every remaining explanation calls for the same action, that action may be settled while its explanation remains open. Results for a supplied list establish what follows within it. To extend the conclusion, establish that the list covers the admissible ways or give an argument that also governs ways omitted from it.
Use that ordinary account directly when it supplies the receiving decision's condition. Assign a CD reading when the work needs comparison on this Scale, a chart position or a test against a declared threshold. Match the narrowing and reopening conditions to :4.2; retain an unresolved reading when the available grounds do not establish them. A level does not replace the surviving alternatives or the reason for their treatment.
The Scale's order describes greater fixity, not a universal improvement. Inquiry may need several alternatives kept live; a particular operational use may need one route fixed under a separately justified rule.
C.2.4 assesses how explicit this content is. Writing out the alternatives and the decision rule can improve articulation without changing closure. New evidence or a changed use can instead reopen one choice while other commitments remain fixed. Preserve the supported observations and surviving choices, then revise only the conclusions affected by the change, using A.16.2 when a publication move is needed.
Characteristic specification
- Assessed subject: the identified episteme edition, with the candidate space, assessment scope and receiving use stated.
- Scale discipline: ordinal.
- What rises: the candidate space or selected route becomes more fixed for the named use. Any binding force comes from its separately established rule.
- What does not follow automatically: truth, trust, formality, or quality.
Starter anchor set
Non-collapse rules
LanguageStateClosureDegree is not:
F;- articulation explicitness;
- gate decision;
- evaluator confidence;
- warrant strength.
A text may be highly explicit but low-closure, or low-explicitness but already high-closure by policy. Those states shall not be collapsed.
Change discipline
For a higher CD reading, identify what the relevant scale anchor adds (:4.2). Depending on the anchor, the grounds can be different predictions for still-live alternatives, evidence excluding an alternative, a selected route, or a rule changing when a choice can be reopened. A rule can strengthen the reopening condition while the same route remains selected. Report only the change the evidence supports. When closure decreases, make the prior fixation, reason for reopening and still-binding remainder recoverable. Use A.16.2 when a published framing, route or use claim must be revised or the applicable fixation rule requires that move; add a note only when the existing material is insufficient.
An ordinary PatternID in governingPatternRef locates the FPF rule relevant to the closure claim. Use a separate relationFunctionClaimRef for its exact defining or constraining ClaimGraph only when admissible interpretation, comparison, migration, publication, or reuse depends on that exact rule identity (E.10, E.4.PFR). Use authoritySourceRef when a non-pattern source carries the relevant authority.
Archetypal Grounding
Tell. Two notes may look equally explicit, but one is still intentionally open while the other is already committed to a single route.
Show (System). An incident cue can be routed to rollback while remaining reopenable if new evidence arrives.
Show (Episteme). A hypothesis sketch can be highly articulated but still low closure because rival explanations remain live.
Bias-Annotation
Explicit closure takes description work; it makes the remaining rivals and grounds of fixation recoverable.
Conformance Checklist
CC-C.2.5-1Closure SHALL be declared independently fromFandAEwhen it matters for routing, docking, or reopening.CC-C.2.5-2Reopen/backoff moves SHALL cite the prior closure state they are relaxing.CC-C.2.5-3A claim that strong closure is binding SHOULD name the applicable guard,governingPatternRef, orauthoritySourceRefthat establishes that binding force.CC-C.2.5-4A closure drop SHALL NOT silently preserve an endpoint-use claim when the supporting route or publication form no longer supports it.
Common Anti-Patterns and How to Avoid Them
- Closure by mood. A sentence sounds decisive, so teams assume high closure. State the fixation and its grounds; add a
CDreading when the receiving use needs that scale. - Irreversible drift. Closure rises informally but no reopening condition exists. State what would justify reopening; use
A.16.2when a published framing, route or use claim must be revised. - Authority smuggling. High closure is treated as if it were automatically a gate or obligation. Check the applicable gate condition for a gate claim; use
A.2.8to check whether the claimed obligation obtains.
Consequences
The benefit is admissible handling of stabilization, commitment, and reopening. The trade-off is making the narrowing and any consequential retreat recoverable.
Rationale
Closure describes how fixed the candidate space is; articulation describes how explicit its meaning is. A.16.0 and its related patterns can use both when each affects the move condition.
SoTA-Echoing
The comparison in :13.1 distinguishes making a diagnostic question explicit from excluding an alternative with a reference-gauge reading. Only the latter narrows that candidate space. In :13.2 a policy instead fixes the route for a horizon; reopening then depends on that policy as well as changed observations. These cases support identifying the source of narrowing and retaining separate AE and CD judgements when both are used.
Relations
- Builds on:
A.18,C.2.2a,C.2.LS. - Coordinates with:
C.2.4,A.16.0,A.16,A.16.1,A.16.2,B.4.1,B.5.2.0. - Constrains: reopen, backoff, and endpoint docking guards.
Worked Examples and Retreat Cases
Explicit but still open
The constructed Pump-alert-21 in C.2.LS:13.1 makes its diagnostic question and the contrasting implications of three alternatives explicit. It has AE4 for that question and CD1 because all three alternatives remain live. When an accepted reference-gauge reading excludes one alternative, the candidate field narrows to CD2; neither remaining route is selected. The question remains fully articulated at AE4.
A practitioner choosing the next diagnostic question can use the stated alternatives and observations directly. A chart or comparison that needs both Scales can use the separate readings. The full case supplies the observations and discrimination rule; the two levels alone would not justify excluding a candidate.
Strong closure under policy guard
An operator rule may be only moderate in AE but high in CD because policy already fixes the next step under the current horizon. This shows why closure is governance-facing, not merely stylistic.
Reopen case
A route may move from CD4 back to CD2 when counter-evidence appears. Make the counter-evidence and changed scope of closure recoverable; use A.16.2 when the publication must be revised.
Authoring and Review Guidance
Author prompt
To assign CD, ask:
- which choice is being assessed, which alternatives does the account leave open for it, and what sets the bounds of that field?
- is one route merely preferred, or fixed?
- if the closure is binding, what applicable guard,
governingPatternRef, orauthoritySourceRefmakes it so? - what would count as an admissible reopen trigger?
Review prompt
An assurance reader should use the applicable anchor in :4.2: identify the contrasts that organize still-live alternatives, the alternatives excluded, the choice made, or the rule that now governs reopening. State the grounds for that reading (:16.2). Confident tone or an appeal to hierarchy alone does not establish that fixity.
Governance note
Whenever a claim about a gate, commitment, or endpoint use depends on CD, the supporting guard, governingPatternRef, or authoritySourceRef should be visible.
Extension and Migration Notes
Local anchor refinement
Contexts may refine the starter closure anchors, but shall keep the ordinal progression and the explicit link to reopen/backoff discipline.
Migration from readiness language
When routing, endpoint use or reopening depends on words such as "settled", "closed", "final", or "open", recover the claimed fixity and its grounds. Use an explicit CD reading when that scale is needed and A.16.2 when a publication move is needed. Sufficient ordinary wording needs no separate level solely to repeat it.
Boundary reminder
CD is not warrant strength and not a gate decision. It speaks only about the local fixity of the current episteme or publication position and its candidate space.
Closure Publication Package Discipline
Minimal closure package
A publishable CD claim should name what has narrowed:
- the rival routes or frames that remain live;
- the route, frame, or interpretation that is currently privileged or fixed;
- any guard,
governingPatternRef,authoritySourceRef, or policy that makes the narrowing binding; - the condition under which an admissible reopen or backoff would occur.
A bare claim such as "now settled" is insufficient when closure affects routing or authority.
Narrowing-source rule
Closure may rise because evidence eliminates rivals, governance temporarily binds a route, or protocol requires fixation under time pressure. State the source of narrowing because different sources imply different reopen expectations.
Partial-closure rule
Closure may be local rather than global. A note can be closed enough for one route while remaining open about broader explanation or classification; a prompt may be fixed enough to hold one question steady while still open enough that rival answers remain live. Publish that locality explicitly.
Continuing and Withdrawn Authority Handling
Authority retention rule
If higher CD carried endpoint expectations, guard claims, or route commitments, a closure drop must say which consequences remain and which are withdrawn. Treat any authority-relation change separately under the pattern that governs that authority relation, as required by A.16.2.
Admissible retreat record
An admissible retreat through reopen, sketchBackoff, or respecify should retain:
- the prior closure state;
- the reason the prior fixation no longer holds;
- the assumption or route being relaxed;
- the still-binding remainder, if any.
This prevents false continuity after retreat.
Closure versus obligation boundary
High CD may coexist with obligations. When prose treats "closed" as "must now be done", use A.2.8 to identify the duty bearer, duty, and instituting rule and basis; retain the applicable governingPatternRef or authoritySourceRef for that obligation claim.
Review Matrix and Reopen Tests
Review matrix
An assurance reader should ask:
- what was narrowed;
- what evidence, choice or applicable rule supports that narrowing, with a
governingPatternRef,authoritySourceRefor guard when one is used; - what would reopen it;
- which gate, release, work, evidence-use, assurance, policy, or adjudication claims depend on the stated closure conditions and remain supported;
- whether the publication distinguishes local closure from whole-case finality.
False-finality test
When a receiving action relies on "final", "settled", or "decided", check the scope and source of fixation and any applicable guard. A final-sounding word alone does not establish those conditions.
Cross-facet reminder
Assess articulation under C.2.4 and representation under C.2.7 independently of closure. Source, evidence, access and action conditions retain their separate claims under C.2.6.
Split-closure review case
A publication may be closed enough for immediate local work use or reliance use while remaining open about broader explanation, long-horizon consequences, or alternative classification. Allow the split when locality is explicit; reject prose that advertises whole-case finality when only one language-state segment is fixed.
C.2.5:End
Sources, Evidence and Conditions for Using an Episteme
Type: Architectural (A) Status: Stable Normativity: Normative unless marked informative
Plain-name. Sources and conditions of episteme use.
Use this pattern when. A claim or cue is being reused, translated or published elsewhere, and the receiving work needs to know what produced it, what supports it, or how someone can use it. An analyst reading an operator's vibration report, for example, may need the operator's observation and its circumstances before treating the report as evidence about a machine.
Sufficient result and non-use. State the relations and conditions that can change the intended interpretation, check or action. Use an already sufficient description directly. A new account, classification or chart is unnecessary when the receiving work can recover those facts and conditions from the available text.
What goes wrong if missed. A report can preserve a sentence while losing access to the observation, model state or interaction needed to use it. An unspecified claim that it is "anchored" leaves the reader to infer whether this means derivation, evidence or a condition of action.
What this buys. The recipient can identify which conclusions remain supported, which action conditions still hold and what must be recovered before the intended use can continue.
Problem frame
A U.Episteme may be produced from observations, refer to a model state, report another document's assertions and participate in an operator's decision. These facts answer different questions about the same episteme. Their importance changes with the receiving use.
Problem
A common label can conceal the particular relation on which a use depends. Knowing that a reading was computed from a trace establishes its provenance; using that trace as evidence for a claim also requires the relevant observation and target claim to match. Publishing the reading in a document supplies access to its expression. A separate interaction may make the reading available for an operator's next check.
Forces
Solution
Recover the claims about sources, evidence, access and action conditions that the receiving work needs. State each with its participants and grounds. Then determine what publication, translation or a changed use preserves, changes or leaves unknown. The result is an ordinary account of those facts and conditions; C.2.1 governs the account and E.24.PUB governs its publication.
Relations, readings and a shared account
A source relation, an evidence-use claim and an operator's use of an alert can hold together. Each keeps its own predicate and grounds. For empirical grounding, apply C.2.1:4.3 to the designated claim coverage and observation or test; apply A.2.4 to the relation between the evidence episteme and its target claim, and A.10 when provenance or bounded reliance must be established. Use C.2.P to distinguish the epistemic objects.
When basedness is hidden by an umbrella word, apply A.6.6:4.0; A.6.P supplies the wider route to an unclear direct relation. Return the recovered assertion or unresolved condition to the receiving use. An ordinary adequate statement can supply this account without a new formal record.
An account collecting several such statements is not one Coordinate. A language-state profile may cite it beside its Characteristic readings. A local summary can occupy a chart slot only when its single measured aspect, subject/input signature and Scale are defined under A.18/A.19, as :15.1 explains. The account's component relations remain recoverable.
Distinctions that can change use
The following are recurring examples of questions to resolve, not a closed taxonomy or alternatives on one Scale. Select a question because its answer can change the intended use.
Recover the relations needed for one receiving use
- Identify the episteme edition, the claim or cue being used and the receiving judgement or action. Keep a source episteme and an analyst's resulting account distinct under C.2.P when their contents or concerns differ.
- Ask which facts the receiving work needs: for example, the source of a reading when diagnosing a calculation, the relevant observation when checking a pressure claim, or an acknowledgement condition when arranging an operator's next check. Obtain only the relations that can change that work. The table in :4.1 helps locate common questions; the subject of the work determines the answer and can require another relation.
- State each needed relation with its participants, applicable scope and grounds. When a trace is used as evidence, establish its relation to the target claim as well as its provenance. When an action is claimed, establish the interaction or governing condition and the relevant performer. Preserve a consequential unknown instead of filling it from a neighbouring fact.
- Compare what the source and recipient can observe, obtain or do. Which claims or actions still have the conditions they require? For a blocked check, identify what is missing. An instrument, a calibration reference, access to a record or authority to act each calls for a different response. For an event that cannot be repeated, seek a contemporaneous record or limit the conclusion to what the surviving evidence supports. Another performer can supply a present check. If the intended result is the recipient’s ability to perform it, examine what the recipient can do on a further occasion under the stated assistance conditions.
For example, an alert can be computed from a measurement, delivered in a report and acknowledged by an operator. Replacing the report file while preserving the accessible assertions can leave the derivation and evidence unchanged. Discovering that the measurement concerned another pump changes its evidential use for the pressure claim. Disabling an acknowledgement channel changes the interaction available to the operator. These changes call for different continuations, so the account must keep their predicates distinct.
Contribution boundary
This pattern organizes the source and use questions; the patterns governing those relations establish the individual claims. Representation factors describe how information is represented; closure describes fixity of the current candidate space. Assess those Characteristics under their own definitions. A required bodily observation or operator participation is stated here or in the cited account, with its contribution to the receiving work.
Several relations and incomplete information
When several relations matter together, name each and explain its contribution. Distinguish their established conjunction from alternatives left unresolved by the available evidence. For example, "the source was either a sensor reading or a simulation" leaves provenance uncertain; it does not establish that both sources contributed. A generic "mixed" label supplies neither the missing facts nor their conjunction.
Bridge implications
A receiving use may need a semantic relation between local senses from different semantic contexts. The recipient's access to a report can differ from the source practitioner's bodily observation or access to model state. That difference may affect the evidence or loss account for the proposed semantic correspondence. Use F.9 to state and test the Bridge and the separate bounded-use claim, with the evidence for the correspondence. State which differences the receiving use can tolerate and which differences the comparison actually found. If observed loss has yet to be established, retain that uncertainty. Use F.9.1 for an optional separate stance note when needed.
Archetypal Grounding
Tell. A felt cue, a controller-side probe score and a textual design note can all help begin an inquiry. Recover the observation, computation or reported assertions that the intended inquiry needs.
Show (System). A console alert is used in an acknowledgement-and-check procedure. Its use account identifies the operator's acknowledgement and how the next diagnostic check is selected. A text display can carry the alert while these action conditions remain essential.
Show (Episteme). A model probe produces a score from an identified internal state. A later prose account reports that result. Recovering the state and probe establishes the derivation; the account's interpretation needs its own stated grounds.
Bias-Annotation
The questions favour explicit source and use claims when they affect a decision. Statements about bodily experience need an obtainable witness or other appropriate basis for the claim being made; describing that experience does not require reducing it to a document property.
Conformance Checklist
CC-C.2.6-1Source, evidence or action relations SHALL NOT be inferred from publication phrasing alone when they matter for source use, reliance or Bridge interpretation.CC-C.2.6-2Embodiment-sensitive or operator-loop cases SHOULD declare the relevant observation, enactment or operator interaction explicitly.CC-C.2.6-3Source and use claims MUST NOT be collapsed into the representation-factor readings governed by C.2.7.CC-C.2.6-4A declaration that several relations jointly matter SHALL identify those relations and their contribution explicitly.
Common Anti-Patterns and How to Avoid Them
- Text-only illusion. Treating a written report as a sufficient account of the observation or interaction that its use needs. Recover the relevant claim and its basis.
- Representation capture. Using symbolic/distributed labels to conceal a source or use condition. State that condition separately.
- Embodiment mystification. Treating bodily or operator-loop cues as beyond explicit publication. Publish the relevant observation or action account and the limits of its recovery.
- Common-word classification. Grouping provenance, evidence and interaction as alternative values because each has been called "anchoring". Establish the predicate before choosing a Scale.
Consequences
The recipient can inspect the facts that make the proposed use possible and return to the particular missing basis. This requires enough source and interaction detail for that use. An account can cite existing adequate descriptions; repeating them or filling every row of :4.1 adds no necessary assurance.
Rationale
A.16.0, A.16.1, B.4.1 and an applicable F.9 use need to preserve different facts when a cue is expressed or used elsewhere. An observation may remain the source after it is reported in text, while the recipient has access only to that report. An operator's interaction can change independently of either. Direct statements preserve these distinctions and explain which continuation changes.
A common nominal summary is useful only when a specified comparison uses it and its criterion consistently measures one aspect (:15.1).
SoTA-Echoing
The following sources contribute particular distinctions and alternatives for the use questions in :4.1.1. They do not establish one universal classification of the listed relations.
Relations
- Builds on:
C.2.1,C.2.P,A.2.4andA.10for the relevant direct claims;A.6.6andA.6.Pwhen the relation needs recovery;A.18/A.19for an independently defined local Characteristic. - Coordinates with:
A.7,A.16.0,A.16,A.16.1,B.4.1,B.5.2.0,C.2.2a,C.2.LS,C.2.7,F.9for any Bridge and bounded-use claim, andF.9.1only for an optional stance note about that claim. - Constrains: source and use accounts in cue publication and Bridge loss notes.
Worked Examples and Bridge-Loss Cases
Bodily observation and its written report
An operator first notices vibration through bodily sensation. A prose report describes that observation. A later analyst who has only the report can recover its assertions but cannot repeat the operator's sensation from the text. Distinguish the source cue from the analyst's episteme; identify which observation circumstances the intended comparison or check needs and whether the report preserves them. Apply F.9 when the use claims a semantic correspondence between local senses from different semantic contexts.
Model-derived score used by an operator
An episteme records a probe score obtained from a model state. A receiving alert reports that score in a procedure that requires operator acknowledgement before selection of the next diagnostic check. Keep both the source/probe derivation and the acknowledgement condition visible. A console display establishes the latter connection only when the required interaction is provided; the displayed score alone does not show that.
Provenance, evidence and interaction in one alert
An alert E asserts that Pump P's pressure is below 90 kPa. The software produced E from trace T. An operator has acknowledged E and selected a diagnostic check under the stated procedure. Inspection then shows that T was a measurement of Pump Q.
The derivation from T remains established. T establishes a measurement of Q. To support the claim about P, look for an applicable relation between the pressures or another source about P. For example, if T establishes 82 kPa at Q and an independently warranted relation for that configuration and time puts P at most 5 kPa above Q, P was below 90 kPa. Without such grounds, T alone leaves the alert's pressure claim open. A commission to measure P directly still requires that measurement.
The acknowledgement and selected check remain facts about the operator's work. Whether to continue, suspend or change that check follows its rule and purpose; neither derivation nor acknowledgement settles the truth of the pressure claim. A report can publish all these statements together with their different grounds.
Authoring and Review Guidance
Author prompt
Start from the receiving judgement or action. Identify the particular claim, source or observation it needs and the conditions of access or interaction that can change its use. State several relations separately when each contributes; leave a consequential missing fact unresolved.
Review prompt
Check that each relation attributed to a published account has the needed participants and grounds. Then change one condition in the case and determine which claim or continuation changes. If the wording cannot distinguish the changed relation from the retained ones, recover the objects through C.2.P and the relation through A.6.6 or A.6.P as applicable.
Bridge note
If a changed source, access or interaction raises a semantic-correspondence question between local senses from different contexts, use F.9 for the Bridge, bounded-use claim and its evidence and loss account. Use F.9.1 only when a separate stance note helps explain that claim.
Local Summaries and Precision Restoration
Local summary rule
A repeated comparison may benefit from a nominal Characteristic summarizing one aspect of the recovered facts. Define that aspect and its subject/input signature, the Scale, the rule mapping the relevant facts to a value, and the receiving comparison that uses it under A.18/A.19. Account for jointly satisfied conditions and keep incomplete information separate from a value. A new relation is not automatically a new value of an existing Scale.
For example, a team comparing access to a designated primary check can define one nominal Characteristic for the tuple (claim, designated check, recipient, time): the recipient can perform that check at that time, or cannot. Establish the capability and access required by the check before assigning either value; otherwise leave the reading unknown. The value helps choose between performing the check and arranging the missing access. This local comparison concerns availability of that check; it does not classify the claim's provenance or evidence quality.
When the summary is used in a language-state chart, declare its Characteristic as its own slot. A changed aspect or Scale changes that chart's meaning under A.19:5.1.2; preserve or establish an explicit mapping when an old position must be compared with it. C.3.2 governs any separate classification using these readings.
Recover a relation hidden by metaphor
When wording such as "the system wants" or "the note suggests" leaves the intended action or claim unclear, identify what the system does, what the note asserts, or who makes the interpretation. Recover the relevant detector, performer or witness through the pattern governing the subject. An ordinary understood metaphor needs no expanded record.
Boundary reminder
Source and use relations answer their stated questions. Representation, articulation, closure and trust retain their own definitions and tests.
Publishing the Required Account
Minimal account
Identify the episteme and the receiving use, then state or cite only the source, evidence, access or action conditions that use needs. Make their participants, grounds and any consequential uncertainty recoverable. Include a source transformation when it changes the claim or its use. For one unchanged relation, its adequate existing description can suffice.
Source-versus-form rule
Distinguish the source episteme, any receiving episteme and their publication forms under C.2.P. State provenance, evidence and use claims about the participants. A prose form alone establishes neither a document as the original source nor a change in the observation or interaction being reported.
Joint-use account
When several relations jointly enable the use, identify their separate contributions and which facts establish each. Retain the known relations while recovering a missing one; uncertainty about one does not erase the others or establish a mixture.
Changed Sources, Access and Use
State the change and its consequence
When a source, access condition or interaction changes, state the consequences for the intended comparison or action. A report derived from an operator's cue may omit timing or bodily load; identify the actual omission and its effect. A change of publication form alone does not establish such an omission.
Bridge-loss rule
If that change raises a semantic-correspondence question between local senses from different contexts, use F.9 to state and test the Bridge and its bounded-use claim, with the loss account; add an F.9.1 stance note only when it helps explain that claim. C.2.6 requires the relevant change to be noticed and represented accurately.
Same-content substitution test
Two cues may be paraphrased into the same sentence while being produced by different observations or available under different conditions. Establish the relations required by the receiving use before substituting one for the other.
Review and Extension Tests
Review matrix
For the stated use, check the original source and relevant transformations, what the recipient can recover or do now, and the grounds for the resulting claims. Include the separate F.9 question when a semantic correspondence is claimed. A joint-use account identifies each relation that matters.
Local extension test
Add another direct relation when the receiving work needs it. Introduce a nominal summary only when :15.1's aspect, mapping and comparison are supplied; a common word for several dependencies is insufficient.
Cross-facet reminder
A changed source or interaction can accompany a change in representation or articulation. Test those changes separately. A model-derived score, a bodily observation or an operator's acknowledgement supplies no reading of vagueness, closure or trust without the corresponding criterion and grounds.
C.2.6:End
LanguageStateRepresentationFactorBundle — How Is the Representation Organized?
Type: Definitional (D) Status: Stable Normativity: Normative unless marked informative
Plain-name. Language-state representation-factor bundle.
Use this pattern when. Use C.2.7 when a comparison or use depends on how an episteme is represented: the distribution of its code, its sparsity or its symbolic structure. Identify the representation and the units being assessed. An ordinary description can suffice when the receiving use needs no factor-bundle claim.
What goes wrong if missed. One representation label such as symbolic, distributed, or encoding basis starts doing too much work: it hides articulation, closure, source or use conditions, Bridge loss, or comparison assumptions.
What this buys. A comparison of representation factors on explicit bases, while keeping articulation, closure and source or use claims independently grounded.
Problem frame
Published position claims in the declared language-state chart over U.CharacteristicSpace must keep representation factors such as locality, sparsity, and symbolicity distinct.
Problem
Using EncodingBasis without its underlying factors collapses several independent choices. That makes comparison brittle and encourages one-factor stories such as distributed = informal or local = precise.
Forces
Solution
LanguageStateRepresentationFactorBundle is a factor bundle, not one scalar characteristic. The core starter set for the comparison questions below is:
LocalityDistributionSparsitySymbolicity
A local alias such as EncodingBasis shall identify the factors and readings it summarizes.
Factors and their readings
LanguageStateRepresentationFactorBundle groups readings of distinct representation Characteristics. Each factor needs its own meaning and Scale, and the declared chart binds these under A.18/A.19. The bundle describes the representation identified in the position claim; a factor reading is distinct from that representation or its publication form.
Minimal factor readings
Before comparing factors, identify the representation and the units being described. A model-side code and its written summary need not have the same factors.
For example, one scheme represents four states by four distinct units, with one unit active at a time. Another represents 28 states by the unordered pairs of eight units, with two units active at a time. Which unit is active identifies the state in the first scheme. Which pair is active identifies it in the second. The active-unit fractions are equal, 1/4 = 2/8, although locality differs. These fractions do not define a universal sparse/dense threshold. Suppose the recipient routes pairs containing unit 2 to one queue and all other pairs to another. Reading unit 2 alone decides the route. Identifying the exact state still requires the pair. Select the value the recipient needs to recover before attributing locality to its code.
In the second scheme, [0,1,0,0,0,0,1,0] and the list of active positions {2,7} express the same code: positions 2 and 7 are active and the other six are inactive. Suppose decoder D1 assigns that pair to state S. Both inscriptions then denote S. If decoder D2 assigns the same pair to state T, the interpretation has changed despite the unchanged list. The shorter inscription alone changes neither the underlying code nor its factors.
When the list travels separately, supply the unit order and decoding rule needed for its next use, directly or through an accessible return to the exact source content. Counting written marks instead of active units requires a different comparison basis.
Why begin with these factors?
The starter set separates three questions that often change how a representation can be read or transferred.
- LocalityDistribution concerns whether one selected unit distinguishes a represented value or whether a combination is needed.
- Sparsity concerns how much of the declared unit set is active or used; its reading depends on what is counted and on the local comparison criterion.
- Symbolicity concerns the explicit tokens and structures available to the receiving interpretation, including the applicable decoding or composition rules.
These questions can matter separately: locating a report, estimating the units used to express it and recovering its interpretation require different information.
The pair code in :4.1 explains why a count of active units cannot settle locality and why a compact inscription needs its decoder. The starter bundle keeps these distinctions available without requiring a complete model of the representation. Select only the factors that affect the receiving question, as :16.1 permits. Preserve the representation and unit basis for each reading; factors of a source model and factors of its explanation concern different representations.
An assessment aimed at reconstruction error, robustness or the effect of a causal intervention may need other Characteristics directly. Use those Characteristics when they answer the question; retain a starter factor only when its reading contributes. The gain from the shared triple is a reusable starting vocabulary for code organization, use of units and explicit symbolic structure. It is not an exhaustive representation-quality model, and it supplies no universal thresholds or empirical claim that the factors are statistically independent. Add a needed factor under :4.3 while keeping the meanings of the factors already used.
Non-collapse rules
Representation organization, articulation and closure are assessed under their own Characteristics; source and use claims retain their direct predicates. A distributed code can be computed from an identified trace; a symbolic note can leave its meaning unclear; a sparse representation can express a claim that remains open. Assess each relevant facet on its own grounds. The representation factors alone establish neither evidence for the claim nor permission to act on it.
Extension rule
Authors may add extra local representation factors only if the extension is published as a factor addition rather than as a new master factor that erases the core factor bundle.
Archetypal Grounding
Tell. A model-state cue can use a highly distributed representation and be obtained from an identified recorded state; a symbolic note can be low articulation if its semantics are still vague.
Show (System). An operator decision aid may mix sparse alert codes and symbolic method-description text.
Show (Episteme). A researcher can describe a distributed activation pattern in a symbolic hypothesis. Assess the model-side code and the hypothesis's expression separately; the new expression alone establishes neither the hypothesis's articulation nor its warrant.
Bias-Annotation
The pattern resists inferring a stage or progression from one representation factor alone.
Conformance Checklist
CC-C.2.7-1LanguageStateRepresentationFactorBundleSHALL be published as a factor bundle, not as a hidden scalar.CC-C.2.7-2Local aliases such asEncodingBasisMAY exist only with an explicit docking to the bundle's factors.CC-C.2.7-3Representation factors MUST NOT silently replace the source and use claims recovered through C.2.6 orLanguageStateClosureDegree.CC-C.2.7-4New local factors SHALL preserve the factor-bundle discipline.
Common Anti-Patterns and How to Avoid Them
- One-factor myth. Treating distributed/local or symbolic/subsymbolic as the whole story.
- Progression collapse. Equating representation shifts with formalization or closure.
- Alias capture. Letting
EncodingBasisor a similar local alias erase the factor bundle.
Consequences
The benefit is cleaner comparison across schools, substrates, and publication forms. The trade-off is the effort needed to make the relevant factor readings explicit.
Rationale
Separate representation factors let the reader compare code organization while assessing articulation and closure and establishing the required source or use claims on their own grounds. The starter triple earns its place through the three receiving questions in :4.1.1. A direct selection of other representation Characteristics is preferable when it answers the task with less irrelevant assessment.
SoTA-Echoing
The triple is an FPF starting choice for :4.1.1, with bounded external comparisons:
Relations
- Builds on:
A.18,A.19,C.2.2a,C.2.LS. - Coordinates with:
C.2.6,A.16.0,A.16,A.16.1,B.4.1,B.5.2.0,F.9for any Bridge and bounded-use claim, andF.9.1only for an optional stance note about that claim. - Constrains: language-state position publication and bridge loss notes around representation shifts.
Worked Examples and Factor Interaction Notes
Distributed but explicit
The pair code in :4.1 is distributed. Publishing its unit order and complete decoding rule can make its interpretation highly articulated without changing the code's locality. A symbolic wrapper with an unspecified decoder would leave that interpretation unresolved.
Symbolic but still low-articulation
A glossary-like note may be fully symbolic while still low in AE because the meaning needed by its receiving use remains unclear.
Mixed representation publication
An operator decision aid may display alert codes beside text explaining a back-end model. Assess the alert code, the text and the model-side code on their stated units. A text summary of a distributed model does not inherit the model's distribution merely by describing it. If one bundle reports several components, identify which readings concern which component.
Authoring and Review Guidance
Author prompt
What must the recipient distinguish, and which units and decoding rule are available to them? Select the questions below that bear on that use.
- how local or distributed is the representation?
- how sparse or dense is it?
- how symbolic or subsymbolic is it?
- which additional factor changes the intended comparison or use?
Review prompt
Check whether the reported factors support the intended comparison. Do not infer unreported factors from one label without a relation that establishes them.
Cross-facet reminder
Assurance readers should also watch for silent replacement of the source and use claims recovered through C.2.6, AE, or CD by representation talk.
Extension and Migration Notes
Local extension rule
Authors may add extra local factors, but each added factor should answer a distinct question rather than duplicating locality, sparsity, or symbolicity under another label.
Migration from alias-heavy prose
Aliases such as EncodingBasis should be unfolded into explicit factor dockings before they are relied upon for comparison, bridge claims, or downstream use.
Boundary reminder
LanguageStateRepresentationFactorBundle describes representational organization only. Admissible use, closure and source or interaction claims retain their own conditions and grounds.
Factor-Bundle Publication Discipline
Minimal representation package
A bundle claim should show the readings relevant to its stated use, using the starter factors and any declared additions. A partial bundle can suffice; identify an omitted or unresolved factor when it could change the comparison or continuation. An alias must not imply that an unreported reading is known.
No hidden scalar rule
Compact overlays such as "sparse-symbolic" are admissible only when they dock to the underlying factor bundle. No compact label may behave as a hidden master score for comparison, bridge comparison, or stage/progression talk.
Alias docking rule
Local aliases such as EncodingBasis are admissible only when their docking to the bundle's factors is explicit and stable. If an alias compresses several factors, the compression should remain visible.
Factor Interaction and Cross-Facet Reading Rule
Interaction rule
Representation factors may be related under a particular code or model. State that relation and its conditions before inferring one reading from another. Locality and active-unit fraction are distinct, as :4.1 shows. Publish the readings needed for the comparison; assess articulation independently under C.2.4.
Cross-facet non-substitution
Representation talk must not silently replace AE, CD, or the source and use claims recovered through C.2.6. A symbolic account of a distributed code may be easier to read while leaving interpretation unclear, closure open or an operator acknowledgement still required.
Bridge reminder
If a representation shift matters in transport across contexts, note that the shift may alter what is preserved or salient. Use F.9 for the Bridge and its bounded-use claim when the receiving use needs a semantic relation between local senses from different contexts; use F.9.1 only for a separate optional stance note about that claim.
Review Matrix and Extension Tests
Review matrix
An assurance reader should ask:
- are all claimed factors visible in the publication or cited source;
- does any alias hide the factor bundle;
- is one factor being used as if it summarized the whole representation state;
- has representation talk started to replace articulation, closure, or source and use claims.
Local extension test
An additional factor is justified only if it captures a distinct representational question that cannot be reduced to locality, sparsity, or symbolicity. The extra factor should extend the bundle, not become a master factor.
Migration test for source terminology
For a source term such as "symbolic", "distributed" or "encoding basis", retain its source-local meaning and scope. If an FPF comparison needs factor readings, establish which factors the source supports and leave the others unresolved. Mere citation of the source term needs no reconstruction into a complete bundle. Test any further cross-facet claim separately.
Bundle-comparison reminder
Representation bundles may be compared across contexts only after the compared factors are explicit. If one context uses a compact local alias and another publishes the full factor bundle, require explicit docking before treating the two descriptions as commensurable.
C.2.7:End
ExtractableStructuralInformation — Structure This Reader Can Recover
Type: Definitional (D) Status: Stable Normativity: Normative unless explicitly marked informative.
Problem frame
Use this pattern when an explanation, pattern text, architecture account or another expressed description must make selected structure recoverable by a reader or observing system. You may be comparing two forms of the same account, deciding whether an added explanation exposes a missing dependency, or checking what an observer can extract with the available preparation and time.
First useful move. Name the relation or distinction the reader needs, identify the available account and its expression, and state the preparation, source access and budget that can change recovery. Compare what the reader can correctly recover under those conditions. A useful result can be two sentences: “Both accounts expose the dependency on a durable write. Only the second account exposes why the reporting replica was chosen; this is an author estimate for readers familiar with replication, with the same source access and three-minute budget.”
The primary subject is extractable structural information: how much selected structure this reader or observer can correctly extract from this episteme through this expression under the stated conditions. Here an episteme is the claim-bearing account; its expression is the arrangement of signs conveying those claims. A reusable form supplies rules that different expressions can follow. Two texts can follow the same form and arrange the same claims differently.
What goes wrong if missed. Complete sentences or a shorter explanation can hide a needed condition. An expert may supply it from memory, so a correct answer can conceal what the account failed to explain. Comparing readers with different preparation or assistance can then make one text appear to expose more structure.
What this buys. One common structural question across forms and readers. The author can identify an additional recoverable dependency, preserve a sufficient explanation, or name the missing source or prerequisite.
Use an ordinary reading correction directly when the missing relation is already clear and no structural comparison is needed. Use the applicable task or learning evaluation when the live question is performance, burden, retention or transfer. This pattern supplies the structural-information characteristic those evaluations may consume; it does not select an explanation-design method or a universal quality score.
Problem
How can an author or evaluator compare structural recovery while distinguishing the account, its expression, the reader and the conditions of extraction?
A diagram may make dependencies visible without changing the described system. An added prerequisite explanation may change the claims as well as their expression. Two complete accounts may expose different structures within a short reading budget. A successful response can also depend on help that was unavailable for another response. The comparison needs to preserve these differences while retaining one meaning for structural amount.
Forces
Solution
Recover the structural question
Begin with what the reader must decide, do or explain. If the structure to assess is given, use it. When the selection is yours, work back from the required answer to the distinctions and connections that support it. Could readers recover every proposed relation and still miss a condition that changes their decision or its grounds? Include that condition and the connection through which it matters. A dependency may determine the order of two actions, for example, and a change of conditions may reverse a design choice. If readers must choose what to model, identify the reasoning they need to recover to choose the relevant parts and distinctions for themselves. A study of structural information may select a structure family for its research question without assigning task utility.
Identify the account and the expression being used. State only the conditions that could change the comparison: relevant concepts and notation the reader can use, available inference or inspection operations, accessible sources and help, and the effort or computation budget. A qualification or course title may locate evidence of preparation; it does not specify the usable knowledge by itself.
Then recover or estimate the selected structure and compare it with the relevant source or correctness criterion. Retain what each expression contributes, what remains missing and what a source return or prerequisite supplies. Reuse a sufficient current result. A new reading trial is useful when its possible outcomes could change the explanation or receiving decision.
Stop when that structural comparison answers the current question. If the selected structure or correctness basis is unresolved, return that gap before assigning an amount. If the practical question is whether a proposed change is worthwhile, use C.11.CRC and C.11 with extraction effort and task value stated separately.
Characteristic and arity
Tech name: ExtractableStructuralInformation. Plain name: structure this reader or observer can extract. ESI is an abbreviation after the full name has been introduced.
This is a relation Characteristic under A.17, with arity three over the ordered tuple:
Its aspect is the amount of selected structure correctly extractable from E through X by O under stated comparison conditions.
E retains its C.2.1 identity. X is an independently identified U.Entity: the sign construction that expresses E, identified under its expression domain's formation and equality rules. If reading order affects the selected operation, those rules must preserve order as part of X. O is the specified reader or observing system. For distributed extraction, identify the observing system and how its participants' contributions produce the recovered structure. An estimate about a reader class states the relevant preparation and variation among its members; it does not treat the class as another individual reader.
Where notation helps, write ESI_C(E, X, O). C fixes the selected structure, correctness criterion, relevant prior knowledge, usable operations, available assistance and tools, access conditions, budget and any other qualification that changes interpretation. These conditions qualify the characteristic of the tuple; they are not an additional bearer or a universal record kind.
E.24.PUB:4.1–4.2 distinguishes the expression X, reusable publication form F and carrier. When a comparison relies on F, identify its rules and how the selected X conforms to them. Several different expressions can follow the same F and yield different recovery results. A conclusion about a reusable form requires a stated selection of expressions and grounds for extending their results to that form. An unpublished draft can also supply X for an ESI comparison; identify the expression and available access directly.
Changing only expression can retain E. Adding, omitting or changing substantive claims can identify another episteme; use C.2.1 and the applicable source-to-target rule for that claim. In an explanation use, E.17.EFP's first screen distinguishes those branches when the difference matters. A comparison intended to isolate a change of form holds the expressed claims constant and identifies the expressions compared.
The presentation carrier belongs in the access conditions when rendering, resolution, searchability or availability affects extraction. Changing access can change the qualified result while E and X remain the same. Attribute that difference to the changed access conditions. Ordinary references to a page, diagram or file are sufficient when they identify the required account and expression without ambiguity.
Correct recovery and the expression's contribution
Correct recovery preserves the selected relation's participants, predicate, polarity, modality and action-changing conditions under the declared source or criterion. State the unit of structure only as precisely as the comparison needs. A qualitative result can name additional and missing dependencies directly; a count must also define its granularity.
Repeated expression of the same relation does not add structural amount merely by increasing words, boxes or arrows. A count of edges in a representation needs a correspondence to the selected semantic relations before it can count those relations.
Keep two questions distinct when the source contains a false or unsupported claim. A reader may correctly recover that the account asserts that claim. Establishing the claim about the described world requires its subject and evidence criteria. For recovery of warranted source structure, an unsupported relation does not count as a correct member. Recovering the assertion and endorsing its content are different results.
Prior knowledge may supply vocabulary, notation and inference used to decode the expression. A reader who independently corrects false text from memory has supplied a correction; the text does not receive credit for that corrected relation. When attribution matters, preserve the initial response and identify actual assistance and source returns before interpreting the observed difference. An explanation can expose already familiar structure: ESI is not restricted to newly learned information.
Positive direction and scale choice
More selected structure correctly extractable under comparable conditions is the positive direction of ESI. If one result includes every correctly recoverable selected relation of another and adds selected relations, it has the positive structural direction. If each exposes something the other does not, retain that trade-off unless a justified scale resolves it.
A scale can be ordinal, a count, a fraction, a nonnegative structural weighting or a bit estimate. Its meaning and method must fit the declared use. A fraction has a fixed denominator; a weighted amount has an explicit structural interpretation. Changing the selected structure, granularity, denominator or weights changes the comparison basis. A task-utility weighting defines a separate score rather than silently becoming structural amount.
An ordered ESI scale declares the positive direction under A.17/A.18. A numeric comparison uses only operations legal for that scale. C.16 supplies the measurement chain when a measurement is claimed: the tuple and measurand, scale, method and model, obtained result and relevant uncertainty. Qualitative comparisons and conditional design estimates remain available without a validated human bit estimator.
A missing observation, unspecified denominator or unresolved criterion is missing basis for a value. It is not an observed zero. A zero has meaning only under a defined scale and a result that supports it.
Reading results, estimates and robustness
A design walkthrough, an actual reading and a formal-model estimate can concern the same characteristic. State which grounds the assertion and the conditions at which it applies.
An observed response establishes the structure recovered in that trial and can ground an estimate of what is extractable. One success does not establish a maximum or reliability for a population; one failure does not establish impossibility. A computational procedure can support an exact bounded result when its operation and selected structure make that result derivable.
Repeat or perturb a reading when the receiving use needs a claim about robustness or transfer. Preserve the first response before giving a changed condition or further help. Success under the new condition strengthens the corresponding changed-condition claim; it does not add structure to the original result by itself.
For a cold-reader check, obtain the instruction from the publication, its declared prerequisites and the knowledge presupposed for the intended reader. The task may supply case facts. A prompt can name a relation and ask readers to find where it holds or apply it to a decision. To assess whether readers can select the relevant relation, leave that choice to them. If the prompt supplies an explanation missing from the publication, the response shows what readers could do with that additional help. Use F.19 to repair the publication before claiming that the publication itself supplied the action.
Absence of human validation restricts empirical claims about people. It does not restrict ESI to computational observers or prevent a conditional author estimate about a human reading situation.
Relation to epiplexity and MDL
A formal structural-information estimate is useful when a mathematical model can represent the selected account, observer and extraction question. Establish that correspondence and what structure it preserves or loses. Use C.29 for an unresolved question about choosing or transferring the representation, or relying on it beyond its supported correspondence.
Finzi et al.'s computational epiplexity selects a time-bounded probabilistic program by a two-part coding criterion. Writing D for the model's data variable gives:
Ties select the shorter program. The model term describes structural information; the residual is time-bounded entropy. Conditional versions allow side information, and practical estimators have their own assumptions. The execution bound and the effort of finding or estimating a model are distinct costs. Finzi et al., v2, §3–4
For an ESI estimate, explain how the episteme E and expression X map to the data D, how the model class represents the observer's available operations, and which selected structure the model term estimates. State how prior knowledge is represented. Conditional model description given side information need not count all familiar structure a reader can recover.
Use epiplexity as a formal specialization where that correspondence holds. Otherwise it motivates the structural-information question without supplying its value. Total code length, text length and arbitrary model size are not interchangeable ESI measures. Ordinary description comparisons may use an adequate domain method without a coding model.
Articulation, effort and useful change
Articulation explicitness (AE, C.2.4) orders progression from a cue to explicit branch-appropriate meaning and stable receiving use. ESI compares the selected structure extractable by an observer under a budget. Two complete descriptions at the same AE level can have different ESI; a reader's difficulty alone does not establish that the semantic branch was left unarticulated.
Measure or compare extraction effort separately. Two expressions may expose the same structure at different cost. Additional selected structure may also be irrelevant to the next task or displace more valuable work. C.11.CRC/C.11 make the marginal choice using the structural result, cost, protected results and receiving value. The chosen result may be the current sufficient explanation.
Use the method appropriate to the proposed change: C.37 for selecting representations, A.6.3.NAR for narrative ordering and source carry-through, or the applicable domain description or instructional method. ESI states the aspect being compared; it does not prescribe adding a diagram, example or lesson.
When the account describes a Method, carry the recovered operations and relations into A.2.2:2.1 if the next question is which of them the holder can use. Keep the extraction result about the episteme, expression and observer: correctly explaining a rule supplies different evidence from performing it with specified support.
E.23.CDI:4.2.1 uses the Method-relative profile to choose learning or provision and to revise an acquisition arrangement when its feedback and receiving result disagree. Stop at the ESI result when structural recovery alone answers the question.
Archetypal Grounding
Architecture account in prose and a diagram
A service account states:
The API returns “stored” only after the Store confirms a durable write. Reports read a Replica updated from Store and may lag by five minutes. A synchronous report query to Store was rejected because report bursts must not delay request writes. If required report freshness becomes one second, reconsider the replica update design.
A second expression adds this diagram while retaining those four sentences:
The reader is an engineer familiar with durable writes and replicas. Select these four relations:
- the acknowledgment dependency;
- the report-source/lag relation;
- the rejected alternative with its reason;
- the reconsideration condition.
Both expressions expose the same claims; the described architecture is unchanged. Compare recovery with the same three-minute budget and source access. The diagram's presence alone establishes no advantage.
Suppose a reading trial recovers three selected relations from the prose and all four from the combined expression. Report that bounded observed difference and identify the additional relation. If a source paragraph was supplied only before the second reading, the comparison also changed assistance. If the same reader encountered both forms, previous reading may contribute. These conditions prevent attributing the whole difference to form alone.
A reader unfamiliar with replication may need a prerequisite explanation. Adding its substantive claims changes the account as well as its expression. If the design requirement changes to one-second freshness, the account supplies reconsideration of the replica update design. It does not supply a replacement architecture. C.33 can record this captured structure and the source return needed for the next design question.
A technical pattern's explicit plan
The plan in C.2.4:13.2 names Maintenance Team 2, Pump P-17, isolation, seal replacement and restoration only after LT-9 passes. For its planning use, the actor, action order and condition are explicit enough for AE4. A maintainer can recover the restore-after-pass dependency even when the note does not explain the LT-9 procedure.
If the selected use is carrying out that test, the applicable procedure must be available as a prerequisite or source return. The plan's articulation result does not teach the missing procedure. The useful ESI comparison names which required test relations the reader can recover with the available source access; it does not replace the planning judgment with a larger or smaller AE number.
An observer with a two-statement budget
Let E describe edges A→B, B→C, A→D and D→C. Expression X1 lists them in that order. Expression X2 lists A→B, A→D, B→C, D→C. Here statement order is part of expression identity. Both expressions conform to the same reusable edge-list form F: one directed-edge statement per line, in any order.
Observer O reads the first two edge statements and returns every two-edge path supported by those edges. Select the paths A→B→C and A→D→C as the structural denominator. X1 gives O the edges needed for A→B→C; X2 gives two edges leaving A and no complete selected path. Under the declared count, the results are 1 and 0 out of two selected paths. With four statement reads, the same procedure recovers both paths from both expressions.
The result follows from the stated algorithm and budget. It compares two X values at the same E, F and O, showing why selecting F alone leaves the supplied material unresolved. This is a bounded computational derivation; a claim about human comprehension or a Finzi estimate needs its own corresponding basis.
A whole-framework architectural rationale
A framework maintainer has several individual pattern-quality results and needs to decide what they establish about the framework as a whole. E.4.FPF:3–4 explains the different questions: E.21 evaluates a pattern; E.2.DA evaluates whole-FPF Pillar adequacy. The selected structure is this relation between the two scopes and the corresponding return.
For a reader who can already use those evaluation questions, the existing rationale and direct links may be sufficient. A less prepared reader may need the questions explained before the links are useful. An ESI comparison checks recovery of that scope distinction and next return under stated preparation and access. Counting the account's headings would not answer it.
Bias-Annotation
Selecting only easy or already familiar relations can make an account look strong while excluding what the receiving use needs. Derive the selection from that use or a declared research question before comparing expressions, and retain consequential losses.
Author familiarity can hide a prerequisite or make a correct expert response seem attributable to the text. Preserve actual help and source use when that distinction affects the conclusion. For a reader class, state relevant variation rather than treating one prepared reader as representative of everyone.
A coding model can privilege structure useful to its model class. Keep that selection and its correspondence visible when applying the estimate elsewhere.
Conformance Checklist
Use this checklist when a characteristic claim or comparison must be relied on or carried to another use. An ordinary local comparison may keep the same information in prose.
Common Anti-Patterns and How to Avoid Them
Consequences
A shared characteristic lets an author compare forms without confusing them with the account's identity or the described architecture. It also lets a domain evaluation consume structural recovery while keeping its other results intact.
The practical cost is making material preparation, selection and access conditions explicit. A measurement or a causal claim about an explanation's effect needs grounds appropriate to that claim; reuse an adequate existing basis.
An additional example, prerequisite, diagram or changed order is useful when it exposes worthwhile structure under the conditions. Sometimes the current expression already supplies the needed structure. Both outcomes are compatible with this characteristic.
Reconsider the qualified result when the account, expression, applicable form rules, selected structure, observer preparation, available operations, help, carrier access, budget or correctness basis changes. Reuse an unchanged result for a matching use.
Architectural Rationale
The three-bearer definition explains why the same account can support different structural recoveries. The episteme identifies the expressed account, the expression identifies the sign construction available for extraction, and the observer identifies who or what can extract the structure. Reusable-form conformance can hold for several such constructions; the two-statement case shows why those constructions need separate recovery comparisons. Qualification conditions retain the resources and access that make each comparison meaningful.
Articulation explicitness, structural amount and task adequacy answer different questions. A plan can state its action condition completely while its reader lacks the prerequisite procedure. A diagram can expose additional relations while remaining insufficient for a design decision. Keeping these results separate preserves both the useful gain and the next question.
A common definition also allows unlike estimators without claiming universal commensurability. A path count and a justified coding estimate can address structural-information questions; their values are comparable only under an adequate scale and correspondence. The source studies therefore inform the aspect and its conditions while domain methods establish the result.
SoTA-Echoing
The computational line in Finzi et al. supplies a bounded-observer structural-information model. Adopt its observer dependence and separation of structural amount from downstream task relevance; adapt its formal use through the correspondence in :4.6. A computational estimate does not by itself establish human reading or learning effects. Finzi et al., v2
MDL supplies the historical model-selection background: compare a model's description with the data description it permits. A small adequate model may explain rich structure, so minimizing total code length and comparing extractable structural amount must retain their different questions. Grünwald's MDL account
For human explanation, current preparation-sensitive research supplies a different constraint. The 2025 expertise-reversal meta-analysis supports treating assistance and prior knowledge jointly, with variation by population and domain. Adapt that result as a requirement to qualify a relevant reader comparison, not as an ESI scale. Tetzlaff et al.
The two 2025 biochemistry populations showed different near-transfer sequencing effects and no significant sequence effect on far transfer. The useful implication is to identify the knowledge and task conditions tested rather than infer a universal instructional sequence from a general expertise label. He, Fiorella and Lemons
At comparable effort, ordinary source-based reconstruction often answers which needed relation became available. Use it for that question. Select a coding estimate, repeated reading or transfer observation only when its distinct possible results can change the claim or next action. Reopen a model correspondence or empirical generalization when new evidence contradicts the qualified use.
Relations
- A.17/A.18 and C.16: arity, characteristic/scale legality and the measurement chain for a measurement claim.
- C.2.1 and E.24.PUB: identity of the expressed episteme and the distinctions among its expression, reusable publication form and presentation carrier.
- C.2.4: degree of explicit articulation and its receiving-use thresholds.
- C.29: choosing or transferring a mathematical representation, or resolving reliance beyond its supported correspondence, when that question remains for a formal structural-information estimate.
- A.2.2:2.1 and E.23.CDI:4.2.1: Method-relative capability and the acquisition or provision choice that can consume a recovered Method structure.
- C.33: captured and missing structure, sufficiency and source return for an architecture use.
- A.6.3.NAR and E.17.EFP: narrative construction and explanation-faithfulness uses that consume structural recovery.
- C.11.CRC/C.11 and C.37: marginal comparison and choice, and selection of useful representations.
- F.19 and A.19.ECS: precise recoverable instruction, and construction of a domain evaluation that may use this characteristic.
C.2.8:End
Repair Claims Based on Declarative Form Alone (Declarative Representation Precision Restoration)
Type: C.2.P precision-restoration child pattern for declarative-representation overread Status: Stable Normativity: Normative unless a section is explicitly informative
Use this when
Use this pattern when the appearance of a representation is being used to establish an operational or classification claim without its required grounds—for example, treating a green dashboard tile as release permission or a graph path as prescribed work.
First useful move. Identify the expression and the claim being inferred from it. Recover what the expression states or represents, then apply the rule for the disputed claim. Return wording that preserves the supported use and identifies what remains unestablished. A sentence can suffice. Use a DeclarativeRepresentationRepair note only when the receiving use needs a separately inspectable repair.
Quick example. A heat-flow graph can express which flow paths a design retains. To decide whether someone may change a valve, check whether the applicable rule permits that person to do so under the current conditions. The graph can still answer the flow question while permission remains unknown.
Use this pattern especially when:
- a graph path or selected flow structure is treated as a prescribed work sequence;
- an
A.10evidence path is treated as permission, release or assurance; - a query plan or program-like expression is treated as a method or as performed work;
- a publication or pattern reference is treated as an actual call, send or application;
- method-like wording leaves the way of doing, its description and an actual episode indistinguishable.
What goes wrong if missed. A reader relies on an effect that the expression does not establish. The green tile may report a status while the required release decision is absent; program-like text may be mistaken for a method without identifying a way of doing.
What this buys. The reader can retain a useful graph, query or publication while resolving the disputed operational claim under its own rule. An object under direct inspection stays distinct from a representation of that object.
Not this pattern when.
- If the graph path,
PathSlice, or flow valuation is already current as graph structure, useE.18directly. - If the evidence-use or status-use classification of an episteme is already the question, use
A.2.4. UseA.10when source recovery, provenance, currentness or bounded reliance must remain replayable. - If the publication face or source-use relation is already current, use
E.17,E.17.EFP,C.2.P, or the direct publication pattern. - If the current claim concerns a semantic way of doing, use
A.3.1; if it concerns the description of that way, useA.3.2. - If the question concerns a reusable operation declaration or its applicability, use
A.6.1. UseE.20when an FPF introduction or revision triggers its mechanism-introduction protocol. - If the current claim concerns planned work or dated work, use
A.15.2orA.15.1. - If the word is only quoted source wording or ordinary navigation prose with no FPF-governed claim, keep it quote-only or ordinary.
Problem frame
A graph can itself be the object under inspection, or a drawing can represent a graph. A table can express claims about a system, or its layout can be the publication form being examined. Recover that distinction before deciding which rule governs the disputed use.
The recurring failure is an unsupported shift from appearance to operational effect. For example, an arrow in a graph can represent a relation without prescribing an action or proving that it occurred.
This pattern repairs that shift. It does not build a general theory of representation. It only restores the exact FPF-governed object, relation, or claim and its subject pattern when declarative-looking wording is overread as imperative or operational.
Problem
Without this repair:
- Graph path becomes work route. A path or path slice in
E.18is treated as an ordered work narrative, even when no work occurrence, work plan, or method description is current. - Evidence path becomes permission. An evidence relation or provenance relation is treated as approval, gate passage, release, safety, or authority rather than as evidence for a named claim or effect.
- Query becomes method. A query, access path, query plan, or dashboard is treated as the semantic way of doing, rather than as a representation, method description, evidence relation, source relation, or ordinary source wording.
- Pattern relation overread as dispatch. A reference to another pattern is treated as an actual call or work occurrence. Ordinary navigation remains adequate when it expresses the intended relation clearly.
- Programming-paradigm label becomes ontology. A label such as functional, declarative or workflow is treated as sufficient to determine an FPF kind.
- Mechanism, method, and work collapse. A method-like expression is repaired to
methodormechanismby vocabulary rather than by the current claim: way of doing, description, formal substrate, law-governed mechanism, plan, occurrence, evidence, or quote-only wording.
Forces
Solution
Repair declarative-representation overread by separating the visible expression, the direct object or relation, and any representation use before naming the subject pattern for the current claim.
The repair order is:
- Read the expression under the conventions used to produce it. Identify the filled row, solver constraint or other construction and recover the assertion it expresses. Use the column definitions, bindings, units or interpretation that determine its meaning. If the same marks admit two readings with different consequences, obtain the convention needed to choose between them or leave that claim open.
- Recover the object or relation at issue. For example, distinguish a graph path, an evidence relation and a reported state value. Use its subject pattern when that rule is already clear; the selection table below helps when it is not.
- Distinguish the direct object from any representation or correspondence use. When the visible expression represents a separately identified object or claim, name the exact relation and target. When the direct object or relation itself is current and no separate representation claim is needed, keep that direct use; do not relabel the direct object as a representation kind. Record
noneonly when the receiving use needs an inspectable account of that distinction. - Recover a source or publication relation when it matters to this claim. Use the pattern governing that relation and its actual participants.
- State the disputed inference. Identify what someone would classify, conclude, plan, carry out or authorize on the strength of the expression.
- Apply the claim's subject pattern. Establish its required grounds or identify what is missing. An actual permission or gate condition comes from that rule.
- Preserve the supported use. The graph may still support a comparison, for example, while leaving action permission unresolved.
- Delimit the unsupported claim. Do not assert the stronger result from appearance alone. Distinguish that missing basis from a use blocked by an applicable rule; a separately established result remains usable under its own conditions.
- Stop or reopen. Stop when the repaired statement supports the intended use or the subject pattern supplies the remaining question. Reopen when a changed source, correspondence or intended use can invalidate that result.
DeclarativeRepresentationRepair note
Use this compact note only when the receiving use needs an inspectable repair of FPF-governed wording:
The ordinary result is the repaired statement and any unresolved condition that changes its use. In the optional note, give the disputed inference and say what the available grounds establish. If a classification remains open, identify the missing criterion or fact. If an applicable rule prohibits an action, identify the rule and the condition that triggers it. Reuse a suitable subject-pattern record without copying it into another repair note.
Use four plain questions before the claim-and-pattern table: What visible thing am I looking at? What direct object or relation is current? What, if anything, does it represent? What further conclusion is being drawn, and do the interpreted content and governing rule establish it?
For a question about format, inspect the formation rules used for the constraint or row. For a question about access, identify the file or other carrier bearing that particular expression.
Subject pattern selection
Dated Work: admission, representation and attribution.
Recover every actual performer and its obtaining A.2.1 system-role assignment through A.13, then let A.15.1 U.Work independently identify the dated occurrence, enacted Method, time, and containing System.
A representation claiming the exact Work keeps every actual performer named or recoverable. Keep the underlying assignment facts recoverable; include an assignment identifier in the representation only when the receiving use needs it.
Add an F.6 relation through that same obtaining assignment only when the representation or receiving use expressly represents precise assignment-bound attribution. Missing or failed F.6 attribution alone leaves the Work intact while its independent A.13/A.15.1 grounds still hold and blocks only that attribution.
Legitimate path and route settlement
A.10 evidence path for <claim, effect, or use> is legitimate when a descriptive source-to-use account is current under A.10. E.18 graph path and PathSlice are legitimate when the graph object, path, slice, crossing, or flow valuation is current. Carrier file paths, URLs, mathematical paths, and quoted source paths are legitimate when their notation, source-use function, or use relation is current.
Repair path wording when it carries an unsupported claim about what happens or what may be done. A clear ordinary reference or navigation instruction can remain unchanged.
For a route or unfolding expression, identify the object before selecting its rule:
Recover the relation the route expression conveys. Direct inspection of a graph object needs no separate represented target or lens account. If a representation or selected structure supports an inference, establish its target and the preservation or loss relevant to that inference under its subject pattern. Use C.29 for a separate mathematical representation or transfer issue; adequate local mathematics or an unambiguous rendering can suffice without a C.29 record.
Method, algorithm, mechanism, plan, and work settlement
For method-like wording, recover the repeatable way of doing and distinguish it from the plan for a particular job and the Work that occurred. A rule for ordering jobs can be applied to many queues. Tomorrow’s selected order can form part of a plan, and the execution record can show a different order. When two expressions serve as models, compare them for the same inquiry under B.5.MR. A change of notation can preserve every admissible answer. A newly chosen method can change the way work is done without establishing refinement between any two models.
Method-description membership. First identify the claim-bearing episteme under C.2.1. Use A.3.2 to determine whether that same individual is a U.MethodDescription. Positive membership requires one admitted U.Method as its exact EntityOfConcern and at least one substantive claim about that Method as a way of doing. Such a claim can state its transformation or enactment concern, applicability, precondition, intended effect or preserved condition, bound, generic participant meaning, or internal composition.
When either condition is false, membership fails; when the available basis does not settle them, leave membership unresolved. Representation form, execution or approval alone does not supply these conditions. Keep any independently established representation, publication, plan, Work, result, formal declaration or evidence use under its own pattern.
A representation or publication change does not decide membership. If claim content, exact subject or effective reference scheme changes, identify the resulting episteme under C.2.1 and apply the test to that individual.
Recover what the source is about and what it asserts:
Cooling contrast. A reusable cooling procedure can be U.Method only after the context-local way of doing, its transformation or enactment kind, transformed referent or structure, preconditions, and intended effects are recovered. “Required cooling effect” alone is claim content, not a method. If a later cooling episode changes the governed loop state, that occurrence remains a separate A.3.4 U.Transformation and needs its own changed referent, boundary, conditions, facts, and continuity or reidentification basis.
When the source label hides method, mechanism, formal-substrate, work, evidence, gate, result, or temporal claims, use E.10.ARCH:3.1 to state the project concern in ordinary words, then identify each exact object and claim separately. Use this pattern to repair only the representation overread and name the subject pattern for the current claim; linked values remain under their own subject patterns rather than becoming one representation-repair claim.
Programming-paradigm and process-model settlement
A paradigm label helps locate the relevant source semantics. Those semantics, together with the claim, determine which FPF question to ask.
The source distinctions below help identify what an expression is being used to claim:
- constructor and process theories distinguish an abstract task from the physical processes proposed to implement it; recover whether the account specifies a transformation or asserts its performance;
- scoped-effect calculi distinguish an operation, its handler and the rules governing scope and effects; recover which object and semantic relation the current claim concerns;
- Declare constraints restrict admissible traces, while object-centric logs record events and object relations; data-ingestion and transformation pipelines are another subject, so identify which of these the account describes;
- e-graphs and monoidal rewriting represent equivalence and composition; inspect that relation before treating an edge or rewrite as execution order.
Use the distinction relevant to the case to recover the object or claim, then apply its subject pattern.
Worked slices
Graph path in a transformation-flow structure
Wording: "The P2W path routes the team from principle to work."
Repair:
Evidence path near release
Wording: "The evidence path authorizes release."
Repair: name the claim or bounded use and recover its grounds under the direct subject pattern. Add the descriptive A.10 source-to-use account when source recovery, provenance, currentness or bounded reliance must remain replayable. Release, permission, or gate passage requires the authority, gate, or release pattern that defines or constrains that claim. This pattern is used only if path wording itself is causing the representation to be overread as a permission route.
Query plan and access path
Wording: "The query plan calls the production work sequence."
Repair: recover whether the query plan represents optimizer choices, expresses claims about an exact method, presents a formal substrate, supplies a source cue or evidence relation, states a work plan, or records an actual query run. If it only represents query-evaluation choices, stop at the representation.
Use A.3.1 for a reusable semantic way-of-doing claim. Use A.3.2 when MethodDescription membership or a use of the method claims is the question; apply the test in 4.4 and retain its positive, negative or unresolved membership result. Use A.15.1 for a performed query run, together with the exact evidence or source-use relation when that later claim is current.
Dashboard predicate
Wording: "The dashboard green path lets the release move."
Repair: recover dashboard face, source relation, status or state bearer, value frame, source currentness, and gate or release claim. The dashboard may be a publication face and source cue. Recover release permission through the applicable gate decision (A.21) or authority relation, including any source contribution required by its governing rule.
Pattern relation
Wording: "This pattern exits to A.10."
Repair: if the current relation is "use A.10 when the source-to-use account must remain replayable", write that declarative boundary. Keep ordinary ownership or navigation wording when its relation and participants are clear. Use exit, receiver, route, owner, home, dispatch, or call language to assert action or control only when the pattern is about an action occurrence, work plan, control mechanism, or communication relation that has those semantics.
Solver algorithm
Wording: "The solver algorithm is the mechanism."
Repair: first identify the direct object or relation, any representation use, and which claim is current. A solver configuration may represent claims carried by a separately identified episteme; apply A.3.2 when MethodDescription membership is the question. File form or executability alone does not establish that membership.
The reusable semantic way of solving may be U.Method; the MILP formulation may expose a formal substrate; a declaration episteme for a reusable operation family, with its required laws and applicability, may qualify as U.Mechanism under A.6.1; a solver run may be U.Work; and a run result may support another claim through its direct evidence relation.
Use A.6.1 to recover and test the proposed declaration. Use E.20 when an FPF introduction or revision triggers its protocol.
Reactor-cooling flow graph
Wording: "The preserved heat-flow path authorizes the valve change."
Repair:
CRISPR guide-selection table
Wording: "The guide-selection table approves the edit."
Repair:
Conformance checklist
Common anti-patterns
Relations
- Builds on:
E.10,E.10.ARCH,C.2.P,A.7,E.17,E.8, andF.19. - Coordinates with:
E.18,E.18.1,A.2.4,A.10,A.19.SPR,A.3.1,A.3.2,A.6.0,C.29,A.6.1,E.20,A.15.2,A.15.1,A.15.4,A.20,A.21,B.3, and direct publication, gate, authority, release, evidence, method, work, and assurance patterns when those claims are made. - Specializes:
C.2.Pfor one recurring case: declarative representation and imperative-metaphor overread. - Used by:
E.10.ARCHapplicability-row distribution and full-pattern text precision restoration when route, path, workflow, call, or dispatch wording hides representation use.
Consequences
- FPF can keep useful graph, path, query, table, dashboard, publication, and pattern-relation vocabulary without banning ordinary words.
- Operational claims receive their grounds and conditions; a supported comparison remains useful while a stronger claim is unresolved.
- Method and mechanism claims become easier to compose because
E.10.ARCH:3.1keeps separately recovered values connected only by the exact direct relations and participant meanings supplied by their subject patterns, without treating source labels as alternate ontology. - The cost is this recovery, with a small note only when the receiving use needs an inspectable repair. Ordinary navigation, source quotation, and already-governed graph or evidence paths do not need the note.
SoTA-Echoing
The sources below support the stated distinctions. The final column gives their use in this FPF repair.
C.2.P.DR:End
Kinds, Intent and Extent, and Typed Reasoning
Type: Typed reasoning discipline pattern Status: Stable Normativity: Normative unless a section is explicitly informative
Use This When
Use this pattern when a claim needs a reusable kind, a subkind comparison, a judgment about whether one exact candidate satisfies one kind, or an optional representation of the candidates that satisfy it in one exact context slice. A kind may be used locally without receiving its own public U.* name; “local” describes the bounded use, not an identity component.
What goes wrong if missed. A source type, practice label, programming class, schema label, mathematical set, or public U.* name starts doing several jobs at once. A source boundary splits one unchanged kind; several kinds inside one source collapse; the kind is confused with its declaration; evidence is treated as membership; a non-applicable request becomes unknown; or a current extension becomes ontology.
What this buys. A practitioner can classify an admissible candidate under a known kind without confusing the kind, its declaration and the judgment. A represented set of current members remains available when a query needs it; the conditions being classified retain their own rules.
Primary EntityOfConcern. One typed-reasoning question: the exact U.Kind individual, its intended candidate domain and membership distinction, any U.SubkindOf comparison needed by the claim, and the C.3.2 candidate question the use asks. The exact KindSignature edition makes its effective designation and interpretation rules recoverable; the scheme and practice/source provenance are not stored on the kind.
First useful move. Write the ordinary conclusion first. For example: For this plant’s operating report, a cooling pump is a pump circulating coolant through the heat exchanger during the reporting interval. The circuit connections and the interval’s flow reading establish that Pump #14 did so. Add a reusable declaration, admissibility detail, explicit judgment, support reference, or extension representation only when a named receiving use needs it.
Not this pattern when. Use E.24.UK when the question is admission of another durable public FPF U-kind. Use the direct subject pattern when the question is whether a physical quality, relation, registration, certification, publication occurrence, Work, or other governed condition obtains. Use A.2.6 for claim, work, or publication scope and C.29 when a separate mathematical representation or transfer question remains.
Problem Frame
U.Kind is the admitted meta-kind whose individuals are reusable intensional classification distinctions. One kind individual is recovered by its declared candidate domain, the membership condition that distinguishes intended members from non-members, and the continuity rule for a material declaration change. A KindSignature states that content for repeated use but is not the kind itself. A current extension can change while the kind continues, and two different intensional kinds can happen to classify the same current candidates.
A practice, source, team, or locality tells a reviewer where meaning may have changed. It does not decide kind identity. When a typed use moves, compare the exact membership distinctions. Reuse the same kind when the candidate domain and operative distinction continue. If they differ, identify two kinds; only then can C.3.3 ask whether an exact directional KindBridge obtains.
When local wording or interpretation also differs, F.9 may relate the corresponding F.17 cells, but it neither creates the kinds nor maps a reference scheme as a whole.
A change to the declaration's effective reference scheme identifies another episteme under C.2.1. Judge its U.Signature membership under A.6.0 and its KindSignature qualification under C.3.2 separately; call it another edition only when the C.2.1 EpistemeEditionRelation obtains. C.3.1 separately decides kind continuity. A changed U.ContextSlice alone creates neither a kind nor a bridge.
Problem
A project often needs classification before it needs another public ontology name. If the kind, its definition, the classified candidate, a record about the candidate, and a displayed set of current members are treated as one object, a label classifies by itself, evidence availability is mistaken for criterion satisfaction, missing information proves non-membership, a table becomes an entity set, or a plan row becomes actual Work. If locality is made an identity key, the same kind also fragments across teams and sources. C.3 keeps each conclusion at its direct pattern.
Forces
Four Objects and a Pre-judgment Check
Keep these four objects separately recoverable:
Before the judgment, C.3.2 returns admissible or not-applicable. Candidate mismatch with the declared ValueKind, or a slice outside declared applicability, is not-applicable and no three-valued judgment is formed. For an admissible candidate, return unknown when the available grounds leave the criterion unsettled.
The kind does not carry scope. A KindSignature episteme may carry its own U.ClaimScope, and a separate classification assertion carries the scope of that assertion. The U.ContextSlice is an evaluation input.
Solution
Use the lightest object that answers the current typed-reasoning question.
Identify what the term must pick out for the work the recipient is doing. “Repair this stool” designates an individual item. For this catalogue, a stool is a seat designed for one person, with neither back nor arms. Those features settle classification under the agreed rule. For a delivery count, establish what counts as one delivered unit. If a complete delivery contains ten seats and exactly two fall outside the stool criterion, the total is eight. Naming those eight also requires grounds for their identities. A furniture specification can concern a design used for many units. A product family groups models on a stated basis. An individual, a design and a kind have different identity criteria. A catalogue row or a plural noun may refer to any of them. Establish what it denotes before using it as evidence of membership.
When choosing the classification is your task, work back from the receiving use. A buyer ordering replacement fittings may need to distinguish two catalogue stools that take different fittings and group a compatible chair with one of them. Try cases on both sides of the proposed distinction. Keep the conditions that change the order or its required justification. Use the existing catalogue kind directly for a request stated in its terms. F.19:4.3.1 explains how to retain the distinction in working prose.
- Recover the kind. Name the candidate domain and the operative membership distinction: what an intended member must satisfy and what separates a relevant non-member. Record the continuity rule used when that distinction changes. Keep practice/source provenance as a cue to compare definitions, not as an automatic identity key. Do not store the current use, ClaimScope, context slice, or reference scheme on the kind.
- Use C.3.1 for subkind and continuity. A
U.SubkindOffact obtains through exact criterion entailment under an aligned interpretation or through exhaustive evaluation over a deliberately closed finite domain. The facts form a preorder. Opposite facts between distinct kinds may express classification equivalence for that applicability; a consumer may order the resulting equivalence groups without identifying the kinds. - Use C.3.2 for declaration and admissible judgment. A repeated condition may justify a
KindSignature. First check candidateValueKindand applicability. Only an admissible application returnstrue,false, orunknown. - Let the governed criterion condition decide. A direct quality, relation, construction, episteme, registration, certification, publication occurrence, legal status, or other governed condition makes the criterion hold when the criterion names it. An observation, record, or source used merely as evidence does not constitute an independently governed condition. Use each condition's direct pattern.
- Keep four outcomes distinct.
not-applicablemeans the judgment should not be formed. For an admissible candidate, a satisfied criterion givestrue, a known failed criterion givesfalse, and an unsettled criterion givesunknown. Missing facts elsewhere in a compound criterion leave an independently settled result intact. A guard may decline use without rewriting any of these results. - Materialize an extension only for use. A query, quantification, comparison, or review may need
KindExtension(k, slice). It represents admissible candidates judgedtrue; notation, rows, or set membership do not create an ontic collection or classification relation. - Keep scope, formality, Work, and publication separate. Formality characterizes the declaration episteme. Scope belongs to claims or capabilities.
U.Workis a kind andW : U.Workis one independently grounded dated work occurrence. Plans, logs, cards, field bundles, carriers, and rows remain their own objects.
Typed reasoning composes with F-G-R and USM in this order: recover kind compatibility; check claim-scope coverage; when the receiving use is candidate-bearing, check classification admissibility and, when admissible, the exact judgment; then apply support, assurance, freshness, and any justified bridge consequence required by the receiver.
Decision Split
When a constructive trace or working model and a target kind or logical representation must be inspected together, explain their correspondence.
The account identifies the source and typed or logical target, what the passage preserves or loses, and the supported reuse with its substitution limits. Name an F.9 Bridge or C.3.3 kind correspondence when the use claims or consumes one. Supply CL or CL^k when the consumed assurance or calibration account requires it. Evidence and proof retain their own subject patterns; a changed slice or notation alone supplies no Bridge.
B.3.5:4.1 provides an optional block for recording that account. C.3 supplies the kind, admissibility, judgment, subkind and kind-correspondence content needed. Use A.22 when the receiving question needs a selected structure, and A.22.CGUS for a remaining question about potential continuations and their conditions; test any required formal qualification under that pattern. For general diagnostic recovery of a subject construction from an inadequate account, use A.7.1.
Archetypal Grounding
Bias-Annotation
C.3 counters lexical, locality, document, and ontology-growth bias. A familiar word, source label, or practice boundary supplies neither kind identity nor membership. A record used as evidence does not create an independently governed condition, while an episteme, status, or relation directly named by the criterion keeps its own governor. The kind/declaration/admissibility/judgment/extension split and the readable first move keep the remedy usable.
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
- Treating a programming type, schema class, source ontology class, regulatory category, or ordinary noun as a durable public FPF U-kind.
- Treating a
KindSignatureas the kind, or attaching its formality and claim scope to the kind. - Using a world-side belongs-to predicate or minting a classification relation merely to state one judgment.
- Treating evidence availability, a schema row, or a publication form as sufficient for classification when the criterion requires a different governed condition.
- Returning
falsewhen the criterion cannot be evaluated. - Treating
KindExtensionor mathematical set notation as ontology. - Repairing a subkind counterexample by silently changing an extension table.
- Treating a plan or work record as a dated work occurrence.
Consequences
Benefits. C.3 supports local typed claims, subkind reasoning, classification, and queryable extensions without premature ontology growth or evidence-created membership.
Costs. Reliance-bearing uses must recover the kind distinction, pin the declaration and slice, check admissibility, and keep not-applicable, false, and unknown distinct.
Rationale
The kind, its declaration, pre-judgment admissibility, one classification judgment when admissible, and a representation of current true members answer different engineering questions and change for different reasons. Keeping them separate lets a kind continue across compatible declaration revisions, lets candidate state change an extension without changing the kind, and lets evidence or a guard change reliance without changing what makes the criterion hold.
SoTA-Echoing
For one supported classification, the ordinary conclusion in C.3:0 can be sufficient. Creating a reusable declaration and an extension for every such answer adds products that the receiver may not consume. Keep the criterion, admissibility and needed support recoverable in the lighter answer. Add the corresponding declaration or extension when repeated interpretation or a set-consuming question needs it.
Detail Map
C.3 is the head pattern for typed reasoning. It leaves each detailed mechanism at its direct neighboring pattern while preserving a discoverable route to that mechanism.
When a case needs detail on declaration, classification, extension, Bridge, kind-use adaptation, abstraction, or applied guards, use the neighboring C.3 pattern that defines or constrains that detail.
Relations
- Builds on:
A.2.6context-slice and scope discipline,A.6.0reusable declaration discipline,C.2.1episteme identity, F-G-R, and direct subject patterns for candidate features. - Coordinates with:
C.3.1throughC.3.5,C.3.A,C.29,E.24.UK,A.8,A.11,F.8andF.18.B.3.5:4.1provides an optional record when constructional and typed or logical content must be inspected together; use A.22 for a needed structure selection and A.22.CGUS for a remaining continuation question or required qualification. UseA.7.1for a general diagnostic return to the subject construction. - Does not replace: direct candidate-feature ontology, A.14 collection membership,
A.2.6scope,C.29representation use, ontic settlement inE.24, U-kind admission inE.24.UK, or naming in Part F.
C.3:End
Kind Identity and Subkind Relations (U.Kind and U.SubkindOf)
Type: Kind identity, subkind relation, and continuity pattern Status: Stable Normativity: Normative unless a section is explicitly informative
Use This When
Use this pattern when work must recover one reusable kind, decide whether one kind is a subkind of another, or decide whether the same kind continues across a changed KindSignature edition.
What goes wrong if missed. A source or practice label becomes an identity key, U.SubkindOf carries dependency or construction, a finite sample is mistaken for a universal order, mutually classifying kinds are silently merged, or a changed declaration is treated as automatically new or automatically harmless.
What this buys. The user gets an operational kind-continuity test, a replayable subkind test, and a preorder of subkind facts distinct from declaration identity, current extension, evidence, bridging, and public naming.
Primary EntityOfConcern. One U.Kind individual recovered through its candidate domain, operative membership condition, intended member/non-member distinction, and continuity rule; or one proposed U.SubkindOf relation between kind participants within declared applicability.
First useful move. Write the two criteria so the reader can test inclusion. If this plant defines a cooling pump as a pump that circulates coolant through a heat exchanger, anything meeting that criterion already meets the pump criterion. Establish that both uses of “pump” have the same meaning. Introduce an occurrence identifier or a formal equivalence grouping only when the receiving use needs it.
Not this pattern when. Use C.3.2 for a declaration, admissibility result, candidate classification, or extension; C.3.3 only for a claimed correspondence between independently identified distinct kinds; and E.24.UK when admitting another durable public kind rather than using an already admitted U.Kind individual.
Problem Frame
A practice/source boundary is provenance and a comparison cue. It affects the continuity decision only when comparison exposes a difference in the candidate domain or membership distinction.
U.SubkindOf is a same-individual dependent kind under U.Relation. Its participants are a narrower kind and broader kind. An effective reference scheme and aligned KindSignature editions make the criteria interpretable and qualify applicability; they are not relation participants or occurrence-identity discriminators. Candidate state, context slice, declaration edition, kind identity, and one obtaining subkind relation can therefore change for different reasons.
Problem
The sentence cooling pump is a pump is useful only when the membership conditions justify it. A current extension table can hide a bad proposal, and different intensional kinds can happen to classify the same candidates. Conversely, a unit rewrite, source move, or clearer declaration need not create another kind. The core needs an obtaining test for the relation and a before/after test for kind continuity without treating a signature, sample, locality, or extension as the kind.
Forces
Core Objects
Direct U.SubkindOf Relation Boundary
A readable sentence such as CoolingPumpKind is a subkind of PumpKind for this declared plant use states that the direct relation obtains. It needs no occurrence identifier when no receiver distinguishes or refers to the occurrence.
The criterion-entailment branch compares the two membership conditions under an aligned interpretation and applicability. For the finite-domain branch, identify the complete domain to which both declarations apply and account for each candidate using the pinned editions. Every narrower member must also satisfy the broader criterion. An unresolved classification that could conceal a counterexample leaves inclusion unestablished. If the declarations apply together only on part of the proposed domain, state that smaller domain on the conclusion.
A candidate that satisfies the narrower criterion and fails the broader one refutes inclusion on the stated domain. Missing observations can leave the finite-domain test open even when the membership rules themselves suffice to prove entailment.
When a receiver needs one occurrence, R_sub is participant-determined by the ordered pair of kind identities. The effective scheme, aligned signatures, and applicability qualify how obtaining is tested and asserted.
A scheme-edition change therefore prompts an alignment and renewed test; it does not create another relation occurrence. If the same participants still satisfy the condition, the same relation continues to obtain. If they no longer do, the prior obtaining claim is no longer current; another assertion may record that change without inventing a scheme-keyed occurrence.
Solution
- Recover each kind before comparing it. For each kind, state the candidate domain, membership condition, intended member/non-member contrast, and continuity rule. Use practice/source provenance to locate the declaration, not to decide identity.
- Check admissibility first. Compare only candidates admissible under both aligned declarations and the stated applicability.
not-applicableforms no C.3.2 judgment. - Select one obtaining branch. Use exact criterion entailment when the membership rules can be compared directly. Use exhaustive evaluation only for a deliberately closed finite domain. State which branch and where it applies.
- Keep observations in their proper role. A non-exhaustive sample, test run, or extension can support the subkind assertion and expose a counterexample. It cannot close an open-domain obtaining claim.
- Keep a preorder over obtaining facts. Reflexivity and transitivity apply. Mutual facts between distinct kinds record classification equivalence for that alignment; they do not imply kind identity. Use the equivalence groups only when a receiver needs a partial order.
- Separate relation, predicate, and assertion. Use the readable relation sentence first. Add
R_sub, a C.2.1 assertion, evidence, or publication only when a named receiver consumes that object. - Diagnose counterexamples at the rule. Repair a false relation proposal, incompatible declaration alignment, or missing distinct-kind bridge. Do not edit an extension row to make the order appear true.
- Decide kind continuity independently. Apply the before/after test in section 6 whenever criterion, candidate domain, assumptions, dependencies, effective scheme, or locality changes. Another
KindSignatureedition neither proves nor denies kind continuity. - Keep scope and Work outside the kind. A kind carries no claim scope. A
W : U.Workremains a dated work occurrence under its direct pattern; keep it distinct from a plan, log, label, classification record, or episteme about W.
Continuity Decision
Compare the old and proposed declarations in this order:
Preserving change. CoolingPumpSignature-3 replaces litres-per-second with an exactly aligned SI expression, preserves the pump candidate domain, cooling-performance discriminator, intended member and non-member probes, and maintenance use. The same CoolingPumpKind continues; new judgments cite edition 3.
Identity-breaking change. A proposed edition replaces physical cooling performance with the presence of schema label CoolingPump. A physical pump without the row changes from member to non-member and a labelled non-performing row can appear to qualify. The operative distinction and candidate domain changed; identify another kind rather than continuing CoolingPumpKind.
Locality change. Journal and grant teams may reuse one ReviewerSystemRole when candidate Systems, required contribution, and acceptance condition remain aligned. If grant review requires a different contribution or admits a materially different candidate boundary, identify another kind. The two labels decide neither case.
Archetypal Grounding
Bias-Annotation
C.3.1 counters hierarchy, sample-as-law, assertion-as-world, locality, and table-repair bias. A stronger-looking edge is not automatically an obtaining relation; a sample does not close an open domain; mutual classification does not identify two intensional kinds; a changed source does not split one; and an extension remains an output representation rather than the place to repair the rule.
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
- Encoding dependency, part-whole, slot filling, construction, system-role assignment, or admission as
U.SubkindOf, or treating a predicate expression, assertion, diagram edge, or table row as the obtaining relation occurrence. - Treating a source hierarchy or public-looking spelling as durable FPF ontology.
- Treating
KindSignatureas the kind or its formality as a property of the kind. - Assuming every signature edit makes a new kind, or that no signature edit can make one.
- Comparing extensions across incompatible editions and repairing a counterexample by changing rows.
- Storing claim scope on a kind.
- Treating a work label or record as an individual work occurrence.
Consequences
Benefits. Local typed compatibility remains small while its consequences for candidate judgments are testable.
Costs. A declaration change that matters to later classification needs an explicit edition and a separate continuity decision.
Rationale
Kind identity, direct U.SubkindOf obtaining, assertion identity, declaration identity, candidate state, and current extension answer different questions and change under different conditions. Their separation lets a kind survive a compatible declaration revision while preventing an assertion or revised criterion from creating an order fact, silently rewriting prior classifications, or hiding a non-obtaining subkind proposal. Keeping the core small also prevents construction, admission, naming, scope, slot discipline, or dependency from being introduced into one hierarchy relation.
SoTA-Echoing
For a subkind question restricted to a deliberately closed finite domain, account for every admissible case in which a narrower member could fall outside the broader kind. The exhaustive branch in C.3.1:4.1 can use established membership judgements or records whose coverage and interpretation establish inclusion throughout that domain. Evaluate any unresolved case that could change the conclusion. Requiring criterion entailment beyond that domain would demand a stronger result than this use needs. A non-exhaustive sample leaves unexamined candidates and cannot supply this warrant. Keep the domain and aligned interpretation on the conclusion; recheck it when either changes. An open-domain claim still needs criterion entailment.
Relations
- Specializes:
A.6.RELforU.SubkindOf: ordered kind participants, criterion-entailment or exhaustive closed-domain obtaining, applicability, lightweight occurrence use, and participant-determined identity; schemes and declaration editions qualify interpretation and assertion rather than occurrence identity. - Builds on:
C.3, A.6.0 declaration identity, C.2.1 episteme and assertion identity, A.2.6/USM context-slice and scope discipline, F-G-R, and C.2.3 formality. - Coordinates with:
C.3.2judgments and extensions,C.3.3correspondence between independently identified distinct kinds,A.2when one local kind is a system-role kind,A.6.5declaration-slot uses that consume an already obtaining subkind relation,C.29representations,E.24.UKdurable U-kind admission, andA.8,A.11,F.8, andF.5when public kind governance is current. - Other governing patterns: Use C.2.1 for subkind-assertion epistemes, whether affirmative, negative, or unresolved; candidate features, classification assertions, kind declarations, context bridges, and public naming decisions retain their own governing patterns.
C.3.1:End
Kind Intent, Membership Judgment, and Extension
Type: Kind declaration and classification pattern Status: Stable Normativity: Normative unless a section is explicitly informative
Use This When
Use this pattern when repeated typed reasoning needs one explicit kind criterion, when one entity or non-entity value must first be checked as applicable and then judged against that criterion in one context slice, or when a named use needs a representation of the candidates currently judged true.
What goes wrong if missed. A kind is confused with its declaration, a practice label splits one kind, a measurement or schema label creates membership, an out-of-domain request becomes unknown, missing information becomes false, a set becomes ontology, or a guard decision rewrites the classification.
What this buys. A practitioner can state an ordinary result, pin the declaration and slice when reliance requires it, distinguish not-applicable from true, false, and unknown, and materialize an extension only for a receiving query or review. A manager can separately change declaration formality, assurance for a relied-on assertion, or claim scope without treating them as one maturity ladder.
Primary EntityOfConcern. One classification use: candidate, kind, KindSignature edition, context slice, pre-judgment admissibility, and—only when admissible—the judgment value.
First useful move. Write the readable result first: For this plant’s operating report, a cooling pump is a pump circulating coolant through the heat exchanger during the reporting interval. The circuit connections and the interval’s flow reading establish that Pump #14 did so. Before evaluating, confirm that a pump candidate and this slice are within the declaration's candidate domain and applicability. Cite support only when the receiving use relies on it; create a reusable signature or extension only for repeated or set-consuming use.
Not this pattern when. Use the direct subject pattern to establish the candidate and the quality, relation, episteme, status, publication occurrence, or other condition named by the criterion; A.14 for membership in a collection; C.3.3 for a claimed correspondence between distinct kinds; C.29 for a separate mathematical representation or transfer question; and E.24.UK for admission of another durable public kind.
Problem Frame
A kind can support useful typed reasoning without acquiring its own public U.* label. Its intent may need a reusable declaration, one candidate may need a current judgment, and a query may need a set representation. These are different objects. Before a judgment exists, the candidate must satisfy the declared candidate ValueKind and the slice must lie within declared applicability. For an admissible candidate, return true when the available facts establish the criterion and false when they establish its failure. Return unknown when they leave the criterion undecided.
The rule about evidence is conditional, not lexical. An observation used merely to support a claim does not create an independently governed quality or relation. But a criterion may directly concern an episteme, an obtaining registration or certification relation, a publication occurrence, legal status, or another governed fact. In that case, determine whether that very condition obtains under its direct pattern; calling the same object evidence in another use does not erase its criterion role. This concept-level rule requires no particular ontology language, schema technology, rule engine, or programming type system.
Problem
The shorthand MemberOf(e,k,slice) is unsafe because readers can take it as an A.14 collection relation, an ontic occurrence, a classification result, a database lookup, or a guard. It also hides whether the request was applicable. C.3.2 restores a declaration, an admissibility result, a three-valued judgment only for admissible candidates, and an optional representation while leaving candidate identity and the criterion's governed conditions with their direct patterns.
Forces
Four Objects and One Applicability Result
ClassificationAdmissibility(candidate, kind, signatureEdition, slice) returns admissible or not-applicable. It is a precondition result, not another kind or membership value. not-applicable means the candidate fails the declared candidate ValueKind/interpretation or the slice falls outside signature applicability; no classification judgment is formed.
Scope is not attached to the kind. A KindSignature episteme may have its own U.ClaimScope; a separate classification assertion has the scope of that assertion; and U.ContextSlice remains an evaluation input.
KindSignature Declaration
Author a reusable KindSignature only when a named receiving use needs the criterion and assumptions to persist across more than one classification. Its claim content declares:
- the exact kind that is its
EntityOfConcern; - the candidate
ValueKindor exact value interpretation admitted as input; - the membership condition in terms of directly governed candidate qualities, relations, constructive grounding, epistemes, registrations, certifications, publications, legal statuses, or other conditions;
- the exact
U.ContextSliceapplicability in which the evaluation may be formed; - the effective reference scheme: designation and interpretation rules for reading that declaration;
- the assumptions and dependencies on which the criterion relies, including any required standards, versions, units, or temporal policy;
- the expression properties needed for its intended use under C.2.3; and
- an optional
ExtentRulefor a named extension-consuming use.
In A.6.0 terms, SubjectKind is the broad candidate kind and RangedValueKind is {true, false, unknown}. not-applicable is returned before this ranged evaluation. ExtentRule is declaration content, not a new ontic relation. Formality characterizes the declaration episteme, not the kind, candidate, truth, or extension.
A changed membership condition, candidate-domain declaration, EntityOfConcern, applicability, or effective scheme identifies another episteme. Recheck its signature membership under A.6.0 and its KindSignature content under this section; claim an edition relation under C.2.1 only when that relation obtains. C.3.1 separately decides kind continuity.
Admissibility and One Candidate Judgment
For exposition, this pattern uses:
A(candidate, kind, signatureEdition, slice) ∈ {admissible, not-applicable}
and, only when A = admissible:
J(candidate, kind, signatureEdition, slice) ∈ {true, false, unknown}
These are local result notations, not newly admitted kinds, A.14 membership occurrences, direct classification relations, or evidence relations. For a fixed candidate, kind, signature edition, and slice, unchanged governed conditions and the same available support and declared dependencies yield the same result; the slice resolves every selector in its declared schema, including any declared version or temporal selector, rather than implicit latest or current.
- Recover the candidate first. An entity is already individuated under its direct pattern. A non-entity value keeps the identity, unit, scale, and interpretation supplied by its governor.
- Pin the inputs. Name candidate, kind, exact signature edition, and exact slice; avoid implicit
latestorcurrent. - Check admissibility. If the candidate does not satisfy the declared candidate
ValueKindor interpretation, or the slice is outside declared applicability, returnnot-applicableand stop. Do not formJ. - Evaluate the governed condition. For an admissible candidate, a satisfied criterion gives
true; a known failed criterion givesfalse. - Keep non-settlement visible. Use the evaluation rule for the criterion. If it requires two conditions and one is known to fail, the conjunction is
falseeven when the other is unmeasured. Returnunknownwhen the available facts leave the criterion unsettled. - Distinguish condition from evidentiary use. A measurement result, source episteme, certification, registration, publication occurrence, legal-status relation, or record may itself be a criterion condition only when the signature says so and that condition obtains under its direct pattern. Its mere use as evidence for some other condition creates neither that condition nor membership.
- Separate guard disposition. A guard checks admissibility, scope coverage, and any judgment as separate predicates. It may decline use on
not-applicableorunknownwithout converting either tofalse.
When a separate claim-bearing classification assertion is current, it is a C.2.1 episteme. Its content designates the candidate, kind, signature edition, slice, admissibility, any judgment, and relied-on support. Its EntityOfConcern is the governed entity about which classification matters; a value classification may stay in another claim's content rather than fabricating a value-shaped entity. The assertion creates neither candidate nor kind.
A domain that needs a durable classification-relation occurrence must supply a separate direct pattern with exact participants, obtaining condition, identity, and relation to these results. C.3.2 does not mint that occurrence.
Extension as Representation
Materialize KindExtension(k, slice) only when a named query, quantification, comparison, review, or publication needs the current true-candidate set.
- Pin the signature edition even though the compact name shows only
kandslice. - State the candidate domain without inventing
U.EntitySet. - Include each admissible candidate whose judgement is
true. Their number answers how many candidates have been individually confirmed. To answer a question about the population total, establish what bounds the number of members. In a complete inventory, confirmed members give one lower bound. Adding every other candidate whose membership remains possible gives one upper bound. These bounds can be loose. If exactly one of two identified units qualifies and the evidence leaves open which one, together they contribute exactly one. An established cap on the entire population can give an upper bound even when its inventory is incomplete. Use a sufficient bound for a threshold, and establish identities and coverage for a complete list. Anunknownjudgement records an inconclusive evaluation. An unevaluated candidate has no judgement yet. An inadmissible request returnsnot-applicablewithout a judgement. - Treat braces, rows, indexes, or database results as representations. They create neither a collection holon, A.14 membership occurrence, direct classification relation, nor criterion condition.
- Use C.29 when interpretation, preservation or transfer of the represented set leaves a separate mathematical question. Otherwise finish the local calculation under the kind and query rules.
Candidate state or a later slice can change an extension without changing the signature or kind. An extension row cannot repair an inconsistent declaration or subkind fact.
Subkind Comparison and Change
Whenever SubkindOfObtains(k1,k2) holds under C.3.1, its practical consequence is checked only where both candidate requests are admissible under the aligned declarations:
For the same candidate and slice, an admissible
truejudgment fork1must not coexist with an admissiblefalsejudgment fork2within the relation's declared applicability.
C.3.1 decides whether exact criterion entailment or exhaustive evaluation over a deliberately closed finite domain makes the relation obtain. Non-exhaustive classifications support its assertion or expose a counterexample; they do not establish an open-domain relation. A not-applicable request is outside the comparison.
Cross-local use first compares kind identities: reuse the same kind directly when its membership distinction continues; only distinct kinds with an obtaining correspondence use C.3.3. A bridge never transfers source classification truth.
Keep these changes distinct:
Required Worked Cases
Physical pump
CoolingPumpSignature-2 admits physical pump candidates and applies in plant slice S-14. Pump #14 is independently identified as a physical pump, so the request is admissible. Governed flow, heat-transfer, and operating-state conditions satisfy the criterion; a calibrated measurement result supports that claim without becoming the pump or its performance. The result is true. A maintenance-query extension may represent Pump #14 but does not create its classification.
Episteme and publication form
Maintenance-instruction episteme MI-22 is admissible for DiagnosticInstructionKind and is evaluated through its claim-bearing content and governed subject. MI-22-PDF-Layout and MI-22-HTML-Layout are different publication forms for the chosen episteme edition; files that bear them are presentation carriers. Arrangement, form, carrier, or encoding alone changes neither the episteme, criterion satisfaction, kind, nor judgment.
Non-entity temperature value
Value 87 °C, with declared scale, unit, interpretation, and time, is admissible for HighTemperatureValueKind when the signature's ValueKind accepts that quantity. It can then be judged against the declared interval without fabricating a value-shaped entity.
Schema label
A row carries label Customer, but the claim asks whether account holder #441 is a contractual customer. If the kind admits account-holder Systems or persons rather than database rows, the row itself is not an admissible candidate. For the account holder, the label may support recovery of the governed contractual relation but does not make that relation obtain. A different row-shape kind could make the row admissible under its own criterion.
Unavailable measurement
For S-15, the maintenance team verifies that the edition of CoolingPumpSignature-2 in use also applies to that slice and that Pump #14 is an admissible candidate. The required flow reading is missing and the remaining evidence leaves the criterion unsettled. The judgement is unknown. The safety guard may decline reliance. The result recorded for S-14 retains its own basis.
Not-applicable request
The value 87 °C is submitted to CoolingPumpSignature-2, whose candidate ValueKind is physical pump. The request is not-applicable; no cooling-pump judgment is formed. Lack of a pump judgment says nothing about whether the temperature value is known.
Registration-defined membership
A KindSignature for RegisteredSupplierKind declares supplier candidates and requires an exact obtaining registration-status relation under the current register rule. Supplier #27 is admissible. If that governed relation obtains, it is part of the membership condition even though a registration episteme may also be used as evidence. A copied row or certificate image alone does not create the relation. This preserves legitimate institutional kinds without treating every record as a world-side fact.
Additional Transfer Cases
Work Boundary
For a Work classification, keep these distinctions:
U.Workis the admitted kind;W : U.Workis one independently grounded dated 4D work occurrence underA.15.1;- a plan, record or description about W remains distinct from W; identify any claim-bearing episteme through C.2.1, separately from its publication form or carrier; and
- further claims about W, such as its resource use or result, follow their direct subject patterns.
A kind may classify an already identified W. A kind symbol, work label, plan, or record never occupies W's individual position, and record existence does not make planned Work actual.
Authoring Rhythm
- Start with one readable classification sentence and its practical use.
- Recover the candidate and the governed criterion conditions before discussing support.
- Reuse an existing signature edition only when it governs candidate ValueKind, criterion, applicability, scheme, and dependencies.
- Check admissibility. Stop with
not-applicablewhen candidate or slice lies outside the declaration. - For an admissible request, return
true,false, orunknownwithout folding in the guard decision. - Create an extension only for a named set-consuming use.
- If a separate assertion is required, give its C.2.1 episteme the EntityOfConcern, content, scope, support use, and edition.
Conformance Checklist
Common Anti-Patterns and Remedies
Consequences
Benefits. Classification becomes inspectable without ontology growth, evidence-created truth, or coercion among non-applicability, uncertainty, and falsity. Repeated criteria can be reused, and set-consuming uses can receive a bounded representation.
Costs. Reliance-bearing uses must pin a declaration and slice, check candidate/slice applicability, preserve unknown, and recover any criterion-bearing status or relation under its direct pattern.
Rationale
The kind, its declaration, pre-judgment applicability, one admissible candidate judgment, and a representation of current true candidates answer different questions. Their separation prevents evidence, locality, time, scope, and notation from rewriting ontology while still allowing a criterion to concern a directly governed episteme, status, or relation when that is the classification condition.
SoTA-Echoing
A refusal can suffice when the receiver needs only the guard decision. If the receiver needs classification or a further action, C.3.2:6 preserves the relevant differences: an unsuitable request yields not-applicable and no judgment; an admissible request yields unknown when the available facts leave the criterion undecided; an admissible candidate's known criterion failure yields false. Collapsing them into false would assert non-membership for a mismatch or missing grounds. Correct the request or seek missing grounds only when a further answer is needed; an adequate negative or unresolved answer can finish the work.
Relations
- Builds on:
C.3,C.3.1, A.6.0 declaration identity, C.2.1 episteme identity, A.2.6 context slices and claim scope, and direct patterns for candidate identity and features. - Coordinates with:
C.3.3correspondence between independently identified distinct kinds,C.3.4local adaptations,C.29mathematical representations, C.2.3 formality, F-G-R evidence and assurance, A.14 collection membership, andE.24.UKdurable U-kind admission. - Does not replace: the direct subject pattern, evidence-use relation, collection membership, claim-scope governor, guard decision, public-kind admission, or a separately justified durable classification-relation pattern.
C.3.2:End
KindBridge and CL^k — Correspondence between Distinct Kinds
Summary. A changed practice, source, team, or scheme first triggers a comparison of kind definitions. If the same kind continues, reuse it without a
KindBridge. When the receiving use is candidate-bearing, evaluate the candidate under the receiving declaration. When two independently identified kinds are distinct and a directional correspondence predicate holds, oneKindBridgedirect relation may obtain. A separate bridge-assertion episteme states direction, paired declaration editions, preservation or loss, evidence, and admitted use, withCL^kwhen the consumed assurance or calibration account requires it. It never transfers source classification truth.
Normativity: Normative in Part C. Identifier C.3.3.
Status: Stable Audience. Engineering managers, architects, assurance leads, editors.
Depends on.
- C.3.1 — U.Kind and U.SubkindOf: kind identity follows candidate domain and membership distinction; subkind facts form a preorder; locality and scheme editions are not identity keys.
- C.3.2 — Kind intent, admissibility, judgment, and extension: check
admissible | not-applicablebefore a freshtrue | false | unknownreceiving judgment. - A.2.6 — USM: Claim scope and selected context slices remain separate from kind identity and kind correspondence.
- C.2.2 — F–G–R: justified bridge penalties affect reliance R, not F or G.
- C.2.3 — Formality assessment: the assessed expression properties belong to the declaration or assertion episteme.
Use boundary. Kind correspondence leaves scope translation and local-sense correspondence to their own patterns when needed. CL^k is an ordinal assessment for a declared kind-correspondence use, not a universal interoperability score.
Purpose and Audience
Typed reuse can fail because a claim's scope changed, because the receiving use employs another kind, or because wording changed meaning. These are separate questions. C.3.3 handles only a claimed directional correspondence between two distinct kind individuals. It lets a team state, for example, that source Vehicle corresponds to target TransportUnit, which distinctions are preserved or collapsed, and what loss a receiving use accepts.
Context
Different sources or practices may use the same kind, or different kinds may coexist inside one source. Names, source labels, compatible schemes, and matching current extensions decide neither. First compare the candidate domains and membership distinctions under C.3.1. Same-kind reuse needs no bridge. Distinct-kind reuse may need a KindBridge when its correspondence predicate can be established. F.9 is additional only when distinct local senses and their bounded use are current.
Problem
- False splitting. A locality change creates two apparent kinds and a bridge even though the membership distinction is unchanged.
- Semantic drift. A different receiving kind is treated as the source kind because names or extensions look alike.
- Hidden order loss. Subkind facts collapse, invert, or become unsettled without being reported.
- Entangled channels. Scope, sense, and kind correspondence are bundled into one score or record.
- Classification transfer. A source judgment is copied as receiving truth without checking receiving admissibility and criterion satisfaction.
- Unreplayable use. An unspecified mapping or implicit latest edition leaves a guard with no stable basis for deciding whether the bridge use is current.
Forces
Solution — Compare Identity, Then Relate Distinct Kinds
- Compare kind definitions. Recover source and receiving candidate domains, membership distinctions, and continuity rules. A changed locality or scheme prompts this check; it does not decide it.
- Stop on same-kind reuse. If the same kind continues, use the declaration edition selected for the receiving use. When that use is candidate-bearing, check admissibility and evaluate the candidate afresh. No
KindBridgeobtains merely because source, practice, team, wording, or scheme changed. - Open a bridge only for two distinct kinds. A
KindBridgeoccurrence is an obtaining direct relation between one source kind and one target kind. Its directional predicate states the correspondence and definedness required by the named receiving use. Identify each kind independently under C.3.1. - Keep the assertion separate. A C.2.1 bridge-assertion episteme designates the relation when needed and carries paired
KindSignatureeditions, mapping rule, selected order-preservation results, loss notes, evidence, and admitted use. IncludeCL^kwhen the consumed assurance or calibration account requires it. A card, row, F.9 relation, or publication does not make the bridge obtain. - For candidate-bearing use, evaluate the receiving candidate. First check admissibility under the receiving signature and slice. For an admissible request, return
true,false, orunknownfrom the target criterion. A supported source fact can determine that answer if its implication for this candidate, criterion and slice is established. Reuse those grounds. Where the mapping loses a difference the target criterion needs, retain the uncertainty or obtain the missing fact. - Route consequences narrowly. When a receiving claim relies on the obtaining bridge, apply only the justified
CL^kconsequence to R. Scope and any sense relation retain their own objects and rules; F and G do not change.
The kinds are the direct relation participants. Scheme and signature editions qualify interpretation, applicability, and the assertion. They do not identify the occurrence. For the ordered kind pair, the direct relation is participant-determined. An aligned scheme-edition change prompts reevaluation of whether the same relation still obtains; it does not mint another occurrence.
KindBridge is the direct relation kind governed here under A.6.REL. This spelling does not by itself admit a public dependent U-kind named U.KindBridge. If admission later matters, E.24.UK must close it separately.
Norms & Invariants (normative)
The KB‑01…KB‑12 rules below govern kind correspondence.
Direct Relation Subject and Scope
KB-01 (Distinct participants and obtaining). One KindBridge occurrence has exactly two ordered participants: an independently identified source kind and an independently identified distinct target kind. It obtains only when its directional correspondence predicate holds within declared definedness. A different locality, label, scheme, or extension supplies no bridge. Signatures, assertions, evidence, CL^k, loss notes, and slices are not participants.
KB-02 (No Scope or sense substitution). A KindBridge maps neither Claim/Work scope nor local wording. Scope translation uses A.2.6 when the receiving claim consumes it. An F.9 relation is added only for a current distinct-sense use. Neither channel is required merely because a kind bridge exists.
No blended score. Scope congruence, sense-relation loss, and kind congruence remain separate. Do not aggregate them into one interoperability score.
Settlement, Assertion, and Identity
KB-03 (Direct settlement). The C.3.3 settlement SHALL make recoverable:
- ordered source-kind and target-kind participants and the proof that they are distinct;
- the directional correspondence predicate, applicability, and definedness; and
- participant-determined occurrence identity for that ordered pair.
The separate bridge assertion states whether obtaining is affirmed, denied, or unresolved; only an affirmative assertion may designate an obtaining occurrence. It also names the declaration and scheme editions used to interpret the predicate, selected source and target subkind facts, preservation/collapse/non-preservation/unknown results, loss, evidence, and admitted use. Include CL^k when the consumed assurance or calibration account requires it.
Another assertion, mapping expression, card, signature, scheme edition, or publication does not create another relation occurrence. A changed interpretation prompts a renewed obtaining test. If the same ordered participants and correspondence continue, the same relation continues; if not, the prior obtaining claim is no longer current.
KB-04 (Fresh receiving classification). With fixed receiving candidate, signature edition, and slice, check admissibility first. not-applicable forms no classification judgment. An admissible request is evaluated reproducibly as true, false, or unknown. A source judgment or bridge assertion may support reliance but is never copied into the receiving result. An unavailable bridge dependency blocks that bridge use without rewriting an independently evaluated receiving result.
Order & Monotonicity
KB-05 (Monotone order). To claim that source fact SubkindOfObtains(k1, k2; sourceRS) survives transfer, identify its two target endpoints. For each endpoint that changes kind, establish the directional KindBridge from source to target. For an endpoint that remains the same kind, establish identity under the two declarations. Establish SubkindOfObtains(k1', k2'; targetRS) using the target criteria or its closed-domain test (C.3.1). If another claim relies on that order, check separately that the premises it needs survive the transfer. Identify a target R_sub : U.SubkindOf occurrence only when its identity matters to the receiving use.
KB-06 (No inversions). A bridge assertion MUST NOT state preservation when the mapped target order is inverted. If SubkindOfObtains(k2', k1'; targetRS) holds for distinct mapped kinds and the required forward fact SubkindOfObtains(k1', k2'; targetRS) is established not to hold, state non-preservation and the loss. If the required target order cannot be settled, state unknown; do not turn non-settlement into either preservation or inversion.
KB-07 (Collapse semantics). A bridge assertion may classify selected source subkind distinctions as collapsed when several source kinds correspond to one target kind. The assertion SHALL designate the affected obtaining U.SubkindOf relations and state the lost properties; the direct bridge relation does not alter either local order.
Congruence & Assurance
KB-08 (Anchor reuse and AT neutrality). CL^k reuses the ordinal anchor semantics of CL but assesses the declared bridge use over kind intent and order. The bridge-assertion episteme labels it kind-congruence. KindAT remains editorial under C.3.5 and independent of CL^k.
KB-09 (Effect on R only). After receiving admissibility has been checked and an admissible candidate has received a fresh target judgment, a claim that relies on both that result and an obtaining KindBridge may apply only the bridge assertion's justified monotone Ψ(CL^k) consequence to R, alongside any independently established scope-relation consequence. A not-applicable candidate forms no judgment; unknown stays unknown; F and G do not change.
KB-10 (Chaining). A minimum over comparable link assessments reports the lowest link rank on their common scale and calibration question. It does not by itself assess the composed correspondence. For a receiving use that depends on preservation through the chain, recover the mappings and intermediate interpretations, establish their joint definedness, and check that the required source distinctions survive every link and the composition. Keep each obtaining KindBridge and its assertion distinct. Return the preserved structure, losses and unresolved requirements; use another correspondence, narrower application or additional basis when an essential distinction is unavailable.
Two locally useful mappings can lose different distinctions. For a finite order test on 0 < 1 < 2, let both monotone maps have values f(0)=0, f(1)=0, f(2)=1 and g(0)=0, g(1)=0, g(2)=1. Each preserves a nontrivial distinction in its own domain, but their composition maps every source value to 0. The lowest link rank cannot establish preservation of the source distinction the final use needs.
Conversely, an established composition that preserves the required common structure remains usable. A local law may derive a composition characteristic from a minimum when its domain and preservation argument justify that law; there is no universal chain-minimum rule. Any numerical consequence for R still requires the receiving model under C.2.2 and B.3.
Loss Notes & Definedness
KB-11 (Loss notes). The bridge-assertion episteme SHALL state which KindSignature invariants are not preserved, which obtaining source U.SubkindOf relations are collapsed or not preserved, and any higher-equality caveats. These claims do not rewrite the source or target kinds.
KB-12 (Definedness and guard use). The bridge predicate and assertion SHALL state definedness. Outside it, a receiving guard declines that bridge use. Independently, receiving classification keeps not-applicable or its admissible true, false, or unknown result; bridge inapplicability rewrites none of them.
Interactions (informative)
With Scope and Sense Relations
A receiving use may need none, one, or several independent relations:
- an A.2.6 scope relation when the claim's admitted extent is translated or compared;
- a C.3.3
KindBridgewhen two distinct kinds are directionally related; and - an F.9 relation when distinct local senses are related for a bounded use.
Open only the channels the receiving claim consumes. Keep their definedness, losses, and R consequences separate.
With Receiving Classification
After same-kind reuse or an obtaining bridge, use the receiving KindSignature edition. Check candidate and slice admissibility. If admissible, evaluate the receiving judgment. If a mapping motivates another signature, author that declaration episteme separately. A source judgment can support a claim but never supplies receiving truth.
With Kind-use Adaptations
For same-kind reuse, select the receiving C.3.4 declaration without a bridge; evaluate the receiving candidate afresh when that use is candidate-bearing. For distinct-kind use, recover the obtaining KindBridge, bridge assertion, receiving adaptation declaration, and any adaptation-correspondence declaration needed for differing constraints or bindings. Source adaptation results are not receiving truth.
With Guards
A typed receiving guard first determines whether the same kind continues or a distinct-kind bridge is current. When the guard's use is candidate-bearing, it checks receiving admissibility and, when admissible, the fresh judgment. It independently checks any scope or sense relation the claim consumes and applies only justified consequences to R. not-applicable, unknown, absent bridge, and guard refusal remain different results.
Authoring, Review & Rating Guidance (informative)
Authoring a KindBridge assertion
- Compare identity before authoring a bridge. For a changed locality, source, team, spelling, or scheme, first repeat the kind-definition comparison under C.3.1. Stop without a bridge when the same kind continues.
- Declare the needed correspondence and order facts. For two distinct kinds, declare only the directional correspondence and subkind facts the receiving use relies on; mark any required but unsettled order result unknown.
- Prefer an existing suitable target kind. If the target already has a suitable kind and declaration edition, relate that kind directly. If a new target declaration is required, author it separately before asserting bridge obtaining; list what the mapping predicate preserves, relaxes, or drops.
- Write loss notes in plain language. Example: “EV vs ICE subkinds collapsed; battery‑health invariants dropped.”
- Fix the definedness area. Bind to target Standards/versions and any environment selectors essential to classification.
- When
CL^kis required, assess it from exemplars. Calibrate on concrete counter-examples and preserved properties; resist optimistic ratings.
Review prompts
- Identity checked first? Same kind reused without a bridge, or two distinct kinds and the obtaining correspondence shown? Add Scope or F.9 relations only when the receiving use consumes them.
- Check the order claims. Any
⊑inversions? Collapses disclosed? - When
CL^kis used, is it plausible? Based on preserved properties, not name similarity? - Loss consequences clear? Do they change the justified inference, require more basis or leave the intended use adequate?
- Definedness area clear? Outside it, does the guard decline this bridge use while preserving the independent classification result?
- Reliance consequence justified? Any numerical change to R needs its receiving model; F and G remain unchanged.
Rating CL^k (rules of thumb)
- High
CL^k: signature equivalence or up‑to‑iso;⊑fragment preserved; only cosmetic losses. - Medium
CL^k: some invariants relaxed or lost; selected subkinds collapsed; the source-order facts required by the use remain preserved. - Low
CL^k: name-only correspondences; properties diverge; order not preserved. Establish that the KB-01 correspondence predicate holds before bridge use, even for a name-only match. Determine which inferences remain justified; narrow the use or repair the correspondence if a required distinction is lost. A numerical R penalty requires the receiving model under C.2.2/B.3.
Worked Examples (informative)
Vehicle → TransportUnit (manufacturing)
Source kinds Vehicle and PassengerCar, target kinds TransportUnit and PassengerTransportUnit, and their declaration editions are independently identified. One KindBridge relation obtains from Vehicle to TransportUnit and another from PassengerCar to PassengerTransportUnit under the pinned scheme editions.
The bridge assertion states that source fact SubkindOfObtains(PassengerCar, Vehicle; sourceRS) is preserved by target fact SubkindOfObtains(PassengerTransportUnit, TransportUnit; targetRS), while the EV distinction is collapsed; it records CL^k=2, the lost battery-health invariants, and definedness limited to registryAPI v1.4 in the selected time window.
A candidate is first checked for admissibility and, if admissible, then classified by the receiving declaration; source classification is not copied. If the receiving claim also relies on an independently established scope translation, that relation's consequence remains separate from the kind-bridge consequence; the kind-bridge consequence leaves F and G unchanged.
Same AuthenticatedRequest kind across services — no bridge
Frontend and gateway services use the same AuthenticatedRequest kind: the candidate request domain, signature-validity condition, and intended member/non-member distinction are aligned. Each service uses its selected declaration edition and evaluates the request afresh. The gateway spelling x-auth may require a C.3.4 vocabulary binding when that wording use is relied on; an F.9 sense relation is added only if that use also needs a relation between distinct local senses. The service boundary and spelling alone create neither another kind nor a KindBridge.
AdultPatient across jurisdictions (clinical)
The obtaining bridge relates source kind AdultPatient to independently identified target kind AdultPerson_Y. In this example the source age threshold is 18 and the target threshold is 21. The assertion gives CL^k=1 and states the jurisdictional editions and the age distinction lost in the correspondence. For an admissible candidate who meets the other target conditions, a verified source classification at 18 or older leaves the target age condition open if both 20 and 22 are possible. A verified lower age bound of 22 on the target date settles it without an exact date of birth. The receiving guard can refuse a use that requires positive classification while the age condition is unresolved. A later request for classification can use any evidence admitted by the target rule that settles the threshold. A separately required birth-date record remains a distinct deliverable.
Anti‑patterns & Remedies (informative)
Conformance Checklist
C.3.3:End
KindUseAdaptationDeclaration — Tailor the Use of an Existing Kind
Summary. Use a
KindUseAdaptationDeclarationwhen a procedure needs a narrower or differently named use of an existing kind without defining another kind. The declaration pins the baseKindSignatureedition, local candidate constraints or vocabulary bindings, intended guard use, and applicability. Check admissibility before returningtrue,false, orunknown. A locality change first triggers kind-identity comparison: the same kind needs noKindBridge; distinct kinds need one only when its exact correspondence predicate obtains.
Normativity: Normative in Part C. Identifier C.3.4.
Status: Stable Audience. Engineering managers, architects, reviewers, and editors.
Use This When
Use C.3.4 when a procedure needs a named local way of using an existing kind without claiming another kind. Typical cases include accepting Vehicle candidates only when they have ABS, using local spelling X-Auth for AuthHeader, or combining a candidate constraint with vocabulary bindings.
First identify the base kind, exact KindSignature edition, receiving use, and meanings of local names or predicates. Write one declaration for candidate constraints or vocabulary bindings and route claim-scope conditions separately. When the receiving use applies the declaration to a candidate, check candidate and slice admissibility, then evaluate that candidate. Stop when the declaration and any required candidate result are sufficient for the receiving use. If the distinction becomes a stable kind, identify it separately and establish any obtaining U.SubkindOf fact under C.3.1.
Do not use C.3.4 merely to rename a kind, represent a catalog row, narrow claim scope, or avoid deciding whether another kind is needed. A vocabulary-only change adds no candidate predicate. not-applicable, unknown, and a guard refusal are different results.
Depends on.
- C.3.1 — U.Kind and U.SubkindOf: kind identity follows the membership distinction;
U.SubkindOffacts form a preorder and kinds carry no Scope. - C.3.2 — Kind intent, admissibility, judgment, and extension:
KindSignatureis a declaration episteme; admissibility precedes the three-valued judgment. - C.3.3 — KindBridge and CL^k: a directional correspondence only between independently identified distinct kinds, plus R-only bridge consequences.
- A.2.6 — Context slices and Scopes: Claim and Work scope over
U.ContextSlice. - C.2.2 and C.2.3: F–G–R and formality characterize the episteme being assessed.
Use boundary. The declaration specifies a use of the base kind. Kind identity, kind correspondence and claim scope retain their own rules.
Purpose
Teams often need a local projection of a widely used kind: Vehicle with ABS for one procedure, or local spelling X-Auth for AuthHeader. Cloning a kind for every use fragments catalogs and creates false bridge pressure. A local-use declaration keeps the base-kind identity, makes constraints and bindings explicit, and gives a guard one versioned episteme to designate.
Context
C.3.1 governs kind identity and subkind relations; C.3.2 governs candidate admissibility and classification judgment. A.2.6 governs a claim's declared scope. A procedure may tailor use for a compliance procedure, product line, or cohort without changing the kind.
Three objects remain distinct:
KindUseAdaptationDeclarationstates one named use of a base kind.KindUseAdaptationJudgmentis the three-valued result for one admissible candidate under pinned declaration and signature editions.KindUseAdaptationCorrespondenceDeclarationrecords how one exact source declaration corresponds to one exact target declaration when their constraints or vocabulary bindings differ.
The third object is a C.2.1 declaration episteme. Its effective scheme makes source, target, and rule designations interpretable. It is not an executable adapter, mapping Method, representation correspondence, obtaining F.9 relation, KindBridge, or target judgment.
Problem
- Kind sprawl. Teams mint near-duplicates for every procedure.
- Hidden constraints. Informal acceptance rules leak into prose and cannot be replayed.
- Scope conflation. Jurisdiction, API version, or another scope condition is treated as part of kind identity.
- Automatic bridge pressure. A changed source or team is treated as proof of another kind and a bridge.
- Collapsed outcomes. A non-applicable candidate, unsettled admissible candidate, and guard refusal are reported as one
unknownorfalseresult.
Forces
Solution — Declaration, Correspondence, and Judgment
A KindUseAdaptationDeclaration is a named, versioned C.2.1 declaration episteme about one local use. The base kind is its EntityOfConcern; its effective scheme gives meaning to declaration names and predicates. Its claim content states:
- the exact base kind and pinned base
KindSignatureedition; - the receiving use and adaptation type: constraint, vocabulary, or composite;
- additional directly governed candidate conditions, when any;
- vocabulary or notation bindings;
- exact candidate and slice applicability plus dependencies;
- scope expectations routed separately through A.2.6; and
- intended guard use and this declaration episteme's formality, when the receiving use requires that assessment.
For a candidate-bearing use, first evaluate adaptation admissibility. A candidate rejected by the base signature's ValueKind, an adaptation-specific candidate requirement needed merely to form the question, or the declared slice applicability is not-applicable; no adaptation judgment is formed. For an admissible request, use:
J_kindUse(candidate, kind, kindSignatureEdition, adaptationDeclarationEdition, slice) ∈ {true, false, unknown}
The judgment concerns the conjunction of the base C.3.2 criterion and every added candidate condition. A known false condition settles the conjunction as false. Relations between the conditions can also decide it: if both inspectors must have signed and exactly one signed, the conjunction is false even while the signer's identity is unknown. Return true when the facts establish all required conditions and unknown when they leave satisfaction unsettled. A vocabulary-only declaration adds no predicate and preserves the base judgment. A guard may decline use on not-applicable or unknown without rewriting either.
An optional pinned-edition representation may list admissible candidates judged true. It is not U.EntitySet, A.14 membership, another kind, or a direct classification relation. Scope conditions stay under A.2.6 rather than becoming kind identity.
When a use moves to another practice, source, or team, compare the base-kind membership distinctions first:
- if the same kind continues, use the declaration and signature edition selected for the receiving use and, when the receiving use is candidate-bearing, make a fresh receiving judgment; no
KindBridgeexists merely because locality changed; - if independently identified kinds are distinct and the use claims a directional correspondence, establish the C.3.3
KindBridge; use the receiving signature and adaptation declaration and, when the receiving use is candidate-bearing, make a fresh receiving judgment; and - if two adaptation declarations differ in constraints or bindings, a separate
KindUseAdaptationCorrespondenceDeclarationmay name source declaration as EntityOfConcern and state target, direction, deterministic rule, definedness, loss, and effective scheme. It creates no bridge or target truth.
A recurring local test can distinguish members of a narrower kind. Identify that kind by its criterion, then use C.3.1 to establish its relation to the base kind. A declaration or catalogue row can record the result.
Norms and Invariants
Definition and Shape
KUA-01 (Definition). A KindUseAdaptationDeclaration SHALL be a named, versioned C.2.1 declaration episteme with exact base kind as EntityOfConcern, effective scheme, pinned base signature, receiving use, adaptation type, candidate constraints, vocabulary bindings, applicability, dependencies, intended guard use, and separate scope expectations. Its formality characterizes the episteme.
KUA-02 (Not a new kind). A declaration MUST NOT introduce a kind or subkind fact. Claiming a new subkind requires an independently identified kind and the C.3.1 test for inclusion.
KUA-03 (Admissibility before judgment). Fixed candidate, kind, base-signature edition, adaptation-declaration edition, and slice first yield admissible or not-applicable. Only an admissible request yields true, false, or unknown; implicit latest and guard-result coercion are forbidden.
KUA-04 (Adaptation type). A vocabulary declaration preserves the base judgment. For a constraint or composite declaration, evaluate the combined criterion for the admissible candidate. Return true when the grounds establish the conjunction, false when they exclude it and unknown when they leave it unsettled. Use known individual conditions and any established joint constraints.
Separation of Channels
KUA-05 (Scope versus candidate). Conditions of the candidate may enter the adaptation judgment. Claim- or Work-scope conditions remain under A.2.6. A declaration may cite both, but a guard routes them separately.
KUA-06 (Guard use). A guard MAY designate a declaration only when its exact edition, base signature, dependencies, applicability, and candidate conditions are recoverable. For a candidate-bearing use, it checks admissibility before the judgment. It makes its use decision separately.
Stable Refinement and Catalog Representation
KUA-07 (Stable refinement). When repeated uses may require one stable membership distinction, review whether another kind would serve them. If that kind and an obtaining subkind fact are established, retain the adaptation declaration for any remaining local use or retire it. Reuse of a vocabulary binding alone adds no membership distinction. Declaration reuse or labeling performs no kind admission and establishes no subkind fact.
KUA-08 (Addressability). Every guard-addressable adaptation declaration resolves to its exact edition, base signature, dependencies, applicability, and intended use. A correspondence declaration also resolves its source declaration as EntityOfConcern, target declaration, direction, deterministic rule, effective scheme, definedness, and loss. A catalog represents those references; it is neither the declaration episteme nor ontology, and consolidation does not merge kind identities.
Cross-local Use
KUA-09 (Identity check before bridge). On a locality change, first compare exact base-kind definitions. Same-kind reuse needs no KindBridge but uses the receiving declaration and, when the receiving use is candidate-bearing, a fresh judgment. Distinct-kind use establishes a KindBridge only when its directional correspondence predicate obtains. Differing adaptation constraints or bindings may additionally require an exact correspondence declaration. Source judgments are never copied as receiving truth; justified bridge consequences affect R only.
KUA-10 (Definedness and fail-closed use). Outside adaptation applicability, return not-applicable and form no judgment. For an admissible request, return unknown when an unavailable dependency leaves the conjunction unsettled under KUA-04; a known false conjunct still gives false. Outside correspondence definedness, the guard declines that cross-local use without rewriting an independently evaluated receiving result.
Invariants and Non-goals
- No Scope leakage. A constraint adaptation adds a test of candidate features. A vocabulary-only adaptation binds a local name and preserves the base judgement. Each declaration identifies the slices in which its constraint or vocabulary binding applies. A separate claim has its own scope G under A.2.6. To use the declarations as models and claim strict refinement, compare them for the same inquiry under B.5.MR. An additional condition already implied by the base criterion can leave every answer unchanged.
- Identity preservation. The base kind remains
k; the declaration does not change itsEntityOfConcern. - Aggregation unaffected. Adaptation and correspondence declarations do not change the applicable F rule under C.2.3 or the justified support-composition model or non-aggregate synthesis under C.2.2; guards route candidate-feature predicates to the adaptation judgment and context predicates to Scope.
Interactions
With Kinds and Subkinds
Use an adaptation declaration for procedural tailoring. If the receiving uses need a stable conceptual distinction as a kind, identify it under C.3.1 and establish any claimed U.SubkindOf relation. Repeated declaration use or a catalog link does not establish that relation.
With Judgment and Declarations
- The base
KindSignatureepisteme supplies the kind criterion and its own F. - The separate adaptation declaration supplies additional candidate-feature constraints or vocabulary bindings and may have its own F.
- The
KindUseAdaptationJudgmentpins both editions and preservesunknown; neither formality assessment belongs to the kind, candidate, or truth value. - An optional extension-like result remains only a pinned-edition representation of true adaptation judgments.
With KindBridge
A locality change first prompts kind-identity comparison. When the same base kind continues, select the receiving signature and adaptation declaration without a KindBridge; evaluate the receiving candidate afresh when that use is candidate-bearing. When two independently identified kinds are distinct and an exact directional correspondence is relied on, establish the C.3.3 KindBridge, its assertion, the receiving declaration, and any needed adaptation-correspondence declaration. Only justified bridge penalties affect R; F, G, admissibility, and classification truth remain unchanged.
With Guards
Guard_KindUseAdaptation designates exact adaptation and base-signature editions, checks admissibility, evaluates an admissible candidate, checks Scope separately, and keeps not-applicable, unknown, and refusal distinct. For distinct-kind cross-local use, it composes with the C.3.3 guard only after the bridge and receiving declarations are recoverable. For same-kind reuse, it performs the fresh receiving evaluation without a bridge.
Anti-patterns and Repairs
Worked Examples
Vehicle@ABSOnly Constraint Use
VehicleABSUse-2026 designates Vehicle, pins its signature, and adds the governed candidate condition that the vehicle has ABS. For this example, the base signature's candidate domain is physical vehicles. A vehicle in the declared slice is admissible. With a true base judgment, the adaptation returns true if the vehicle is known to have ABS, false if it is known not to have ABS, and unknown if that condition is unresolved; a non-vehicle input is not-applicable.
Surface, rig, and time conditions used only to bound the claim remain Scope. If ABS becomes a stable classification distinction, recover another kind and test its subkind relation separately.
AuthenticatedRequest@Frontend Vocabulary Use
FrontendAuthHeaderUse-2026 binds authHeader to local spelling X-Auth and adds no candidate condition. Its judgment therefore equals the admissible base judgment.
Moving the same request kind to another team requires a fresh receiving evaluation when the receiving use applies the declaration to a request, but no KindBridge merely because the team or spelling changed. If two independently identified request kinds differ, establish any bridge separately.
AdultPatient@Clinic Composite Use
ClinicAdultPatientUse-2026 pins the base adult-patient signature. For this illustrative composite use, it binds local spelling ClinicAdultPatient to the base kind's AdultPatient designator and adds the candidate condition ageAt(patient, slice) >= 21, with age expressed in years; the chosen clinic and claim window remain separately governed scope/applicability values.
At an applicable slice, a person in the declared candidate domain is admissible; with a true base judgment, unavailable birth-date information needed to decide the age condition yields unknown.
In Jurisdiction Y, first compare the exact patient-kind membership distinctions. If the same kind continues, use the Y declaration and evaluate afresh without a bridge. If the threshold or interpretation makes a distinct target kind and a directional correspondence is relied on, establish the KindBridge. A separate adaptation-correspondence declaration may then state how the two exact use declarations differ. Neither object transfers source truth.
Authoring and Review Guidance
A card can express an adaptation or correspondence declaration, or point to its exact edition and show the content needed for the receiving use. Keep the declaration's claim-bearing content distinct from the card's arrangement as publication form and from its carrier under C.2.P. A pointer-only card needs an accessible source for the declaration content required by the use. The declarations retain all applicable conditions in section 6; changing their presentation does not change the base kind.
Rules of thumb:
- Keep candidate conditions small and governed.
- Check ValueKind and applicability before the three-valued judgment.
- Put claim-scope conditions in Scope, not kind identity.
- Treat locality as a comparison cue. Require a bridge only for distinct kinds with an obtaining correspondence.
- If several teams reuse one stable conceptual constraint, review whether another kind is warranted; reuse alone establishes none.
Reviewer questions:
- Are the exact base kind and declaration editions recoverable?
- Is the type—constraint, vocabulary, or composite—correct?
- Are candidate conditions, applicability, and ClaimScope separated?
- Does evaluation distinguish
not-applicable,true,false,unknown, and guard refusal? - On locality change, was kind identity compared before any bridge was claimed?
- For distinct-kind use, do the bridge predicate, receiving declaration, any required candidate judgment, any adaptation correspondence, and only justified R consequence remain separate?
- Does a stable conceptual distinction warrant another kind, or is the declaration sufficient?
Conformance Checklist
C.3.4:End
KindAT — Intentional Abstraction Facet for Kinds (K0…K3)
Use.
KindATis an informative planning facet on one localU.Kind. Its anchors—K0 Instance, K1 Behavioral Pattern, K2 Formal Kind/Class, and K3 Up-to-Iso—describe how the kind is intended to be treated. They guide planning and catalog search; they supply no classification or assurance result.
Normativity: Informative for anchors, heuristics, examples, and guidance; normative only for the usage rules that prohibit guard/composition use and constrain placement.
Status: Stable
Audience: engineering managers, architects, editors, and assurance leads.
Depends on.
- C.3/C.3.1: kind identity by candidate domain and membership distinction, obtaining
U.SubkindOfrelations, and kind continuity. - C.3.2: the separate
KindSignaturedeclaration episteme, four-input classification judgment, and optional pinned-edition extension representation. - C.3.3: the obtaining
KindBridgerelation and its separate bridge-assertion episteme carrying loss, evidence, admitted use, and anyCL^krequired by the consumed assurance or calibration account. - C.3.4: the
KindUseAdaptationDeclarationepisteme andKindUseAdaptationJudgment. - A.2.6, C.2.2, and C.2.3: Claim/Work scope, F–G–R, and formality assessed on the episteme that owns the expression.
Purpose
Teams need a quick answer to a planning question: is this local kind intended as a curated instance-like cohort, a behavioral pattern, a formal invariant-bearing kind, or a kind considered up to structural equivalence? The answer can guide where declaration rigor and assurance effort are likely to pay off. An ordinary description of those planning needs is sufficient when a shared tag adds no useful distinction.
KindAT records how one local kind is treated for catalogue and planning work. Comparing the levels of abstraction of two models requires a shared inquiry and interpretation (B.5.MR). Suppose one-day and three-day schedules both satisfy a service policy. A revised model keeps the earlier commitments and limits the interval to at most two days. If the one-day schedule is still feasible, the three-day schedule shows which previously possible answer the model excludes. The kind of service order can keep the same K tag throughout.
Use the governing objects for their respective decisions:
- the local kind and its order remain under C.3/C.3.1;
- the
KindSignatureremains a declaration episteme whose expression properties may change under C.2.3; - for an admissible request,
J(candidate, kind, signatureEdition, slice)remainstrue,false, orunknown; - any
KindExtensionremains a pinned-edition representation of true candidates; and - bridge and kind-use adaptation objects retain the ontology assigned by C.3.3 and C.3.4.
Orthogonality and rationale
- G says where a claim or Work use applies.
- a local kind says what the declaration and claim are about.
- a KindSignature declaration episteme states the reusable criterion; assess its formality under C.2.3 when the receiving use needs that assessment.
- R says how well a receiving claim or use is supported.
- KindAT is only a catalog and planning tag on the local kind.
Anchors K0…K3 (informative)
K0 — Instance-level
Intent. The local kind is used for named exemplars or a curated cohort.
Cues. The reusable criterion, when one is needed, relies mainly on direct identity features or an enumerated bounded candidate domain.
Non-example. A stable invariant-bearing distinction belongs nearer K2 even if few candidates are currently known.
Planning. Prefer slice-bound judgments and assurance over the current candidate domain. When reuse depends on matching named candidates, make that correspondence explicit. Any KindBridge assessment follows the actual distinct-kind correspondence and its preservation evidence.
K1 — Behavioral pattern
Intent. The local kind is recognized through repeatable behavior or role-like performance rather than a mature formal invariant set.
Cues. A KindSignature may use controlled prose, behavioral obligations, or executable acceptance predicates.
Non-example. A kind with stable explicit predicates and order relations belongs nearer K2.
Planning. Invest in making the signature criterion evaluable and in testing behavioral diversity. For a bridge that compares behavioral distinctions, its assertion states the demonstrated preservation and loss.
K2 — Formal kind/class
Intent. The local kind has explicit invariants, relations, and a reviewed position in a local kind order.
Cues. A reusable KindSignature declaration episteme pins predicate-like criteria, dependencies, and reference scheme; judgments are replayable under editions.
Non-example. An informal cohort or role cue does not become K2 merely because it is stored in a schema.
Planning. Develop the expression properties the receiving use needs under C.2.3; plan assurance across relevant subkinds and boundary cases. KindBridge assertions may support medium or high CL^k only from demonstrated signature/order preservation.
K3 — Up-to-Iso
Intent. The kind's governed criterion is invariant under a declared isomorphism or equivalence notion.
Cues. Structural equivalence, rather than individual identity, is load-bearing in the signature and receiving use.
Non-example. A class whose candidate identity matters beyond the declared structure is not K3.
Planning. For a use that relies on equivalence, make the equivalence witnesses explicit and establish receiver acceptance. High CL^k is justified only when the obtaining bridge and its assertion demonstrate preservation of the relevant equivalence structure.
Manager heuristics (informative)
These are planning cues, not default F values, R values, classification results, or CL^k assessments.
Misuse and antidotes (informative)
- “Higher KindAT means wider G.” Only the scope governor changes G.
- “Gate on KindAT.” Use the classification, scope, evidence, and policy predicates required by the receiving guard.
- “Depth in
U.SubkindOfdetermines KindAT.” The facet concerns intentional stance, not graph depth. - “KindAT belongs on the claim or signature.” The tag is on the local kind; a catalog may represent that assignment.
- “A reused kind-use adaptation declaration has been promoted automatically.” If the distinction is conceptual and stable, separately identify a local kind and establish any obtaining
U.SubkindOfrelation under C.3.1. - “KindAT rates quality.” Formality belongs to the relevant episteme; assurance concerns a target claim and a named receiving use under B.3.
Usage rules (normative)
AT-01 (Facet, not Characteristic). KindAT SHALL be treated as the informative catalog and planning facet defined here. It has no algebra or threshold and MUST NOT appear in guard predicates or composition math.
AT-02 (Placement). If recorded, KindAT SHALL characterize one local U.Kind under an effective reference scheme. A catalog row may represent that assignment. KindAT MUST NOT be attached to a claim, capability, KindSignature episteme, candidate, judgment, or extension as a substitute for its own governor.
AT-03 (No F–G–R effect). Editors SHALL NOT imply that a higher KindAT widens G, raises the signature episteme's F, increases R, or changes a classification value. Any such sentence MUST name the declaration, scope, evidence, or receiving-use change.
AT-04 (Bridge neutrality). Editors assign KindAT from the kind's intentional stance. The bridge-assertion episteme may record an informative anchor comparison, but CL^k remains a separate assessment of the admitted bridge use from demonstrated signature/order preservation and loss.
AT-05 (Catalog representation). When KindAT is represented in a catalog, the catalog SHOULD identify the local kind and effective reference scheme. If the catalog also uses a signature, subkind relation, kind-use adaptation, bridge or extension, it SHOULD reference the corresponding declaration edition, relation or representation separately. Absence of a tag means “not set”, not K0.
Authoring and review guidance (informative)
Fast rubric
- Concrete exemplars or a bounded cohort suggest K0.
- Behavioral obligations with few stable global invariants suggest K1.
- Explicit invariant-bearing criteria and a reviewed local order suggest K2.
- An explicit, load-bearing equivalence notion suggests K3.
Review questions
- Is the tagged object the local kind rather than its signature, card, candidate, claim, or extension?
- Does the anchor describe the kind's intentional stance rather than the current number of candidates?
- Are proposed F and R changes stated as planning decisions over their owners rather than effects of KindAT?
- Does a stable distinction expressed in an adaptation declaration require a separately identified kind and independently obtaining subkind relation?
- When cross-context use relies on a KindBridge, does it recover the occurrence and separate bridge assertion instead of inferring congruence from the tag?
Integration notes (informative)
- C.3.1/C.3.2. KindAT may guide work on the signature declaration and assurance plan; it changes neither kind continuity nor the four-input judgment.
- C.3.3. KindAT may suggest what preservation evidence to seek. The bridge assertion carries loss, evidence, admitted use, and any
CL^krequired by the consumed assurance or calibration account. - C.3.4. Repeated adaptation-declaration use is a review cue only. A new local kind and any
U.SubkindOfrelation are established independently. - A.2.6. Scope remains G on the claim or Work use. KindAT never supplies coverage.
- C.2.3. The relevant declaration or claim episteme owns F. KindAT can motivate investment but cannot assign the value.
Worked mini-examples (informative)
- K0.
Account_US_GAAP_2025_Q1_Cohort: suppose the approved quarter roster defines this cohort. Check the account identifier against that edition of the roster. Counting the confirmed entries answers a separate question. - K1.
RetrySafeRequest: the declared criterion permits repeating a request after a lost reply without duplicating its intended effect. Check that property for the admitted failures, including a reply lost after the first execution completed. - K2.
Account: use explicit posting and balance invariants, test relevant subkinds, and evaluate each candidate with the signature edition and slice. - K3.
UndirectedGraphup to node relabeling: state the equivalence notion and require equivalence witnesses that preserve it. Any KindBridge use additionally needs the distinct-kind correspondence and preservation basis required by C.3.3.
Conformance checklist (normative)
C.3.5:End
Typed Guard Macros for Kinds + USM (Annex)
Summary. These guard macros combine C.3 declaration compatibility, C.3.2 judgment for candidate-bearing use, RoleMask and KindBridge declarations/relations, and A.2.6 Scope without collapsing them. A claim quantified over a kind can be checked at declaration level; applying that claim or a capability to one candidate additionally requires an admissibility check and, for an admissible request,
J(candidate, kind, signatureEdition, slice).true,false, andunknownremain classification values, while allow/refuse remains a separate guard disposition. KindAT never appears in a guard.
Normativity: Normative for macro obligations, evaluation order, three-valued/fail-closed discipline, and the conformance checklist; informative for decision trees, examples, and implementation-like skeletons.
Status: Stable
Audience: engineering managers, editors, reviewers, assurance leads, and authors of regulatory, evidence, ESG, and Method–Work checks.
Depends on.
- A.2.6 USM:
U.ContextSlice, Claim/Work scope,Gamma_time, scope bridges, and SpanUnion. - C.3/C.3.1: local kinds and obtaining
U.SubkindOfrelations. - C.3.2:
KindSignaturedeclaration epistemes, candidate/slice admissibility,J(candidate, kind, signatureEdition, slice),true/false/unknown, and optional extension representations. - C.3.3: obtaining
KindBridgerelations and separate bridge-assertion epistemes carrying loss, evidence, definedness, admitted use, and anyCL^krequired by the consumed assurance or calibration account. - C.3.4:
KindUseAdaptationDeclaration,KindUseAdaptationCorrespondenceDeclaration, andJ_kindUse. This annex abbreviates them asRoleMask,MaskAdapter, andJ_mask, respectively;roleMaskEditionnames the adaptation-declaration edition. ThusJ_mask(candidate, kind, kindSignatureEdition, roleMaskEdition, slice)retains C.3.4's five inputs. - C.3.5: KindAT as an editorial facet forbidden in guards.
- C.2.2/C.2.3 and Part B: F–G–R, formality on the owning episteme, bridge consequences, and scope congruence.
- A.15/A.15.1: the separation of capability, plan, actual Work occurrence, and every episteme about it.
Purpose and audience
Use this Annex when a receiving action must check one or more of these without blending them:
- the declaration-level compatibility of a claim's quantified kind with a consumer's expected kind;
- the classification of one exact candidate under one exact signature edition and slice;
- Claim or Work scope coverage;
- cross-context kind and scope bridges and their R consequences;
- a RoleMask declaration and exact masked judgment; or
- an actual capability use or Work occurrence whose input/output candidates are typed.
When a receiving rule already makes the needed predicates and disposition clear, apply it directly; no macro notation is required. The practical gain of these macros is a readable refusal reason. “The kinds are incompatible”, “the candidate is known not to satisfy the criterion”, “classification is unknown”, “scope does not cover”, “a bridge is unavailable”, and “the guard refuses use” remain different outcomes.
Problem
Using one two-valued “membership defined” question for several jobs hides the candidate and signature edition, conflates unavailable support with known failure, and makes refusal look like a classification result. A record or bridge assertion can then appear to establish target truth. In Method–Work use, a JobSlice, plan or log can likewise be mistaken for the performed Work occurrence.
C.3.A distinguishes two levels:
- declaration level: which exact local kind and
KindSignatureedition a claim quantifies over, and whether that kind is compatible with the receiver's expected kind under C.3.1 order or an obtaining KindBridge; and - candidate-use level: whether one exact target-side candidate satisfies the receiver's exact declaration in that exact slice, with the claim-kind consequence supplied only by the already established C.3.1 order or obtaining KindBridge.
Shared outcome model
All guards obey these invariants.
-
Exact declarations. A kind designator never substitutes for the exact
KindSignatureedition needed by the use. -
Candidate check for candidate-bearing use. A universally quantified claim or proof can be checked for declaration compatibility without a wildcard candidate. Actual application, test attachment, capability input/output use, or other candidate-bearing action pins the candidate, checks admissibility, and evaluates the four-input judgment only for an admissible request.
-
Admissibility, then three classification values. Check candidate and slice admissibility under the pinned declarations before any C.3.2 or C.3.4 judgment below. An inadmissible request gives
not-applicable; no judgment is formed. For admissible inputs,truemeans the criterion is known to hold;falsemeans it is known to fail;unknownmeans the evaluation cannot settle because evidence or a declared dependency is unavailable. -
Separate guard disposition. A guard returns an action disposition such as allow or refuse. An inadmissible request causes refusal without a classification judgment. Both
falseandunknownnormally cause fail-closed refusal, but the guard MUST preserve which classification value it consumed. -
Scope separation. Scope coverage is a USM predicate over a named slice. It does not classify the candidate or repair kind compatibility. Scope translation enters only when exact local senses require it under A.2.6; a changed locality or scheme alone does not trigger it.
The Scope Bridge shorthand below denotes the obtaining F.9 Bridge, while a separate affirmative C.2.1 claim states this translation's direction, rule, and permitted loss. Reliance uses the current A.10 disposition or, when a named assurance claim is current, a B.3 result supporting this same use.
-
Bridge separation. An obtaining KindBridge relation connects exact, independently identified distinct source and target kinds when its directional correspondence predicate holds under C.3.3. Its separate bridge assertion supplies mapping, loss, evidence, definedness, admitted use, and any
CL^krequired by the consumed assurance or calibration account; neither object creates a target kind, signature, or judgment.A locality change alone supplies no such relation. Same-kind reuse selects the receiving declaration edition and, when the receiving use is candidate-bearing, a fresh receiving judgment.
-
R-only consequences. Justified scope- and kind-bridge consequences affect R only. They do not change F, G, or classification truth.
Normative guard macros
Names such as Guard_TypedClaim are editorial handles. A context may alias them only when the same objects, values, and refusal distinctions remain recoverable.
Guard_TypedClaim — declaration-level admission
Intent. Decide whether claim C, quantified over local kind k_claim, may enter a receiving use restricted to kind k_receive in TargetSlice, without claiming anything yet about an unnamed candidate.
Guard_TypedClaim(C, k_claim, claimSignatureEdition, k_receive, receiveSignatureEdition, TargetSlice, thresholds?) SHALL:
- recover the exact
KindSignaturedeclaration episteme editions whose respectiveEntityOfConcernvalues arek_claimandk_receive, and whose evaluation domains and effective reference schemes cover the declared use; when both roles use the same kind and edition, state that identity rather than duplicating the declaration; - establish declaration-level kind compatibility:
- the kinds are identical or
SubkindOfObtains(k_receive, k_claim; effectiveReferenceScheme)holds under C.3.1, with an identifiedR_sub : U.SubkindOfoccurrence only when occurrence identity is needed; or - for a required directional correspondence between independently identified distinct kinds, an obtaining KindBridge relates exact source
k_claimand targetk_receiveunder the paired source and targetKindSignatureeditions, and a separate current bridge assertion states the mapping, applicability, loss, evidence, admitted receiving use, and anyCL^krequired by the consumed assurance or calibration account;
- the kinds are identical or
- require
U.ClaimScope(C)to cover the exactTargetSlice, preserve every declared slice selector, and name an explicitGamma_timewhen membership depends on time under A.2.6; - apply only the justified bridge consequences to R;
- check evidence freshness separately when the admission implies reliance; and
- check the policy-required expression properties on the exact claim or declaration episteme under C.2.3; a threshold is meaningful only on its named, justified local scale.
The subkind direction above is contravariant only for restricting a universally quantified claim: a claim over Vehicle may enter a PassengerCar-restricted use when PassengerCar is a subkind of Vehicle. It is not a generic compatibility direction for producer outputs, operation arguments, mutable positions, or arbitrary typed slots; each such use states its own variance rule.
The inclusion or correspondence must cover the receiving candidates for which the claim will be used. Inclusion established on one closed roster supports restriction to that roster. Applying the rule to an additional candidate needs grounds covering that candidate. For distinct kinds, establish that the mapping covers the required target cases and preserves the premise and consequence the receiver needs. An unrestricted target use requires that ground across its entire domain.
This guard MUST NOT invent an anonymous candidate or infer a candidate classification from declaration compatibility.
Guard_CandidateUse — apply a typed claim to an exact candidate
Intent. Decide whether claim C, quantified over k_claim, may be used for exact target-side candidate candidate in a receiving use restricted to k_receive.
Guard_CandidateUse(C, candidate, k_claim, claimSignatureEdition, k_receive, receiveSignatureEdition, TargetSlice) SHALL:
- identify the candidate under its direct governor before classification;
- perform the declaration, kind-compatibility, and Scope/time checks in
Guard_TypedClaimsteps 1–3 for the same editions and slice; - check receiving admissibility; refuse on
not-applicablewithout forming a judgment, otherwise evaluateJ(candidate, k_receive, receiveSignatureEdition, TargetSlice); - continue candidate-bearing use only on
true: for a proper subkind, the already establishedSubkindOfObtains(k_receive, k_claim; RS)supplies the monotone claim-kind consequence; for a bridged use, rely only through the obtaining KindBridge and its current assertion, without inventing a source-context candidate judgment; - refuse on known
falsewhile retaining that value; - refuse on
unknownwhile retaining the missing dependency or unavailable support reason; and - for a
truereceiving judgment, complete the R, freshness, and policy checks inGuard_TypedClaimsteps 4–6 before allowing use.
Evidence may support a classification assertion, but record presence, bridge presence, or guard invocation MUST NOT make the candidate satisfy the receiving criterion. When k_claim and k_receive are identical under one declaration edition, record that identity and evaluate the candidate once.
Guard_TypedJoin — compose typed producers and consumers
Intent. Compose producer A, which declares output kind k_A, with consumer B, which expects input kind k_B.
Guard_TypedJoin(A, k_A, edition_A; B, k_B, edition_B; TargetSlice) SHALL:
- pin both declaration episteme editions;
- establish output-to-input compatibility in the covariant flow direction:
- the kinds are identical or
SubkindOfObtains(k_A, k_B; effectiveReferenceScheme)holds; or - for a bridged flow, an obtaining KindBridge maps
k_Ato exact, independently identified distinct target-side kindk_A', its separate assertion carries the current mapping and loss basis, andk_A'is identical tok_BorSubkindOfObtains(k_A', k_B; targetReferenceScheme)holds;
- the kinds are identical or
- compute serial scope as the intersection of the two governed scopes and require coverage of
TargetSlice; - when a produced candidate enters B, check admissibility and evaluate
J(candidate, k_B, edition_B, TargetSlice)only for an admissible request; continue only ontrue, retainingnot-applicable,false, andunknownas distinct refusal grounds; and - route justified bridge consequences to R and check freshness before allowing use.
Declaration compatibility alone MUST NOT classify a future or actual output. Scope widening MUST NOT repair a type mismatch. The universal-claim variance rule in Guard_TypedClaim does not reverse this producer-to-consumer direction. For a promise covering every permitted output, check that compatibility holds over the producer's full permitted output domain. An inclusion proved on one batch covers that batch.
Guard_MaskedUse — exact RoleMask use
Intent. Use exact candidate candidate under a named RoleMask declaration in TargetSlice.
Guard_MaskedUse(artifact, candidate, kind, kindSignatureEdition, roleMaskEdition, TargetSlice) SHALL:
- recover the exact C.3.4 RoleMask declaration episteme, its base kind, pinned base signature edition, intended use, candidate-feature constraints, bindings, dependencies, and definedness;
- for cross-context use, compare base-kind identity and recover target declarations; when this use requires a correspondence between distinct kinds, establish the KindBridge relation and assertion under C.3.3 and recover any needed
MaskAdapterdeclaration; - check artifact scope separately through USM;
- check receiving admissibility; refuse on
not-applicablewithout forming a judgment, otherwise evaluateJ_mask(candidate, kind, kindSignatureEdition, roleMaskEdition, TargetSlice); - continue only on
true, refuse while preserving knownfalse, and fail closed while preservingunknown; - keep context predicates out of the candidate-feature criterion; and
- apply any justified bridge consequence to R and complete required freshness and policy checks before allowing use.
A mask name is not a kind synonym. Repeated mask use can trigger review for a separately identified local kind and independently obtaining U.SubkindOf relation; no guard or catalog action performs that admission.
GuardSpanUnionTyped — parallel support lines
Intent. Publish SpanUnion for the same typed claim supported by independent lines.
For each line, the guard SHALL:
- recover the same governed claim, quantified kind, and signature edition;
- perform the declaration, kind-compatibility, and Scope/time checks in
Guard_TypedClaimsteps 1–3 for that line's slice; - when a line's evidence is candidate-specific, bind each exact candidate, check admissibility, and retain its judgment when admissible;
- complete that line's R, freshness, and policy checks, preserving its bridge consequences;
- provide the USM independence justification; and
- include no slice outside the union of covered line scopes.
If lines quantify over different kinds, normalize through separately justified kind relations or publish distinct claims; do not hide the difference in SpanUnion.
GuardXContextTyped — cross-context typed reuse
Intent. Reuse claim C from a source context in target TargetSlice while keeping scope translation, kind correspondence, and target classification separate.
Guard_XContext_Typed(C, sourceKind, sourceSignatureEdition, targetKind, targetSignatureEdition, TargetSlice, candidate?) SHALL:
- compare source and target kind identity and establish declaration-level compatibility under §4.1 steps 1–2; if that compatibility relies on a directional correspondence between distinct kinds, recover the obtaining KindBridge and its assertion with pinned scheme/signature editions, mapping, definedness, loss, evidence, admitted use, and any
CL^krequired by the consumed assurance or calibration account; - recover the independently identified target
KindSignatureedition; - when the receiving claim requires Scope translation, recover the obtaining Scope Bridge and its applicable congruence assessment, the separate affirmative translation-use claim, and the current reliance branch under A.2.6;
- require Claim scope, translated when needed, to cover the exact
TargetSlicewith all its declared selectors, including anyGamma_timerequired by A.2.6; - when the receiving use is candidate-bearing, check receiving admissibility and evaluate
J(candidate, targetKind, targetSignatureEdition, TargetSlice)only for an admissible request; continue candidate-bearing use only ontrue, retainingnot-applicable,false, andunknownas distinct refusal grounds; - apply only justified scope- and kind-bridge consequences to R and complete required freshness and policy checks; and
- make the separate allow/refuse decision.
A source judgment may support reliance but MUST NOT be copied as target truth. For declaration-level use, omit step 5 and complete the use without forming a candidate judgment.
Evaluation semantics and order (normative)
E-01 (Order). Recover exact declarations and kind compatibility first; check Scope coverage second; when the receiving action is candidate-bearing, check admissibility and evaluate the candidate judgment only for an admissible request; then apply R consequences, freshness, and policy thresholds before the separate action disposition.
When composing macros, reuse checks whose inputs and receiving use are unchanged, and apply the same justified bridge consequence only once on the receiving R path.
E-02 (Determinism). With fixed candidates when any, kind/signature editions, slices, bridge/assertion editions, dependencies, and time selectors, the judgments and guard predicates MUST be reproducible. Implicit “latest” is forbidden.
E-03 (Admissibility and three values). Every C.3.2 or C.3.4 classification consumed by a guard MUST retain true, false, or unknown. Keep not-applicable outside the classification range. Missing evidence or an unavailable declared dependency in an admissible evaluation MUST NOT be coerced to false.
E-04 (Fail-closed without truth rewrite). A required check returning false, unknown, or not-applicable, a missing declaration, non-obtaining relation, unavailable bridge assertion, or uncovered Scope causes refusal. The refusal is not itself a classification value or an assertion that the relevant world-side relation fails to obtain.
E-05 (Bridge composition and consequence). Apply C.3.3:KB-10 to a relied-on chain. Establish joint definedness and preservation of the source distinctions needed by the receiving use from its mappings and intermediate interpretations. A minimum of comparable link ranks is a link summary, not that preservation result. The receiving R path records each relied-on bridge and assertion and applies only its justified consequences; neither F nor G nor a judgment value is modified.
E-06 (Predicate separation). Declaration compatibility, candidate admissibility, candidate classification, Scope coverage, evidence freshness, bridge applicability, capability fit, and action disposition SHALL remain separately inspectable predicates.
Conformance checklist (normative)
Proven-equivalent aliases
A context-specific guard alias is equivalent only when all required objects, inputs, classification values, bridge distinctions, and disposition boundaries can be recovered. Similar wording or the same final allow/refuse bit is insufficient.
Bridge consequences
When a receiving R model uses penalties Phi(CL_scope) and Psi(CL_kind), each penalty is monotone non-increasing in its own governing congruence scale. The model must justify that numerical consequence for the receiving use. This Annex prescribes no numeric form and performs no arithmetic on F or G.
Decision trees (informative)
D1 — Admit a quantified claim.
- Pin the quantified claim kind, receiving kind, and both exact signature editions.
- Require the receiving kind to be identical to or a subkind of the claim kind, or establish the receiving use's required correspondence between distinct kinds through the exact source-claim to target-receiving KindBridge relation and assertion.
- Check Claim scope against the exact TargetSlice, preserving its declared selectors and any
Gamma_timerequired by A.2.6. - Apply R consequences and freshness/threshold checks.
- Return the separate action disposition. Do not ask for a candidate unless the receiving use applies the claim to one.
D2 — Apply the claim to a candidate.
- Identify the candidate under its direct governor.
- Complete D1 steps 1–3: declarations, kind compatibility, and Scope/time.
- Check receiving admissibility. On
not-applicable, refuse without a judgment; otherwise evaluate the four-input target judgment. - On
true, use the established order or bridge for the claim-kind consequence and complete D1's R, freshness, and threshold checks before allowing use; onfalse, refuse as known failure; onunknown, refuse and retain the non-settlement reason.
D3 — Compose or cross a context.
- Pin source and target declarations.
- Recover declaration compatibility through identity, the required subkind relation, or an obtaining KindBridge with its separate assertion; recover Scope Bridge separately when Scope translation is required.
- Check the serial or translated scope.
- For a candidate-bearing receiving use, check admissibility of the output or other candidate and evaluate it under the target declaration only when admissible.
- Apply R consequences and decide separately.
D4 — Publish a union.
- Complete the relevant D1/D2 checks per line.
- Demonstrate support-line independence.
- Publish only the supported union; retain line-specific classifications and bridge consequences.
Guard anti-patterns and remedies (informative)
Worked examples (informative)
E1 — Same-context braking claim. A policy quantified over Vehicle pins VehicleSignature@v4; the receiver pins PassengerCarSignature@v3. Declaration admission establishes SubkindOfObtains(PassengerCar, Vehicle; plantVehicleScheme).
Applying the policy to VIN-17 evaluates J(VIN-17, PassengerCar, v3, S-plant); on true, C.3.1 monotonicity supplies the Vehicle-side consequence needed by the universal claim. If axle data are missing and the remaining facts do not settle this VIN’s classification, the judgement is unknown and the guard separately refuses the use.
E2 — Cross-plant reuse. An obtaining KindBridge relates source Vehicle to target TransportUnit; its assertion records a collapsed EV/ICE distinction and CL^k=2. The target signature is independently authored.
Plant B evaluates the candidate under its target edition. Suppose the source rule requires every mapped vehicle to carry the same identifier. It can apply where the mapping covers the target units and preserves that identifier requirement. A source rule specific to battery-electric vehicles needs to distinguish those vehicles. If Plant B's classification combines them with vehicles powered only by combustion, establish the propulsion type from another admissible source or restrict the rule to a supported subset. The CL^k rating alone supplies neither this distinction nor a numerical R decrement.
E3 — API adapter. A producer declares Request; a consumer expects AuthenticatedRequest. Declaration compatibility fails until an adapter and target declaration are recovered. For request req-884, suppose the other grounds establish neither the required condition nor its failure. If key-validation support is unavailable, the target judgement is unknown and the consumer refuses. An established failure of another necessary condition would already settle the target judgement as false.
E4 — Masked clinic use. The guard designates the exact AdultPatient@Clinic RoleMask declaration, base signature edition, candidate and slice. The mask label and a record entry can identify them. Age classification requires evidence for the applicable criterion. Suppose age of at least 21 is required. A verified age of at least 22 at that time satisfies the age condition. If the available record establishes only age of at least 18 and both 20 and 22 remain possible, the age condition is unknown.
Rationale
One final allow/refuse bit is operationally convenient but omits the distinctions behind the decision. Keeping declarations, candidate judgments, Scope, bridges, evidence, and disposition separate lets a reviewer see which repair is needed and prevents a guard from becoming a hidden relation, assertion, or evidence-to-truth converter.
C.3.A:Annex A - Regulatory and compliance alignment [A/I]
Purpose and fit
Regulations name categories such as Adult person, Class II medical device, Personal data, and Lease. A local context needs both a faithful category correspondence and explicit jurisdiction/version/time applicability. The kind channel answers “about what”; USM Scope answers “where and when”; neither answers whether one exact local candidate satisfies the target criterion.
Normative obligations
C-REG-1 (Regulatory declarations). Each used regulatory category SHALL be an exact authority-context local kind with a separately identified KindSignature declaration episteme edition. Any F value characterizes that episteme, not the kind.
C-REG-2 (Kind correspondence). Cross-context category use SHALL first compare authority and local kind definitions. If the receiving policy relies on a directional correspondence between distinct kinds, it SHALL recover an obtaining KindBridge relation between exact authority and local kinds plus a separate bridge assertion with mapping, pinned editions, preservation/loss, evidence, definedness, admitted use, and any CL^k required by the consumed assurance or calibration account.
C-REG-3 (Scope). Jurisdiction, effective dates, grace periods, and other contextual applicability conditions SHALL be Claim scope over exact context slices with explicit Gamma_time.
A product-family or platform distinction belongs in Scope only when it is a context-slice dimension of the claim; when it classifies the target entity, recover it as an exact kind and, for candidate-bearing use, an exact candidate judgment. A direct candidate feature remains with its own governor and SHALL NOT be treated as a Scope condition.
C-REG-4 (No synonym shortcut). A legal label, translation row, or policy card SHALL NOT substitute for the KindBridge relation, its assertion, or the target declaration.
C-REG-5 (Exact candidate use). Whenever a policy is applied to candidate candidate, the guard SHALL check receiving admissibility first. For an admissible request it SHALL evaluate J(candidate, localKind, localSignatureEdition, localSlice) and retain true, false, or unknown; otherwise it retains not-applicable without a judgment. Declaration compatibility alone is insufficient.
C-REG-6 (Consequences). Justified kind- and scope-bridge consequences SHALL affect R only. They SHALL NOT alter F, G, or the candidate judgment.
C-REG-7 (Editioning). For a law change that changes the criterion, identify the new declaration episteme and check its signature qualification and any edition relation under C.3.2:5. A change in claim applicability changes Scope. C.3.1 decides kind continuity. Guards SHALL pin editions and time and SHALL NOT rely on “latest”.
C-REG-8 (Local adaptation). A local nuance MAY use a RoleMask declaration. If it becomes a stable conceptual distinction, the context SHALL separately identify any new local kind and establish its obtaining subkind relation; mask reuse does not perform that change.
Regulatory guards
Guard_RegAdopt(P, candidate, authorityKind, authoritySignatureEdition, localKind, localSignatureEdition, S_local).
- Recover the authority/local declarations and establish their declaration-level compatibility under §4.1 steps 1–2; recover the KindBridge relation and assertion when the use requires a correspondence between distinct kinds, and check bridge applicability.
- Check P's governed scope and explicit time against
S_local. - Check local candidate admissibility; on
not-applicable, refuse without forming a judgment; otherwise evaluateJ(candidate, localKind, localSignatureEdition, S_local). - Continue only on
true; retain knownfalseorunknownbefore refusing. - Apply only the justified R consequence of each required, applicable bridge.
- Check freshness of relied-on regulatory and candidate support before allowing use.
Guard_RegChange(change, impactedDeclarations, impactedScopes).
- Decide whether the change alters criterion, reference scheme, applicability, or more than one.
- If the declaration’s criterion, interpretation or applicability changes, identify the required edition under C.3.2:5 and use C.3.1 to settle kind continuity. A change confined to the policy’s jurisdiction or effective period can leave the entire classification declaration in force.
- Update Scope independently when jurisdiction/version/time coverage changes.
- Reassess whether the receiving use now requires a correspondence between distinct kinds, and, when it does, the obtaining KindBridge relation and its assertion's mapping, loss, evidence, admitted use, and any
CL^krequired by the consumed assurance or calibration account. - For an affected candidate use, apply the declaration in force in the receiving slice. Preserve judgements already made under their own editions and slices. If the only changed decision concerns where the policy applies, reassess its coverage and reuse classification results whose inputs and grounds still hold.
Guard_RegXContextUse(P, candidate, sourceKind, targetKind, targetSignatureEdition, S_target). Use Guard_XContext_Typed to check that the candidate and slice are admissible before evaluating the target criterion. If the admissibility test returns not-applicable, the guard refuses the use without a classification judgement. For an admissible request, a missing target dependency yields unknown if the available facts leave the criterion unsettled. The bridge assessment supplies neither the missing fact nor the judgement.
Worked examples [I]
Adult dosage across jurisdictions. In this example the authority kind AdultPerson@RegY uses threshold 18 and the hospital kind AdultPatient uses 21. The obtaining KindBridge and its assertion state the boundary loss and CL^k=1. The hospital evaluates its target signature edition in the dated formulary slice. For candidate P-44, suppose the only age information is that P-44 met the authority threshold that day. Ages of 20 and 22 are both compatible. The target age condition remains unknown, and the guard refuses a use requiring positive classification. An attested bound of at least 21 would settle the hospital's age condition under these rules.
GDPR and CCPA. Two source kinds relate to independently identified product-context kinds through separate bridges/assertions. Each policy has its own jurisdiction/time Scope. A data item is governed by a fresh target judgment; an alias table is support, not classification.
Product classification and shipping. Evaluate the product against its category in the target signature edition. The shipping policy uses that judgement together with the particular shipment, destination, intended use and date. Establish the kind correspondence or scope translation where each is needed. If the shipment exceeds the policy’s residual-risk threshold for automatic release, the policy may call for manual review even when the product classification is settled.
IFRS and US GAAP Lease. Each authority kind and local corporate kind remains independently identified. The bridge assertion records the short-term-exception loss. Test planning targets boundary candidates under pinned target declarations rather than treating one shared label as truth.
Guidance and migration [I]
- Inventory regulatory claims, exact category declarations, and applicability slices.
- Recover or author target
KindSignaturedeclaration editions; keep F on those epistemes. - Compare kind identity; establish KindBridge relations and separate assertions with loss and admitted use only for required correspondences between distinct kinds.
- Rewrite candidate-bearing guards to pin candidate, local kind, signature edition, and slice.
- Preserve
unknownand record refusal separately. - Route Scope through USM and bridge consequences through R.
- Use RoleMask declarations for local procedural tailoring; separately establish a new kind/order relation only when the distinction becomes conceptual and stable.
Manager's compact pattern [I]
- Where and when? Claim scope over exact context slices.
- About what? Exact local kind and signature declaration; KindBridge relation/assertion for a required correspondence between distinct kinds.
- Which exact thing? Fresh target
J(candidate, kind, signatureEdition, slice). - Can we act? A separate guard disposition after scope, judgment, bridge, freshness, and policy checks.
C.3.A:Annex B - Assurance lanes and evidence design [A/I]
What typed assurance adds [I]
Verification (VA) can establish a consequence for all members of the declared kind under the stated assumptions. Validation (LA) can test particular candidates and boundary cases. On reuse, check the types and meanings the argument connects (C.2.2:4.6). If the evidence depends on a tool, check separately whether the version used was qualified for that operation. A classification judgement still needs the candidate facts and the applicable criterion.
Normative obligations
EA-1 (Declaration and candidate binding). Every VA/LA artifact SHALL cite the governed claim, exact quantified kind and signature edition, and assumed Scope. Candidate-specific evidence SHALL additionally name each exact candidate and its four-input judgment.
EA-2 (Subkind coverage). A claim over kind k SHALL justify coverage over relevant obtaining subkind relations and paired signature editions. RoleMask rows may cover named procedural uses but SHALL NOT silently stand in for a stable subkind.
EA-3 (Three values in evidence use). A test, observation, or proof may support a classification assertion. An unavailable evidence dependency yields unknown when it leaves the criterion unsettled. Preserve any independently settled result, including a known false conjunct under C.3.4. Failed retrieval itself MUST NOT be recorded as candidate false.
EA-4 (Independent unions). SpanUnion SHALL include a support-line independence account and preserve per-line candidate judgments and bridge consequences.
EA-5 (Bridges). Cross-context evidence use SHALL recover a Scope Bridge when the receiving claim requires Scope translation and a KindBridge relation/assertion when it relies on a correspondence between distinct kinds. It SHALL use the independently identified target declaration and, when the receiving use is candidate-bearing, evaluate the target candidates afresh. Consequences affect R only.
EA-6 (Freshness). Evidence windows and tool/declaration editions SHALL be explicit and tied to the governed slice. Expiry causes refusal or unknown at the predicate it disables; it does not widen Scope.
EA-7 (Tool qualification). Tool qualification SHALL remain distinct from checking the types and meanings used in an inference, proving a claim, establishing candidate facts, classifying a candidate and deciding whether to rely on the result.
EA-8 (No scope-by-wording). More general wording, more matching candidates, or additional evidence-matrix rows SHALL NOT widen G. A ΔG+ change requires the new support or sufficiently congruent bridge basis required by A.2.6; otherwise retain or narrow the declared Scope.
Evidence matrix [I]
A matrix can make declarations and slices jointly inspectable. A sufficient ordinary test record may carry the same content, directly or through accessible exact returns.
Rows plan declared distinctions; they do not classify every candidate. A proof-only row may remain declaration-level when it proves a universal claim. A test or monitoring row becomes candidate-bearing and records exact judgments for the exercised candidates.
VA lane [A/I]
- VA-1. A proof carrier SHALL cite the exact claim, quantified kind,
KindSignatureedition, and assumed scope slices. - VA-2. A proof of a universal claim need not invent a candidate; application to an actual candidate uses
Guard_CandidateUseseparately. - VA-3. Cross-context proof reliance SHALL recover the target declaration and any required bridge channels, with their loss and R consequences.
- VA-4. If the proof check relies on a particular kernel, establish that the kernel used is qualified for this check. The declaration retains its expression properties, and classifying a candidate still requires the applicable criterion.
Example: a proof over PassengerCarSignature@v4 assumes a dry-road slice. Reuse at Plant-B requires kind-identity and scope settlement, with bridges only where required. Application to VIN-17 then uses the Plant-B target signature and target judgment.
LA lane [A/I]
- LA-1. Each test or monitoring campaign SHALL make its exact declaration editions, context slices, tested candidates, and their judgments recoverable, directly or through accessible exact returns.
- LA-2. Boundary probing SHALL distinguish criterion boundaries from Scope boundaries.
- LA-3. A KindBridge assertion that records collapsed distinctions SHALL prompt assessment of their effect on the intended claim and use. Where those distinctions matter, repair coverage, restrict the supported use or refuse reliance. The bridge assertion does not alter target truth.
- LA-4. Freshness and SpanUnion independence SHALL remain explicit.
Example: rows PassengerCar and LightTruck use pinned signature editions; columns cover dry/wet slices. The tested VINs are exact candidates. If the missing sensor reading leaves one VIN’s classification unsettled, record unknown for that VIN.
Qualify a tool for its intended use [A/I]
Qualify provers, checkers, measurement pipelines, and classifiers separately. A classifier output can support an assertion about J; the tool neither becomes the candidate nor makes the governed criterion hold. Version drift may make the support unavailable and hence produce unknown for a candidate-bearing use.
- B.6-1. When a relied-on proof, observation or correspondence between meanings depends on tool qualification, the receiving guard SHALL recover the tool version used, the operations and conditions its qualification covers, and its qualification status at the time that matters for this use.
- B.6-2. Missing or weaker tool qualification MUST NOT be hidden by lowering the owning episteme's F or widening G. The receiving policy may require additional independent support, reduce or condition R, or refuse the use while preserving the exact unavailable-support reason.
Evidence guards
Guard_EvidencePlan_Typed SHALL check the plan's exact declaration editions and context slices, its bridge/assertion needs, and the candidate-selection, freshness, independence, and tool qualification required by the planned use. These inputs may be supplied directly or through accessible exact returns. A planned classification does not count as a candidate judgment.
Guard_EvidenceAttach_Typed SHALL make every evidence unit's binding to its exact claim/use, declaration edition, slice, candidate and judgment for candidate-specific evidence or a candidate-bearing receiving use, support relation, freshness, and bridge consequences recoverable, directly or through accessible exact returns. It SHALL preserve unknown and the separate attach/refuse disposition.
Anti-patterns and remedies
End-to-end example [I]
A two-plant braking claim pins the PassengerCar declaration and Plant-A scope. VA proves the quantified claim over that declaration. LA tests exact VINs in dry/wet slices and records their judgments. The test record identifies the tool version used and which measurements its qualification covers. Comparing the meanings of the measured quantities at the two plants requires its own evidence.
Plant-B reuse recovers the target declaration and any required bridges with their loss and R consequences; each Plant-B candidate is evaluated afresh before evidence is attached.
C.3.A:Annex C - ESG and Method–Work guards
ESG obligations (normative)
When a state transition publishes or relies on a claim quantified over kinds, the ESG guard SHALL:
- pin the claim, exact quantified claim kind, receiving kind, and both needed
KindSignatureeditions; - establish the correct subkind restriction direction or, for a required correspondence between distinct kinds, the exact source-claim to target-receiving KindBridge relation and separate assertion;
- check Claim scope against the exact slice with all its declared selectors, including any
Gamma_timerequired by A.2.6; - when one or more actual candidates are part of the transition, evaluate the four-input judgment for each candidate under the receiving kind and preserve all three values;
- when a RoleMask is used, recover its declaration edition and, when the transition is candidate-bearing, evaluate the exact masked judgment;
- apply justified bridge consequences to R only;
- check the policy-required expression properties and required freshness on their owners; and
- return a separate state-transition disposition.
ESG MUST NOT widen G to hide incompatibility, treat a label as a candidate judgment, or convert unknown to false.
Method–Work obligations (normative)
Open this Method–Work guard when capability, method and WorkPlan are part of the prospective or actual Work use. A transformation identified under A.3/A.3.4 may already be a sufficient result: natural, spontaneous or formal change, for example, can be described without claiming Work.
Apply A.13 when the claim concerns scale-free agency. If the receiving use also claims that particular Work occurred, establish that occurrence under A.15.1. Identify who acts and what changes. For A.12:4.2 Reflexive Split, establish two distinct participants and the part relation of each to their common containing holon. Where the work instead concerns joint dynamics without separable acting and affected participants, use the applicable dynamics, interaction or causality account.
Method–Work use has two different boundaries.
Prospective entry. Before execution, a guard may decide that a holder capability, method, intended U.WorkPlan, JobSlice, and candidate inputs are sufficient to start. That decision SHALL NOT claim that Work already occurred. The holder's actual ability, ability statements or currentness assessments, fit predicates, WorkPlan, JobSlice, and entry record remain distinct.
Actual result or acceptance. When the check concerns performed Work, the guard SHALL identify exact W : U.Work as an independently grounded, world-side, dated 4D Work occurrence under A.15.1. Keep W distinct from the U.Work kind, JobSlice, capability and plan. A log, result record or assurance record about W remains a separate object; when it is a claim-bearing episteme, its content designates W.
A conforming Method–Work check SHALL:
- pin every expected input/output local kind and signature edition;
- for cross-context candidates, compare kind identity and recover target declarations and any required bridges;
- require the capability's governed Work scope to cover exact JobSlice with explicit time;
- for every actual input candidate and, after execution, every actual output candidate relied on, check admissibility and evaluate the four-input judgment under its pinned expected declaration only when admissible, preserving all three values;
- use exact RoleMask declarations and masked judgments when procedural tailoring is used;
- apply any justified bridge consequence to R, then check capability measures, qualification/currentness, and fit as separately governed predicates before deciding entry or acceptance;
- before execution, return only an entry disposition and keep W absent;
- after execution, ground result or acceptance in the independently identified W;
- keep W, inputs, outputs, JobSlice, capability, plan, logs, and assertions distinct; and
- refuse fail-closed on
falseorunknownwithout rewriting either value.
Ready-to-use skeletons
ESG_TypedGate(Claim, claimKind, claimSignatureEdition, receiveKind, receiveSignatureEdition, TargetSlice, candidates?). When the transition is candidate-bearing, apply Guard_CandidateUse to each with both declaration editions; otherwise apply Guard_TypedClaim for declaration-only use. Reuse shared checks and bridge consequences under E-01, then return the separate transition disposition.
MethodWork_EntryGate(Capability, WorkPlanRef, JobSlice, inputCandidates, inputDeclarations). Apply §C.2's prospective checks in E-01 order and check required freshness. Return “entry allowed/refused”. Do not create or identify W.
MethodWork_ResultGate(W, JobSlice, actualInputs, actualOutputs, declarations, ResultRecordRef?). First recover the independently grounded dated W under A.15.1. Apply §C.2's declaration, Scope, admissibility, and input/output judgment checks in E-01 order, then complete any acceptance predicates. Keep any ResultRecordRef as a reference to a separate episteme whose content designates W.
Worked examples [I]
ESG braking policy. The claim pins VehicleSignature@v4 and the dry/wet TargetSlice. The consumer is restricted to PassengerCar, and SubkindOfObtains(PassengerCar, Vehicle; plantVehicleScheme) holds under the paired exact declaration editions.
For VIN-17, evaluate J(VIN-17, PassengerCar, passengerCarEdition, TargetSlice)=true; C.3.1 monotonicity then supplies the Vehicle-side classification needed by the universal claim. If the unavailable brake-configuration data leave this classification unsettled, the judgement is unknown and the transition guard separately refuses the use.
Risk-score Work entry and occurrence. ComputeRiskScore capability is considered for request req-884 in JobSlice api-v2.3/eu-west/t-204. The entry guard evaluates the request under the pinned AuthenticatedRequest signature. If true, it may admit execution; no Work occurrence yet follows.
After execution, actual W = RiskScoreRun-2026-07-22T10:03Z-884 : U.Work is independently grounded as the dated world-side occurrence. RiskScoreRunLog-884 is a separate episteme designating W. The output score value is a separate candidate evaluated under its declared output kind and signature edition.
Cross-context plant use. The source claim and source kind cross via separate Scope and KindBridge channels. Plant-B recovers its own target declaration and evaluates TransportUnit candidate TU-9. Bridge assertions affect R; they do not classify TU-9 or create the later Work occurrence.
Anti-patterns and remedies
C.3.A:End
Decision Theory (Decsn-CAL)
Type: Calculus (C) Status: Stable Normativity: Normative unless marked informative
At a glance. C.11 is the choice-calculus pattern for the moment when options exist and the working question is which option to choose, including whether another probe is worth its cost before commitment.
The alternatives may include a policy or a trajectory. Use C.43 when the practitioner must also construct alternatives, connect a choice to later observations and actions, or revise the course during execution.
Problem frame
When ongoing Work still needs a next action. If admissible actions still have to be recovered from ongoing Work, its domain Method, and current facts, use A.15.7 first. Return here only when a current chooser and OptionSet exist and comparison can change the result; a mandatory response or one familiar cue need not be expanded into an option set.
Use this when. Use this pattern when one DecisionSubject already has an OptionSet in hand and the question is how to choose among those already-available options under uncertainty, preference, causal or subjunctive dependence, and bounded probing or computation.
First useful result. Identify who makes this choice, the alternatives, what each could lead to, and the rule used to compare them. State which option or options the rule supports, why all fail, what further inquiry is worth doing, or where the remaining question belongs. When another party decides, that comparison can serve as advice. To report an actual decision, establish who made it or which authorized procedure put its result into effect.
When a choice is still open, give the recipient the comparison available now, the missing grounds and who can settle it. A supported recommendation or an unfinished comparison can be delivered before the holder decides.
First-minute questions.
- Who or what makes this choice? If several participants contributed, to whom is the decision attributed?
- What options are already on the table now?
- What current basis is being used to compare them?
- What could a further observation or calculation change about the choice or what the recipient can rely on?
- Is this still local choice, or has the question moved to a neighboring problem—for example, search, pool policy, selector-result declaration, publication availability, or enactment?
When advice still hides the decision. Use C.11.DUA when a recommendation or evidence demand still needs its receiving question, demanded work or requirement merits recovered. Return here when a current chooser and options support a local choice. An already adequate choice uses this pattern directly.
Typical reroutes. For candidate construction or development, use C.38–C.40 as specified in §4.4; that section also distinguishes pool policy, selected-set publication and enactment. Use A.13 when agency itself is at issue, or A.18/A.19 when Scale bindings or a declared CharacteristicSpace or reusable predicate are the unresolved question.
What the chooser gets. The same options can yield different recommendations if they indicate an outcome, bring it about, or depend on what another chooser will do. Identify the dependence that matters, then check which alternatives the current facts support and what further inquiry could change.
Not this pattern when. Do not start here when the current question is still generating candidate options, setting exploration or exploitation policy over a candidate pool, declaring selector-facing set-result content, making that result available to an audience, or sequencing execution under an operational plan.
Problem
Many systems have options on the table but still lack one explicit doctrine for what makes one option rational to choose. They mix together at least four different questions.
One question is still generating candidate options, variants, or open-ended search directions. Another question is governing how broadly a candidate pool should be explored or exploited before narrowing. A third question is planning, sequencing, replanning, or enacting the chosen option once a choice has already been made. The fourth question, and the one governed here, is choosing among already-available options under uncertainty, dependence, and bounded deliberation.
A second distortion appears when decision theory is reduced to a slogan about expected utility. Real choosers face evidential and causal distinctions, subjunctive or success-first cases, probe costs, information value, computation value, and situations where the chooser is not just one isolated individual.
Without an explicit place for choice calculus, search, candidate-pool policy, and planning get treated as the same question, while the doctrine of choosing among live options disappears behind generic talk about rationality.
Forces
Solution
Causal-use hook for choice records
When a choice depends on an effect, intervention, counterfactual, causal-policy, or off-policy claim, the ChoiceResult keeps the decision question local and cites the C.28 support result as one basis.
Omit the tail when causal support changes neither the comparison nor the chosen result. If causal wording changes the result, include the tail or downgrade the wording. A C.28 result does not choose, permit, or deploy an option; C.11 uses it with the other decision premises and may still select, defer, or abstain.
Primary EntityOfConcern and admissible choice result
C.11 governs theory-side choice among already-available options. Its selected decision result states what should be chosen from the current OptionSet, including whether further probing, information gathering, or computation is rational before the choice is fixed.
The OptionSet choice question begins only after an option set already exists. It does not form complete ways of obtaining one result, govern open-ended generation of options, or govern the execution order of a plan after a choice has already been made.
Decision discipline over a live option set
Apply the ChoiceRule to the available OptionSet. A completed comparison yields one explicit ChoiceResult under that rule. If the basis still leaves the requested single choice open, §4.2.4 explains what to return to the holder now.
-
Identify the chooser. Name the
DecisionSubject. UseDecisionSubjectGranularityto identify the unit to which the choice is attributed in this arrangement, for example a controller, a committee or an organization acting through a delegate. If attribution to a component or its containing system changes the result, establish their actual relation and which decision each makes. Distinguish the chooser from participants who provide evidence, advice or execution. UseA.13for a disputed claim about agency and the applicable assignment or mandate for a claim about who can decide. -
Freeze the current option set. Establish which alternatives the commission allows and which it asks you to compare. Use the available alternatives in the current comparison as one
OptionSet. A choice among the offers received can close on those offers. For a claim about the whole permitted domain, establish what an omitted option could change. One offer shown to meet the requirements can settle a request for a workable option. To claim a greatest option under the governing preference, establish that it is at least as preferred as every eligible alternative. Under a partial order an eligible option is nondominated when no eligible alternative strictly dominates it. To name all options attaining a best value, account for other alternatives that could tie. If the rows are only labels or fragments and the question is several complete ways to obtain the same result, stop here and applyC.38. UseC.39when a way to obtain the result still needs explanation, orC.40when usable material needs feasible variation and examination. UseC.18when generation-account or archive/front claims are needed. -
Make the comparison basis explicit. State the preference or measure, current facts and uncertainty, and the
OutcomeModelused to estimate each option’s consequences. Identify the use for which the options must qualify and any condition that excludes an option regardless of its score. Apply the same interpretation, horizon and material assumptions across options. The comparison is not usable if some options are being judged under one belief state and other options under one later, unmarked update. If rival outcome models give different rankings, compare all relevant options under each credible model. Show which conclusions hold across them and which depend on an unresolved assumption. A new measurement may update a parameter or the model. A disputed preference or mandate requires its own decision. -
Choose the dependence layer that governs the case. Start from the evidential baseline when the choice is being compared through likely outcomes under the current
BeliefState. Add oneInterventionModelwhen taking one option changes the world through intervention rather than mere observation. Add oneCounterfactualModelplus oneSubjunctiveDependenceRelationwhen the case depends on one predictor, one structurally linked chooser, or one decision-procedure coupling that intervention talk alone does not capture. Use the least-committing dependence layer that still covers the live case, and do not switch layers across options without saying so explicitly. -
Compare a live inquiry alternative before commitment. Consider another observation or computation when a plausible outcome could change the choice, the claim supporting it or the use for which it is admissible. Identify whose time or other resources the inquiry consumes. A mandatory check has the force and scope of its governing rule. Apply the probe-worthiness test to that alternative, stating the feasible probe, budget, cost and relevant information or computation value at the precision the choice needs. Otherwise this step adds no investigation or recording requirement. An empty form field does not activate inquiry, and no new search is required to certify that no other inquiry exists. Compare the best attainable next probe for this choice. A longer sequential experiment design can be a separate question. If an inquiry is binding, apply its governing terms to eligibility or use. When no worthwhile feasible probe remains, apply the choice rule to the evidence available. The cost of more inquiry can justify stopping. Selecting an option still needs grounds for its eligibility and preference. If a binding prerequisite remains unmet, keep the affected choice or use open, narrow it where the rule permits, or select an eligible alternative. If the expected gain warrants a feasible probe, return
probe againand name it. After someone performs the probe, use its actual result to update the comparison. An inconclusive observation can leave the ranking unchanged. If one choice result is already fixed and the remaining probe would change only execution-path description, call-plan ordering, enactment budget, or checkpointing of that chosen option, stop treating the probe as local choice doctrine and applyC.24. -
Apply the
ChoiceRuleand report what it supports. A completed comparison has one result form:choose now,reject current set,probe again, orreroute because this is no longer local choice. An open comparison can still inform the holder under §4.2.4. Forchoose now, give the option or retained set and its grounds. State whether the alternatives in a retained set are equally preferred or incomparable, and which further decision, if any, remains. If the result isreject current set, name the reason no current option survives under the present basis and, when more work follows, the neighboring question that now takes over. If the result isprobe again, name the next probe and the exact comparison defect it is supposed to repair. State the conditions under which the result follows from the rule. If another person or system must adopt the recommendation, the result can go to them now.
Receive heterogeneous premises without relabeling them as signals
An interesting observation, information-gain estimate, capability result, objective or reward, articulated former cue, or C.17 characterization enters this pattern only under its exact source/result identity. Use E.10.LRN first when learning wording still hides that identity. Preserve the source and support needed for the bounded premise use. Reuse an adequate existing account; use A.10 when the source-to-use account remains implicit or must be replayable. An applicable RelianceDisposition qualifies only that premise use.
An objective, reward, preference, or loss enters as the EvaluativeMeasure, PreferenceOrder, or ChoiceRule input it actually supplies and is not evidence by numerical form. When an A.16.1 cue pack remains current as a source or provenance episteme, preserve that source identity and route the separately articulated endpoint result through its direct claim owner. A C.17 novelty, surprise, use, or creativity characterization remains a characterization and does not license a move by itself.
No separate premise-qualification result sits between those owners and C.11. Use the qualified inputs in the live option/probe comparison and return only the existing ChoiceResult. When the missing comparison basis is specifically what a finite candidate contributes relative to the current configuration, use C.11.CRC to construct that ordinary comparison claim and return here. If every premise and the finite comparison are already explicit, proceed directly.
When a later outcome must first be compared with an earlier decision's actual relied-on basis, use C.11.DBR to localize the material difference and preserve unaffected uses. Return here when the current chooser, finite OptionSet and comparison basis are available. This prepares a particular reconsideration; it adds no qualification layer for otherwise ready premises.
Well-formed comparison state
Well-formedness constraint: a live C.11 comparison state is usable only when the decision record states the following, subject to the stated condition on inquiry:
- one
DecisionSubjectat oneDecisionSubjectGranularity; - one current
OptionSet; - one current comparison basis through
PreferenceOrderorEvaluativeMeasure, plus oneBeliefStateand oneOutcomeModel; - one active dependence layer for the current comparison, unless the record explicitly says that comparison is still being reopened;
- when an inquiry alternative is live, an account of its feasibility, cost and possible contribution to the choice; when a decision or later use needs a retained inquiry limitation or reason, that minimum content in the same decision result.
An inactive inquiry item requires no empty value, no no-probe statement and no waiver. The chooser, option set, shared comparison basis and applicable dependence layer remain necessary.
The comparison is still unfinished, not yet wrong but not yet closeable, when any of the following remains true:
- the comparison moves between a participant’s preference and the collective decision without establishing whose rule applies;
- the option set is still changing while the record also claims to rank the options;
- one option is being judged under one belief state and another under one later update that is not itself declared as the next probe result;
- one heavier dependence layer is invoked for rhetorical force, but the record never states what defect of the lighter comparison it repairs;
- the record asks for more information without identifying a feasible inquiry and a plausible result that would change which option qualifies, the preference between options, or the claim on which the recipient could rely.
Minimal admissible decision semantics
This minimal choice doctrine does not settle every decision-theory dispute, but it already supports some semantic distinctions by value and rules out others.
The comparison can use the following when their basis is explicit:
- an incomplete or partially ordered
PreferenceOrder, with unresolved alternatives retained explicitly, a worthwhile further probe, or an unfinished comparison; incomparability alone does not justifyreject current set, which requires the current rule to exclude every option; - one
EvaluativeMeasurefor magnitude, threshold, or trade-off-sensitive cases, so long as the measure being used now is explicit enough to explain why the current result follows under it; - one temporary unresolved criterion conflict, so long as the record says whether the present comparison is using one priority order, one threshold, one explicit trade-off measure, or one unfinished state that still blocks closure;
- one explicit
BeliefStaterevision, so long as it enters the comparison as one named probe result or one named model repair rather than as one silent basis shift; - a collective
DecisionSubjectwhen the decision is attributed to that body under its actual decision arrangement. If members disagree, establish the aggregation or delegation rule needed for the collective result.
The following are not admissible in this C.11 body:
- silently totalizing one partial preference relation just to force
choose now; - silently switching from one criterion mix or one belief state to another across options;
- treating unresolved cross-scale or cross-collectivity conflict as a settled local ranking before the aggregation question has been resolved;
- using a record's form as a substitute for a stated comparison doctrine.
A partial order can leave alternatives incomparable. Equal preference means that the rule treats two alternatives as tied. If the receiving task accepts a set, the retainedTieSet field can list the retained alternatives and state which of these conditions holds. If the task needs one course, identify the applicable tie-break or the decision still needed.
Probe-worthiness rule
Use the following three conditions to appraise an optional next probe for this choice. An independently binding condition continues to govern the affected option or use.
- the probe fits inside the remaining
ProbeBudget; - the expected improvement in the decision or its warranted use from the information or computation justifies the full
CostToProbe, including delay and work displaced; - a plausible result could change which option qualifies or is preferred, resolve a tie, reveal that the current set fails, change what use is warranted, or move the question elsewhere.
This is the local or myopic probe-worthiness rule for C.11: judge the best next feasible probe over the current OptionSet, not one whole non-myopic experiment design over longer horizons. Later sequential or non-myopic OED may strengthen this doctrine, but they do not relocate the local-choice question outside C.11.
Remaining uncertainty alone gives no reason to probe. State what plausible outcomes would change for the chooser. If every attainable result leaves the decision and its warranted use unchanged, this decision supplies no gain from that optional inquiry. An independently required check still follows its own rule.
If no worthwhile probe is available, state what the current basis and ChoiceRule support. If they leave the choice unresolved, name the missing premise or authority.
If the next probe would no longer change which option survives but would only change how one already-chosen option gets enacted, budgeted, or checkpointed, the question has already crossed to C.24.
ChoiceRule versus ChoiceResult
ChoiceRule and ChoiceResult are not the same kind of thing.
ChoiceRuleis the doctrine or operator that says how the current comparison basis, dependence layer, and any applicable probe-worthiness value support oneChoiceResult.ChoiceResultrecords the conclusion reached by applying that rule to the present comparison.
A usable ChoiceResult gives the selected disposition and the grounds for it. It can inform a later decision or take effect directly under a delegated decision procedure. To establish which occurred, use the record of that decision or procedure.
The record has four result forms:
choose nowreject current setprobe againreroute
A fifth soft result such as "keep thinking", "stay with the current view", or "the case is still complex" is not a conforming output. It is one unfinished state that still needs to be typed.
For choose now, the emitted ChoiceResult should show:
- the selected option or retained tie-set and the comparison basis supporting it;
- the outcome of a live inquiry comparison or a limitation needed by the decision or its recipient, when applicable.
For reject current set, the emitted ChoiceResult should show:
- that no member of the current
OptionSetsurvives under the present comparison basis; - the reason each option fails the stated conditions. Some options may fail for different reasons;
- the next neighboring question only when more work now follows, such as new option generation or one explicit escalation path.
For probe again, the emitted ChoiceResult should show:
- the next probe;
- the comparison defect that probe is expected to repair;
- the reason the probe is still worth its cost.
For reroute, the emitted ChoiceResult should show:
- the neighboring question and the pattern used to answer it;
- the reason this is no longer local choice among already-available options.
Closure rule over the current OptionSet
The comparison may close as choose now only when all of the following are true together:
- the current
OptionSetis stable enough that no further options are being invented for this comparison; - the common basis and rule identify an eligible preferred option, or a set that the receiving task accepts. State whether the retained alternatives tie or remain incomparable;
- no feasible optional probe offers enough expected improvement in the selection or its warranted use to justify its full cost, and the selection meets the applicable prerequisites;
- the record still concerns local choice rather than pool policy, selector-result declaration, publication availability, or enactment.
The comparison may close as reject current set only when all of the following are true together:
- the current
OptionSetis explicit and stable enough to reject as the present choice set; - the current comparison basis is explicit enough to show why no member survives under the present basis;
- no feasible optional probe offers enough expected improvement to justify its cost in reconsidering that rejection or its warranted use;
- the result is still one local choice conclusion rather than a pool-policy, selector-result declaration, publication-availability, or enactment result.
Assess the probe conditions above from the current basis and any live inquiry alternative. They are conditions of a warranted choice, not a requirement to produce a separate account of omitted checking.
The comparison should close as probe again only when all of the following are true together:
- one next probe is named by value;
- that probe fits the remaining
ProbeBudget; - its expected contribution to the choice justifies its
CostToProbe; - at least one attainable result could change which option qualifies or is preferred, which claim or use is warranted, or where the question belongs. State what an inconclusive result would leave open when that matters.
The comparison should close as reroute when the record shows that the decision question has changed:
- to
C.38when labels or fragments must first become complete-enough ways of obtaining the same result; - to
C.39when a way to obtain the result still needs explanation; - to
C.40when usable material needs feasible variation and examination; - to
C.18when the question concerns the generation account or archive/front stewardship; - to
C.19when the question is now how broadly to keep exploring or exploiting one candidate pool; - to
C.24when one choice result already exists and the next task is now sequencing, enactment, or execution-path probe work; - to
G.5when the next task is declaring or naming selector-facing selected-set content; when that result already exists, useE.17for its source-backed publication face and return to source andE.24.PUBfor the publication occurrence and audience availability.
If no closure condition holds, send the comparison the available basis supports. Identify the eligible alternatives, the missing comparison or priority, and who can settle it. For a requested single choice, retaining several candidates can inform the decision holder while their choice remains open. Record a ChoiceResult when the rule supports one of the four forms.
Minimal decision-record form
A minimal C.11 decision record has the shape below. Include ProbeDecisionValue only for a live inquiry alternative or an inquiry judgement whose content the decision or recipient needs. Include only the relevant information or computation value. A retained prose reason or limitation alone does not activate that block: put only the needed statement in ChoiceResult.
Exact syntax is unnecessary. The chooser, options, shared comparison basis, ChoiceRule and ChoiceResult are required. Inquiry content is required only under the stated condition; an absent block requires no placeholder or omission explanation.
Use branch language only when it changes the comparison being performed.
Resource-aware choice over an archive or front
- When the
OptionSetis drawn from aFront,Archive, or their composition, identify that source basis. - Apply one declared decision lens over that source family rather than assuming an unstated universal winner rule.
CostToProbe,ValueOfInformation, andValueOfComputationbelong to that lens-side choice doctrine.- They may justify another probe, one changed local comparison outcome, or one stop decision, but they do not rename the current
DominanceSetand do not declare one shortlist-family result. - If one candidate remains worth probing because its expected information value still exceeds its expected cost, say that explicitly as one lens-side choice judgement.
- If one archive point remains worth probing because it may change the frontier later, keep that as one resource-aware choice claim, not as evidence that the point is already on the current front.
- For an actual
A.19.SelectorMechanismapplication, retain its set-returning declaration and bindings:A.19.SelectorMechanismremains the cited set-return floorSelectionSlotremains the selector output floor- if later selector-facing set-result declaration is required, that set-returning floor may support one
Shortlistor oneRankedShortlistinG.5rather than one forced single winner
Dependence layers and neighboring repairs
Classical evidential baseline
Stay with the classical evidential baseline when the question is to compare already-available options through preferences, utilities or desirabilities, beliefs, and likely outcomes under uncertainty.
In this baseline, the options are being compared as evidence about what consequences are likely if they are chosen. This is the ordinary default when intervention structure, predictor-coupling, or context-sensitive non-commutativity do not affect the comparison.
Typical choice results are:
choose nowbecause the current sharedBeliefStateandOutcomeModelalready make one option or tie-set survive, and no still-feasible probe is worth its cost;probe againbecause one further observation, measurement, or comparison pass could still change the ranking without requiring a heavier causal, subjunctive, or context-order repair;reroutebecause the current decision question is no longer comparing one fixedOptionSet, but has become search, pool policy, selector-result declaration, publication-availability, or enactment work.
The baseline remains unfinished when the comparison cannot keep a shared BeliefState and OutcomeModel across the options, or when a limitation already calls for a further dependence layer and the comparison still uses only the evidential baseline.
Causal repair
Switch on one InterventionModel when taking one option changes the world through intervention rather than merely signaling which outcome was already likely.
The comparative question changes: the working question is no longer only what this option indicates about the outcome, but what this option causes in the outcome structure.
Typical choice results are:
choose nowbecause, under the declared intervention structure, one option now causally dominates or remains the survivor and no remaining feasible probe can change that comparison with enough expected gain to justify its cost;probe againbecause an uncertainty about the effect of an intervention prevents the current comparison and the proposed probe could change which option is preferred;reroutebecause the intervention-use question has already moved from local choice into enactment planning, protocol design, or one neighboring question.
The causal repair is incomplete if the comparison still treats options only as evidence after invoking causal language, or if an InterventionModel is named without stating what defect of the lighter evidential comparison it repairs.
Success-first or subjunctive repair
Switch on one CounterfactualModel plus one SubjunctiveDependenceRelation when Newcomb-like, blackmail-like, or other predictor-coupled cases remain under-described by the older evidential-versus-causal split.
What changes here is that the comparison must stay answerable to linked decision procedures, predictors, or structurally similar choosers rather than only to direct intervention on one local event.
Typical choice results are:
choose nowbecause, under the declared counterfactual or subjunctive structure, one option survives once the predictor-coupled comparison is made explicit;probe againbecause one further model clarification, predictor assumption check, or decision-procedure comparison could still reverse the current survivor relation;reroutebecause the governing decision question is no longer settling local choice doctrine but has become one wider characterization, negotiation, or enactment question that only borrowed predictor-coupling language.
If that coupled structure is not live, do not activate this branch. If a predictor-coupled or success-first repair is named but the linked structure that changes the comparison is still unspecified, the branch is not yet load-bearing in the current decision.
Active-inference neighboring repair
Bring the active-inference line into view when the chooser is embodied, online, and socially coupled, and when the decision cannot be understood as one disembodied choose-then-act moment.
The comparison is no longer over one frozen snapshot alone. It must now ask whether an additional observation, coupled update, or socially mediated or role-expectation clarification changes what should be done now.
This minimal choice doctrine makes that social-expectation pressure explicit, but it does not yet operationalize one full ROE or SocialExpectationRegime object model inside C.11. If the case hinges on that machinery, the decision record should state that this local C.11 floor does not settle it.
Typical choice results are:
probe againbecause one further embodied observation, coupled update, or explicit role-expectation clarification can still change the state estimate enough to reverse the current survivor relation;choose nowwhen further delay would close the window for an eligible option and the current comparison supports choosing it;reroutebecause the question has already become enactment sequencing or agent-characterization work rather than local choice.
C.11 keeps the choice question visible there, but A.13 still governs the narrower question of what kind of agent or agential system is in play. For measured characteristic and evidence claims, use A.17 for Characteristic identity and arity, A.18 for Scale and Coordinate bindings, A.19 for the declared space and reusable predicate, C.16 for the measurement account, and A.10 for bounded reliance on the evidence. C.24 still governs sequencing and enactment once a choice result has already been fixed.
Do not invoke this line only because one agent is acting in the world. Invoke it when embodied coupling, online updating, or explicit social-expectation pressure changes what the chooser should do now from the current OptionSet.
Quantum-like neighboring repair
Bring the quantum-like line into view when context effects, order effects, response-replicability tension, or incompatible-question structure change the comparative state enough that one simple commutative probability reading no longer fits.
What changes here is the practical structure of comparison. One order of questioning or one framing path may produce one different survivor relation from another. The comparison must therefore either stabilize the comparison under one declared order or show why one more clarifying pass is still needed.
This minimal choice doctrine keeps the branch at that measurement-sensitive recognition point. It does not yet claim one full quantum-like state-space package inside C.11; it claims only that the live comparison may need one explicit measurement-class or order-sensitive repair rather than one plain commutative reading.
Typical choice results are:
choose nowunder one declared order or framing because rival orders no longer change which option survives;probe againbecause one framing-sensitive comparison pass, one further question order, one response-replicability check, or one explicit measurement-class clarification could still reverse the survivor relation;reroutewhen the current decision question is no longer deciding among live options but has become one selector-result declaration, publication-availability, or enactment problem that only borrowed order-effect language rhetorically.
Do not promote this line to the unmarked default unless those repaired limitations are live in the case.
Do not invoke this line merely because a case feels psychologically subtle. Invoke it when one changed order, framing, response pattern, or incompatible-question structure changes the comparison state or the survivor relation in the live choice.
Use the evidential comparison when its dependence assumptions suffice for the decision. If two approaches recommend different options, identify the assumption responsible and what could decide between their uses.
These approaches can model the same choice under different assumptions. A causal or order-sensitive model may reverse a recommendation. To claim that one model strictly refines another, compare their feasible possibilities for the same inquiry and interpretation under B.5.MR. Establish what the second preserves and which previously possible answer it excludes.
Reroute as soon as the question stops being local choice
Use C.11 while the question remains: from this current OptionSet, what should the DecisionSubject choose, and is another probe worth its cost before commitment?
A completed ChoiceResult answers the local comparison. Return it to the continuation that needs it, preserving the evidence, information and action conditions used in the choice. Under C.43, a changed goal, available means or later observation can reopen the appropriate contribution without requiring the whole control task to become one fixed option set. The local next-probe rule below is not a general solution for all sequential information acquisition.
Reroute immediately when the question changes:
-
If the current rows are labels or fragments and the question is how several complete ways could obtain the same result, leave this pattern and work in
C.38first. -
If a way to obtain the result still needs explanation, use
C.39. If usable material needs feasible variation and examination, useC.40. UseC.18when the question concerns the generation account or archive/front stewardship. -
If the options already exist but the question is how broadly to keep exploring or exploiting the candidate pool, leave this pattern and work in
C.19, where the next useful output is one explicit pool-policy result rather than one localChoiceResult. -
If one option is already chosen and the question is how to sequence, budget, or enact that choice, leave this pattern and work in
C.24, where the next useful output is one enactment-facing call plan orCheckpointReturn. -
If the question has shifted from deciding to declaring or naming selector-facing selected-set content, leave this pattern and work in
G.5. Its next useful output may be aShortlistorRankedShortlistwhen alternatives remain for later choice, aJointUseSetwhen every named member is included for one bounded use, a narrowed handoff, abstain, or escalation. None is one more localChoiceResult.If that result already exists and the current question is presentation or availability to an audience, use
E.17for the source-backed publication face and return to source andE.24.PUBfor the publication occurrence and availability.
ProbeBudget covers the resources available for further observation or deliberation while the choice or reliance on it can still change. ValueOfInformation and ValueOfComputation express the expected improvement compared with deciding now. Once the choice and its permitted use are settled, the budget for performing or sequencing the selected work belongs to C.24. C.19 and C.24 may use these measures in their own decisions.
Outside this pattern remain candidate generation, pool-wide exploration policy, selector-facing set-result declaration, publication availability, and execution planning.
Minimal inventory and mathematical floor
The minimum usable inventory for this pattern is:
- subject and option objects:
DecisionSubject,DecisionSubjectGranularity,OptionSet; - evaluative and epistemic objects:
PreferenceOrder,EvaluativeMeasure,BeliefState,OutcomeModel; - dependence and comparison objects:
InterventionModel,CounterfactualModel,SubjunctiveDependenceRelation,ChoiceRule,ChoiceResult; - probe and bounded-resource objects:
ProbeActionSet,ProbeBudget,CostToProbe,ValueOfInformation,ValueOfComputation.
The applicable objects from this inventory are required because the decision record must carry one explicit path from the OptionSet through the ChoiceRule to the emitted ChoiceResult.
Always explicit versus conditionally activated objects
A completed C.11 decision record makes the following objects explicit. For a comparison sent while the decision is open, identify the usable premises and the missing basis under §4.2.4.
DecisionSubjectandDecisionSubjectGranularity;OptionSet;- one evaluative basis through
PreferenceOrderorEvaluativeMeasure; BeliefState;OutcomeModel;ChoiceRule;ChoiceResult.
The following objects activate when the case needs them:
InterventionModelfor causal repair;CounterfactualModelplusSubjunctiveDependenceRelationfor success-first or predictor-coupled repair;- For a live inquiry, identify its
ProbeActionSet,ProbeBudgetandCostToProbe. UseValueOfInformationfor new observations,ValueOfComputationfor additional reasoning over available evidence, or both when the inquiry needs both.
Use these names or ordinary wording that makes the applicable objects and their roles in the comparison recoverable.
Immediate lexical commitments:
- the default chooser term is
DecisionSubject, notAgent; DecisionSubjectGranularityidentifies the unit to which the decision is attributed. A committee may decide for itself or on behalf of an organization under a mandate. Where scale matters, identify the actual part and whole. B.5.MR compares the levels of abstraction or refinement of two models for a given inquiry; the same chooser can use both models.- use
A.6.Pif a relation's predicate or participants remain unclear, andA.6.5when its argument or slot structure needs clarification.
Local plain glosses for the load-bearing inventory:
DecisionSubject: the system making the decision, which can be a person, an automated controller, a committee or another organized body;DecisionSubjectGranularity: the unit to which the choice is attributed in this arrangement. A committee member can submit an individual ranking and the committee can decide by vote;OptionSet: the concrete options already on the table now;PreferenceOrder: the current better-than / worse-than ordering over those options for this decision subject;EvaluativeMeasure: the explicit utility-style or desirability-style scoring measure used when the case needs magnitudes, thresholds, or trade-offs rather than only one ordering;BeliefState: the current uncertainty-bearing state about the world, the case, and the likely consequences of the options;OutcomeModel: the model that maps options plus the current uncertainty picture to the consequences that matter for this choice;InterventionModel: the part of the model that says how the world changes because one option is actually taken;CounterfactualModel: the model used to compare relevant non-actual alternatives or alternate decision procedures;SubjunctiveDependenceRelation: the dependence between this choice and one predictor, one linked chooser, or one structurally similar decision procedure when intervention talk alone is not enough;ChoiceRule: the rule for evaluating the current alternatives and any worthwhile probe, and for reaching one of the four result forms;ChoiceResult: the disposition that follows from applying that rule, with the grounds the receiving use needs;ProbeActionSet: checks, measurements, simulations or questions available while the choice or its warranted use can still change;ProbeBudget: the money, time, attention or delay available for those inquiries;CostToProbe: the resources an inquiry would consume, including the work or benefit displaced by its delay;ValueOfInformation: the expected improvement in the decision from using the possible observation results, compared with choosing on what is known now;ValueOfComputation: the expected improvement in the decision from performing the proposed additional computation. Compare either gain with the inquiry’s cost on the same basis and count that cost once.
A person may make a decision, an organization may decide by an agreed procedure, or an automated controller may choose under delegated limits. An analyst, AI assistant or tool can also supply comparisons while another party decides. Identify the actual arrangement before attributing the decision.
This floor is enough to keep choice doctrine inspectable and stable. It does not yet assume one full branch-specific quantum-like package or one cross-scale geometry-heavy package.
Boundary on multilevel and social-expectation doctrine
DecisionSubject and DecisionSubjectGranularity identify who makes the choice and whether it is an individual or collective decision. Several participants can contribute evidence or carry out the chosen course. Attribution to a larger system requires grounds for its decision, such as a collective rule or an authorized delegate acting within a mandate.
This minimal choice doctrine does not fully settle the broader questions that can accompany that chooser-bearing scope, including:
- collective aggregation doctrine over conflicting preferences or criteria;
- conflicts between the objectives of a participant, a group and a containing organization whose decisions are linked in this case;
- one full
ROEor social-expectation structure for socially scaffolded choice; - one full multilevel or geometry-heavy formal package for those cross-scale or cross-collective questions.
If one of these questions matters to the choice, retain the unresolved aggregation, social-expectation or multilevel premise and obtain its needed support. The local choice record alone does not settle it.
Minimal decision tuple and finish condition
A C.11 decision record is complete only when it states:
- who or what is choosing:
DecisionSubjectat oneDecisionSubjectGranularity; - what is currently choosable:
OptionSet; - how the options are compared:
PreferenceOrderorEvaluativeMeasure, plusBeliefStateandOutcomeModel; - which heavier dependence layer is active when the case needs it:
InterventionModelfor causal repair, orCounterfactualModelplusSubjunctiveDependenceRelationfor success-first or predictor-coupled repair; - what comparison doctrine currently governs the case: one explicit
ChoiceRule; - when an inquiry alternative is live, an account of its feasibility, cost and possible contribution to the choice; when a decision or later use needs a retained inquiry limitation or reason, that minimum content in the same decision result.
- what the current comparison concludes: one emitted
ChoiceResultthat says choose now, reject the current set, probe again, or reroute. That result must name either the selected option, the retained tie-set, the rejected current set, or the next probe or reroute named by value.
The result is usable when ChoiceResult follows from the stated comparison basis, ChoiceRule and any applicable probe decision value.
A C.11 pass is finished enough for practical use when all three conditions hold:
- the current comparison basis is explicit enough to explain the stated
ChoiceResult: why an option or tie-set survives, no current option survives, or probing or rerouting is needed; - any live inquiry alternative has been resolved sufficiently to support the choice, and the result retains any inquiry reason or limitation needed by this decision or its recipient;
- the next question is explicit:
choose now,reject current set,probe again, orreroute.
An inactive inquiry item adds no placeholder or omission account to this tuple or finish condition.
If the current rule leaves alternatives tied or incomparable, report which and what that means for the receiving task. A result that retains several eligible options can supply a request for a set. Choosing one of them requires a rule or a decision by the designated chooser.
If those conditions are missing, retain the unfinished comparison and identify the premise or decision still needed.
Archetypal Grounding
System grounding
Tell. A research team already has three experiment plans on the table. The option set exists. The question is to decide which plan to run and whether one more measurement is worth the delay.
Comparison. The team can choose among three feasible plans. It first minimizes the predicted number of missed defects per 100 reference cases. It breaks ties by schedule. Its current estimates use two equally likely calibration states.
On this basis, A averages 5 misses, B averages 3.5 and C averages 4. If the team must choose now, its rule selects B.
Probe. A one-day calibration run can identify the state before the plan is fixed. The team could then choose B in the first state and C in the second, for 2.5 expected misses per 100. The improvement over choosing B now is one expected miss per 100. The day fits the budget, and the team judges that avoiding this loss is worth the delay.
Result. The current ChoiceResult is probe again. Once the run identifies a state, apply the comparison in that row. If the run becomes unavailable while the estimates and rule still hold, B remains the supported choice on the present evidence.
Show. Use C.38 if the team must turn labels or fragments into complete ways of obtaining one result. Use C.39 if a plan still lacks an explanation of how to obtain that result, or C.40 if the team can vary a plan and needs to examine the difference. Use C.18 when the team needs a generation account or archive/front claim; C.19 for exploration policy over the plan pool; and C.24 for the run sheet and execution order after choice.
Episteme grounding
Tell. A planning team has three models for predicting restart delays. It needs one forecast for the staffing decision due tomorrow. For this use the team judges a model by its mean absolute timing error on comparable held-out shifts.
Comparison. Under the agreed mix of shifts the three models miss restart time by 6, 3 and 5 minutes on average. All three can supply the forecast in time. The current comparison supports model 2. Choosing it for staffing does not establish which mechanism caused the delays.
When another comparison matters. Suppose tomorrow’s roster includes a previously unrepresented shift. Applying all three models to the same set of comparable logged shifts could change their order. Check whether those logs can be obtained before the roster is fixed and whether the resulting change in staffing would justify the effort.
Result. For the original mix, choose model 2 on its lower error. For the changed mix, return probe again if the log comparison is feasible and worth doing. If the logs are unavailable and the previous order cannot be supported for the new mix, report that limit to the planner.
Show. This pattern governs neither open-ended hypothesis generation nor operational rollout. Those questions stay outside this pattern even when the decision later feeds them.
Collective and contextual grounding
Tell. A clinical board must decide whether to escalate a patient now or order one more test. The board is the chooser, not one isolated individual, and the result shifts when the case is discussed in prognosis-first versus risk-first order.
Show. C.11 keeps the case legible by typing the chooser as one DecisionSubject at explicit DecisionSubjectGranularity, keeping the available actions as one current OptionSet, keeping one explicit BeliefState and OutcomeModel around those actions, and asking whether another test belongs in the ProbeActionSet with enough expected value to justify its CostToProbe.
Show. The changing patient state and the board's update process matter when they change feasible actions or estimated consequences. The observed order effect requires an explicit comparison-order account. Neither collective choice nor an order effect alone establishes that active-inference or quantum-like formalism is needed; use a branch only for the remaining limitation it addresses.
Show. If the order effect still changes which option survives, the comparison should say that directly and keep the comparison unfinished. The next comparison move is then either one framing-stabilizing probe or declaring the comparison order under which the current result will be judged. The record should identify that comparison instability instead of describing it only as mixed intuitions.
Show. An admissible output here therefore looks like one of three concrete records:
ChoiceResult = probe again with a comparison under both stated question orders because it can distinguish a missing prognosis premise from a framing-dependent ranking; the expected decision gain justifies the feasible delay;ChoiceResult = choose now and escalate because the receiving decision accepts the declared risk-first comparison, and no further feasible test offers enough expected decision gain to justify the harmful delay;ChoiceResult = reroute to C.24 because the board has already chosen escalation and the next task is now treatment sequencing rather than local choice.
Show. Record which comparison the board used, its conclusion and whether the board acted on it. When the ranking still depends on an unresolved framing choice, retain that uncertainty.
Bias-Annotation
This pattern is intentionally biased toward Prag and Onto/Epist discipline.
It prefers a clear decision-theory EntityOfConcern, an explicit neighboring-question split, and a minimal mathematical floor over a less constrained notion of rationality that allows more variation in its formulation.
That bias can feel too strict when the chooser, option set, or dependence structure is still changing. Address this by keeping the unfinished state explicit, rather than broadening the pattern into a general rationality account: retain a tie-set, a probe again result, or state the neighboring subject pattern that governs the question.
The family map also remains plural: causal, success-first, active-inference, and quantum-like repairs stay visible without being overpromoted into one default doctrine.
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
A result is usable when its recipient can recover the choice and the grounds on which it follows.
Consequences
Rationale
A live option set and a live choice among that set are not the same question as generating options, governing a candidate pool, or sequencing execution. Keeping that distinction explicit is what makes the doctrine usable rather than ceremonial.
DecisionSubject covers a person, a technical system or an organized body for the particular decision. Use A.13 when a claim about agency itself needs examination.
A minimal mathematical floor is necessary because choice doctrine without one stable object stack quickly turns into verbal drift. But a pattern also fails if it keeps only the object names and never shows how those objects discipline an actual choice. That is why Solution here is procedural: it must carry the path from OptionSet through one ChoiceRule to one ChoiceResult, including the stop-or-probe decision, rather than only one survey of neighboring theories.
The receiving work can use the selected option, retained alternatives or next probe together with the grounds and limits of that choice.
At the same time, this pattern should not treat a full quantum-like or geometry-heavy package as settled merely because those neighboring lines exist.
SoTA-Echoing
Practical reading of this alignment:
-
In ordinary current-option cases, start with the classical evidential baseline and use it to emit one explicit
choose now,reject current set,probe again, orrerouteresult under one sharedBeliefStateandOutcomeModel. -
Failure of a simple Bayesian or passive-read-like model is not yet evidence that QL is necessary. Try richer classical, causal, performative, instrument, active-sensing, or representation-abstraction rivals before keeping the QL branch as load-bearing.
-
If intervention structure changes the survivor relation, state that explicitly and switch to causal comparison rather than treating it as a correlational comparison.
-
If predictor-coupling or structurally linked choice procedures remain load-bearing, keep the subjunctive layer visible and say what linked structure could still reverse the current result.
-
If another measurement, comparison pass, or search pass is being considered, treat its value and cost as part of the current decision doctrine rather than assigning them only to later planning.
-
If the chooser is embodied, online, and socially coupled, or if context and order effects change the comparison state, keep those repaired limitations visible by naming the observation, social-expectation clarification, order stabilization, response-replicability check, or measurement-class clarification that could still change the current
ChoiceResult.In those cases, state when fuller
ROE, quantum-like state-space, or multilevel doctrine still sits outside this local choice body. -
If the quantum-like line is activated, treat it as one measurement-sensitive mathematical lens or representational repair, not as one claim that the chooser or world is physically quantum.
-
If none of those heavier repaired limitations is live, stay with the lighter branch rather than adding a branch label that does not change the next action.
Relations
C.11.MV constructs a Method-option worth comparison, preserving the receiving result, present alternative, learning and transition effort, recurring support and distribution of burden. It uses this choice discipline and C.11.DUA where a demanded inquiry or requirement itself needs appraisal.
- Builds on:
A.6.P,A.6.5,A.10,A.13,A.18,A.19; coordinates with:E.10.LRNfor learning-word recovery,C.11.CRCfor a missing finite configuration-relative comparison claim,C.17for bounded characterization - Read next when candidate work is needed:
C.38for forming complete ways of obtaining one result,C.39for a missing explanation of how to obtain it, orC.40for feasible variation and examination of usable material. UseC.18when generation-account or archive/front claims are needed. - Other neighboring questions:
C.19for one explicit pool-policy result over exploration or exploitation governance;C.24for one enactment-facing call plan orCheckpointReturn;G.5for the selector-facing result kind for the receiving use—retained alternatives, all-member joint use, narrowed handoff, abstain, or escalation; andC.28when the choice result depends on causal-use support - Keeps outside: candidate generation, pool-wide exploration or exploitation policy, selector-facing set-result declaration, publication availability, and execution sequencing
- Aligns with: classical evidential decision theory, causal decision theory, success-first or subjunctive repair, bounded-resource metareasoning and probe-cost doctrine,
C.28causal-use question/rung/support vocabulary, active-inference-adjacent decision work, quantum-like contextual repair where context or order effects occur, and multilevel mathematical-lens pressure at the minimal-floor level only
Quantum-like choice-boundary note
Use C.11 first when the question is local choice, belief, preference, comparison, question order, option-menu framing, or an incompatible-question effect inside one decision situation. Quantum-like wording is retained only when it adds a concrete residual order-effect, probe, frame, or comparison reading that ordinary decision or probability language would hide.
Decision repair sequence:
- State the current
DecisionSubject,OptionSet, comparison basis, andChoiceRule. - Ask whether the apparent QL issue changes the local choice result: choose now, reject current set, probe again, or reroute.
- If the issue is question order, incompatible questions, response-replicability tension, non-shared comparison frames, or CbD-style variable identity / joint-availability trouble inside the current option set, keep the QL repair inside C.11.
- If the current option set itself is suspect, incomplete, generated by the wrong frame, or still needs expansion/reframing/search, leave C.11. Use
C.39for a missing explanation of how to obtain the result orC.40for feasible variation and examination of usable material. UseC.18when a generation account or archive/front claim is needed;C.19for live-pool policy;A.19for the declared CharacteristicSpace or reusable predicate; orB.5.2for abductive hypothesis formation, according to the question that needs repair. - If the issue changes boundary state, bridge/export faithfulness, coordinated-work evidence, measurement admissibility, or viability envelope, apply the pattern governing that claim.
- Emit one
ChoiceResultand one next question; do not leave the QL branch as a theory label.
When the question leaves local choice, apply the subject pattern instead of stretching C.11:
Useful outputs:
choose nowunder a declared order/frame;probe againbecause another question order, response-replicability check, or frame test could still change the survivor relation;reroutebecause the live problem is no longer local choice;- no QL wording when ordinary uncertainty, preference conflict, or option-generation work is enough.
C.11 may cite C.26 as the common quantum-like modeling lens only for the residual question after the local-choice split is applied. It is not the whole-cluster pattern for boundary, bridge/export, distributed-evidence, viability, or state-representation coarsening claims.
C.29 mathematical-lens use relation
C.29may supply a lens-supported prediction, distinction, obstruction, diagnostic boundary, or rival-lens note that a decision record can cite. If the output is aChoiceResultor local choice record, useC.11to state and test the decision. AnyG.5selector-result declaration andG.9benchmark result remain separate. When one of those results must be available to an audience, useE.17for its source-backed publication face and return to source andE.24.PUBfor the publication occurrence and availability.C.29does not select the option by mathematical elegance.
C.11:End
Configuration-Relative Contribution Comparison
Tech name:
ConfigurationRelativeContributionComparisonPlain name: compare what this finite change adds to the current configuration
Type: C-pattern
Status: Stable
Use: construct a finite contribution comparison for a named choice, design or control question
Use This When
Use this pattern when a proposed finite change is being justified by “what it adds,” but its contribution has not been compared with the current configuration under the conditions that matter to the receiving decision. The change may add, replace or remove something; it may also acquire information before a later commitment.
First useful result. Tell the decision maker what would change with the proposal over the period they care about, compared with continuing without it. Include who gains or loses, the resources needed and any later choice it opens or closes. State which findings are estimates and what would change the comparison. This account is a C.2.1 episteme. It supplies a premise for a choice (C.11), for planning later actions (C.43), or for the receiving design Method.
Cheap exit. If an existing domain calculation or A.19/C.11 account already states the same finite baseline and change with the material comparison conditions and reopen condition, use it directly.
Not this pattern when. Do not use it for a source-only comparison with no realizable configuration change; for a purely causal question under C.28; for a mathematical-lens question already answered by C.29 and a field Method; for an archive/front relation under C.18; or as a substitute for field-specific finance, optimization, operations, engineering, experimental-design, or capability-development calculation.
Problem Frame
Candidates are often described by an isolated score, average benefit, frequency, local gradient, shadow price, or expected information gain. The receiving decision concerns something else: whether a finite change is worthwhile from this current configuration, over this horizon, for these affected Systems, under these resource and authority constraints.
The same candidate can contribute differently when prerequisites, complements, substitutes, bottlenecks, thresholds, congestion, implementation capability, uncertainty, reversibility, and future options differ. A good isolated characteristic can therefore coexist with a dominated configuration change, and a weak-looking local result can preserve valuable options or reveal a critical blocker.
Problem
Without an explicit configuration-relative comparison, practitioners make at least six transfers:
- a finite change is approximated by a derivative outside its valid region;
- a shadow price for one active constraint is multiplied into the value of an asset that changes several constraints;
- a candidate is compared with an empty system instead of the actual current configuration;
- several result and resource coordinates are hidden inside one scalar;
- interaction and common-cause overlap are counted as independent contribution; and
- information, option value, or reversibility is treated as realized benefit.
A.19 provides comparison mechanisms and characteristic spaces, C.29 governs mathematical lenses, and C.11 governs the choice. The remaining recurring practitioner action is to construct the finite comparison claim that those patterns can consume.
Forces
Solution
Construct the smallest finite counterfactual comparison that can change the answer to one named choice or construction question.
- Name the receiving question. State who needs the comparison, the choice or construction it informs, when the result must be available, and which consequence could change that work. For a local
C.11choice, identify its deciding System,DecisionSubjectand currentOptionSetor option-set question. For ongoing control, policy design or development, identify the continuation, rule, direction or means being compared and its receiving Method; a local option set need not already exist. - Establish the baseline. Identify the actual or currently relied-on configuration
S0. Describe how it would operate without the proposed change over the comparison period, including changes already expected. Set the system boundary, identify the affected Systems, who holds or benefits from what, and what is held fixed. When an omitted affected System could change the answer, use A.1.CSD and bring back consequences applicable to this configuration, change, period and evidence window. Give the grounds for using a historical, empty or ideal baseline if the question calls for one. - Name the finite change. State the addition, replacement, removal, intervention, or probe
Δ, the realizable candidate configurationS1, admissibility conditions, implementation capability, planned transition work, reversibility, and excluded variants. - Fix the horizon and scenarios. Compare both configurations over the same interval under the same external scenarios. Preserve differences in timing or later conditions caused by the proposed change. Where the evidence covers different intervals, explain how it supports that comparison or leave the affected consequence unresolved.
- Declare result coordinates. Name the result vector whose coordinates can change the decision and the protected coordinates that may not be silently scalarized. Include affected-System consequences and distributional differences when they matter to the comparison.
- Declare resource coordinates. For the action, transition, information acquisition, and computation, name the attention, capital, time, material, energy, and other resources consumed or expected to be consumed in this field case. State required authority separately as an admissibility condition. Keep costs of evaluating and realizing the change distinct.
- Recover constraints and interactions. State active and potentially activated constraints, complements, substitutes, thresholds, congestion, downstream effects, common causes, overlaps, and double-counting risks.
- Recover option effects. State whether the finite change opens, closes, delays, preserves, or makes irreversible later options. Keep information value and option value as decision inputs, not already realized operating results.
- Qualify evidence and uncertainty. Identify source claims, currentness, uncertainty, sensitivity/robustness results, transfer limits, rival explanations, and
A.10reliance dispositions where an evidence-bearing claim is used. - Write the comparison claim. State what
S1contributes relative toS0only under the declared coordinates, horizon, scenarios, constraints, interactions, and evidence. Use dominated, non-dominated, beneficial, harmful, or indeterminate wording only when the stated relation supports it; do not force one scalar winner. - Route mathematical near-misses. Apply the distinction in
C.11.CRC:4.2; the finite comparison may consume a derivative, sensitivity, shadow price, variational, or inference result without becoming identical to it. - Return to the receiving work. Supply the comparison to the named question from step 1. Use
C.11when a local chooser and options are ready: it combines the claim with preferences, belief state, outcome model, probe worth and other premises to obtain aChoiceResult. UseC.43when the consequence must first be connected to conditional future action, realization or revision; a subject design or development Method can consume the comparison directly. State the smallest configuration, horizon, evidence, resource, constraint or alternative change that reopens this comparison.
Lightweight comparison form
ConfigurationRelativeContributionComparison@Context is a form name for one comparison episteme. It is not a root U-kind, universal delta value, selector result, or decision.
An ordinary comparison sentence or existing calculation can carry this result when the material premises are recoverable. Omit a field when it cannot change the comparison and that omission is apparent from the bounded case. The current configuration and receiving question remain necessary.
Mathematical and model-routing distinction
This is why calculus of variations matters without becoming the default interpretation of marginality. When the candidate is a whole trajectory, field, function, control, or shape, pointwise finite-coordinate reasoning can miss the coupled admissible deformation. Use C.29 when the mapping from the world or domain model to that mathematical object, its preservation/loss or its transfer limits remain at issue. An adequate domain formulation can be used directly. Specialist practice governs derivation, discretization, solver choice, optimality and sufficiency checks, and validation.
Recognition and assurance split
Recognition. A user can recognize a conforming comparison when S0, finite Δ, S1, boundary, horizon, result and resource coordinates, interactions, uncertainty, and receiving decision are visible.
Assurance. Trust in the numbers and relations remains separate. Field evidence must support the baseline and candidate behavior; implementation capability and the basis for planned transition work must be credible; causal claims use C.28; source reliance uses A.10; for an assurance claim, use B.3 with the named target claim and assurance use; authority and permission use their direct patterns. This pattern creates none of those results.
Worked Slices
Flood-pump modernization
Station FPS7-C19 currently supports bounded discharge use. Compare two ways of operating it over the coming service interval. In S0 the station continues with its existing diagnostics and scheduled maintenance. In the proposed S1 it uses an additional bearing-temperature sensor. The change Δ includes placement, cabling, controller changes, access for maintenance, calibration and operating procedure. Both forecasts use the same loading scenarios. Neither forecast establishes that the sensor has been installed.
Result coordinates include discharge continuity, failure detection, maintenance access, recoverability, and evidence continuity. Resource coordinates include outage time and maintenance burden plus the resource costs of installation work, calibration, and observation.
Two gaps affect this proposal. Establish whether the proposed location can be reached during an allowed maintenance window. To assess earlier warning, compare the proposed signal with the existing diagnostics under the loads where their indications might differ. A load observation alone cannot establish access. Price each attainable check against the choice it can change (C.11.DUA). If access is impossible at this location, compare a feasible placement or retain the current configuration.
Capability-development programme
For the same later assignment, compare continuing with the present programme and normally available support against adding the proposed practice. Include prerequisites, practice time, the support provider's time and work displaced elsewhere, and the expected time to competent use. Determine what the participant did and what an assistant supplied. A jointly completed artifact can meet the immediate need. To judge a gain in the participant's capability, compare earlier performance with performance on a later comparable task under the support conditions promised for later work. Transfer, retention and detection of critical errors need observations suited to those claims.
Operations constraint intervention
A positive shadow price for one bottleneck supports a local statement about relaxing the formulated active constraint. A finite machine addition also changes labor, maintenance, setup, downstream capacity, energy, resilience, and perhaps the active constraint. The S0/S1 comparison can therefore disagree with shadow-price multiplication without making the shadow price useless.
Capital allocation
Compare the portfolio over the decision horizon with and without the proposed project. If it would displace another investment, include what that investment would have produced in the feasible alternative. Compare shared resources, cannibalization, risk concentration, financing constraints, staging, irreversibility and information available before later commitments. Corporate Finance supplies the NPV, real-options, scenario and portfolio Methods. Return the finite comparison to that practice or to C.11 when it informs a local choice.
Bias Annotation
- Scalar bias: do not hide protected coordinates or distributional effects inside one score.
- Smoothness bias: do not replace a realizable finite change with a derivative outside its validity region.
- Additivity bias: inspect complements, substitutes, overlap, thresholds, congestion, and common causes.
- Baseline bias: compare what would happen with and without the change over the same horizon. A current snapshot can establish the starting configuration.
- Physics-prestige bias: a variational or thermodynamic form does not establish a physical mechanism or decision authority.
- Information-as-result bias: expected information and option value are inputs to choice, not already realized benefit for the intended recipient System.
Conformance Checklist
- Is the receiving choice or construction question named, including who needs the comparison and what it can change?
- Are
S0, finiteΔ, and realizableS1explicit? - Are system boundary, affected Systems, horizon, scenarios, and evidence window compatible—and, when a missing bearer could change the comparison, was
A.1.CSDused before freezing this coordinate? - Are result and resource coordinates explicit, with protected coordinates not silently scalarized?
- Are implementation capability, planned transition work, reversibility, and excluded variants recoverable?
- Are constraints, interactions, overlap, thresholds, congestion, and downstream effects considered where material?
- Are future option effects distinguished from realized results?
- Are evidence, uncertainty, sensitivity/robustness, transfer limits, and unsupported overreads visible?
- Is each derivative, sensitivity, shadow-price, functional-variation, variational-inference, or evolutionary-variation result used only for its question?
- Does the output remain a comparison claim returned to its named use, with
C.11retaining anyChoiceResultsought under that calculus? - Is the smallest reopen condition stated?
Common Anti-Patterns and Repairs
Consequences and Reopen Condition
Benefits. Finite additions, removals, replacements, probes, and investments become comparable without requiring additivity, smoothness, or a universal scalar. Domain calculations can be reused while their validity regions remain visible. The result supplies a stable comparison input to local choice, design or control while leaving the receiving judgement and its authority to their governing practice.
Costs. Practitioners must name a baseline, vectors, interactions, and uncertainty that an isolated score could hide. Some cases remain indeterminate until field evidence or implementation capability is available.
Revisit the comparison when changed conditions could alter its claim. A newly available substitute, tighter resource constraint, changed horizon or loss of relied-on evidence can require a new assessment; retain unaffected premises and claims.
Rationale
The pattern is narrow because its result is neither a new value ontology nor a decision. The reusable action is to construct a finite configuration-relative claim for a receiving choice or construction. The claim can remain useful while alternatives or a conditional policy are still being developed. A.19 remains the comparison-mechanism owner, C.29 the mathematical-lens owner, and C.11 the choice owner.
SoTA Echoing
Relations
- Builds on:
C.2.1,A.10,A.19,C.16,C.27,C.28, andC.29. - Supplies: a finite comparison claim to
C.11for local choice, toC.43for a connected control continuation, or directly to a field-specific portfolio, programme, intervention, architecture or experiment Method. The receiving use determines which connection is needed; these are not mandatory successive stages. - Coordinates with:
A.1.CSDwhen affected-System consequence coordinates are missing;C.18when the candidate changes the possibility space;C.19for pool governance;B.3for assurance;A.15for the distinction between transition plans (A.15.2) and performed transition Work (A.15.1); and the direct field practice for calculation and validation. - Keeps outside: universal marginal value, a new delta kind, domain formulas and thresholds, causal proof, assurance, permission, selected-set declaration, and
ChoiceResult.
C.11.CRC:End
Reconsider a Decision When Its Relied-on Basis Changes
Type: Method pattern Status: Stable
Problem frame
Use this pattern when an earlier choice, commitment or recommendation is known, and a later observation, outcome or changed condition may alter the basis for one use of that result. A team may have completed an intervention while its expected benefit failed to appear. A model may remain a correct account of its original observations while no longer supporting the next prediction. The question is: what does this later result change about the earlier choice, and what can remain in use?
First useful move. Name the earlier result and the use now in question, then recover the premise that supported that use. Compare the later result with that premise under comparable conditions. A finding that nothing material changed can finish the work. A material difference supplies a bounded input to a separately made current decision.
The reader needs access to the earlier basis, the later result and the subject knowledge needed to interpret them. Ask the responsible contributor for an unavailable premise or interpretation; a decision log alone does not supply it. Ordinary prose in the existing discussion can carry the answer. Use a retained account only when another person or later action needs it.
Use the direct way when it already suffices. A subject Method that already compares the earlier basis, later outcome, affected uses and current alternatives needs no additional C.11.DBR pass. Use C.11 directly when the current chooser, options and comparison basis are ready. Use A.15.7 for a next action during ongoing Work when no earlier-basis reconstruction is needed. Begin with A.10.1 when a source changed and its actual receiving uses are still unknown.
Within ongoing control, return the affected and preserved uses to C.43. A changed observation may require another immediate action; a changed model, goal or means can require rebuilding a different contribution. This Method establishes which relied-on basis changed and where that matters. The receiving control account connects that result to an available intervention and its execution. An adequate subject feedback Method can already supply this comparison without a separate reconstruction.
Problem
Feedback can miss its receiving decision in two opposite ways. A later result stays in a report while the team continues relying on a defeated premise. Or one changed condition reopens every related decision, including commitments and completed work whose grounds remain sound.
Neither recency nor surprise identifies the affected choice. The missing connection is between a later result and the basis used by an earlier result for a particular use. A comparison must establish that connection before it can prepare reconsideration. Recording a change, diagnosing its cause, choosing a response and performing that response are different contributions.
Forces
Solution
Recover the reliance, compare the later result, separate affected from preserved uses, and supply the smallest current question that follows. These operations form one reusable preparation Method because the comparison cannot bear on a choice until its reliance and conditions are known. They can return to one another as missing facts emerge; their presentation below does not prescribe a work schedule.
Recover what supported the use
Start with one earlier result, its issuer or chooser and the receiving use now in question. Find the assumption, expected consequence, evidence-use claim, comparison value, constraint or condition on which that use depended. Recover the subject, configuration, horizon and any condition for reconsideration from the decision, calculation, instruction or responsible contributor.
Test the proposed reliance by explaining its role: did it make an option admissible, support its preference, promise an outcome, or limit where the result could be applied? A nearby document or a plausible retrospective explanation is only a candidate basis. Use the actual rationale, derivation or established subject relation to settle it. If it cannot be recovered, return that uncertainty; do not infer what was intended from the eventual outcome.
Keep separate grounds separate. A choice of operating arrangement may rely on an expected time saving while a distinct service obligation remains binding. A failure of the saving premise does not establish failure of that obligation. Conversely, if both uses relied on one failed contribution, both belong in the reconsideration.
Qualify and compare the later result
Establish what was observed, where, when and under which conditions, and what interpretation its evidence supports. A.10 supplies claim-specific reliance; C.16 and C.27 supply measurement and temporal qualification when these are missing. Completed intervention work establishes completion, not its intended consequence. Use C.28 when a proposed response depends on causal attribution.
Compare the later result with the recovered basis at the same relevant subject, configuration, population, measure and horizon. For each difference in conditions, ask whether a justified transformation makes comparison possible, whether the earlier claim simply concerns another use, or whether the comparison remains unresolved. A weekly total and a per-case rate, for example, answer different questions until volume and scope are aligned.
Apply the condition that matters to this decision. It may be a threshold, a change in preference, lost feasibility, a violated assumption or an agreed condition for return. It need not be numerical. If the original basis contains no applicable criterion, recover what the current use requires; do not silently invent a retrospective threshold and attribute it to the old chooser.
The comparison has three useful outcomes:
These are findings about the named use. A large numerical difference can leave a choice intact while a small one crosses its constraint. An unknown cause need not prevent acting on a supported consequence. End an optional investigation for this choice when its attainable answers cannot improve what the recipient can decide or rely on. A separately binding investigation remains due under its governing terms.
Bound what changes and preserve the rest
Begin with the known use. Add another receiving decision only when its direct reliance on the changed premise can be established and its disposition matters to the present question. Follow that dependence until no further receiving conclusion can change. A graph, citation or organizational adjacency can suggest where to look but cannot establish the dependency.
State the inspected boundary. Retain decisions and claims whose independent grounds and conditions remain satisfied. Leave an uninspected branch uninspected rather than declaring it unaffected. This reconsideration is bounded by the receiving question; A.10.1 owns unknown-receiver discovery after source change.
Keep historical occurrences and current applicability distinct. Earlier work happened even if its expected benefit failed. An earlier recommendation remains a recommendation, and its original limited claim may remain true after another use loses support. If an authority, permission or work claim is itself challenged, obtain its direct judgement; do not revise it through a conclusion about expected benefit.
Prepare and obtain the current decision
For a material branch, give the recipient a reconsideration statement: the earlier result and use, changed basis, qualified later result, material difference and remaining uncertainty, and the preserved boundary. Add the candidate chooser and known alternatives when available. Name a missing input or authority instead of inventing it to complete a form. This claim content can fit in a few sentences.
Choose the continuation by what is now missing:
- Use C.11 when one chooser and a finite current OptionSet exist with a comparison basis. C.11.CRC can construct a missing finite contribution comparison. C.11.DBR is not an extra qualification layer for every premise entering C.11.
- Use the direct subject Method for strategic, asset, organization, operating, maintenance, human-development or other subject decisions. Its consequences, evidence requirements and authority remain in that practice.
- Use A.15.7 for a next-action question in ongoing Work. Use the applicable problem-framing or construction Method when the question or means must change before options can be compared.
- State which decision or authority is missing when only an analyst's recommendation is available. A recommendation can be the useful completed contribution while the authority holder's decision remains outstanding.
The person, controller or authorized body that holds the decision uses the receiving Method. Return its decision to those who depend on it. An analyst who has only made a recommendation can deliver that recommendation now and identify the decision still outstanding. If an overview is needed, cite the decision once made, using the terms of the receiving practice.
Carry the result to its next use and stop
Explain what remains usable, what current result has been obtained and which contribution is missing. If the decision calls for planned work, the responsible planner can use A.15.2 to prepare or revise the WorkPlan. Establish separately who may carry it out, who is available and what was actually performed.
Stop when the included uses have a supported preservation answer, an obtained current decision, or a precise unresolved contribution. Stop earlier when a narrower answer completes the receiving question. Preserve a reason and a condition for another return only where later reliance needs them. A new observation, changed use or newly discovered dependency can reopen the comparison it could materially change; independently preserved results remain usable.
Archetypal Grounding
A completed change and a missing saving
A team chose arrangement A for one recurring report. Its recorded comparison expected two hours of preparation instead of six under arrangement B, for the same report content and weekly volume. A also needs one hour of support per report; B needs none. A separate decision requires that the report still be delivered each Friday. The rollout of A was completed. The first later dashboard reports six hours of preparation for A, and somebody proposes reversing the whole change.
Recovering the basis separates the expected saving from the Friday obligation and from the completed rollout. Inspecting the dashboard then reveals that it combines two reports; the earlier comparison concerned one. Six is not yet a comparable observation. The useful answer is the scope mismatch, with the delivery obligation and rollout history preserved. No causal or reversal conclusion follows.
Change the evidence. A corrected, qualified observation now establishes five hours for the same content and volume. The current decision rule requires at least two hours of total effort saving relative to B, including A's support burden; all other relevant conditions are stipulated unchanged. A now requires six hours in total and supplies no saving, so it fails that condition. This challenges continued reliance on the expected saving, but does not establish why the shortfall occurred.
The responsible manager receives the comparison, with Friday delivery and rollout history preserved. A supported adjustment C takes three preparation hours plus one support hour and meets the unchanged delivery and quality constraints. C takes four hours in total against six each for A and B. On the stipulated preference for lower total effort, the analyst can recommend C. The manager can choose it on those grounds without another study of A's cause.
The manager's actual choice, any necessary plan revision and the later performance of C are separate results. If the manager has not decided, the analyst returns the comparison and recommendation, not an implemented change.
Change the observation instead. Had the qualified observation been three preparation hours for A with its other conditions unchanged, its four total hours would provide the required two-hour saving. Preserve the earlier choice for this use; the fact that three differs from the original estimate of two does not require a new choice.
If C was already an available improvement, comparing it would be a separate live opportunity, not proof that the original basis had failed.
A model remains valid for one use
A laboratory qualified a linear calibration for readings from 0 to 10, with a maximum conversion error of 0.2. A qualified observation at 15 under the same operating conditions has error 0.6. That observation refutes the proposed bound of 0.2 over the entire interval from 0 to 15. The qualification and earlier converted results in the interval from 0 to 10 still stand.
Now the task needs readings up to 15 at the same error bound. The old calibration cannot cover that task by itself. Seek a qualified alternative for the full range or justify two calibrations whose combined coverage and behavior at the join meet the task. A proposal to narrow the task needs a decision by its holder. Changing a coefficient or extending the range alone establishes neither the error bound nor model refinement (B.5.MR). If coverage remains missing, state which readings can still be used.
The model file need not have changed. A.10.1 would be appropriate for a different situation: a corrected calibration source whose actual consumers must be discovered.
Keep subject returns distinct
The shared preparation stops before these domain consequences and decisions. It supplies neither a strategic pivot rule nor a maintenance trial or capability-development assessment.
Bias-Annotation
- Hindsight: an outcome makes a convenient old rationale seem obvious. Recover actual reliance and state what cannot be recovered.
- Recency: later information is treated as stronger without checking its conditions. Compare qualified claims.
- Escalation and sunk effort: completed work is used to defend a failed premise, or failure is used to erase history. Keep both facts.
- Selective preservation: only convenient dependencies are inspected. State the bounded receiving question and unresolved branches.
Conformance Checklist
- The account identifies the known earlier result, its current receiving use and the role of the relied-on basis.
- The later result is qualified and compared at a commensurate scope and horizon, or the decisive mismatch is explicit.
- Materiality follows the use's constraint, comparison or applicability condition; a no-material finding can finish without a new choice.
- Affected uses follow established reliance; preserved results retain their grounds, and uninspected branches remain outside the conclusion.
- A material change identifies a decision to reconsider. An overview may cite the holder's recorded decision or report the available recommendation and identify the choice still open.
- Recommendations, causal claims, authority, planning and performed work keep their own grounds and limits.
- The method stops at the useful result or precise gap and adds no duplicate pass to an adequate direct subject Method.
Common Anti-Patterns and How to Avoid Them
Consequences
The reader can preserve a decision for one use while reopening another, retain completed work while challenging its expected benefit, and give a chooser a specific comparison question.
The cost is recovering actual reliance and aligning the observations. Poor records or inaccessible contributors can leave that recovery unresolved. A short statement suffices for a small use; a broad dependent set can require more work. The Method does not promise that a replacement or favorable outcome exists.
Rationale
Reconsideration needs a result before current choice: an explanation of what changed in the earlier basis and where that matters. The comparison is relative to a receiving use, so a change can defeat one application while leaving another intact.
The cheap preservation branch and direct subject exit make the generalization useful without prescribing a universal review cycle. Qualifying the later observation before interpreting its effect prevents novelty from becoming authority. Completing an overview after the current decision preserves the direction in which the results are obtained and used.
SoTA-Echoing
For the working question in §5.1, a decision log preserves why A was selected but does not align the two-report observation; a current choice comparison needs that alignment already supplied. The present Method obtains it, preserves the independent delivery obligation and supplies only the remaining choice question. Once a direct subject treatment supplies those operations, adding this Method again brings no corresponding gain. This comparison concerns the stated use, not a claim of measured superiority over decision records or adaptive management.
Relations
- Uses A.10, C.16 and C.27 for the evidence, measurement and time qualifications needed by the particular comparison; C.28 for a causal claim when one is required.
- Prepares C.11 or the direct subject decision Method from a known earlier decision use. C.11.CRC supplies a missing finite contribution comparison; C.11.DUA compares attainable inquiry, a narrower answer and continued work.
- Coordinates with A.10.1: unknown receivers after source change belong there. Enter here only for a surviving known decision-use question not already settled by the subject return.
- Receives an E.18.1 return when its distinction between what still follows and what no longer follows, together with the smallest reopened continuation, leaves an earlier-decision question. P2W retains its accepted problem-side entry.
- Returns to A.15.7 for live work steering and A.15.2 for separately needed planning. A.15.1, A.2.8.PER and A.21 retain performed-work, permission and gate claims.
- Uses C.2.1 when the receiving use needs an identifiable retained account.
C.11.DBR:End
Make Advice and Evidence Demands Worth Their Burden
Tech name:
DecisionUsefulAdvicePlain name: make advice and evidence demands useful to the recipient's decision
Type: C-pattern
Status: Stable
Problem frame
Use this when a report, expert answer, review comment, rule or checklist asks somebody to act, wait, investigate or supply evidence, but it is unclear how that demand helps their present question or whether they can carry it out.
An engineer reports that a discharge criterion was below the incoming background level. A colleague examining the criterion can use an attributed report and the distinction between total level and the operation's contribution. An archive search would add work without improving that answer. If the colleague intends to publish who imposed the criterion, a contemporary record may be worth seeking. An answer about which inquiry is worth making can be complete while the authorship of the criterion is still unknown.
The broad activity is decision support through advice. This pattern addresses the branch in which advice turns uncertainty or a requirement into consequential work for its recipient. Its subject is the advice statement or requirement being composed or examined, including the action or inquiry it demands. An empirical claim, recommendation, method instruction and binding requirement retain their different meanings and force. The method recovers the demand, appraises it and returns a usable continuation.
First useful result. Give the supported answer or repaired advice, the continuation available to the recipient, and any limitation that changes its use. The result can be an explanation, recommendation, bounded inquiry, proposed requirement revision or an impasse. Ordinary prose is sufficient.
Ordinary non-use boundary. If the question, support and feasible answer are already clear, answer directly. Use A.10 directly for a source-reliance question, C.11 for an already adequate local choice, or C.19.2 for economical application under an undisputed declaration. This composition adds value when somebody otherwise has to reconstruct the connection between those questions. Domain inference, experimental design, permission and enactment remain with their own methods and authorities.
Problem
A limitation is often followed by a demand that does not follow from it. “This study cannot establish the effect” can be true while “therefore commission the ideal experiment before doing anything” is infeasible or inferior to a bounded action. Conversely, an inexpensive check may distinguish a damaging commitment from a useful alternative. Skipping it merely because the present answer is convenient repeats the same failure to compare consequences.
The difficulty starts before a well-formed option set. Advice such as “validate thoroughly” can hide an investigator, access requirement, delay, claimed guarantee and transfer of cost to somebody else. The recipient has to recover that work before choosing whether it is worthwhile.
A further failure appears at completion. A method may allow work to proceed without more inquiry yet require a probe block, waiver or explanation of every omitted check before accepting the answer. Producing that account is itself work. Sometimes a retained reason protects a later use; sometimes it adds nothing to the completed decision.
Forces
Solution
Start from what the recipient is trying to settle. Recover the work hidden in the advice, compare what that work can change with what it costs, and give the best warranted continuation. Follow that continuation through the required result and stopping point.
Recover the receiving question and demanded work
Identify the question the recipient needs answered and the horizon at which an improvement matters. Read an existing instruction as something somebody would have to perform: what would they believe, do, stop doing, postpone or spend if they followed it?
Recover the performer, resources, access, time or authority that affects feasibility. A study ending after the decision window serves a different use. For a request for help, establish whose result is wanted. For a shared deliverable, identify the work each participant can supply. If the recipient is to acquire an ability, identify what they must later do under the support conditions intended for that use and what performance would show it. Successful joint work alone warrants the claim about the deliverable (Library Reference §9.3.1).
For a text review, examine the connected passage that creates the demand, including its steps, mandatory fields and completion clauses. “Consider further evidence” and “obtain it before proceeding” impose different work. A final template can impose work that the preceding prose made optional.
Keep three answers distinct:
Resolve these questions when relevant to the receiving use; they are not three mandatory output fields. A source limitation qualifies the attempted claim. It does not by itself choose evidence acquisition.
Compare attainable inquiry with its receiving value
If the relevant distinction or possible result is not yet expressible, use B.5:4.1.2 to develop grounds from an accessible episode, source or attempt. The immediate comparison can concern a bounded exploration likely to improve the description or reveal a dependency. It need not promise a particular discovery. Return with the basis obtained before demanding a specific action-changing answer. Appraising an already formulated inquiry and constructing its missing basis are different contributions.
For an optional inquiry, identify a plausible result that would change the conclusion, choice or warranted use. Compare it with using present knowledge, taking a smaller reversible step, using a different source, narrowing the claim, deferring or stopping. For a binding inquiry, identify the work due, its recipient and the evidence or report owed under the governing rule. Compare permitted ways of meeting it. A duty to measure and report can hold whatever the reading is.
Include the attainable inquiry's cost, delay, displaced work and downside. Compare what it adds to the present configuration and horizon; C.11.CRC supplies that missing comparison when needed. A large uncertainty is not sufficient reason for a large study. Ask how resolving it could improve this use. A cheap probe that can change an irreversible commitment deserves a different answer from an expensive check whose plausible results leave the useful answer unchanged.
Qualitative or ordinal judgement often suffices. A numerical value-of-information or computation model is useful when it improves the choice; do not demand probabilities or a common scalar merely to fill a form. If plausible differences in judgement could reverse an important recommendation, examine those differences at proportionate effort. Keep distinct consequences visible when a scalar would hide the trade-off.
An inquiry can be useful even if the physical action stays the same. It may establish which claim is warranted, meet an applicable evidence condition or make a proposed use admissible. Conversely, inability to obtain evidence does not make the stronger claim true or admissible.
For exploratory research, name the explanatory, model-building, hypothesis-discriminating or option-creating gain at the research horizon. A study can earn its cost through those gains before an operational intervention is selected. The direct research method must still connect its attainable observations to that gain.
Compare a proposed change with continuing the work
For optional advice to change a way of working, include continuing without that change in the comparison. A useful new contribution can be considered while the present work succeeds and its constraints remain unchanged. Recover what the recipient could gain, on which horizon, and what following the advice would require.
When the gain depends on a constituent action, use B.1.5.EW to recover the encompassing work performed through it, the contributions needed below it and the conditions that change its use. Use B.1.5.RS for a proposed replacement in those wholes. Include a second receiving whole or an intermediate connection while it can reverse the advice. Use an already adequate account directly.
Compare the full continuation: understanding and discussing the advice, obtaining support, learning, adaptation, coordination, execution and displaced work. Use C.11.CRC when the finite comparison or interactions are missing. If several changes compete for the same recipient capacity, compare their combined demand with ongoing commitments and changes already selected. Individually useful changes may require a smaller set, deferral or a different combination.
Keep deciding to think further, presenting advice, deciding to change and performing the change separate. Cheap computation can justify frequent consideration while interruption and transition costs justify infrequent advice or intervention. Consider timing and a smaller change when those alternatives matter. A change in performer capability or adaptation cost can justify reconsidering the encompassing way of working.
Use the available basis before asking the recipient for more work. Unclear benefit alone gives the recipient no obligation to prove an optional suggestion useless or fund an expensive evaluation. Obtain more information only when a plausible attainable answer can change a worthwhile continuation.
If presenting the advice has no proportionate basis, continue without presenting it; no rejection record is required merely to finish. Preserve a reason or reconsideration condition only for a receiving use that needs it.
An existing obligation still constrains the available continuations. Examine its merits and possible revision through section 4.3 when that question is live.
Appraise a disputed requirement and its current force
When the requirement itself is in question, identify the quantity, activity or claim it governs and the problem it is meant to address. Examine its basis and uncertainty, threshold, protective contribution, burden and distribution of error costs. Include delay and displaced risks: reducing one named risk may increase another.
Compare retaining, tightening, relaxing or replacing the requirement only where these are meaningful alternatives. Resolve labels such as “safety” or “security” to the particular hazard, exposure, action or evidence condition. Apply the same proportionality question to the evidence demanded to justify the requirement itself.
A limit below background calls first for distinguishing the incoming background, the activity's attributable increment and the total resulting level. A total-exposure objective can justify reducing background too. Neither “below background” nor “required for safety” settles the merits without that contribution comparison.
Keep the merits judgement separate from the currently available action set. A binding requirement continues to govern until the relevant authority or agreement changes it. Advice can recommend a lawful exception, a revised criterion or another activity and identify the feasible route to that change. If the current requirement makes the proposed action unavailable and no acceptable alternative or timely revision is feasible, state that impasse.
When applying apparatus under C.19.2, compare eligible applications under the current result, guarantee and constraints. A cheaper method that drops a required guarantee is a different proposal. Resume under a revised declaration only when that revision is available through its governing authority or agreement.
Return the advice through first useful completion
Give the supported answer and available continuation in language the recipient can use. Retain uncertainty, attribution, a reason or a return condition when it changes that use. Where a substantive decision record is needed, put this content in the same result.
For an external-text review, return a concrete repair: identify the demanded work, the consequence or burden that makes it defective, and the smallest replacement that reaches a warranted result. Keep a useful evidence demand when the comparison supports it. A finding for another author can be durable without becoming a per-sentence ledger.
Follow the selected method through its completion requirements. Keep the recommendation or decision and its useful basis. An inactive inquiry creates no empty field, no waiver and no requirement to explain why checking was omitted. If a later recipient would otherwise overgeneralize the result, infer fresh evidence or miss a material limitation, retain the minimum explanation in that result. A short useful explanation alone need not instantiate a complete probe account.
Apply the same appraisal to any prescribed record. If it has binding force, distinguish its usefulness from what the current procedure requires and use the available revision route. This appraisal does not grant permission to disregard it.
Finish when the recipient has the warranted answer and continuation for the present question, including a justified stop when necessary. Reopen when the receiving use, possible consequence, feasible means, source basis or information opportunity changes materially.
Keep recognition and assurance about the same advice
The opening and worked situations help the reader recognize a consequential demand and repair it. The checklist below tests that same advice and the work it demands. It does not introduce a separate admission record or require every use to undergo a new review.
Use A.10 for the evidence boundary, C.11 for a choice result, D.4 for ethical mediation and the direct assurance method for a stronger guarantee when that question is live. Advice that a step is sensible supplies neither permission nor evidence that the step was performed.
Archetypal Grounding
These illustrative situations state their conditions for the comparison. Reported experience remains attributed; the numerical example uses stipulated decision inputs, not measurements from performed experiments.
A useful distinction now and corroboration for a later claim
Present use. An engineer reports a criterion below incoming background. The recipient wants to understand whether that is necessarily a design mistake. An archive search would take weeks. Its plausible results could change the historical account but would not change the requested distinction.
Completed answer. “On your account, the threshold was below the incoming background. A criterion for the operation's added contribution and a criterion for total exposure govern different quantities; the latter can require reducing background too. Compare the protection gained with the feasible treatment burden before judging the threshold.”
That answer supplies the requested distinction and identifies the engineer as the source of the threshold report. The answer itself completes this use; a second account of why archives were not searched adds nothing.
Changed use. The recipient now plans to attribute responsibility in a report. Suppose an accessible contemporary record can identify who set the criterion and what it governed. A one-hour record check is feasible and could settle the attribution.
Completed advice. “Before assigning responsibility in the report, check the contemporary record for who set the criterion and what it governed. If that remains unresolved, present the engineer's account as their reported experience and leave responsibility unassigned.” The recipient has a feasible next check and a way to finish the report if the record leaves the attribution unresolved.
A completed choice with conditional inquiry content
A research team chooses among feasible plans A, B and C. Two calibration states, s1 and s2, have equal weight. The team's rule minimizes expected residual loss; shorter duration breaks equal-loss ties. The following are cardinal comparison inputs.
Completed decision. Retain T0 and: “The research team chooses B now among A, B and C for this experiment. At equal state weights the expected losses are 7, 2.5 and 4.5. We minimize expected residual loss, with duration breaking ties.” B is preferable in either state. The current basis supports this choice; no unresolved inquiry alternative needs a separate account.
A completion form demanding unused probe budget, cost and information/computation fields would add omission accounting to that result. Conditional inquiry content leaves the decision and its basis intact.
Opposed condition. Change the loss model to T1 below. A specified one-day calibration can distinguish the states perfectly, an available technician can perform it, and two days remain. The team accepts that delay for a reduction in expected loss.
Completed first choice. Retain T1 and: “Probe again with the one-day calibration. Choosing B now has expected loss 3; choosing A after s1 or B after s2 has expected loss 1.5. The available day and technician make the probe feasible, and the team accepts the delay for that reduction.” This is a completed C.11 probe-again result. The measurement remains to be performed. An actual s1 result would then select A; s2 would select B.
Transfer requiring a reason. Return to T0, now sent to a colleague considering B for another calibration range. Retain: “B was selected using the existing model T0 and its two calibration states; distinguishing those states would not alter that choice. For another calibration range, reopen the comparison using evidence applicable to that range.” The explanation changes the colleague's use: reopen the comparison instead of treating B as generally preferred. It belongs in the same decision result and needs no empty probe block.
Repair a study demand in an external report
A consultancy report says, “Do not introduce the revised teaching exercise until a large randomized study proves its effect.” The teaching team must choose materials for four sessions next month; that study would take two years. The present synthesis does not establish a causal learning effect. The team's immediate question is whether the revised materials can meet the existing course requirements with less preparation.
For this illustrative comparison, both formats meet the content and participant-protection requirements, and no binding rule requires the proposed effect study for this bounded use. The remaining uncertainty is whether the revised materials run with the room's existing tools. The team requires that readiness to be established before selecting them. A thirty-minute setup rehearsal by the instructor in that room can settle it; it creates no student exposure.
Use the team's stipulated planning estimates: the current exercise requires ninety minutes of preparation per session; the revised materials require thirty if compatible, otherwise the team retains the current exercise. The same room and tools will serve all four sessions. The team regards compatibility and incompatibility as equally plausible and prefers less instructor preparation once the course and protection constraints are met. The rehearsal fits the available preparation window.
A successful rehearsal makes the revised materials available and saves four hours across the sessions; a failed rehearsal leaves the current exercise selected. With equal weights, the expected preparation saving is two hours before the rehearsal's half-hour cost, or ninety minutes net. Retaining the current exercise uses no rehearsal time but forgoes that opportunity. These estimates warrant this rehearsal; its small size and reversibility alone would not.
Completed advice. “Run the half-hour setup rehearsal to settle compatibility with this room's tools. If the revised materials work, use them for the four sessions; otherwise retain the current exercise. Under our stated estimates, that information saves an expected ninety minutes of instructor time after its cost. Keep the learning-effect claim open.”
The advice gives the team a decision it can now make about the rehearsal. Compatibility with the stated room and tools becomes known only after the rehearsal is performed. If domain review identifies an unacceptable participant risk or a binding evidence prerequisite, meet the governing condition before using the materials.
Opposed cost condition. Suppose the same discriminating rehearsal now requires three hours of instructor time, still within the available preparation window. Its expected gross preparation saving remains two hours. The completed advice changes: “Retain the current exercise for these four sessions. The proposed rehearsal would cost an expected hour more than it saves. The current materials meet the course requirements; the comparative learning-effect claim remains unresolved.” A longer study remains a separate research proposal whose attainable contribution and burden must be justified for its own receiving use.
Earn a long research horizon
A research group has two rival explanations of a recurrent anomaly. Neither yet supports an intervention. For one instrument setting their predicted ranges overlap in the middle and differ at the ends, even after allowing for measurement error. The instrument is available for four weeks. A broader survey would exhaust the year's allocation without separating the predictions.
The group must choose a design for its next instrument. A result compatible with only one model would change that design. The group judges the chance and value of resolving this design choice worth the four weeks, after comparing the best other use of that time. A result in the overlap would leave both models in contention.
Completed advice. “Use the four weeks for the measurement. Compare the observed result, including its uncertainty, with both prediction ranges. If only one accommodates it, use that finding in the design choice. If both do, retain both explanations. If neither does, check the measurement and reconsider both models. The broader survey offers no comparable basis for the design choice.”
The attainable gain is a better basis for instrument design. No immediate operational intervention is required. If instrument uncertainty makes every attainable result compatible with both predictions, or a more valuable use of the four weeks becomes available, reconsider the allocation.
Bias-Annotation
The recipient's urgency can make cheap action seem preferable even when a decisive check is available. Conversely, the advice issuer can enjoy the benefit of confidence while transferring investigation costs and delay to others. Compare those consequences explicitly where they differ across affected parties.
An operational vocabulary can undervalue explanation and future options. State the receiving horizon and research gain before judging inquiry. A safety label can also suppress appraisal of a criterion, while a cost argument can hide its protective contribution. Preserve both the merits question and the requirement's current force.
Conformance Checklist
The following conditions concern the advice being composed or examined. They require no separately completed checklist.
Common Anti-Patterns and How to Avoid Them
Consequences
Advice becomes a usable answer to a receiving question. The practitioner can continue with qualified knowledge, select worthwhile inquiry, repair a requirement proposal or recognize an impasse. A reviewer can repair an external report without requiring FPF-formatted input.
The method costs the attention needed to recover hidden work and compare its consequences. That cost is unnecessary for an already clear bounded answer. It also cannot make unavailable means available or settle domain evidence and authority questions by general reasoning alone.
Architectural Rationale
The recurring problem is the connection between a qualified claim and the work an answer asks somebody to perform. A.10 supplies bounded reliance; C.11 chooses among available options and tests probe-worthiness; C.11.CRC compares a finite contribution; C.19.2 limits apparatus effort under a declaration. Here those contributions are composed after recovering the receiving question and demand, including appraisal of the declaration itself when disputed.
This gives advice authoring and advice review the same method. Both start with what the recipient would do and end with usable advice. A wording repair under F.19 may make a demand clear while leaving its cost or merits unresolved. A decision calculation may be correct while answering a different question from the one the recipient needs.
The smaller route is direct use of the existing patterns or a short composing entry. Prefer it whenever it reaches the same warranted completion at comparable effort. This companion is useful where recovery of implicit work, requirement appraisal and the connected completion would otherwise have to be reconstructed.
SoTA-Echoing
Value of information already supplies an answer to whether uncertainty is worth reducing. The additional work here is recovering the question hidden in advice and carrying the answer through completion. The following comparisons adopt source contributions at the effort the receiving problem needs.
Compared with demanding the ideal evidence package first, this method can reach a warranted explanation or bounded continuation with less work. Compared with immediate action without inquiry, it retains a feasible check that changes a choice or admissible claim. Those are conditional gains: the quality of the domain models, the available observations and the recipient's use remain decisive.
Relations
- Uses:
A.10for source-reliance boundaries;C.11for local choice and probe-worthiness;C.11.CRCwhen a finite contribution comparison is missing;C.19.2for economical application under the current declaration. - Coordinates with:
C.18when useful options or hypotheses must be generated;C.28and direct domain methods when the advice depends on causal inference;D.4for ethical mediation; the applicable assurance method for a stronger guarantee. - Used by:
F.19when clear wording leaves a questionable substantive demand;E.19when advice usefulness, feasibility or evidence burden is the live review question;E.11.PUAwhen a pattern's demanded work obstructs its first useful result. - Leaves with their owners: authorization, requirement enactment, domain research design, actual Work and its evidence. The recommendation supplies a reasoned continuation; those contributions determine whether and how it can be carried out.
C.11.DUA:End
Compare the Worth of a Method Option
Type: Method pattern Status: Stable
Primary working result:
keep,revise,replace,branch, orstopfor one Method decision, or a retained set with its limitation when the choice remains unresolved. The result compares the receiving contribution and full burden of the current alternatives, then states what the selected action preserves, creates or changes within its qualification window. Bounded evidence may inform another Method-repertoire decision without transferring this decision or a broader claim.
Use This When
Use this pattern when a practitioner must decide what to do with an admitted Method or status-preserved candidate relative to current alternatives. Use the available basis at its supported strength: observations, credible estimates, domain judgement or a structural constraint can support a present choice. Begin with the decision and alternative set: keep the present Method, revise it, replace it, branch by situation, or stop the Method change or use.
The first useful result is a practical-worth comparison for the named Method or candidate, decision situation and receiver. It makes the alternatives, material consequences and burdens, their support and limits, and the available action recoverable. Include the status quo, stopping or the next best use of scarce Work when it can change the decision. An existing domain comparison that already supplies this basis can be used directly.
Here practical worth is Plain practice wording for the situated judgment that a Method-related course is worth keeping, changing, replacing, branching, or stopping relative to named alternatives. It is not a universal score, Method admission, fit result, causal attribution, or organization-wide adoption decision.
Use C.11.DUA §4.3 directly when the whole question is a requirement’s justification; C.39.RQ §4.4 applies it to a situated Method criterion. Return to C.11.MV when a Method-related course also needs comparison. Do not use this pattern when no current alternative or decision receiver is named.
Conformance, familiarity, one favourable occurrence or Method popularity alone does not establish comparative worth. A.3.1.QM qualification, A.3.1.CC coherence and A.3.1.FT fit can inform the present comparison; each retains its own scope. A missing stronger empirical result can limit this decision without erasing an independently supported current choice.
Working Distinctions
Problem Frame
A Method can be coherent and fit a situation yet still demand too much capability, coordination, tooling, confidentiality exposure, recovery Work, or opportunity cost relative to another course. Conversely, a burdensome Method can be worth keeping where it prevents a consequential failure that a lighter alternative misses.
Teams often hide that choice in one score or in Method familiarity. The score erases who bears the burden, which Systems and direct relations are required, what evidence supports each result, and whether an alternative should be retained for another situation. Practical worth must therefore remain a situated comparison with visible trade-offs.
Problem
Conformance can become value, fit can become superiority, one positive trial can become effectiveness, and immediate participant preference can become long-term worth. A new Method may be compared with an undefined status quo, while stopping or spending the same effort elsewhere is omitted. Capability, support, authority, confidentiality, recovery, and maintenance demands disappear into a weighted total whose assumptions cannot be recovered.
The resulting decision is hard to challenge and harder to revise. Later evidence cannot show which cell changed, a non-dominated alternative is discarded, and a local keep becomes an organization-wide adoption claim.
Forces
Solution
C.11 and C.11.DUA supply choice and the appraisal of demanded work. The comparison here constructs their Method-option input: the attainable receiving contribution, all relevant bearers of learning, transition and recurring burden, feasible alternatives and the condition under which keeping the current way is preferable.
Compare the feasible alternatives for the same receiving decision. Identify who would obtain each result and who would carry each burden. If the decision rule settles the trade-off, give the supported action and the follow-up in §4.1. If deciding between alternatives requires a preference or authority the evaluator lacks, return the alternatives that remain and identify whose decision is needed.
Pattern-Use Unfolding
-
Name the decision and receiver. State who will use the result, which Method or candidate account is under decision, the receiving practice and domain result, qualification window, authority boundary, and the action that
keep,revise,replace,branch, orstopwould change. The evaluation result supplies no permission or authority by itself. -
Construct the current alternative set. Include feasible Methods or variants, a support or capability change when it can address the same problem, the status quo, and stopping or the next-slot use of scarce Work when material. Preserve admission and candidate status for every alternative.
-
State the comparison basis. Name the domain results and material burdens, side effects, opportunity costs and reversibility conditions, with the evidence supporting them. Identify required capability, System participation, Work and direct relations separately from the Characteristic values used in comparison. Add situation-specific characteristics where needed. State non-compensable stops and who bears each consequence. When the merits of a protective requirement are disputed, use §4.5; otherwise compare under the applicable criteria.
-
Assemble the available basis by alternative. Use current A.3.1.QM qualification, credible domain estimates or judgement, and structural constraints alongside A.3.1.WT actual Work, A.3.1.CC corrections and A.3.1.FT fit or transfer results where available and relevant. Distinguish observed results from estimates, recover the source and conditions of each relied-on claim, and retain the missing basis that limits the decision. Absence of evidence for one alternative is not evidence of its inferiority.
-
Preserve the receiving result and Method contribution boundary. Record the domain result separately from Method task completion, support use, report, or self-assessment. Distinguish direct observation, association, contribution, and causal claim. If contribution is unknown, compare the observed or estimated bundle at its supported strength and carry the attribution limit where it matters to the recipient.
-
Compare capability, System, Work, and relation demands explicitly. For every alternative, name holder-capability inputs, participating Systems, Agent-performed Work, direct relations, assignments, permissions, authority, access, provider and confidentiality boundaries, maintenance, recovery, and evidence Work that changes the decision. Do not let availability or assignment substitute for capability.
-
Keep trade-offs visible. Identify gains and burdens for each affected party. If improving one party's result increases another's burden, apply the agreed priority or return the non-dominated alternatives to the holder of that decision. A recommendation can be complete before the holder decides. Use A.19.CPM when the trade-off calls for comparing multi-criteria profiles under a declared rule and A.19.SelectorMechanism when the receiving work calls for a set-valued selection.
-
Compare branch and further-inquiry options where they matter. Ask whether alternatives serve different situation families and what information, retained material, capability, support or rollback a branch needs. Identify who will select the method in use and whether they can distinguish the relevant situations in time to act. For a guarantee, the course selected from the available information must work in every compatible situation covered by the commitment. For a risk-limited commitment, assess the resulting rule against its risk criterion. For an undertaking that covers concurrent work, also establish that the branch choices can be carried out together with the support actually available. Select a further trial or inquiry through C.11.DUA only when its obtainable result could change the choice enough to warrant its full burden, including displaced Work and the available window. An adequate current choice can finish directly.
-
Return the strongest supported action.
keepretains the present course for the named window;revisenames the decision or maintained result to reconsider under step 10;replacenames the selected alternative and transition conditions;branchnames the situation discriminator and retained alternatives;stopnames the stopped use or change, what happens to current Work, and any unresolved trade-off. If the available basis cannot distinguish the possible actions, return the retained set and the limitation, together with any supported present course. Select a further probe only under step 8. A stronger choice can remain unresolved when no worthwhile inquiry is feasible. -
Route each action according to its kind. For
revise,replace, orbranch, name the receiving C.39.RP repertoire choice, C.39.RQ requirement, C.32.MWA.PP architecture decision, B.1.5.CM construction result, or A.3.1.MV entry and the change needed there. If an existing claim is contradicted, identify that claim and reopen only its owning result. If this is a first situated choice or a new applicability distinction, create the needed decision or entry through its owner; do not invent an earlier claim or contradiction.keepadds no repair return.stoppreserves any unresolved trade-off and stopped use without inventing a correction owner.If the comparison exposes a possible A.3.2.DE description, C.37.MP Method representation profile-row, A.15.8.MS support, or directly governed lower representation contradiction, return the bounded evidence to A.3.1.CC; A.3.1.CC identifies and rechecks the maintained correction owner. A worth result does not rewrite that owner automatically.
-
Return evidence to another Method-repertoire decision only when needed. Supply the situated Method-worth evidence and limits needed for a compatible decision in engineering or another subject practice. This return transfers neither the C.11.MV action, authority, Method rank, causal claim, nor evidence beyond its source-use conditions.
-
State refresh conditions. Reopen when alternatives, domain consequences, capability or support demands, direct relations, evidence reach, costs, reversibility, situation fit, or the receiving decision window changes.
Record the Result
Ordinary prose or an existing comparison can carry this content. Use the table to recover a material missing position; it does not require another row after an adequate result.
What Changes in Practice
Teams stop asking whether a Method is “best” in the abstract. They can see what it changes in the receiving practice, what it demands from people and Systems, which side effects and opportunities matter, what the evidence reaches, and why several alternatives may remain.
This also makes stopping constructive. A stop can protect scarce Work or avoid an unrecoverable commitment while retaining the evidence and conditions needed to reopen the decision later.
Minimal Constructed Worth Replay
Calibration-Team-A is choosing how to review reports through day D30. Until now, reviewers have used a lightweight tolerance checklist (ALT-L) for both uniform-unit and mixed-source reports. Their current procedure permits a tolerance decision on the displayed numbers without first establishing common units. An admitted alternative, M-Unit-Review-1 (ALT-M), requires the reviewer to establish common units, compare the normalized values against the declared tolerance and stop release if a mismatch remains unresolved. A manual conversion table is available. Use description edition 4 with the representation repair from A.3.1.CC. ALT-S keeps release on hold when neither review can support it. A difference in checklist wording alone would not establish a difference in the permitted review actions.
The evidence for ALT-M includes the two earlier admitted A.3.1.WT occurrences of M-Unit-Review-1 performed by the same qualified Reviewer-17: seven minutes on the ordinary same-unit report and twelve minutes plus one table lookup on the mixed-unit report; both decisions were correct and the mixed-unit release was stopped. A separate constructed matched replay of ALT-L, admitted as W-L-Rep-1 and W-L-Disc-1 under the same holder, team, report pair, and window, took five and six minutes. It accepted the ordinary report but failed to expose the mixed-unit mismatch.
The replay supplies no causal claim beyond these occurrences and no evidence about another holder or automated support route. The two earlier A.3.1.WT observations are not trials of the later representation repair in A.3.1.CC.
Recommend a branch through D30. Where source units may vary, use ALT-M so that the release decision depends on establishing common units. Where the input source guarantees common units, retain ALT-L. If the source units cannot be established and the formal review cannot resolve them, retain ALT-S. The two matched reports support the observed differences in review decisions and effort. Use of the repaired representation has yet to be observed.
In this constructed case the team holds the review decision, has the reviewer and table available, and adopts that recommendation for the D30 window. Record the new applicability entry APP-UnitReview-TeamA-D30 through A.3.1.MV, including the three alternatives and the source-unit condition. The team changed when reviewers use each procedure. The earlier observations still concern the procedures used on those reports.
This result covers the stated conditions through D30. For a later reliance, use A.10.1 to identify which premises remain applicable and which qualification, condition or evidence gap needs renewal. A real expiry still limits use; new field investigation is selected for its useful obtainable answer, not merely because the date advanced.
This result does not show that ALT-M is universally superior or caused every correct decision. It demonstrates why lower burden alone cannot compensate for the current release stop. The merits of that stop are a separate question when they are disputed; use §4.5 for that appraisal.
An Adequate Choice Before the Review Window Closes
Suppose the current report sources have an inspectable same-unit schema, the reviewer and support remain qualified for this use, and the existing comparison covers the present release question. Keep ALT-L for that bounded case and preserve the ALT-M branch whenever source-unit variability matters. Another forty-minute replay would displace review of a newly changed report before the deadline without changing this choice. The recommendation is complete on the available basis.
An Unresolved Alternative with No Obtainable Probe
A proposed automated conversion route might reduce later review effort, but its contribution remains unknown and no permitted test access is available before the current decision. The existing ALT-M route still supplies the required review under its stated conditions. Retain the automation proposal for a later comparison, stop the unsupported present change, and continue the qualified ALT-M course. That decision leaves the automation comparison unresolved. Further investigation can be chosen when its obtainable answer is worth the full burden; retaining the proposal creates no trial commitment.
Appraise a Disputed Protective Requirement
Use C.11.DUA §4.3 when the live comparison asks whether a protective criterion, threshold or evidence demand is justified. Identify the protected person, system or result and the particular unwanted outcome. Examine the governed quantity or activity, the measurement and causal basis needed for this question, the threshold’s value choice, and the additional protection under the present conditions. Compare that protection with the burden, delay and displaced protective Work, using the domain methods needed to judge them. Return supported retention, tightening, revision, replacement or removal only among meaningful alternatives.
In the unit-review case, normalization exposes a consequential mismatch that the tolerance-only checklist missed. That supports preserving the protection where source units can vary. Whether every report needs the same extra operation still depends on the unit guarantees, what the operation adds and what its burden displaces. The label safety and the fact that a rule is currently binding settle neither that merits question nor the authority to amend it.
Name the disputed protective rule and explain whether it should be retained or changed, what protection your advice preserves, and why the evidence supports it. If you propose a change, identify who can approve it and when it can take effect. End the appraisal with the supported advice and its relevant limits. Approval of a proposed change is a separate decision. Until the rule is changed, select only a way of working that complies with it. If none can deliver the needed result by the deadline, report that consequence. A later change of the rule does not undo the missed opportunity.
Archetypal Grounding - Field Evidence and Alternatives
The SSFD study reports 95 examples of evidenced individual benefits across 41 detailed workplace-project reports and distinguishes direct and indirect contribution. Those reports can supply domain-result, process, capability, burden, and evidence-strength positions for a worth comparison. They do not by themselves supply a common current alternative, comparable counterfactual Work, SSFD-only causality, or one universal decision.
Project 11 supplies substantial reported outputs from one by-wire steering application; Project 32 supplies a representation difficulty. Preserve both. A practical-worth result may retain SSFD for one situation, revise its representation or support for another, or return missing alternative evidence. Do not average the difficulty and benefit counts into a score.
The sustainable-design study supplies three families of separate workshop observations: The Natural Step, Whole System Mapping, and Biomimicry across 23 workshops, more than 172 qualified respondents, 27 companies, and three industries. Companies chose which workshops they took; most received two or more, order varied, and participants could differ.
The source reports immediate participant-perceived value of activities and mindsets and warns against theoretical-only recommendations. It does not itself supply one named receiver's feasible current-alternative set, a common decision, matched alternative Work, or a counterfactual basis.
Use those observations for the participant experiences they document. For a present choice, identify the feasible options, recipient and decision rule. A binding constraint or an already established difference can settle that choice. If it depends on a comparative effect on later work, account for voluntary participation, unmatched timing and participants, immediate self-report and any missing basis for transfer, causation or combining workshops. Return what can be decided with those limits.
The Halogen case can supply evidence about project demands, practitioner skill sets, organization, and cyclic adaptation. It can justify a branch or variant question when conditions differ, but one firm's history does not make its repertoire universally preferable.
EC-417 Early Stop
The full EC-417 application supplies the receiving problem, alternatives, calculations and the distinct limits used below.
In the constructed EC-417 case, the team must release a controller firmware and harness change on D0. Its software-integration slot is D-21, but signed supplier pinout evidence is expected at D-8. The present question is which proposed way to coordinate that release is worth pursuing under the current constraints.
C.32.MWA.PP §5.3 supplies the timing and capacity comparison:
B2 moves preparation work to another role. Verify that its holder can absorb the 1.60 hours in the required window, is qualified to prepare the difference, and can access the relevant evidence under the supplier's confidentiality conditions. Include the work that allocation displaces and how project, user and provider Systems exchange, check and recover the records through the repository, PLM and CI systems. Three A.15.8.MS support tasks establish only the results observed on those tasks. A three-release WorkPlan describes intended work.
Keep C-EC-Release-v2 as a candidate. Preserve the missing A.22 selection basis, untested AI-provider and feedback branches, and unsupported CUR-EC417-CadenceEffect-1.
With evidence at D−8, A misses the original slot. B exceeds the safety engineer's limit. B2 reduces the estimated demand on that engineer from 4.40 to 2.80 hours, within 3.20 hours. Recommend at most three trials if B2 also meets the supplier and trial-entry conditions. The responsible holder can select a trial on those grounds. If it cannot proceed, defer or stop, or seek a decision to change the slot for A. Arrival of signed evidence by D−21 would reopen A for the original slot.
An actual release or trial entry still needs every applicable capability, assignment, permission and authority condition. If the question is demonstrated release worth, return the missing performed-release and alternative evidence. The observed support-task results, with the tailoring operation still undefined and its permission/authority stop untested, do not establish that stronger result.
Use situated evidence in another Method-repertoire decision
Supply evidence to another Method-repertoire decision only when that receiving use is compatible and the following content is recoverable:
The receiving practice remains responsible for its Method repertoire and compatibility claims. The contribution supplies evidence for that decision; it does not direct the recipient to take the C.11.MV action.
Bias-Annotation
Conformance Checklist
- One decision receiver, Method or candidate subject, receiving practice, domain result, action, and qualification window are named.
- The alternative set contains only current feasible courses and includes status quo, stop, or next-slot Work when material.
- Admission and candidate status are preserved for every Method-related alternative.
- Consequences, burdens and their bearers, applicable Characteristic values, required participants and relations, and evidence strength remain distinct in the comparison basis.
- Qualification, observations, estimates and structural or domain judgement retain their sources, applicable conditions and limits; stronger observed, fit, transfer and causal claims have the evidence they require.
- Missing evidence is not converted into inferiority, and one favourable occurrence is not converted into causality.
- Trade-offs, non-dominated alternatives, branch discriminators, and non-compensable stops remain visible.
- When requirement merits are disputed, the recommendation compares the protective contribution and burden, retains the threshold or evidence basis needed for that judgement, and separates it from current force and feasible amendment.
- The result is a supported
keep,revise,replace,branch,stop, or retained set with its limit and any supported present course. A further probe is selected only for a useful obtainable answer worth its whole burden. - A
revise,replace, orbranchresult names the owning result under §4.1 step 10 and distinguishes a new situated choice from correction of an identified contradiction;keepadds no repair return;stoppreserves any unresolved trade-off and stopped use. - A possible description, representation or support contradiction returns as bounded evidence to A.3.1.CC under §4.1 step 10 for the applicable owner-specific check.
- Any evidence transferred to another repertoire decision states its source-use conditions and non-transfer boundary.
Common Anti-Patterns and How to Avoid Them
Consequences
Practical-worth decisions become revisable accounts rather than Method rankings. Teams can act while retaining branch conditions, evidence gaps, and stop alternatives, and later evidence can reopen the comparison position that changed.
The cost is resisting a convenient scalar answer. Some comparisons leave several alternatives open or justify another probe, and the team must name burdens, capability and support demands, side effects, and opportunity costs that were previously hidden.
Rationale
Worth is relational: a Method-related course is worth something for a receiver, situation, domain result, alternative set, and time. Coherence and fit can be necessary inputs to the decision, but they do not settle it. A visible trade-off is more actionable than a score whose composition and evidence reach are hidden.
The five actions distinguish maintenance from change.
Keeppreserves a situated course and adds no repair return.revise,replace, andbranchidentify the owning decision or entry and distinguish a newly needed choice from correction of an existing claim.stopprotects against unsupported or dominated continuation while preserving any unresolved trade-off and stopped use without inventing a correction owner.
None supplies implementation authority.
SoTA-Echoing
Relations
- A.3.1.QM can supply a bounded qualification for the present choice. A.3.1.WT supplies actual Work, observations, results, burdens, adaptations, and evidence gaps. A.3.1.CC supplies coherence corrections; A.3.1.FT supplies fit, failure, applicability, and transfer limits. None supplies the practical-worth action.
- For
revise,replace, orbranch, C.39.RP owns repertoire choices, C.39.RQ requirements, C.32.MWA.PP architecture decisions, B.1.5.CM construction results, and A.3.1.MV maintained repertoire/applicability entries. Name the result to create or change; identify a contradicted claim only when one exists.keepadds no repair return;stoppreserves any unresolved trade-off and stopped use. - A possible A.3.2.DE description, C.37.MP Method representation profile-row, A.15.8.MS support, or directly governed lower representation contradiction returns as bounded evidence to A.3.1.CC, which identifies and rechecks the maintained correction owner.
A.19.CPMandA.19.SelectorMechanismgovern non-trivial comparison and selected-set semantics.A.2.2governs capability inputs;A.10evidence use;A.15.1actual Work; direct governors remain authoritative for Systems, Agent-performed Work, relations, assignments, permissions, authority, access, and results.- A compatible Method-repertoire decision in another practice may use the bounded Method-worth evidence. This use transfers neither action nor authority and does not establish composition of the supplying and receiving disciplines.
- A.3.1.MI may consume the resulting bounded decision when introducing or revising a Method in practice; later observations can reopen C.11.MV without retroactively changing the evidence used here.
C.11.MV:End
C.13 — Constructional Mereology (Compose‑CAL)
Status: Stable Type: Pattern
At a glance. Use C.13 when a practitioner must show how entities and relations that obtain form one whole, collection, aspect, or measured portion. The account names the relations and conditions on which that construction or delimitation depends.
Use this when. Use this pattern after the direct relation patterns have identified the participants and relations that obtain, when you need a compact, inspectable account of how they assemble a whole, form a collection, distinguish an aspect, or delimit a measured portion. Reuse an existing ordinary account when it already supplies those facts and their construction rule; the notation is optional.
First useful move. Start with the construction or delimitation the recipient needs explained. When its extent needs choosing, use the relation the question concerns. In an assembly, vary how the same parts are arranged or how a part is used in its enclosing whole. For a collection or a measured portion, consider how a change in membership or boundary would affect the answer. Follow the relevant relation as far as the conclusion or its required grounds depend on it (C.39.SA:4). Once the relevant participants and relations are established, state the construction or delimitation rule. Use sum, set, or slice where the notation helps.
What goes wrong if missed. A readable component, belongs-to, or aspect statement may lack its construction account; or the opposite mistake occurs and a diagram, list, or Γ_m expression is treated as if it created a whole, a relation, or a holon.
What this buys. A compact three-form construction discipline that keeps integrated assembly, collection, aspect, and measured-portion accounts distinct while leaving relation conditions, whole identity, evidence, and public relation names with the patterns that define them.
Not this pattern when. Not this pattern when the current question is only relation vocabulary, evidence or assurance without a structural claim, epistemic representation, a selected dependent U.Structure, temporal phase without an aspect claim, public-kind admission, or transformation composition without a direct transformation-composition governor.
Intent
Provide three forms of construction account: how constituents assemble an integrated whole, how entities form a collection under that collection's own belongs-to rule, and how an aspect or measured portion is distinguished from its bearer or whole under its own rule. The calculus records how already identified entities and obtaining relations support that named object.
Also known as “Γₘ mereology” and “constructor-based composition”.
Layer. calculus. Depends on. A.14 and the direct relation patterns for participants, obtaining conditions, and occurrence identity; the direct kind pattern for the candidate whole and its identity or reidentification rule. Consumed by. A.1 when candidate holon recognition needs constructive assembly, B.3.5 when a named assurance use needs a structural grounding account, and subject patterns that need a compact construction narrative.
Compose-CAL keeps exactly three narrative forms—sum, set, and slice. A materialized construction trace is a C.2.1 episteme about the construction facts. Its claims can designate entities, relation occurrences, rules and identity conditions. Designation alone does not make the trace a participant in the construction; any such participation needs its own direct relation basis.
Problem Frame
FPF needs both readable structural relations and a recoverable account of how constituents assemble a whole, entities form a collection under its belongs-to rule, a facet distinguishes an aspect, or a boundary delimits a measured portion. A relation name alone may leave the construction opaque. A constructor expression alone can commit the opposite error by treating syntax as the source of entities, relation obtaining, or whole identity.
Problem
A bare list of ComponentOf, collection-specific belongs-to, or AspectOf claims does not say which assembly, collection rule, or facet makes them one construction. A bare sum, set, or slice expression also leaves the construction unexplained: the same constituents can participate in different assemblies, a collection need not be an integrated holon, and an arbitrary facet label does not establish an aspect. The construction account must therefore name world-side facts and preserve the candidate's direct identity or reidentification rule.
Forces
- Parsimony vs truth. Three construction forms are easier to reuse than an open constructor catalogue, but no form may replace a missing direct relation or assembly rule.
- Readable statement vs constructive account. Practitioners need ordinary component, belongs-to, and aspect statements; reviewers may also need to inspect how those relations support the described construction.
- Input set vs assembly. The same entities can be assembled through different obtaining relations and can therefore yield different wholes.
- Continuity vs extensional snapshots. Constituents and part-relation occurrences can change while the same whole continues when its direct reidentification rule permits the phase change.
- Construction vs evidence. A trace states the construction account; evidence and assurance separately support or warrant the claim content.
- Cross-domain reuse vs owner bypass. Systems, epistemes, methods, and work occurrences can all need constructive grounding, but their direct patterns retain kind, part, and identity authority.
Solution
Solution sketch
Choose the form by the relation you need to explain. Use sum for an assembly, set for membership in a collection, and the appropriate slice for an aspect or a bounded portion. An installed pump can belong to a fleet and be a component of a skid. Each claim needs its own facts.
The familiar argument lists are readable shorthand, not complete ontological signatures. A complete use names the described object and states the direct facts beside the shorthand. Input order does not matter to the list of designated inputs, but assembly relations, rules, facets, portion boundaries and identity conditions do matter. The same input set under a different assembly can yield another whole.
A construction may obtain in the world even when no trace episteme has been written. When another piece of work needs an inspectable trace, materialize a C.2.1 episteme whose claim content names the described object; its constituents, entities that belong, bearer, or containing whole; the relation occurrences that obtain; the construction rule; and the identity conditions.
Creating, editing, publishing or losing the account does not, by itself, establish a change in the construction it describes. A register update that also changes belonging must be assessed under the collection's constitutive rule.
Do not add a fourth constructor merely to carry time, execution order, parallelism, representation, evidence, or a domain label. Keep those claims with their direct patterns beside the construction account.
A measured-portion account
Use the portion profile of slice when the receiving work needs to inspect a PortionOf claim. First identify the portion x and whole y under A.14:5.1. State what delimits x within y, why they are the same kind of stuff or content, the extensive measure μ and its unit or declared conversion, and the conditions under which both identities and the relation hold. An amount alone leaves the particular portion unidentified. These facts support the ordinary sentence “x is a portion of y under this measure and boundary rule.”
Γ_m.slice(y, portionBoundary) is optional shorthand for that account. Its second argument identifies the delimiting rule; it is not the aspect profile's facet argument. The result describes x and its already established PortionOf relation. For ProperPortionOf, also check 0 < μ(x) < μ(y) under A.14:5.1; additivity alone does not establish it. For a joined portion, apply A.14's disjointness and admitted-join conditions before adding measures. A boundary, unit or measure change reopens the affected account; the direct identity rule decides whether the portion or whole continues.
To infer a value for the whole from observations of its parts, say how the characteristic of the whole depends on those parts under the observation conditions. To obtain total mass by adding the portion masses directly, the portions must cover the whole without overlap and their masses must be comparable. A local temperature reading with its uncertainty can establish that a limit was exceeded at one spot at the time of measurement. To establish that the whole stayed below the limit, cover the unobserved material and times with observations or a justified bound. C.16 gives the measurement method for the chosen characteristic.
In this constructed example, Billet-20 is an identified 20 kg stock. EndPortion-5 is the same material between its left end face and a specified cross-section inside that billet; its mass is 5 kg under the same measurement conditions. The account names both objects, those boundaries and the identity conditions of the stock and portion, so it supports ProperPortionOf(EndPortion-5, Billet-20) and the ratio 5/20 = 25%. An AspectOf or integrated ComponentOf claim requires its own relation grounds. If the indicated cross-section is unknown, “5 kg of Billet-20” gives an amount but leaves this particular portion unresolved. After physical separation, inspect the new boundaries and stock identity before carrying the earlier relation into the new state.
The same operation can describe identified report text spanning pages 1–10 of a 120-page edition under fixed pagination. That account establishes the text extent; a claim that those pages form a logical constituent needs its separate content relation. Repagination requires recovering the same text span and recalculating its page measure. The text may remain the same while its page-count account changes.
When B.3.5 is elected for the PortionOf assertion, link tv:groundedBy to this C.2.1 account and declare validationMode=axiomatic. Evidence needed to establish its boundary, mass or other facts remains subject to the receiving assurance question. Without that profile election, the direct relation remains usable under A.14.
Normative Standard (high‑level)
- C13-N1 — Direct facts first. A C.13 trace is conformant only when every named constituent or collection entity and every named part, collection-belonging, or aspect occurrence is independently identified under its direct pattern.
- C13-N2 — Whole identity stays direct. The candidate whole follows its direct identity and reidentification rule. Equality of an input list or trace does not decide whether the existing whole continues or a new whole must be identified; a permitted constituent replacement can preserve one whole, while the same constituents under another assembly can form another whole.
- C13-N3 — Construction form.
sum,set, andsliceare the only C.13 forms. Reordering or duplicate designation does not change the listed inputs; it does not merge distinct entities or relation occurrences and does not erase assembly, facet, or portion-boundary differences. - C13-N4 — Mereological discipline. Direct part patterns and A.14 govern acyclicity, antisymmetry, transitivity where applicable, recurrence, and occurrence identity. C.13 does not strengthen a direct relation by calculus convention.
- C13-N5 — Trace separation. A materialized trace is a C.2.1 episteme. Its notation or publication alone establishes neither the described entities and relations, their construction and identity rules, nor holonhood.
- C13-N6 — A collection account is not component assembly.
setsupports a collection account; it does not imply integrated assembly,ComponentOf, system agency, or A.1 recognition. Usesumonly when constructive part relations and assembly are independently grounded. - C13-N7 — Elected assurance grounding stays with B.3.5. A direct structural Working-Model edge may be published without a C.13 trace. When that publication elects B.3.5 or a named current requirement demands the profile, follow B.3.5 for the required
tv:groundedBylink and declaredvalidationMode; C.13 supplies the reconstructible trace content. Evidence, warrant, currentness, and reliance remain separate, and neither the link nor the mode replaces the direct construction or identity facts. - C13-N8 — Subject owners remain authoritative. Method, work, and discipline construction may use C.13 only after their direct patterns identify parts and whole-forming relations. A selected dependent
U.Structuredoes not become a whole, collection, aspect, or holon by selection or name. - C13-N9 — Transformation composition stays fail-closed. A C.13 entity-construction trace, method decomposition, work decomposition, common changed referent, or temporal subdivision establishes neither transformation parthood nor a composite transformation. Without a direct transformation-composition governor, retain the independently identified changes and the blocker; infer neither composition nor atomism.
Scope, applicability, terms & notation
Use Compose-CAL when the claim concerns the construction of one system, episteme, method, work occurrence, discipline, collection, aspect, or measured portion and its direct part patterns already supply the required participants and obtaining relations. Do not apply it merely because a sentence contains part, a diagram groups nodes, or a selected structure organizes relations.
Γ_m— the three-form notation for a C.13 construction narrative.- construction trace — claim content that names the object being described; its inputs; the direct relations that obtain; its construction rule; and its identity conditions; when materialized, it is a C.2.1 episteme.
- constructive part relation — a world-side relation occurrence governed by its direct pattern; it is designated by the trace but not created by it.
- assembly rule or method — the rule explaining the construction: assembly from constituents, collection belonging, aspect distinction under a facet, or portion delimitation under a measure. The account cites that rule separately from the object it describes.
- identity or reidentification rule — the direct rule that identifies the described object and says which changes preserve or end it.
Alias readiness. These are common readable projections only when their direct relation meanings match the case:
ComponentOfmay accompany asumconstruction;- an ordinary belongs-to sentence may accompany a
setconstruction; the collection's own pattern supplies its meaning, occurrence history, and any recurrence rule; AspectOfmay accompany asliceconstruction;PortionOfuses the measured-portion profile ofslicein :4.1.1, with A.14's same-stuff/content, measure and boundary conditions;ConstituentOfneeds the direct logical or content-part relation; material mixtures use the pattern for their portion or component relation.
The readable label and the trace answer different questions. The label states which direct relation obtains; the trace explains how the relation set supports this construction. Neither one is evidence merely by being present.
Structural CT2R Typing-Grounding Use
When a target kind or logical representation must reuse both a Γ_m construction account and an independently grounded Working-Model relation, explain their correspondence and which mereological structure the target preserves or loses. The StructuralCT2RTypingGroundingUnfoldingStructureBlock from B.3.5 is an optional way to record that account.
Use A.22 when the receiving question needs a selected structure and A.22.CGUS when potential continuations and their conditions are at issue. C.13 supplies the subject-side construction relations. Recover any required bridge, kind intent, proof, empirical evidence and admissible reuse under their governing patterns. Apply A.7.1 when an inadequate working account must be diagnosed against the subject construction.
Use this split especially when a readable relation label such as ComponentOf, a collection's belongs-to predicate, AspectOf, ConstituentOf, or RepresentationOf is being reused beyond what its sentence warrants. The label does not by itself prove constructive grounding or a wider structural projection. Name the construction trace and Working-Model relation, the target kind or logical representation, any bridge used, the structure preserved and collapsed, and the proof or evidence relation required by the stronger claim. If the evidence instead diagnoses a mismatch that requires revision of the working ontology, apply A.7.1.
Archetypal Grounding
Pump Skid: Integrated Assembly, Not A Parts List
PumpSkid #7 is assembled from identified Pump, Motor, Baseframe, Manifold, enclosure, pipe, cable, and connector entities. The direct mechanical, electrical, and fluid part-relation patterns identify the fastening, coupling, enclosure, terminal, flange, and seal occurrences. The skid assembly method or rule states how those facts form the candidate, and the skid reidentification rule says which replacements preserve PumpSkid #7.
The shorthand Γ_m.sum{Pump, Motor, Baseframe, Manifold, ...} is truthful only with those facts beside it. It records an integrated-assembly construction; it does not make the relations obtain. The composition sustains skid-level boundary, interface, load-envelope, and operating characteristics not attributable to one constituent alone.
A more detailed account can establish more about the same motor, such as the torque it can deliver. To compare two models of skid performance, use the same question about the skid in both. State how the model of the coupled parts predicts the whole's performance. Under B.5.MR:4.2 a strict refinement preserves the earlier assertions in every admissible configuration, admits at least one configuration, and rules out a previously possible performance answer.
The same parts unconnected on a pallet do not form the skid. A bill of materials and drawing are epistemes about intended or possible assembly. They do not substitute for actual part relations, assembly, or identity. The same part set connected in a different assembly may constitute another whole.
Collection And Aspect
A fleet register can identify one collection and the vehicles that belong to it under the registration rule. Γ_m.set{Vehicle-1, Vehicle-2, ...} reports that collection only after the fleet identity, registration rule, and each belongs-to fact have been established. If the same candidate later passes all six A.1 matters, a separate sum account may report its independently grounded constructive parts and assembly.
Reuse the filled Reactor-7 case in A.14:5.3: Γ_m.slice(Reactor-7, thermal-boundary) reports ThermalEnvelope-7, the insulation panels, seals, and boundary interfaces picked out by the thermal-boundary rule, the obtaining AspectOf occurrence, and the enclosure identity and ending conditions. A permitted panel replacement can preserve the aspect while requiring a current trace; dismantling the enclosure, changing the facet rule, or reidentifying Reactor-7 ends the reported occurrence.
Episteme, Method, Work, And Discipline Holons
A theory episteme may use C.13 only after C.2.1 identifies the episteme and the direct semantic-part rule in A.14 and any applicable claim-composition rule identify its constituents, their integration and the assembly and reidentification rule. When the question is what their grounds support together, use B.1.3:4. C.2.1 claim-graph content helps constitute the episteme but does not thereby supply a C.13 part predicate. For example, a definition, diagram or evidence item enters a sum trace only when a direct part predicate independently obtains; stating, showing or supporting the theory alone is insufficient. The trace then reports the claim-bearing assembly rather than creating it.
A composite Method may use C.13 after B.1.5 or its direct Method-composition pattern identifies submethods, whole-forming relations, constraints and the whole Method's reidentification rule. For the operational parts of a particular performance, A.15.1:4.1a and B.1.5.EW:4 ask what work was done through each constituent. For instance, an inspection Method can permit either checking a current calibration certificate or performing a new calibration. If the inspector checked an existing certificate, that check belonged to the inspection Work. Whether the earlier calibration itself belonged to a larger Work depends on how that work was constituted. A recipe, plan, work breakdown or common performer can help locate the performance. Establish the actual part relation from the work performed.
A discipline may use C.13 only after C.20 and its direct relation owners identify the constituent entities required by the case, the whole-forming relations that obtain among them, and the discipline identity or reidentification rule. Do not infer parts from C.20 card positions or from objects merely associated with the discipline. For example, a practice, organization or canon item enters a sum trace only when a direct part predicate independently obtains.
Selected Structure And Transformation Stops
A selected BoundedModelUseStructure organizes identified model-use relations for a use. Selection does not by itself establish constructive parthood, agency, holonhood, or an MHT. C.13 can discuss construction of an underlying system, episteme, method, or work whole when its direct facts are available; it does not turn the selected relation organization into that whole.
Mounting, wiring, fluid-connection, and configuration changes may each be independently identified U.Transformation occurrences. A skid entity-construction trace does not make those changes constituents of one composite transformation. If a use requires transformation contribution, parthood, composite identity, or holonhood and no direct governor supplies the relevant compatibility and reidentification law, retain the separate changes and the blocker. Missing composition facts establish neither composition nor indivisibility.
Scope Justification
The same three forms work across mechanical, biological, informational, Method, Work, and Discipline cases: sum reports integrated assembly, set collection belonging, and slice either an aspect or a measured portion under its distinct profile. Using them does not replace the patterns that define the particular entities and relations. Order and timing remain with Method and temporal patterns; evidence and warrant remain with assurance patterns; public kind admission remains with E.24.UK; the direct identity rule decides whether the existing whole continues or a new whole must be identified, with B.2 used when reidentification is at issue.
Bias-Annotation (cognitive anti-patterns and counter-moves)
Conformance Checklist (normative, calculus‑level)
The following regulate a C.13 use.
Common Anti-Patterns and How to Avoid Them
- Constructor as public relation. A
Γ_mtrace is shown as the relation the working reader should use. Keep the direct relation in ordinary prose and use the trace only for the construction account. - Trace as cause. A diagram, formula, parts list, or trace publication is said to create the assembly. Recover the actual inputs, obtaining relations, assembly, and identity rule.
- Item in a collection treated as a component. A
setconstruction is used to infer integrated assembly structure. Keep collection belonging distinct; usesumonly after constructive part relations and assembly independently obtain. - Same parts, same whole. Two assemblies with the same component names are treated as one whole. Compare their obtaining relations, assembly rules, boundaries, and identity conditions.
- Temporal constructor drift. A phase, schedule, or assembly order is modeled as a Compose-CAL constructor. Keep temporal and method claims in their own planes.
- Method or work shortcut. Recipe steps, plan items, or WBS rows are called parts of an actual method or work occurrence. Use the direct method- or work-composition owner first.
- Transformation shortcut. Changes of constituents are called parts of one transformation because the entity was assembled. State which transformation-composition governor is missing and retain the separately identified changes.
Consequences
Benefits
- Inspectable construction. A practitioner can recover which inputs, relations, and rule support the described construction or delimitation.
- Identity clarity. The account distinguishes the input list from the assembly and keeps whole reidentification with the direct identity rule.
- Human-first use. Ordinary component, belongs-to, aspect and measured-portion sentences remain readable; notation is optional shorthand for a named construction use.
- Plane separation. Order, time, evidence, representation, kind admission, and receiving reliance keep their direct owners.
- Cross-domain reuse. The same three forms can describe system, episteme, method, work, discipline, collection, aspect and measured-portion cases without claiming one universal part relation.
- Composition limits. Missing direct part or transformation-composition governors remain visible instead of being hidden by a trace.
Costs and mitigations
- A truthful trace needs more than a parts list: relation occurrences, assembly, and identity conditions must be named. Reuse a direct pattern's existing facts rather than duplicating them.
- A world-side construction can obtain before anyone writes a trace, and its direct relation claim can remain usable without an assurance profile. When its publication elects B.3.5 or a named current requirement demands it, materialize the required trace and validation mode without treating publication as the cause of construction.
- The same inputs can support different assemblies, and one whole can survive permitted replacements. Always carry the direct reidentification rule.
One-line takeaway.
sum,set, andsliceexplain identified construction facts in a form another reader can inspect.
Rationale (informative)
Why exactly three forms?
sum, set, and slice keep three recurring practitioner questions apart:
sumasks how identified constituents and constructive relations assemble an integrated whole;setasks which entities belong to a collection under that collection's identity and belongs-to rule;sliceasks which aspect is distinguished under a facet rule, or which measured portion is delimited under a boundary rule. These profiles keep their different relation conditions.
The forms are intentionally small, but their inputs do not determine ontology by themselves. An assembly is more than a set of part names; a collection is not automatically a holon or agent; and a facet label is not an aspect occurrence. The direct patterns provide relation obtaining and occurrence identity, while the candidate's direct pattern decides whether the existing whole continues or a new whole must be identified.
Why traces remain epistemic. A construction can obtain before anyone writes its account. A materialized trace is claim-bearing content used to inspect, communicate, or support that account. Evidence and assurance may warrant the claim, G.11 may govern the selected edition's currentness, and the account may be relied on in receiving work or declined for that use. None of those epistemic results changes the world-side assembly by itself.
Why order, time, selected structure, and transformation composition are outside. Method order and temporal extent answer different questions from parthood. A selected U.Structure is an organization of relations for a use, not automatically another holon. Entity construction also supplies no law by which several actual changes compose into one transformation. Keeping these stops explicit prevents an economical notation from becoming an ungoverned ontology.
SoTA-Echoing
A.14:14 supplies the source decision used by the set and slice forms. For set, it requires the collection's own belongs-to rule and blocks both automatic parthood and the conclusion that separate parthood is impossible. For the aspect profile of slice, it requires an obtaining AspectOf relation with its bearer, facet rule, and identity conditions. The portion profile instead uses A.14:5.1's same-stuff/content, extensive-measure and boundary conditions; it introduces no AspectOf claim. An ordinary change to a collection rule or occurrence, aspect or bearer, facet rule, portion boundary or measure, identity condition, or materialized trace reopens the affected construction account and the claim it reports.
Relations
Builds on
- A.14 and the patterns for each relation. They define and test the participants, conditions, recurrence, and occurrence identity for component, collection-specific belongs-to, aspect, portion, constituent, and other constructive relations.
- C.2.1. It governs the identity of a materialized construction-trace episteme.
Coordinates with
- B.3.5. Governs the required trace link and validation mode only for a structural Working-Model edge covered by an elected profile or named current requirement. Its publication and assurance apparatus does not create the construction or decide whole identity.
- A.1 and B.2. A.1 consumes constructive assembly as one component of holon recognition; B.2 consumes construction and direct reidentification facts when the question is whether the existing whole continues or a new whole must be identified. C.13 decides neither public-kind recognition nor whole reidentification.
- B.1.5 and direct method-composition patterns. Supply submethods, whole-forming relations, constraints, and whole-method reidentification before a method construction is narrated.
- A.15 and direct work-mereology patterns. Supply work parts and whole-forming relations before a dated work construction is narrated.
- C.20 and direct discipline-composition relations. Supply discipline constituents, whole-forming relations, and identity conditions before a discipline construction is narrated.
- A.22 and A.1.1. Keep selected relation organization and bounded model use distinct from construction of the underlying holon.
- A.3.4. Identifies each actual bounded change independently; C.13 supplies no transformation-composition law.
- C.29. Keeps formulas, graphs, diagrams, and traces as representations of independently recovered objects when representation is at issue.
- A.7.1 and A.22.CGUS. Handle diagnostic return and the wider structural projection from construction to a target kind or logical representation when those uses are at issue.
Constrains
- A pattern that relies on a constructive whole states that whole, its constituents, the part relations that obtain, the assembly rule, and the identity rule rather than relying on a list or diagram.
- Collection, aspect, portion, method, work, discipline, selected-structure, and transformation cases keep their direct owners and stop at missing governors.
- New construction forms require a separate parsimony and use argument; ordinary time, order, evidence, representation, and domain labels do not justify one.
Provides
- the three construction narratives
Γ_m.sum,Γ_m.set, andΓ_m.slice; - a plain-language discipline for connecting those narratives to direct facts without making the trace their cause;
- explicit unassembled-collection, selected-structure, existing-whole continuity or new-whole identification, and transformation-composition stop conditions.
C.13:End
Measurement & Metrics Characterization (MM‑CHR)
Status: Stable Type: Pattern
Use this pattern when. Use C.16 to make a reading interpretable or to construct the model needed for a proposed measurement. Start with what is being measured and how the procedure relates it to an indication.
What goes wrong if missed. Raw output, indication, actual subject state, measurement result, diagnosis, and criterion verdict collapse into one number; model and calibration assumptions disappear; uncertainty is laundered away; and a dashboard or evidence link is mistaken for work, result, assurance, or decision authority.
What this buys. A measurement model that connects indications to what is being measured, or a stated ambiguity that changes what to do next. For a performed measurement, the resulting account identifies the attributed values, uncertainty, conditions and work needed to interpret them.
Intent (Normative)
Name. Measurement & Metrics Characterization (MM‑CHR).
Use this when. A reading needs interpretation, or a proposed measurement needs a model. Ask what quantity or Characteristic is sought, how the procedure produces an indication, and what can be inferred from it under the measurement conditions.
What changes in practice. The practitioner constructs or recovers the relation that makes a reading informative about the subject. This can reveal an influence to include, an ambiguity to preserve, or an arrangement to change. When reporting a performed measurement, connect the attributed values and their uncertainty to the method, model, calibration and work that obtained them. A later diagnosis or decision uses that interpreted result.
Not this pattern when. Use A.17 for the Characteristic, A.18 for scale-operation legality, C.16.P while measurement wording is still ambiguous, A.19.CPM for comparison, A.19.SelectorMechanism for selection, C.28 for causal use, A.10/G.6 for provenance, B.3 for assurance, G.4 for an acceptance declaration, G.11 for currentness, and C.11 for a decision result. C.16 supplies none of those results by implication.
Local designators. MeasurementSpecification, MeasurementMethod, MeasurementModel, MeasurementWork, MeasurementResult, and MeasurementResultEpisteme name objects in one case; they are not new public U-kinds or universal relation types. MeasurementMethod is one U.Method; MeasurementWork is one dated U.Work; MeasurementResultEpisteme is one C.2.1 episteme.
Objects used together. A measurement specification fixes the Characteristic, Scale and applicable unit, and states which method, model and interpretation conditions govern the result. A reading attributes values to a subject under those conditions. Section :5.5.1 supplies ReadingContent when a reusable interface needs to classify that content; an episteme stating it has its separate C.2.1 identity. A unit supplies the adopted quantity reference and conversion meaning. An evidence stub is an optional compact locator into A.10/G.6 provenance. These objects can be expressed together while retaining the distinct specification, value, claim, episteme and performed measurement.
Scope and result boundary (Normative)
C.16 covers construction and use of measurement models. For a performed measurement, it keeps the following parts of the result account recoverable:
- the entity or ordered tuple being measured and the quantity or other Characteristic attributed to it under the stated conditions;
- one Characteristic and one Scale, with Level or Coordinate and Unit when applicable;
- the reusable measurement specification,
U.Method, measurement model, calibration requirements, and uncertainty treatment; - dated measurement work with performer, actual bindings, resources, and time stance;
- the value or set of values attributed to the measurand together with relevant information, including uncertainty and interpretation basis; and
- direct comparability within the declared basis.
C.16 does not turn an instrument message, file, dashboard tile, ledger row, or evidence citation into a measurement result. It does not own the actual subject state, diagnosis, criterion verdict, acceptance action, assurance claim, causal conclusion, or decision.
Problem Frame
A measurement is often compressed to subject → value. That abbreviation hides the measurand, the quantity or characteristic intended to be measured, the model relating inputs to an output quantity, the calibration basis, the work occurrence, and the uncertainty carried into later use. It also makes raw instrument output, a displayed indication, an attributed measurement result, a diagnostic interpretation, and a criterion verdict look like one object.
The failure becomes visible when two readings are compared, when a detector output is treated as the state of the subject, or when a dashboard value is reused as evidence, assurance, acceptance, or decision authority. C.16 restores the measurement-specific objects before any receiving use is judged.
Forces
- Interpretability vs convenience. A compact value is easy to carry; a usable result needs its measurand, Characteristic, Scale, model, calibration, uncertainty, and time stance.
- Model dependence. Measurement may use corrections, calibration coefficients, influence quantities, and inference. Omitting them from an account leaves those dependencies unchanged.
- Cross-domain reuse vs scale coercion. Physics, software quality, architecture, survey, and judging cases need common discipline without one common scale.
- Repeatability vs occurrence identity. A reusable method and operation declaration do not establish that measurement work occurred or that actual participants were bound.
- Result vs later interpretation. A value attributed to a measurand is not by itself a diagnosis, conformance verdict, causal conclusion, assurance claim, or decision.
Solution - Construct and interpret a measurement (Normative)
To develop a measurement model, begin with §§5.1-5.4. A proposed relation can supply a conditional calculation or expose an ambiguity before any measurement is performed. An existing model that answers the question can be used directly.
When interpreting a performed measurement, recover one ordinary sentence:
Dated measurement work
Wapplied methodMto measurandx, using modelf, calibration basisK, and actual input bindingsX, and obtained output quantity valueywith stated uncertaintyu; epistemeEstates that measurement result under its declared Characteristic, Scale, unit, time stance, and interpretation basis.
If a fact needed for that interpretation is unavailable, state which conclusion remains undetermined and what information could resolve it.
Name the measurand and measurement subject
M‑SUB‑1. Identify the measurement subject and what you intend to determine about it. For a quantity, the measurand specifies which quantity of that subject under which conditions. If choosing the measurand is your task, begin with the recipient's decision. C.39.SA:4 helps when several composition scales or levels of model abstraction could serve that decision. Which subject bears the consequential property, at what places and times? Hold a proposed reading fixed and vary another condition allowed by the task. The pressure at one port may settle whether the local valve reaches its opening threshold. The same port reading can accompany different peak pressures elsewhere in the vessel. A vessel-wide limit needs a relation to those pressures, observations covering them, or a sufficient bound. A justified bound can settle the limit while the exact maximum remains open. The measurement scale and unit can stay the same across these different questions. For a non-quantity Characteristic, identify the subject, the Characteristic and the Scale on which a value is attributed.
M‑SUB‑2. Preserve arity. An entity Characteristic has one subject; a relation Characteristic has the exact ordered tuple required by A.17. A relation reading is not silently rewritten as a unary property of one participant.
M‑SUB‑3. Distinguish the measurand from the actual subject state. A measurement result attributes values under a method and model; it does not make the physical, social, architectural, or epistemic state identical to the result episteme.
Fix Characteristic, Scale, unit and time stance
M‑CSLC‑1. One measurement specification binds exactly one Characteristic to exactly one Scale. A reading names its Coordinate or value on that Scale, and a Level when the Scale supplies that named category or tier.
Specification identity and resolution. Identify the specification by its operative content: the selected Characteristic and Scale, applicable unit convention, method/model requirements, applicability, calibration and uncertainty rules, and time interpretation. A description expresses that content; its filename or template label does not identify it. A changed requirement produces a changed specification.
An application can instead bind a new specimen, instrument, calibration instance or observation window under an unchanged specification when its rules permit that variation. Retain those actual bindings in the measurement account. A DHCMethodRef, when used as a measurement-definition reference, resolves to this exact specification and any requested description edition, rather than to the enacted Method or dated Work.
Start writing before every choice is settled: identify the wanted distinction, possible Scales, alternative models and the observations that would discriminate them. Revise that account as the model is tested in :5.3.1. A draft with an unresolved choice remains useful inquiry material; a consumer that depends on that choice needs it before using the draft as a completed specification.
M‑CSLC‑2. When units apply, name the quantity kind and presentation Unit. Conversions are admissible only when they preserve the quantity kind and the Scale supports the operation. Nominal and ordinal labels do not acquire interval or ratio arithmetic by being encoded as numbers.
M‑CSLC‑3. Use the Scale's order to interpret the Characteristic: a higher temperature value means hotter. When a later evaluation asks which value is preferable, state its preference under A.17/A.18. A measurement or magnitude comparison needs no preferred direction.
M‑CSLC‑4. State the time stance: instantaneous or as-observed at T, aggregated over window W, or another specified temporal basis. A later value does not silently replace an earlier result.
Separate method, description, model, calibration, and work
M‑METH‑1. MeasurementMethod is one U.Method. Its U.MethodDescription may state generic participants, parameters, effects, and measurement conditions; it contains no actual-participant slots and does not claim that measurement occurred.
M‑MODEL‑1. MeasurementModel relates input values and relevant influences to the values attributed to the measurand. In quantity measurement, these are input, influence and output quantities. Identify the model version, assumptions, corrections and domain of validity. Recover what its formula, software function or other expression represents. Use C.16.MR to construct that relation from the indication-producing procedure when it is missing or unsuitable; :5.3.1 connects the construction to interpretation and its next use.
M‑CAL‑1. Name the calibration basis required for the use: reference standard or comparison basis, dated calibration work and result when required, calibration coefficients or corrections, applicable interval, and uncertainty contribution. A calibration certificate or ledger row cites these facts; it does not establish them by being stored.
M‑WORK‑1. MeasurementWork is one dated U.Work. First recover every actual performer's A.13 core for the measurement action, including the same obtaining assignment; then independently admit the Work under A.15.1 from its performance history, at least one obtaining enactsMethod relation, temporal extent, and at least one obtaining locally declared containing-system relation. Add F.6 afterward only when the measurement claim also needs precise assignment-bound attribution. Name the measurand through its direct subject relation or an A.6.1 operation-application binding. Name another enacted Method, resource, or concrete participant only when the measurement claim uses its independently obtaining relation or binding. A plan, compatible signature, method description, instrument type, or retained reference establishes none of those actual facts.
Construct the measurement relation
- Start with what is being measured and why. Specify the subject, Characteristic, conditions and required range of interpretation under §§5.1-5.2. Separate what is already known from values the proposed measurement must resolve.
- Follow how the indication is produced. Describe the procedure connecting the subject to the indication. Recover the measurement principle, applicable calibration relation, or combination of both that connects the quantities. Include intermediate conversions when they change the answer. Physical laws, assessment models and instrument-specific relations come from the relevant subject knowledge; B.5:4.2 helps recover their construction.
- Include influential conditions. Consider how the apparatus interacts with the subject, what it samples or averages, and its resolution and operating range. Include an influence when its omission could change the interpretation needed for this use. Explain a correction through the relation that gives its direction and magnitude. Retain an unknown influential quantity as unknown, using available bounds or distributions when justified.
- Determine what the relation resolves. With actual or proposed indications, derive the compatible sought values and their uncertainty under §5.4. If different sought values can produce the same indication, identify that ambiguity. Work a small case or limiting case to expose an omitted influence, inconsistent units or a failed inversion. C.16.IR constructs the joint cases and distinguishes feasible alternatives from bounds that may include unattainable values.
- Choose the useful return. Supply the interpreted value, interval or conditional result when it answers the question. Otherwise identify which change could resolve the remaining ambiguity: refine the relation, change the measurement arrangement, obtain an applicable calibration or narrow the conclusion. Choose further observation by the distinction it can resolve and the work it demands, using C.11.DUA when that choice needs deliberation. C.16.RM compares changes to models, arrangements and calculation, then carries the selected repair through the interpretation while retaining the wanted quantity.
When an observed discrepancy matters, compare its plausible sources in the subject account, measurement relation and actual arrangement. Change the contribution that can alter the answer; sometimes removing an unwanted influence from the arrangement is more useful than modeling it in greater detail. A model-development result states the relation and what it would establish. A claim about a performed measurement also identifies the work and obtained result under §§5.3-5.5.
Instrument resolution describes how small a change in the measured quantity can produce a perceptible change in indication. Extra display digits alone establish neither that capability nor a smaller uncertainty in the result. When comparing two measurement models, hold the question, subject and conditions fixed or establish how the observations bear on the same quantity. Apply B.5.MR:4.2 if the claim is that the second model strictly refines the first. A change of instrument may supply new evidence for that comparison.
Recover input quantities, output quantity, and uncertainty
M‑IO‑1. Name each actual input quantity used by the model, including indications, repeated observations, environmental or other influence quantities, reference values, calibration coefficients, and applied corrections. Name the output quantity whose value is attributed to the measurand. These are measurement-model roles, not a universal work input-output ontology.
M‑UNC‑1. State the uncertainty associated with the attributed value or values whenever it affects interpretation or use. Identify the contributing input uncertainties, correlations or covariance when relevant, propagation method, coverage or interval interpretation, and significant model inadequacy. An uncertainty number without its interpretation is not complete.
M‑UNC‑2. Propagation follows the declared measurement model. Linearized propagation, sampling, interval, set-valued, or another method is admissible only under its own assumptions. Combining provenance pointers is not uncertainty propagation, and more cited grounds do not monotonically guarantee lower uncertainty.
State one measurement result and one result episteme
M‑RES‑1. MeasurementResult is the value or set of values attributed to the measurand together with relevant information needed to interpret them. At minimum, recover the measurand, Characteristic, Scale, attributed value or values, Unit when relevant, uncertainty, method, model, calibration basis, time stance, and measurement work.
M‑RES‑2. MeasurementResultEpisteme is one C.2.1 episteme. Its ClaimGraph states the C.16 result, subject, interpretation basis, polarity or domain status when applicable, and uncertainty. The episteme is not the measurand, actual subject state, raw output, indication, diagnosis, or criterion verdict.
M‑RES‑3. When exact work and governed actual changes first establish the episteme's identity and that inception matters, A.15.PROD supplies the local entity-identity inception claim. C.16 does not introduce a work-to-result relation.
Reading content and the range required by a receiver
A Coordinate supplies a value on a Scale. To use it as a reading, recover what receives that value, which Characteristic is attributed, and under which interpretation and conditions. Write the attribution with its basis: observation, measurement model, prediction, content evaluation, normalization, scoring or another stated derivation. These bases support different conclusions. A conditional calculation may already answer the question; a claim that measurement occurred additionally needs :5.3–:5.5. A receiving comparison or transformation also supplies its own admission and operation rules.
An ordinary sentence or existing table can express this content. Obtain the missing subject, Scale or interpretation by comparing the available records and their premises. Keep unresolved alternatives visible while doing so. A new form or finished result episteme is unnecessary merely to develop the account. When a claim-bearing episteme is needed, identify it through C.2.1 and distinguish its content from its EntityOfConcern, effective scheme, expression and external evidence.
The recurring typed range. ReadingContent is a C.3 kind individual over the NonemptyClaimContent entities identified in C.2.1:4.1.1–:4.1.2. Its KindSignature uses that candidate domain and the following membership criterion under the FPF interpretation supplied by A.17, A.18 and this section: the content makes one identifiable principal attribution of a value, admissible value set or distribution on one declared Scale to one Characteristic of an identified subject, ordered subject tuple or specified modeled case. It includes the qualifications that determine that attribution's meaning, including the time/population interpretation, conditional premises and uncertainty interpretation when used. A supporting equation or referenced input can concern other Characteristics; a bundle of independent principal attributions is a profile of readings.
This content-fixed criterion needs no extension enumeration. For a judgment, pin the candidate, this declaration edition and a C.3.2 slice that resolves every selector in its declared schema, including any declared version or temporal selector. It applies where the candidate retains the interpretation of C.2.1, A.17 and A.18 used here; an unestablished interpretation alignment cannot support that classification.
The criterion is not a measurement-validity test. A false or insufficiently supported attribution can be a member. A bare Coordinate, a graph encoding and a document carrier are outside the candidate content-entity domain and give C.3.2 not-applicable, not false. An interpreted claim that a measurement is required, without any value attribution, is a non-member. Missing knowledge needed to recover the attribution leaves membership unknown.
A missing reading is an input status; it supplies no zero, category or fictitious Coordinate. An explicitly stated set of possible Scale values can be a reading when its interpretation is given, even if it leaves the receiving question unresolved.
Identity and replacement. Apply C.2.1's content-identity rule to the whole selected attribution and its asserted qualifications. A change of subject, Scale meaning, attributed values, conditional premises or an asserted basis changes that content. A new encoding or wording can preserve it through an established interpretation correspondence. New external evidence for the unchanged claim need not change its content; adding or changing an assertion about that evidence does.
The ReadingContent kind itself continues while its candidate domain and membership distinction remain the same; changing one reading does not change the kind.
For a proposed replacement, first resolve its content and check membership. Then check the receiver's actual subject, basis position, Scale, scope/window, input domain and evidence conditions. Two members need not be equal or interchangeable. A unit conversion can provide a usable replacement on another Scale through its declared map; it does not make the two Scale-specific Coordinates identical. Preserve the original reading and the conversion or derivation used. An unresolved required condition blocks that dependent use, while a satisfied input test still does not perform the transformation or establish its result.
Profiles and calculation. A profile keeps each component's reading content and the position or contributor by which the receiver uses it. Equal numeric values do not merge different subjects, basis positions or contributions. When the operation depends on multiplicity or order, retain it in that existing profile/grouping basis; do not reduce the operands to a bare set of numbers or content values.
A copied assertion about one contributor supplies no additional physical contribution. Distinct observation occurrences or repeated uses count separately only when the declared aggregate meaning requires them. Repeating a calculation can return equal content while remaining another application under A.6.1; neither a new application nor a new result-episteme identifier creates another contributing subject.
A reference to a reading resolves to the selected content, with its interpretation and any required source edition. A reference to its result episteme instead resolves that episteme; extracting its relevant content is a separate, explicit use.
USCM reads the admitted Coordinates and conditions from ReadingContent and returns content attributing its declared Score. ULSAM does the same for a lawful aggregate, retaining the grouping basis; CPM consumes the interpreted profiles to return its own comparison result. UNM transforms a Scale-specific Coordinate; retaining its result as a reading also retains the subject, target Scale and transformation basis. UINDM chooses positions and supplies no missing reading. These are direct uses of shared rules.
Keep comparability and scoring bounded
M‑CMP‑1. Direct comparability is conservative: two readings cite the same resolved measurement specification, Characteristic, Scale and unit semantics, compatible model and calibration regime, and a compatible time or population basis. Similar labels or units are insufficient.
M‑CMP‑2. Cross-template conversion, normalization, scoring, aggregation, comparison, selection, or cross-context transport names its method, declaration, and loss or uncertainty consequence under the pattern for that operation. Name an F.9 Bridge when cross-context semantic correspondence is required.
M‑SCORE‑1. A Score is another declared Scale reading. Its scoring method and actual application remain under their direct Method, Work, and operation-binding patterns. A score does not overwrite its source measurement results.
Route provenance and later use outward
An evidence stub may carry a type-of-ground and identifier that lead to the exact A.10/G.6 provenance path. The path can cite the method description, model, calibration, work, inputs, output, result episteme, source publications, and transformations. Neither the stub nor a graph edge establishes those objects or their obtaining relations.
A later comparison, diagnosis, criterion evaluation, acceptance action, or decision is a separately governed use of the measurement result. Recover its actual operation and any result under the direct pattern. Ground dated U.Work when the account asserts it or the receiving use requires it. The use relies on the result episteme through an exact premise, reference, operation-argument, decision-use, or other direct relation. Currentness belongs to G.11; bounded reliance to A.10 or B.3 under their entry conditions.
Lexical and neighboring-pattern discipline
Use measurand, measurement subject, Characteristic, Scale, Level, Coordinate, value, Unit, measurement method, measurement model, calibration, uncertainty, measurement work, and measurement-result episteme with their defined meanings. Plain-register metric, reading, score, and output are acceptable after first-use mapping. Do not use measurement result, evidence, validation, or verification as umbrella terms for several governed objects.
Key relations. C.16 uses A.17 and A.18 for Characteristic and Scale legality; A.6.1 for declaration-local positions and operation bindings; A.13 for each actual performer; A.15.1 for independent admission of the dated Work; and F.6 afterward only when precise assignment-bound attribution is needed. If claim-bearing source wording still says only “role,” use E.10.ROLE first, then use A.2 or A.2.1 only when an exact local system-role kind, classification, or assignment has been recovered. C.2.1 covers the result episteme; A.10/G.6 provenance; G.11 currentness; B.3 assurance; and the pattern for the next diagnosis, acceptance, causality, comparison, selection, or decision question.
Scale-type admissibility quick reference (Informative)
A.18 governs operation admissibility by Scale type; CG-frame admissibility profiles govern their declared mechanized uses. This table summarizes those distinctions. A Scale-compatible calculation also needs the statistical or subject-model assumptions that justify its use.
Reminders (informative; see A.18 for normative rules).
- G‑1 (Order). Ordinal transforms preserve the declared order and distinctions.
- G‑2 (Differences). On interval or ratio, Δ is meaningful; on ordinal or nominal, it is undefined.
- G‑3 (Ratios). Ratios of raw values require a ratio Scale. Ratios of derived differences require their own quantity interpretation and model conditions.
- G‑4 (Unit coherence). Interval or ratio arithmetic presumes compatible units (or a declared conversion).
- G‑5 (Target polarity). If polarity is targeted, comparisons use distance‑from‑target semantics as declared by the relevant subject pattern, template, and cited method or mechanism.
Provenance and use semantics (Normative)
What an EvidenceStub is and is not
An evidence stub is an optional compact locator from the reading claim to an exact provenance path. It may identify a source publication, calibration record, instrument output, model edition, work occurrence, transformation, or other ground, but A.10/G.6 govern the path and its citations.
- Several stubs form a list of locators, not a measurement algebra. Their union is not uncertainty propagation and does not guarantee stronger warrant.
- A provenance edge may be asserted only after its direct source relation, work fact, participation, production, representation, or citation relation is independently established.
- A later user states the exact relied-on claim and local
RelianceDisposition; use B.3 when a named assurance claim is at issue. Mere availability, citation, or graph membership does not establish actual use.
Measurement-result boundaries (Normative)
Keep the following objects distinct even when one carrier displays several of them:
The carrier, dashboard, ledger, criterion clause, and evidence path may represent or cite several rows. None collapses their identities or establishes another row by presence alone.
Archetypal Grounding
Calibrated detector receiver. The detector emits raw counts. Its processing yields an indication of 41.8 kPa. The measurand is gas pressure at port P over the stated sampling window; Characteristic is Pressure; Scale is a ratio quantity scale; Unit is kPa. Measurement model PressureModel-4 uses counts, reference offset, temperature, and calibration coefficients as inputs and pressure as output. Dated measurement work names its performer, detector, port, resources, bindings, calibration basis, and uncertainty propagation. The C.16 result attributes 41.8 kPa ± 0.6 kPa to the measurand under that basis; one C.2.1 episteme states it. The raw counts, displayed indication, actual pressure, result episteme, a later leak diagnosis, and a pressure-limit verdict remain different objects.
Internal-combustion-engine test bench. One dated test-bench work occurrence binds the engine, dynamometer, fuel batch, ambient conditions, method, model, and calibration records. Torque, exhaust temperature, and emissions are three Characteristics with separate Scales and result epistemes; their input quantities, output quantities, covariance where relevant, and uncertainties remain separately recoverable. Aggregation work may later construct a declared performance summary, and evaluation work may apply an emissions criterion. Neither the summary nor the pass/fail verdict is the torque or emissions measurement result.
Architecture coupling. The measurement subject is the exact ordered module pair under a declared dependency census window, not either module alone. The Characteristic is Coupling on an ordinal Scale. The method description defines generic dependency classes; dated work binds the actual codebase edition and pair. The result episteme states the Level and basis. A later release decision may rely on it, but the dashboard tile and decision record do not establish the census work.
A voltmeter changes the voltage it reads
The sought quantity is the open-circuit voltage E of a source. Model the source as an ideal voltage E in series with resistance R_s; the connected voltmeter has input resistance R_m. The meter closes the circuit. Ohm's law gives current I=E/(R_s+R_m), and the indication is V=I R_m. The measurement relation is therefore E=V(1+R_s/R_m).
For E=10 V and R_s=R_m=1 megohm, the indication is 5 V. The known resistance ratio recovers the open-circuit value as 10 V. The difference comes from the measurement interaction.
If both E and R_s are unknown, one indication leaves several pairs compatible with it. For V=5 V, R_m=1 megohm and an available bound 0.8≤R_s≤1.2 megohm, the conditional voltage interval is 9≤E≤11 V. That interval may answer the question. When a narrower answer is needed, a second indication with a different known input resistance supplies another equation, provided the source stays unchanged and the circuit model still applies. Repeating the original arrangement supplies the same relation and leaves this ambiguity.
These calculations use an ideal circuit. For an obtained measurement result, include uncertainty in the indications and resistances and any model inadequacy that affects the use.
Interpreting an assessment of independent performance
The sought quantity p is the fraction of a population able to perform a specified action independently under stated conditions. An applicable assessment calibration supplies a, the probability of a positive test when the capability is present, and b, the probability when it is absent. Partitioning the population by that capability gives the expected positive fraction q=ap+b(1-p).
With a=0.9 and b=0.1, the relation is q=0.1+0.8p. An observed positive fraction 0.7 gives the estimate p=0.75 under this model. Sampling uncertainty, uncertainty in the calibrated rates and their applicability determine how precisely that estimate can be used. When a=b, the expected positive fraction is independent of p, so this test supplies no such distinction.
Now allow hints during the assessment. The earlier a and b may no longer describe the procedure. With the changed rates unknown, the positive fraction alone no longer determines p. If the independent-performance claim is still needed, return to that performance condition or obtain a calibration applicable to the changed procedure. The model explains which inference is available; it uses the subject's account of the capability and its assessment.
Carry a reading into a calculation and revise its basis
A retained account attributes temperature 20 °C ± 1 °C to Pump P over window W under a named specification and model; this example admits every additive error from −1 to +1 °C and has no further constraint on temperature. Its content is a ReadingContent member. Relying on it as a measurement result still needs the performed-measurement and uncertainty grounds above. The Coordinate 20 on the Celsius Scale, the interval, the reading content and an episteme stating that content answer different questions.
A declared scoring rule reads that Celsius position and assigns g(c)=c/40 on [0,40], with higher temperature preferred in this illustrative use. With the required admission grounds, it returns score content for P at W: 0.5, bounded by [0.475,0.525] under this interval-propagation rule. That content can pass to a comparator accepting this score Scale and bound. The scoring calculation supplies no second temperature measurement. If an upstream UNM instead supplies 293.15 K, the number cannot fill the Celsius argument. Apply the declared inverse conversion, or select a kelvin-input scoring rule whose correspondence is established.
Now comparison of the written model and calibration record reveals that the measurement needs an exact +2 °C correction not included before. With the same admitted reading errors, the corrected account attributes [21,23] °C, and this same scoring rule gives [0.525,0.575]. Recover whether the specification already permitted this calibrated correction or must itself change. Preserve the earlier assertion and its basis, then reconsider the scores and comparisons that relied on it. A new window W2 instead needs a new attribution for W2; it can leave the Characteristic, Scale, specification and scoring rule unchanged. These returns follow from the general identity and replacement rules, not from the displayed numbers alone.
For the question “which temperatures are compatible with this account at W?”, 19 °C was admitted only by the earlier bound and 23 °C only by the corrected one. The correction revises the model-based attribution. A further observation at W, justified under the corrected model, could instead exclude some of its previously compatible temperatures while preserving its other assertions.
Bias-Annotation
Conformance Checklist (Normative)
For a proposed model, apply the subject, scale, model and applicable calibration checks. Apply work and result checks when asserting a performed measurement, and later-use checks when making that later claim.
- Subject. Identify the entity or ordered tuple, the quantity or other Characteristic and its conditions. When the measurand had to be chosen, explain which difference in the receiving decision justified that choice. For an inference from a part to the whole, give the relation or bound that carries it.
- CSLC: Characteristic, Scale, Level or Coordinate, applicable Unit and time stance are interpretable. Add a preference rule only for a use that judges which values are preferable.
- Method/model: the method, model version, inputs, output quantity, assumptions and validity domain are recoverable. When the relation had to be constructed, §5.3.1 explains how the procedure produces the indication and what sought values it can distinguish. Keep a proposed model separate from a claim of performed measurement.
- Calibration: applicable calibration work/result, reference basis, coefficients or corrections, validity interval, and uncertainty contribution are cited when required.
- Work: every actual performer has the A.13 core; the dated
U.Workis independently admitted under A.15.1; F.6 is added afterward only when precise assignment-bound attribution is needed. The measurand relation or A.6.1 binding is present; further enacted Methods, resources, or participant bindings are present only when the measurement claim uses them. - Result: one C.16 measurement result attributes value or values to the measurand with uncertainty and relevant information; one C.2.1 episteme states it.
- Separation: raw output, indication, actual subject state, result, result episteme, diagnosis, verdict, and decision are not collapsed.
- Comparability: direct or transformed comparison names its basis and does not upgrade the Scale or mint a common scale.
- Provenance/use: A.10/G.6 provenance, G.11 currentness, bounded reliance, assurance, and later work remain under their subject patterns.
- Boundary: no method description, plan, signature, carrier, ledger row, evidence edge, or stored reference is used to infer actual participation, work, or result identity.
Common Anti-Patterns and How to Avoid Them
- Template as occurrence. A reusable measurement specification, model, signature, or calibration procedure is treated as proof that work occurred. Ground dated work and actual bindings.
- Generic result field. A record has
result=...without saying whether it is output, indication, measurement result, diagnosis, verdict, or decision. Name the direct result kind and governor. - Evidence algebra. Evidence locators are unioned as though idempotence or count determined uncertainty or warrant. Use measurement-model uncertainty propagation and A.10/B.3 reliance separately.
- Scale drift. A template id is reused as though changed Scale, model, unit, or calibration requirements preserved the same specification. Publish a successor and state the relation; do not mutate historical readings.
- Arithmetic on ordinal. Encoded levels are averaged or ratio-compared. Stay with order-preserving operations or introduce a separately governed scoring method and Scale.
- Multi-Characteristic stuffing. One reading carries a vector while pretending to be one measurement. Create separate results and declare any later aggregation.
- Result-to-verdict shortcut. A value inside a tolerance is called accepted without performed criterion evaluation. Ground the separate evaluation work, exact clause application, verdict episteme, and later decision.
Consequences
Benefits. Measurement results become interpretable and reusable while remaining distinct from subject states and later judgments. A practitioner can inspect the measurand, Scale, method, model, calibration, work, uncertainty, episteme, and provenance, then enter the smallest pattern for the next question—comparison, diagnosis, acceptance, assurance, causality, or decision.
Trade-offs. The chain is longer than a dashboard field. Model assumptions, calibration status, and uncertainty can make a formerly crisp number conditional or set-valued. That cost is the information needed to avoid false precision and hidden result substitution.
Failure containment. Missing model validity, stale calibration, ungrounded work, absent actual bindings, or unreported uncertainty narrows or blocks the measurement claim.
Rationale
Measurement is not merely reading a carrier. It is performed work under a method and model that attributes one or more values to a measurand and supplies the information required to interpret those values. That architecture explains why indication, actual subject state, measurement result, result episteme, diagnosis, and verdict must remain distinct.
SoTA-Echoing
These sources contribute different parts of measurement interpretation. Their quantity, model and observation concepts retain their source meanings; correspondence to an FPF concept requires F.9 only when the use depends on that semantic mapping.
Value and result distinctions. VIM3 entry 1.1, entry 1.9 and entry 1.27 distinguish a quantity, its adopted unit reference and a quantity-value scale; entry 2.9 includes attributed values and the information needed to interpret a measurement result. The measurand (entry 2.3), indication (entry 4.1) and resolution (entry 4.14) locate three different questions in the measurement. FPF also uses nominal Scales and non-measurement evaluations; VIM terminology does not classify all such readings or admit an FPF kind. ReadingContent is the bounded FPF interface construction in :5.5.1.
Constructing and revising the model. GUM-6:2020, §§7, 9-10 and 12, supplies the distinction between the measurement principle, effects of implementation and adequacy for use. Section :5.3.1 turns that distinction into a construction and an ambiguity test. The circuit and assessment cases work this instruction using their stated subject models.
Dounas-Frazer and Lewandowski (2018), §2, distinguishes models of the phenomenon and measurement equipment and allows revision of either model or either physical arrangement. C.16 adopts those different returns. Its stopping question is the intended use of the measurement: a sufficient interval can end the work, while a consequential discrepancy can require further investigation. B.5 and C.11.DUA supply the wider inquiry and choice.
Relations - Placement (Informative)
Measurement-relation construction. C.16.MR derives a relation from the sought property, measuring arrangement and consequential influences, and obtains a first conditional result. This construction can support a proposed measurement. For a result from an actual measurement, use the conditions above for the performed work and its interpretation.
Indication interpretation. C.16.IR begins with an available measurement relation and determines which values or comparisons an indication supports. Use it when influential unknowns, lost distinctions or uncertainty can change the requested answer. It can return a sufficient bound while other quantities remain unknown.
Measurement repair. C.16.RM begins when an existing measurement disagrees with an expectation or cannot resolve the distinction needed for use. It chooses which model, arrangement or calculation to change, recovers a result from the available observations when possible, and tests the changed contribution at the scope of the receiving question.
Architecture measurement boundary. C.32.P2S, C.32.PAD, and C.32.ADA may cite C.16 readings only after the characteristic, bearer, scale, coordinate, value, unit when relevant, and admissible use are declared. C.16 readings do not become architecture characteristics, decision criteria, eval programs, evidence, gates, or decision authority by themselves.
Structural-information measurement boundary. C.33, C.34, and C.35 may name captured structure, lost structure, similarity, preservation, entropy, epiplexity estimate, compression, generated-carrier adequacy, or search-output context. When a claim about any of those uses a value, score, coordinate, threshold, dashboard reading, or eval result, state the measurement construction and admissible-use assertions under the C.16 and evaluation/criteria predicates, with their subject patterns used as locators.
Precision-restoration relation. C.16.P is the first-stage wording-use restoration pattern for characteristic, scale, coordinate, score, metric, axis, dimension, and related characterization wording when the measurement object is not yet recoverable. Once the wording is resolved, use C.16 for the measurement-chain question or the pattern for the recovered non-measurement claim.
C.27 temporal-claim relation.
- C.27 may flag: a rate/rate-change reading whose admissible use depends on admissible measurement construction, evidence, sampling window, or finite-difference method.
- This pattern keeps: measurand and measurement-subject identity, method, model, calibration, input/output quantities, uncertainty, dated work, measurement result, result episteme, comparability basis, units, sampling window, and provenance routing.
- Non-admissible use: a rate-change label is not a measurement template, and temporal words such as velocity, acceleration, throughput, cadence, or recovery speed are not admissible measures by themselves.
- Neighboring-pattern use: when load-bearing, the claim cites
baseCharacteristicRef, the relevant measure reference, sampling window, construction method such asDHCMethodRef, andC16RouteRef; C.27 keeps only the temporal-claim adequacy question.
C.28 causal-use relation. C.16 governs measurement construction, result interpretation, uncertainty, and direct comparability. C.28 governs the causal-use relation when the same result episteme is used to claim effect, intervention success, causal fairness, policy optimality, counterfactual comparison, off-policy causal evaluation, causal-RL evaluation, or causal method superiority. A C.16-admissible measurement result is therefore not by itself admissible for causal use under C.28.
Evidence, currentness, and assurance. Use A.10 and G.6 for source recovery and provenance for the method, model, calibration, Work, inputs, result episteme, and later use. Use G.11 for currentness and B.3 when a named assurance claim is at issue. Evidence, provenance, currentness, and assurance do not by themselves establish the C.16 measurement result.
Kernel. MM‑CHR imports the canonical Characteristic vocabulary and the CSLC discipline fixed by A.17 and A.18; it does not redefine them. CharacteristicSpace reasoning (for change) lives in the patterns that consume MM‑CHR readings.
Using patterns. KD‑CAL, Arch‑CAL, G.4, and other consumers cite C.16 measurement-result epistemes and then apply the direct patterns for their comparison, evaluation, acceptance, aggregation, or decision use, grounding dated U.Work when the account asserts it or the receiving use requires it. They do not produce a measurement merely by naming a template, score field, criterion, or evidence profile.
External terminology. A standard can be cited and used in its own sense. When a C.16 use depends on correspondence between different local senses, use F.17 to identify the terms and F.9 to establish the needed Bridge. That correspondence does not itself establish a measurement result.
Measurement and probe note for quantum-like readings
For a sensor indication, survey response, dashboard value, score or model coordinate, establish what the receiving question asks about it. A question about a proposed probe can be answered by a model of what the probe changes. A claim about an observed effect of probe order needs grounds for the performed comparison. For example, if the first survey item reveals information needed to answer the second, their wording and order show how the two administrations differ. Whether that information changed respondents' answers requires evidence about their responses.
Use the ordinary subject or measurement account as far as this question requires. For an obtained measurement result, recover the measurand, subject, Scale, Unit where applicable, time, method, model, calibration, actual work and uncertainty that support its interpretation (§5). Distinguish the indication, attributed value and any later verdict. Carry the provenance needed for the specific reliance. For a model prediction, state what follows under its premises.
When the ordinary account leaves a named contextual or probe-order obstruction that changes an inference or action, C.26 tests its mathematical and practical significance. For a performed measurement that needs an inspectable note, the following fields can record its basis.
C.29 mathematical-lens use relation
If a mathematical lens depends on a measurement, recover the C.16 measurand, Scale, model, calibration, work, uncertainty, result episteme, and comparability basis first. C.29 may then state the lens-use admissibility claim; it does not construct the measurement, make values comparable, or provide provenance. A.10/G.6 retain provenance and B.3 retains assurance.
C.16:End
Construct a Measurement Relation
Type: Method pattern Status: Stable Normativity: Normative
Problem frame
Use this pattern when you have an indication or a proposed measurement procedure, but the relation between that indication and the property you want to determine is missing or unsuitable. A connected voltmeter can change the voltage. A test response can depend on both capability and the assistance offered. A counter can report the remainder after a wrap rather than the total number of events.
First useful move. Follow one indication back through the procedure that produces it. Name the sought property, the conditions under which it is sought, and one interaction or transformation that can make the indication differ from it. Express the relation supplied by that interaction. For an ideal source with open-circuit voltage E and series resistance Rs, a voltmeter of resistance Rm reads V=E*Rm/(Rs+Rm). This shows why the indication alone need not determine E.
The Method constructs a measurement relation: how sought values, the measuring arrangement and relevant influences produce an indication, or how those contributions jointly constrain the sought value. The relation may use laws, an applicable calibration, an empirical response model or their combination. It can be used to plan a measurement or interpret an obtained indication.
The worked constructions concern quantities and event counts; they require elementary algebra and, in the assessment case, conditional probabilities. Subject Methods supply the physical laws, instrument response, assessment meaning or computational operations. C.16 supplies the Characteristic, Scale and performed-measurement account. This construction can return a conditional relation before any measurement is performed.
When a usable relation already supplies the needed answer under the current procedure and conditions, use it. If the relation is available but several sought values fit the indication, the next work is to determine what that indication can resolve. If an existing measurement disagrees with a prediction, compare possible changes to the models and arrangements.
Problem
An instrument display, test outcome or log value is produced by a procedure. Its meaning depends on that procedure and on the conditions in which the property is sought. Substituting the display for the sought value can therefore answer a different question.
A correction copied from another arrangement can also fail. Its sign, magnitude and uncertainty depend on the relation that produced it. Unknown influences can leave multiple compatible values even when the displayed number is stable.
The working difficulty is to construct the relation at the level of detail needed to interpret the indication for the intended use. This requires selecting consequential influences and checking whether familiar formulas apply to the stated conditions.
Forces
Solution
Fix the property and conditions of interest. Follow the production of the indication, combine the relations that connect it to the sought value, and retain the influences that can change its interpretation. Obtain a first conditional result or expose the missing relation.
Fix what the measurement is meant to determine
Specify the subject, Characteristic, Scale and conditions needed to distinguish the sought value. Include timing, location, population or averaging interval where they change the question. Use C.16:5.1-:5.2 for that specification.
Distinguish values that can differ across stages of the procedure. The open-circuit voltage before connecting a meter and the terminal voltage during measurement are different sought quantities. Either can be a legitimate target; choosing one determines the required relation.
Name the receiving question. Estimating a value, deciding whether it exceeds a limit and detecting a change can tolerate different remaining uncertainty. This choice helps decide which influences are consequential.
Follow the production of the indication
Trace the procedure from the subject through the interactions, transformations and sampling that produce the indication. For each consequential stage, identify the relation supplied by the physical principle, known operation, applicable calibration or response model.
Start with relations already available for the actual or proposed arrangement. A supplied conversion can be enough. When it is missing, derive it from the subject account or identify the empirical relation that must be learned. B.5.RC helps recover a construction from an unfamiliar description; B.5.FM helps formulate the first model.
Retain an intermediate value when another relation uses it or when it exposes a consequential interaction. In the voltmeter example, current connects the source equation to the meter indication. Eliminate that current only after the relations use the same current and conditions.
For a statistical procedure, describe how response probabilities depend on the sought property and relevant conditions. The expected response and one realized sample are separate inputs to inference. Learning or transporting the response model requires the subject's estimation and applicability work.
Include the influences needed by the question
Ask which features of the arrangement can change the interpretation at the intended range or decision boundary. Relevant candidates include loading the subject, offsets, limited resolution, sampling, timing, saturation and dependence on environmental or operating conditions. Follow the procedure to choose among them.
Describe a consequential effect where it enters the relation. An offset at an instrument's input can propagate differently from an offset added to its displayed output. A known correction follows from that relation; an uncertain effect retains its uncertainty. Several effects can be modeled together when their combined contribution is sufficient for the use.
Retain an unknown influential value as unknown. Available bounds or a justified probability model can support a useful conditional answer. Do not assign zero solely because the influence has not been measured.
Compare omission with inclusion for the conclusion the user needs. An ideal limit can check the relation. To omit an influence under the actual operating conditions, show why including it leaves the required conclusion intact. For a value estimate, a bound on the change over the operating range can show that the change fits the accuracy needed. Retain the conditions and any tolerance used.
Compose the relation and obtain a first result
Combine relations only after their shared quantities, units, stages and conditions agree. Use substitution, a joint equation system, a conditional response calculation or another operation appropriate to the supplied model. C.29.2 helps construct the computation when obtaining the consequence is itself difficult.
A direct formula for the sought value is convenient but not required. Keep an implicit relation when solving it needs additional information or when several sought values remain compatible. A computational procedure can express the relation when its steps and interpretation are available.
Work one small case. Supply the known values, retain unknowns, derive a sought value or compatible set, and substitute back where that tests the construction. Inspect an informative limit: a meter with negligible loading, identical assessment response rates, or an interval too long for a wrapping counter.
Carry uncertainty that can change the conclusion. C.16:5.4 gives its interpretation requirements; the mathematical propagation Method depends on the relation. Algebraic solvability alone does not determine whether a near-threshold decision is resolved.
Return the relation and its next useful use
Return the relation with the meanings of its inputs and sought value, the conditions that make it applicable, and the first result it supports. Preserve this information in the calculation, model or explanation being used; a new record is needed only when the receiving work requires one.
If the relation leaves a consequential ambiguity, name the indistinguishable values or the missing contribution. The next measurement should be selected for how it can resolve that distinction. C.11.DUA helps compare the value and cost of obtaining it.
When a discrepancy calls for repair, keep open the distinct possibilities of changing the subject model, the measurement model, the subject arrangement or the measuring arrangement. Additional mathematical detail is one option. Removing an unwanted interaction or restoring the specified procedure can be simpler.
Reopen the affected part when the target conditions, measuring procedure, calibration range or consequential influence changes. Use B.5.MPC.R when that change also affects connected physical, mathematical and computational work.
Archetypal Grounding
Recover an open-circuit voltage from a loaded indication
The sought value is a source's open-circuit voltage E. The supplied ideal circuit represents the source by E in series with resistance Rs. Connecting a voltmeter of input resistance Rm completes the circuit. Assume positive resistances, a stable source and a meter whose indication equals its terminal voltage.
The common current I satisfies
Eliminating I gives
For Rs=Rm=1 megohm and V=5 volts, the inferred open-circuit voltage is 10 volts. Substitution returns I=5 microamperes and the original 5-volt indication. The two voltage drops explain both the sign and magnitude of the correction.
If the desired quantity were instead the connected terminal voltage, V would already be the value in this ideal model. The measurement question determines which result is needed.
As Rm becomes large relative to Rs, loading becomes small and V approaches E. For a use that needs only a bound on this effect, E-V=V*Rs/Rm supplies it. If Rs/Rm is at most 0.001 and V is 5 volts, the loading correction is at most 0.005 volts. Whether that is negligible depends on the receiving question and the other uncertainties.
If Rs is unknown, the same relation can expose an unresolved contribution. With V=5 volts, Rm=1 megohm and Rs between 0.8 and 1.2 megohms, E lies between 9 and 11 volts under the ideal assumptions. The interval may be enough. A performed measurement additionally accounts for indication and resistance uncertainty and relevant model inadequacy.
Build the relation for an assessment of independent performance
The sought quantity p is the fraction of a stated population able to perform a specified action independently under specified conditions. An applicable assessment account supplies s, the probability of a positive response when that capability is present, and f, the probability when it is absent.
Partition the population by the capability. The probability of a positive response is
For s=0.9 and f=0.1, q=0.1+0.8*p. A sample positive fraction of 0.7 gives the estimate p=0.75 obtained by substituting the sample fraction. Its use also depends on sampling uncertainty, uncertainty in s and f and whether those rates apply to this population and procedure. The formula describes the expected response; the sample proportion estimates it.
If s=f, q is independent of p and the test does not distinguish the two conditions through this response. The failure is visible in the constructed relation. Further estimates of that same marginal positive rate cannot identify p through this relation.
Now change the procedure by offering hints. The earlier rates may cease to apply. Recover rates for the changed procedure or restore independent performance if that remains the target. The subject's capability and assessment Methods establish those conditions; the general construction shows where they enter the inference.
This population estimate does not identify which particular people or agents have the capability. That different question requires an individual assessment account.
Interpret a wrapping event counter
A modeled counter increments once for each event and retains a value from 0 to 15, returning to 0 after 15. Readings immediately before and after an interval are r0 and r1. The sought quantity is the number N of intervening events. Assume no reset, no missed increment and readings taken at the stated interval boundaries.
The counter operation gives
Here d is the least nonnegative remainder of r1-r0 modulo 16, and k is a nonnegative integer. With r0=14 and r1=3, d=5: the compatible counts are 5,21,37 and so on.
If the interval is known to contain fewer than 16 events, N=5. Without that bound, subtracting the displayed numbers or returning the modular difference loses possible complete wraps. A wider counter, a recorded wrap count or shorter observation intervals can supply the missing information for a later measurement.
To infer elapsed time, another relation is needed: the counter must count timing ticks with an applicable tick-duration account. A counter of arbitrary work events alone measures no duration. This identifies a missing relation.
Bias-Annotation
The examples have small, supplied subject models. Real instruments can require dynamic response, correlated influences or numerical inversion. Assessment rates can depend on population and conditions. A familiar model should therefore be reused with its applicable range and interpretation.
The construction can be shared across human, organizational and machine measurement practices. Its concrete laws and response models remain those of the subject. The general steps do not establish that a statistical assessment measures an individual's capability.
Conformance Checklist
- CC-C16.MR-1 - Sought property. Subject, Characteristic and target conditions identify what is to be determined.
- CC-C16.MR-2 - Indication production. The procedure's consequential interactions and transformations have relations with recoverable meanings.
- CC-C16.MR-3 - Shared values. Composed relations agree on units, stages and shared quantities.
- CC-C16.MR-4 - Influences. Consequential effects enter where they act; unknown effects retain their relevant uncertainty.
- CC-C16.MR-5 - First result. A case supplies a sought value, compatible set, conditional result or missing contribution.
- CC-C16.MR-6 - Interpretation. A statistical expectation, sample estimate and obtained measurement retain the qualifications needed by the use.
- CC-C16.MR-7 - Useful return. Further work addresses a distinction that can change the receiving result.
- CC-C16.MR-8 - Changed conditions. The construction identifies which relation must be reconsidered when the procedure, range or target changes.
Common Anti-Patterns and How to Avoid Them
Consequences
The reader obtains a relation usable for interpreting or planning a measurement. Intermediate interactions, response assumptions and unknown influences become parts of a calculation whose result can be inspected and revised. A missing contribution becomes a specific next task.
The cost is recovering the procedure and the subject relations it uses. A complete metrological or statistical investigation can be much larger than the first construction. The Method permits a useful conditional result or interval before that larger work.
Architectural Rationale
The result is a relationship for interpreting an indication with respect to a specified property. A model of the subject can predict its behavior while leaving the measuring procedure undescribed. Connecting that procedure to the target property is an independently useful construction.
Construction, resolving ambiguity and repairing an inadequate arrangement cooperate but can start from different inputs and return different results. A measurement relation can be available before its data are obtained. It can also reveal that a particular indication leaves the target unresolved.
Physical interaction, mathematical expression and computational readout participate together. Their roles explain why revising a measuring instrument can require revising an equation or interpretation, and why a new display format alone may leave the underlying ambiguity unchanged.
SoTA-Echoing
The working question is to connect an indication to the property needed by the task. The selected construction follows the indication-producing procedure and includes its consequential interactions and transformations. An alternative is to reuse a supplied direct calibration or conversion. That alternative saves derivation when it already applies to the current arrangement and range. The present Method is needed when a changed interaction, procedure or target leaves that relation incomplete.
The voltmeter case exhibits the trade-off. Reading V directly is sufficient for the connected terminal voltage in the ideal model. Recovering the open-circuit value adds the source resistance and loading relation; it changes the answer from 5 to 10 volts in the worked case. At a sufficiently small resistance ratio, an available loading bound can justify the simpler use. More detail is valuable through that changed interpretation.
GUM-6:2020, §§7, 9-10 develops measurement models from a principle, empirical information and effects of implementation. Adapt: :4.2-:4.4 combine those contributions, locate an influence where it acts, and retain unknown effects when no useful correction is available. The guide concerns quantity measurement; the assessment and counter constructions here use their separately stated response and operation models.
Dounas-Frazer and Lewandowski (2018), §2, distinguishes the investigated phenomenon, measuring equipment and their models. Adapt: :4.1-:4.2 recover the measurement contribution, while :4.5 preserves the separate model and arrangement repairs. The source studies experimental physics education.
The common procedure is a methodological synthesis. The worked cases are constructed under their stated conditions. Reopen the construction choice when a supplied calibration achieves the same interpretation with less work, or when a newly consequential interaction or response condition defeats the existing relation. Detailed uncertainty propagation, calibration design and model identification remain substantial specialized Methods.
Relations
- C.16 identifies measurement, Characteristic, Scale, indication, model and obtained result.
- B.5.RC recovers a needed construction from its description; B.5.FM develops a first model.
- A.3.3.CC constructs compatible variables and constraints; A.3.3.TR constructs a rule for change where a dynamic measurement requires one.
- A.3.3.PI retains information for a future prediction; a measurement relation supplies its observation contribution.
- C.29.2 constructs the computation needed to obtain a modeled result.
- B.5.MPC.R revises connected physical, mathematical and computational contributions.
- C.11.DUA decides whether reducing a remaining ambiguity is worth the work.
C.16.MR:End
Determine What a Measurement Indication Can Resolve
Type: Method pattern Status: Stable Normativity: Normative
Problem frame
Use this pattern when an available measurement relation leaves open what can be inferred from an indication. A loaded voltmeter reading can fit several source voltages. A saturated sensor can establish that a quantity exceeds one threshold while leaving a higher threshold unresolved. A stored counter remainder can correspond to several event counts.
First useful move. Try to construct two cases that fit the same indication and stated conditions but give different answers to the receiving question. If an ideal instrument reports the square of a signed displacement, a reading of 9 square metres fits both +3 and -3 metres. It determines the magnitude but leaves the direction unresolved. Restricting displacement to nonnegative values would resolve that difference only when the subject conditions support the restriction.
The Method determines what the indication resolves under the available relation, domain and uncertainty assumptions. Its result can be a value, a compatible range, a settled comparison, two remaining alternatives or an incompatibility among the premises. C.16 supplies the subject, Characteristic, Scale, performed measurement and result meanings. C.16.MR constructs a missing measurement relation.
The examples require elementary equations, inequalities and integer remainders. More difficult relations can require subject-specific inverse methods, probability models or validated numerical computation.
Apply an already suitable conversion or analysis function directly when it supplies the answer with adequate uncertainty. Use this construction when ambiguity, omitted influences or the meaning of a computed range can change that answer.
Problem
A measurement relation can predict the indication without allowing the sought value to be uniquely recovered. Several quantities may jointly affect the same reading; a conversion can erase distinctions through clipping, averaging, squaring or wrapping. An algorithm can return one solution while other compatible solutions remain.
The receiving question also matters. An interval may settle an operating threshold even when a point value is unavailable. Conversely, precise arithmetic on one arbitrarily selected solution can hide a decision-changing ambiguity.
The problem is to recover the alternatives allowed by the indication and its conditions, then determine which differences matter to the intended use. This requires keeping the observation, model assumptions, computed approximations and supported conclusion distinguishable throughout the calculation.
Forces
Solution
Construct the cases compatible with the indication and relevant conditions. Obtain the sought values represented by those cases, retaining the distinction between feasible examples and bounds that contain all solutions. Use that result to answer the question.
Fix the sought distinction and available relation
State the quantity or property being sought, its subject and conditions, and the question that depends on it. Distinguish the state before measurement, the state during interaction and the reported indication when they differ. For the voltmeter case in :5.1, the target is open-circuit source voltage, while the reading is the voltage with the instrument connected.
Recover the relation used to interpret the indication, including the domain of each unknown, relevant calibration, influences and error assumptions. Keep a known input as known and an unknown influence as a variable. Derive a missing contribution with C.16.MR before treating it as part of the available relation.
A domain restriction needs its subject meaning. Positive length, a known rotation range and a passive resistance bound can exclude mathematical solutions when those conditions hold. An inconvenient branch is a reason to inspect the model, not a reason to drop that branch.
State whether the receiving question needs a point value, an interval, a threshold decision or a comparison. This selects which remaining differences matter.
Construct the jointly compatible cases
Substitute the obtained indication into the relation. Combine that condition with the admitted domains and available information about influences. For repeated or multiple readings, state which subject values and parameters are shared and which may change between them.
With bounded error, include the unknown error in its stated range. If a relation gives an ideal indication f(x,z), x is the sought value, z contains influential unknowns and r is the reported reading, a common model is:
r = f(x,z) + e, with e in the supplied error range.
Use that equation only when additive error describes the indication convention. An error before clipping, an error after clipping and a missed event can require different relations.
A mathematical expression for the compatible sought values is:
S(r) = {x : there exist admitted z and e for which r = f(x,z) + e}.
An implicit relation can be used in the same way without converting it to a single-valued function. Retain dependencies among unknowns: independently combining their separate ranges can introduce cases excluded by their joint relation.
A probability model supplies a different additional contribution. Values inside a bounded-error range have no assigned likelihood merely from membership in that range. A confidence interval or posterior distribution requires its corresponding statistical method and assumptions.
Obtain values or bounds without losing alternatives
Choose the lightest adequate computation. Solve a small equation, eliminate an influential variable, enumerate a finite domain, or obtain a bound. A numerical formulation can use C.29.2 when its construction is nontrivial.
During elimination, preserve the conditions of each operation. Dividing by an unknown expression can discard its zero case. Squaring an equation can admit additional roots. Test the retained cases in the original relation and domains. Finding one root establishes a possible case; uniqueness requires an argument covering the admitted domain or a sufficient justified restriction.
For several unknowns, obtain the sought value from each compatible joint case. The other unknowns can remain unresolved when every compatible case gives the same sought answer. Section :5.1 gives a source voltage determined while its internal resistance remains unknown.
When a complete solution is costly, distinguish two computational results:
- A feasible example is one case that satisfies the original relations and conditions. Two such cases can establish a consequential ambiguity.
- An outer bound contains every compatible sought value but can also contain values that no case realizes. If the whole bound satisfies an inequality, that inequality holds for every compatible value. A bound spanning both sides of a threshold alone does not establish that both outcomes are possible.
For example, let an ideal indication obey r = x² with x ≥ 0, and let r = 9. A computation that has retained only the outer bound 0 ≤ x ≤ 4 leaves values on both sides of the threshold x = 2. The original relation, however, admits only x = 3. Thus x > 2 is resolved, and no feasible witness with x ≤ 2 exists. The coarse bound left the calculation unfinished; it did not establish ambiguity in the indication.
Use a validated enclosure method when the conclusion depends on retaining every solution through numerical calculation. Ordinary sampling can discover a counterexample but can miss another branch. When computation has not established existence, completeness or a needed bound, retain that limitation in the answer instead of interpreting solver termination as the missing result.
Test what is resolved for the receiving question
For a point-value question, test whether all compatible cases agree on that value. For a threshold or other condition, test whether it holds throughout the compatible set or a sufficient outer bound. Establish that the premises admit a case before using such a bound as the interpretation of a measurement; substitution of an available feasible case may suffice.
To establish that the readings leave two answers possible, find two cases that satisfy all the readings, shared parameters and domain conditions and give different answers. If the calculation has only produced a bound spanning both answers, try to find such cases or sharpen the bound. Report which possibility is still untested when neither calculation succeeds.
Separate a proved empty compatible set from a search that has not found a case. An empty set shows that the combined readings, model and conditions are inconsistent. Locate a consequential conflict; the cause may be a recording error, a changed subject, an unsuitable error bound or an inadequate measurement relation. A failed search instead limits the computational result.
A point result under ideal data can broaden when input uncertainty is restored. Recompute the relevant range when a reading, calibration coefficient or domain changes. Numerical sensitivity and branch ambiguity can require different remedies.
Choose the useful return
Return the value, range or comparison with the conditions that support it. Stop when it is sufficient for the intended use. A lower bound can be more useful than an unnecessarily precise estimate.
When two compatible cases give different answers, choose a further observation or relation by what it could distinguish. In :5.3 another counter with a different modulus distinguishes possibilities that repeating the same remainder preserves. When the remaining obstacle is an unfinished calculation, first see whether solving the existing relations or tightening a bound will settle the decision.
Use C.11.DUA when the value and effort of reducing the ambiguity need comparison. It can be reasonable to act with the remaining uncertainty or leave the question unresolved. B.5.MPC.R coordinates a needed revision across the physical, mathematical and computational contributions.
Archetypal Grounding
One source, two loads and an unidentified influence
Suppose a source has unknown open-circuit voltage E ≥ 0 and finite internal resistance Rs ≥ 0. An ideal meter of known input resistance Rm > 0 reads the connected voltage:
V = E*Rm/(Rs + Rm).
The relation and passive-source assumptions are supplied. The question is whether E is at most 8 V.
A reading of 5 V with Rm = 1 MΩ gives:
E = 5 V*(1 + Rs/(1 MΩ)).
Rs = 0 and E = 5 V fit the reading; Rs = 1 MΩ and E = 10 V also fit it. The threshold question is unresolved. The reading does establish E ≥ 5 V under the stated domain.
Now a meter with Rm = 3 MΩ reads 7.5 V. Assume E and Rs stayed unchanged between the readings and both meters obey the given model. The two equations share those two unknowns. Express E from each equation and equate:
V1*(1 + Rs/R1) = V2*(1 + Rs/R2).
Hence, when the denominator is nonzero,
Rs = (V2 - V1)/(V1/R1 - V2/R2).
The supplied readings give Rs = 1 MΩ and E = 10 V. Substitution recovers both 5 V and 7.5 V. The answer to E ≤ 8 V is now no.
The inference depends on the shared-source condition. If E changes between readings, replacing it with E1 and E2 leaves the original inference unsupported. A second number is useful through the relation connecting its measurement to the first.
For two zero readings under the same model, E = 0 follows while Rs can be any finite nonnegative value. A zero denominator in the derived formula is a reason to return to the original equations. Full recovery of Rs is unnecessary for this voltage result.
A saturated reading can settle one threshold
An ideal instrument reports r = min(y,10), where y ≥ 0 is a quantity expressed on the instrument's stated scale. For r = 10, every y ≥ 10 is compatible.
If the working question is y ≥ 8, the reading settles it. If the question is y ≥ 12, y = 10 and y = 13 are compatible cases with different answers. Reporting y = 10 as the determined quantity would lose that distinction.
Now suppose the reported reading follows r = min(y,10) + e, with -0.2 ≤ e ≤ 0.2. This error is applied after the clipping operation. For the same r = 10, the ideal clipped value can lie from 9.8 to 10. The compatible set becomes y ≥ 9.8: y = 9.8 with e = 0.2 is a feasible endpoint, and every y ≥ 10 fits with e = 0.
The reading still settles y ≥ 8. It no longer settles y ≥ 10, since y = 9.8 and y = 10 are both compatible. This is a bound-based result; no probability was assigned to the error. Repeated readings do not narrow that bound merely by being repeated. A narrower result needs a relation that supports the reduction, such as a justified stochastic error model or a smaller error bound.
Combining counter remainders
A device starts at zero, increments once per event and stores N modulo 16. No reset or missed event occurs. A supplied count bound is 0 ≤ N < 100. The stored value 5 permits:
N in {5,21,37,53,69,85}.
These six values come from N = 5 + 16*k with an integer k in the range allowed by the bound. They all satisfy the original operation. The reading settles N < 90, while N < 30 remains unresolved; 21 and 37 are witnesses.
Suppose a second counter starts at the same event boundary, counts the same events, and stores N modulo 17. It reports 4. Test the six candidates against this second relation: only N = 21 remains. The physical event correspondence and initialization make the mathematical intersection appropriate.
If the supplied bound is relaxed to N < 300, both 21 and 293 satisfy the two remainders. The earlier uniqueness conclusion therefore depended on the bound. If the counters cover different event intervals, use separate counts and recover the relation between them before combining their readings.
Bias-Annotation
The cases use supplied, simple relations. Real inverse problems may be high-dimensional, unstable or sensitive to uncertain model structure. Establishing complete bounds can then require considerably more work than exhibiting one ambiguity.
The set-based branch treats uncertainty as admitted possibilities. Statistical inference and decisions under risk can distinguish those possibilities by likelihood, loss or preference when their own assumptions are supplied. Use the result form appropriate to the question; a demand for uniqueness can make an already useful measurement appear inadequate.
Conformance Checklist
- CC-C16.IR-1 - Receiving distinction. The target property, subject conditions and required conclusion are stated.
- CC-C16.IR-2 - Joint premises. Readings, domains, influences, errors and shared-value assumptions enter the available relation with their meanings preserved.
- CC-C16.IR-3 - Inverse operations. Elimination or numerical calculation retains the needed branches and tests candidates in the original relations.
- CC-C16.IR-4 - Result strength. Feasible cases, outer bounds, proved emptiness and incomplete computation have distinct interpretations.
- CC-C16.IR-5 - Supported answer. Uniqueness or a threshold answer follows across the compatible cases; claimed ambiguity has feasible witnesses with different answers.
- CC-C16.IR-6 - Uncertainty meaning. A range, confidence statement or probability distribution keeps the assumptions that justify that form.
- CC-C16.IR-7 - Useful return. Work stops with a sufficient result or identifies a contribution capable of changing the unresolved answer.
Common Anti-Patterns and How to Avoid Them
Consequences
The practitioner can identify what information the measurement supplies for a question. A useful bound can end inquiry; an explicit pair of alternatives can direct a better observation or a model repair.
The cost grows with the relation and the conclusion claimed. A single ambiguity witness can be cheap, while excluding every alternative may require a global argument or specialized computation. The Method retains that difference so the amount of work follows the receiving use.
Architectural Rationale
Indication production and interpretation have different directions. A procedure can map many subject conditions to the same output. Interpreting that output therefore asks which of those conditions remain possible and which distinctions the receiving question needs.
This explains the separate contribution from C.16.MR. A measurement relation can already be available while its inverse use is unresolved. Constructing compatible cases makes that relation usable without requiring a new model or an estimate for every parameter.
The same reasoning connects physical measurement, mathematical inverse relations and computational information loss. Their subject premises differ. A source's loading law, an instrument's clipping order and a counter's event semantics supply the cases to which the common reasoning applies.
SoTA-Echoing
Practice question. What does an indication determine when an available relation has influential unknowns, lost distinctions or consequential input uncertainty?
Selected answer and alternative. Adopt a question-relative compatible-set calculation or sufficient bound when those features can change the answer. The alternative is direct use of a suitable calibrated analysis function, or a justified estimate with its uncertainty. Keep that cheaper route when it settles the same question. In :5.1, using one loaded reading as E returns the wrong target; two elementary compatible cases expose the loss at comparable effort. In :5.2, a threshold result needs less work than a point estimate. The selected method trades a convenient single output for the alternatives needed by the decision.
Measurement-model contribution. JCGM GUM-6:2020, §11.4 and §§13.5–13.6 distinguishes calibration and analysis functions, permits directly estimating an analysis function, and retains implicit measurement models under a stated local-uniqueness assumption. Adapt that direction distinction in :4.1–4.3. When uniqueness is unavailable, :4.4 returns the alternatives needed by the question. The document supplies quantitative measurement-model practice.
Computational contribution. IBEX 2.9, Contractors, Introduction and Forward-Backward describes filtering a domain while preserving its feasible solutions. Adopt that preservation requirement for the bounded numerical branch in :4.3. A surviving outer domain can still contain infeasible points, which motivates the feasible-witness distinction. Symbolic elimination and finite enumeration suffice for the worked cases.
Connecting the common question, target projection and sufficient-result return is a methodological synthesis. The cases are constructed under their stated conditions. Specialized inverse, statistical and decision methods contribute where the question needs their further operations or guarantees. Reopen the comparison when a cheaper analysis settles the same use with equivalent uncertainty, when a proposed bound misses a compatible branch, or when a changed measurement relation invalidates the result.
Relations
- C.16 supplies measurement subject, Characteristic, Scale, model, actual work, uncertainty and obtained-result meanings.
- C.16.MR constructs a missing or unsuitable measurement relation.
- C.29.1 examines the correspondence and preserved distinctions when mathematical representation itself is in question.
- C.29.2 constructs a needed computational formulation and its accuracy conditions.
- A.3.3.PI retains information for a future question under a dynamics account; the present Method interprets an available indication.
- B.5.MPC.R coordinates a needed revision across physical, mathematical and computational contributions.
- C.11.DUA compares the useful effect and effort of obtaining additional information.
C.16.IR:End
Repair a Measurement Model or Arrangement
Type: Method pattern Status: Stable Normativity: Normative
Problem frame
Use this pattern when a measurement disagrees with an expected result or cannot distinguish the cases needed for the work. A detector combines a beam with background light. A voltmeter changes the voltage it measures. A program turns correctly recorded counts and times into an incorrect rate. Each situation offers a different place to repair.
First useful move. Put the wanted quantity beside the indication and recover the relation used to connect them. Mark the contribution that prevents the wanted answer. If a detector responds to beam plus background, repeating that reading leaves their separate contributions unknown; a reading with the beam removed can make the difference between beam-on and beam-off readings usable.
The Method selects and carries through a change to a measurement model, its computation or the actual arrangement. The useful result is a corrected interpretation, an adequate changed measurement, a conditional design result or a remaining limitation with its effect on the work. C.16 supplies the measurement, model, performed work and obtained-result meanings.
The examples assume elementary algebra, bounds and units. The subject law, calibration and operating conditions come from the relevant physical or other domain practice. Apply an established correction or repair procedure directly when the fault and applicable remedy are known. Use the present Method when the choice of what to change remains consequential.
Problem
A disagreement can arise from the measured subject, its description, the instrument, the instrument model or the calculation. Improving the wrong contribution consumes effort while leaving the wanted answer unavailable. Repeated observations can reduce some random effects and preserve a loading error, background contribution or computational mistake.
A change can also alter the quantity being measured. Cooling a specimen to stabilize an instrument may change the specimen property whose value was wanted at the original temperature. Better agreement in the changed situation therefore needs a relation back to the intended use.
The problem is to locate a useful repair, carry it through the affected relations and establish what the revised result supports. Explaining the historical cause of a fault and obtaining an adequate measurement can require different amounts of work.
Forces
Solution
Follow the disagreement from the wanted result through the indication-producing relations. Compare changes at the contribution that matters, choose a sufficient available repair, and use the revised relation to obtain the result needed by the work.
Align the question and observations → locate the consequential contribution → compare a model, computation or arrangement change → carry through the selected repair → determine what the revised result supports.
The starting point can be a known fault, an unexplained discrepancy or a proved inability to distinguish the needed cases.
Align the comparison and target
State the quantity or distinction sought, its subject and operating conditions. Put the expected result and obtained indication on a comparable basis: same quantity, time interval, units and relevant preparation. If the expectation concerns an unconnected source but the reading concerns a loaded source, retain that difference in the relation.
Recover how the indication was produced and interpreted. Separate an observed disagreement from an assumed explanation. A temperature-dependent response may come from the subject, sensor or their interaction; the correlation alone leaves those contributions to be distinguished.
Use C.16.MR when the measurement relation itself is missing. Use C.16.IR to find which alternatives fit an available relation and which distinction remains unresolved. A sufficient bound may already answer the question and end the repair attempt.
Locate the contribution that could change the answer
Follow the target-to-indication relation far enough to identify the consequential influence or incompatible operation. Ask what each candidate explanation would require changing and what present information bears on it.
The table locates choices; the domain Method supplies the physical law or diagnosis that makes one choice appropriate. A common influence can cross several rows. For example, a change in illumination can alter both the subject response and the detector response.
For a suspected computational fault, run a supplied or constructed input whose answer follows from the intended relation. Compare intermediate values, units and branch choices until the divergence is located. A successful calculation on that input checks implementation of the relation. Compare the relation with the actual subject separately when its adequacy is in question.
Compare changes by their predicted contribution
For each available repair, change the corresponding part of the relation and determine what would become knowable. An added parameter can represent a missing effect while introducing an unknown that one reading cannot determine. A physical change can remove the effect or make its influence small enough to bound.
A difference measurement is useful when the target changes between two observations in a known way while the influences being cancelled remain sufficiently stable. A larger input resistance is useful when the loading relation shows how much it reduces the consequential error. Repeating an unchanged observation needs its own error or dynamics account to support a narrower conclusion.
Prefer an available comparison or a small reversible change when it can settle the repair choice. If two explanations predict the same indication under every condition of the proposed test, a deterministic reading cannot distinguish them. For probabilistic predictions compare likelihoods for the recorded value, or the probabilities assigned to its reported interval or category, as appropriate. A result possible under both explanations can still favor one. Use C.11.DUA to compare the gain from an attainable observation with its effort and delay.
The needed conclusion sets the reach of diagnosis. Replacing a defective assembly and obtaining the required response may restore the measurement without identifying which internal component failed. Retain the unresolved cause if later reliability, maintenance or causal explanation depends on it. Use C.28 when the required result is a causal attribution.
Carry the selected repair through the measurement
Keep the target and receiving question fixed while revising the selected contribution. If an alternative target or operating condition is acceptable, make that change part of the receiving decision and derive the relation for the new use.
For a model repair, include the consequential influence and obtain its value, range or relation from available information. An added correction term needs that basis. Carry any remaining uncertainty into the sought result.
For a computation repair, correct the operation and recalculate the affected results from retained inputs. Reuse observations whose meaning and conditions still fit. Identify outputs that depended on the incorrect operation so that subsequent comparisons use the repaired result.
For an arrangement change, derive how preparation, interaction, indication and calibration are affected. An attenuator can move a detector out of saturation while introducing an attenuation factor and its uncertainty. A shield can remove ambient light while changing alignment. Retain the conditions that make the new relation appropriate.
A planned change supports a conditional result. Once performed, use the resulting observations and conditions to interpret that measurement. Older observations can still answer their original question or be reinterpreted through a supported revised relation.
Test the repaired contribution at the needed scope
Choose a comparison capable of exposing the defect the repair addresses. For computation, compare with an independently obtained result for the same mathematical input. For an omitted influence, use a relevant reference condition, intervention or already available contrast. For a claimed operating range, include the range boundary or change of behavior on which the claim depends.
Read the result through the repaired measurement relation. Propagate the uncertainty that could change the receiving answer, and use C.16.IR when compatible alternatives or an outer bound require interpretation. A closer numerical match can still leave a consequential ambiguity.
When the comparison leaves the needed answer unresolved, return to the remaining contribution. A further repair may be justified, or the useful result may be a narrower claim, a changed use or a stop. Do not expand a successful local comparison into a wider range claim without the corresponding basis.
Return the usable result
Give the corrected value or comparison, the conditions under which it applies and the consequence for the receiving work. Keep the reason for a changed interpretation where another user could otherwise continue using the old one.
Stop when the result is adequate for that use. Further investigation may have value for another question, such as broader operating conditions or the origin of the fault; it is a separate continuation whose gain should be clear.
Archetypal Grounding
Separate a beam from background and detector offset
The wanted quantity is the optical power I from a beam at a detector. A supplied linear response model is:
r = g*(I + A) + b.
Here A is the ambient optical power reaching the exposed detector, g = 2 mV/mW is a known response coefficient and b is an electronic offset. The detector is unsaturated, and g and b remain unchanged across all three observations. A remains unchanged between the two exposed readings; the opaque cap blocks both optical contributions:
The capped reading gives b = 2 mV. Subtracting only this offset from the beam-on reading and dividing by g gives 5 mW. That is I + A; it leaves the wanted beam contribution mixed with ambient light.
Use the exposed beam-off reading to cancel both unchanged contributions:
I = (r_on - r_off)/g = (12 - 6) mV / (2 mV/mW) = 3 mW.
The same observations give A = (6 - 2)/2 = 2 mW. Substitution reconstructs all three readings. The repaired interpretation supplies I using observations already available.
A possible arrangement repair is to shield the receiver from ambient light while preserving the beam at the detector. Under A = 0 and the same g and b, a new beam-on reading of 8 mV would also give I = 3 mW. That value is conditional until the new measurement is performed.
Suppose instead that the shield removes ambient light but transmits a known fraction alpha = 0.9 of the original beam. With the same g and b, a new reading of 7.4 mV gives (7.4 - 2)/2 = 2.7 mW for the transmitted beam. Recover the original beam through the changed relation:
I = (r - b)/(g*alpha) = (7.4 - 2)/(2*0.9) = 3 mW.
If the transmission fraction is unknown, this new reading alone leaves the original beam unresolved. The earlier on/off observations still support their result of 3 mW. The shield's changed interaction must be included in any conclusion drawn from the new measurement.
For a question I ≤ 3.2 mW, suppose each on/off reading has an arbitrary additive error between -0.1 and 0.1 mV, while g and the shared background remain fixed. The difference gives 2.9 ≤ I ≤ 3.1 mW, so the bound settles the question. Drift in A between on and off would add a further term. Alternating readings reduces that uncertainty only with a supplied account of the drift.
Reduce loading enough to settle the question
A source has open-circuit voltage E and internal resistance Rs between 0.8 and 1.2 MΩ. A meter with input resistance Rm reads:
V = E*Rm/(Rs + Rm), hence E = V*(1 + Rs/Rm).
With Rm = 1 MΩ and an ideal reading V = 5 V, the compatible bound is 9 ≤ E ≤ 11 V. It leaves the question E ≤ 10.5 V unresolved.
One option is to bound Rs more closely and correct the first reading. Another is to reduce loading in a new measurement. Use a meter with Rm = 100 MΩ. Suppose the same passive-source law and resistance range hold at the new reading. If its ideal indication is 9.9 V, the open-circuit voltage at that reading satisfies:
9.9*(1 + 0.8/100) ≤ E ≤ 9.9*(1 + 1.2/100),
so 9.9792 ≤ E ≤ 10.0188 V. The new reading alone settles E ≤ 10.5 V at that time without determining Rs. To answer the question about the earlier time as well, establish that E stayed the same or bound its change tightly enough for the threshold. If both E and Rs stayed the same, the two ideal readings can also be solved jointly by C.16.IR:5.1.
Restore a supplied reading-error bound of ±0.1 V for the new meter. The positive factors give:
9.8*1.008 ≤ E ≤ 10.0*1.012,
or 9.8784 ≤ E ≤ 10.12 V at the new reading. The threshold is still met. For a lower threshold the stated reading error could change the decision. The conclusion depends on the meter's operating range and the loading model.
Recompute a rate from retained observations
A flow logger records cumulative pulses and elapsed timestamps in milliseconds. The supplied calibration is one pulse per litre, with no missed pulses or resets in the interval. The receiving question concerns average volume flow over that interval.
Both endpoint timestamps are taken at pulse events. The retained endpoints are 120 pulses at 10,000 ms and 180 pulses at 40,000 ms. A program subtracts the timestamps, divides the count difference by 30,000 and labels its output 0.002 L/s.
The pulse difference represents 60 L. The timestamp difference represents 30 s. Evaluating the intended relation gives:
average flow = 60 L / 30 s = 2 L/s.
The program's division obtained 0.002 L/ms and then attached the wrong time unit. Convert milliseconds to seconds before division, or convert the resulting rate afterward. Both repairs give 2 L/s. Recompute other affected intervals from their retained counts and timestamps.
A reference input of 10 pulses over 2,000 ms should give 5 L/s. A program result of 0.005 L/s reproduces the conversion fault; 5 L/s confirms this operation for that input. This calculation checks the implementation. Whether the actual instrument counted every litre and its clock tracked elapsed time requires the relevant calibration and observation when those premises are disputed.
The useful result is a repaired average rate from existing observations. Repeating the physical flow measurement is unnecessary for the identified unit-conversion fault while those observations and their calibration remain usable.
Bias-Annotation
The examples separate influences with simple relations and supplied conditions. Coupled effects, nonlinear response, drift, hysteresis or uncertain model structure can require richer diagnosis and uncertainty Methods. The pattern identifies where those contributions become necessary. Use the relevant domain Methods for their detailed treatment.
The method favors a repair sufficient for the receiving use. A fault investigation concerned with reliability, causal explanation or a wider operating range can require work beyond restoring one useful reading. Keep that broader question explicit when it governs the effort.
Conformance Checklist
- CC-C16.RM-1 - Comparable question. The wanted quantity, actual indication, expectation and relevant conditions support the comparison being made.
- CC-C16.RM-2 - Repair location. The proposed change acts on a contribution capable of altering the unresolved answer.
- CC-C16.RM-3 - Choice basis. Available comparisons, domain relations or proportionate further inquiry support the selected repair.
- CC-C16.RM-4 - Target continuity. The revised arrangement retains the intended target or a stated relation to the accepted changed use.
- CC-C16.RM-5 - Dependent results. Changed models, computations, calibration and observations are carried through the affected interpretation.
- CC-C16.RM-6 - Adequate test. The comparison can expose the addressed defect and supports the range or result claimed.
- CC-C16.RM-7 - Useful completion. The returned result and remaining uncertainty answer the receiving question or state the consequential limitation.
Common Anti-Patterns and How to Avoid Them
Consequences
The practitioner can change the contribution responsible for an inadequate measurement and recover a usable result. Available observations may suffice; an equipment change can simplify the inference; a computational repair can restore results without a new physical experiment.
A targeted repair also limits what is established. A successful local result may leave the fault's cause, a wider operating regime or the uncertainty of a different quantity open. Those limits direct further work when its receiving question becomes current.
Architectural Rationale
Measurement joins a subject, an interaction, a mathematical relation and a calculation used to interpret an indication. The same unwanted effect can be addressed by representing it, reducing it in the arrangement or changing how the indication distinguishes the target. Keeping these alternatives together helps choose a repair by its effect on the wanted result.
C.16.MR constructs a relation and C.16.IR determines what it resolves. This Method begins when that existing use is inadequate and chooses what to change. B.5.MPC.R supplies the wider repair across physical, mathematical and computational contributions. Here the distinction between subject and instrument, cancellation conditions, loading reduction and reinterpretation of retained observations makes that repair actionable for measurement.
The result can be epistemic, such as a corrected interpretation, or include an actual physical alteration. Its status follows the work performed and the premises available. That distinction permits a useful design calculation before equipment is changed and a useful recomputation after an implementation fault is found.
SoTA-Echoing
Practice question and selected answer. How should one change a measurement whose current model, calculation or arrangement fails the receiving question? Select a contribution-specific repair, compare what it would resolve, and stop at an adequate result. The alternatives are a known applicable correction, a more detailed model, or more observations under the unchanged procedure. A known sufficient correction remains the cheaper route. The selected comparison is useful when repetition preserves the defect or extra model detail adds an undetermined influence. In :5.1 the already available on/off contrast resolves the beam contribution; in :5.2 changing loading yields a sufficient bound without identifying every parameter. Changing equipment can cost more than calculation, so its value depends on availability and the wanted conclusion.
Model and apparatus alternatives. Dounas-Frazer and Lewandowski (2018), §2 and §§4.1–4.3 explains iterative comparison and revision of physical-system and measurement-system models and apparatus. Its optics cases distinguish controlling unwanted light, avoiding saturation and revising response models. Adapt these alternatives into :4.2–4.4's choice of repair location and conditions. The paper's educational investigations support these modeling moves in experimental physics; they do not establish the effectiveness of the present cross-case Method.
Adequacy and computational comparison. JCGM GUM-6:2020, §12 distinguishes comparisons with reference situations, computational tests using generated data, fit over the intended range and the adequacy of a simpler model. Adapt those distinctions in :4.2 and :4.5: select a comparison for the contribution and claim being repaired. Section :4.3 applies C.11.DUA to its attainable value and cost. The numerical cases are constructed demonstrations using the instrument laws and conditions stated in each case. This source supplies the comparison distinctions, not those instrument laws or performed observations.
The integrated repair sequence, target continuity and return through a useful conclusion are methodological synthesis. Reopen the choice when a cheaper remedy settles the same question, a changed condition invalidates the correction, or a comparison reveals a consequential influence left outside the model.
Relations
- C.16 supplies measurement, model, work, result and uncertainty meanings.
- C.16.MR constructs or revises the measurement relation needed to interpret an indication.
- C.16.IR obtains compatible values and determines what the relation and readings resolve.
- B.5.MPC.R coordinates repair across the wider physical, mathematical and computational connection.
- B.5.RR carries a changed premise through affected reasoning.
- C.29.2 constructs or repairs a needed computational formulation and its accuracy conditions.
- C.28 supplies a causal account when identifying the cause is the required result.
- C.11.DUA compares further inquiry and its attainable contribution to the receiving use.
C.16.RM:End
Restore Precision to Characteristic and Scale Wording
Type: Characterization precision-restoration pattern Status: Stable Normativity: Normative unless explicitly marked informative
Plain-name. Characteristic-scale wording repair.
Intent. Restore the meaning of a characterization claim when a compact word hides what is characterized, how it is assessed or compared, or what the result supports. Once those relations are clear, use the applicable subject pattern.
Use this when
Use this pattern when a word such as metric, strong, indicator or threshold carries a characterization claim whose basis is unclear. A score may hide its Scale, an indicator its relation to the property indicated, or an apparent improvement the comparison that supports it.
First useful move. Ask what is being characterized, what the sentence asserts about it, and which basis the reader needs to interpret or use that assertion. Repair only the missing contribution. An ordinary sentence can be sufficient.
When the meaning and applicable conditions are already recoverable, use the subject pattern or ordinary domain explanation directly. The neighboring patterns in §6 govern their own questions; use C.16.P only for the unresolved wording.
Problem frame
Working texts often use compact characterization words. A problem arises when a term implies a comparison or further conclusion while hiding the construction that could support it.
The repair question is:
What does the sentence assert about which subject, what supports its intended comparison or use, and which rule governs any further FPF claim?
Problem
The same compact word can denote different constructions. An “axis” can be a coordinate in a model, a Characteristic or an ordinary explanatory direction. A score can supply a value for a comparison without establishing a wider claim about evidence, readiness or permission.
The repair must recover the meaning needed by the receiving work. It should not require a new measurement, characteristic space or formal record when an ordinary explanation or an already applicable subject rule settles the question.
Forces
Solution
Rewrite the claim so the reader can recover its subject, the relevant characterization or comparison, its basis and the use it supports. Keep necessary applicability and stop conditions.
Recovery sequence
- Read the trigger in context. Recover what the expression asserts and who needs to understand or use it. If a decision depends on its interpretation, identify that decision. Distinguish the question asked from a comparison the available data happen to support. Retain the source wording when the repair needs an inspectable comparison.
- Recover the bearer. Name the subject being characterized. For a relational Characteristic, retain the tuple required by its definition.
- Recover the construction. Use §5 to identify what the wording means. A trigger can be ordinary prose; its appearance does not create an FPF kind.
- Use a recovered subject rule directly. Once the construction is clear, apply its subject pattern from §6 or the relevant domain practice.
- Repair what remains hidden. Recover the contributions needed by this claim: for a proposed measurement, its Characteristic, Scale and method; for an observed value, its observation and interpretation; for a computed or assigned value, the model or rule and the inputs it uses; for a calculation, the operations that the Scale admits; for a comparison, its reference or comparator set; for a threshold, its predicate and applicability; for an indicator, the indicated property or claim and the direct relation, including distortion risk when it is a proxy. Keep necessary use and stop conditions. If an indicator relation is required but no governing rule can state or test it, return
missing-governorrather than storing anindicatorRolelabel. Use A.6.RCD:0.1 to distinguish that absent rule from missing case facts or a failed positive test. Add explanatory non-use guards only under the F.19:4 test. - Separate further claims. A measurement or score may contribute to evidence, a decision or another use. Apply the governing conditions for that additional claim; repairing the characterization alone does not establish it.
- Return a useful statement. Say what the reader can now recognize, compare or conclude, or which missing contribution leaves the claim unresolved. Type-correct wording with no useful interpretation is insufficient.
When the receiving work needs a separate inspectable note, the following scaffold can retain the repair. Include only fields relevant to that claim:
A local rewrite is sufficient when the sentence conveys the needed meaning and conditions. An ordinary PatternID can locate the governing rule. Include relationFunctionClaimRef only when the use depends on the identity of its defining or constraining ClaimGraph. Include indicatorRelationDisposition when an indicator or proxy relation is under repair, following §4.1. An explanatory nonAdmissibleUse guard is optional under F.19:4's independent-ground, plausible-reader, contribution and smallest-clear-correction test; its absence needs no entry.
Trigger split
Adjacent Claim Governance Named by Value
Archetypal Grounding - Worked cases
Worked comparison. A report calls a new implementation “stronger.” The receiving question is how typical response time changed across four specified, matched workloads. The agreed summary is the median, here the mean of the two middle ordered times. Response time is elapsed duration on a ratio Scale. The earlier times are 80, 100, 120 and 140 ms. The corresponding new times are 70, 90, 110 and 130 ms. For these four cases, median response time fell from 110 to 100 ms.
If the requirement is a median of at most 105 ms on these cases, the new result meets it. If each case must finish within 105 ms, the 110 and 130 ms cases violate it, even though every paired case was 10 ms faster. The same records settle both questions. They provide no comparison for other workloads without grounds for transfer. A change in the tested workload mix would also require grounds for comparing the two editions.
Bias-Annotation
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Consequences
Benefits. The reader can interpret the characterization, identify the supported comparison or use, and locate a missing contribution without constructing a larger measurement account unnecessarily.
Trade-offs. Some compact phrases become longer because the bearer, characteristic, scale, threshold, proxy, or subject pattern must be named. The gain is that measurement, quality, mathematical-lens, evidence, assurance, gate, decision, and causal-use claims do not hide inside one scalar word.
Stop condition. Stop when the intended characterization, its basis and applicable conditions are recoverable. An adequate ordinary statement can finish the repair. Continue to a subject pattern only for a remaining question that needs it.
Rationale
A compact word can summarize a construction once the reader can recover it. Restoring only the missing subject, relation or condition lets that shorthand remain useful while making its implication inspectable.
SoTA-Echoing
Measurement and comparison practice distinguishes characteristics, scales, measures, indicators and their intended uses. These sources contribute different parts of that recovery:
Relations
E.10catches hidden characteristic and scale wording and selects this pattern only when construction is hidden.E.10.ARCHdefines the shared wording-use recovery order and applicability row.A.17,A.18, andC.16govern characteristics, scales, values, measures, and measurement use.A.19governs characteristic-space construction.C.25governs Q-bundles.C.16.Qgoverns quality-term or evaluative characterization wording.E.21governs pattern-quality evaluation characteristic spaces.C.29governs mathematical-lens use.- Evidence, assurance, gate, work, decision, causal-use, release, benchmark, and publication patterns define or constrain their own claims.
C.16.P:End
Quality-Term Precision Restoration
Type: Characterization precision-restoration pattern Status: Stable Normativity: Normative unless explicitly marked informative
Plain-name. Quality-term precision restoration.
Intent. Recover what a quality statement means for its subject and intended use. Distinguish the sense, interpretation basis and kind of result before using the statement for evaluation or another decision.
Use the form already defined for a known endpoint. When unresolved alternatives must remain inspectable, qualityTermAscription(...) supplies a transitional form with explicit participants and conditions. Its detailed fields do not apply to every occurrence of the word “quality.”
Philosophical, representation-learning, engineering and search uses can retain their different meanings. A recovered action question, relation, requirement or commitment follows its own subject rule.
Use this when
Use this pattern when wording such as quality, good, fit, high-quality, quality characteristic, quality improved, or an evaluative -ility claim hides the intended quality meaning or evaluative use.
First useful move. Identify what the quality statement concerns and what “quality” means here. For an evaluation, also recover the criterion used to judge that subject. Use the sentence and its receiving question to distinguish the plausible interpretations. If the meaning and applicable endpoint are already clear, use that endpoint directly.
Lowest sufficient use. Keep ordinary praise and source-local description ordinary when no further FPF-governed claim is made. A supported ordinary explanation can settle the ambiguity. Preserve an insufficiently articulated cue under A.16.1 or its possible continuations under B.4.1; do not invent the missing interpretation to complete a record. Use the transitional form only when the remaining ambiguity must be inspected or reused, with the required interpretation basis in §4.1.
Practical gain. The reader can recover what is described or evaluated, what any comparison establishes, and what the resulting claim supports. A shared quality word no longer substitutes for the relation between different traditions or for a separate evidence or decision claim.
Not this pattern when.
- If the issue under repair is hidden characteristic, scale, score, metric, coordinate, threshold, or comparison construction, use
C.16.Pfirst. - If the claim being made is already a Q-bundle, pattern-quality coordinate, relation construction, action invitation, evidence, assurance, gate, work, decision, causal-use, release, or source-use claim, apply the pattern for that claim directly after any needed quality-word repair.
- If the word is ordinary praise or source-local wording with no FPF-governed use, keep it ordinary, quote-only, or reduced-use rather than publishing a quality-term repair.
Problem frame
Quality statements, including evaluation claims such as “the model quality improved” or “the architecture is high-quality”, can concern different subjects and senses. For example:
- Phenomenal character or qualia when the experienced quality itself is the topic of description rather than an externally measured characteristic.
- Preconceptual fit or felt rightness before stable EntityOfConcern characterization.
- Latent and distributed fit signals in learned representations, world models, or active inference loops.
- Explanatory merit of a theory, problem frame, or conjecture.
- Architectural-description fitness and compression merit of an architecture description or architecture model under a declared viewpoint.
- Engineering quality families such as reliability, maintainability, security, evolvability.
- Usefulness and selection value in open-ended search, novelty–quality–diversity, or portfolio selection.
- Control adequacy of a policy, model, or controller in a closed loop.
The failure modes are recurrent:
- Sense elision. One quality word hides several non-equivalent senses.
- Bearer confusion. The subject of the quality statement is unclear: record, publication carrier when the carrier itself is described or evaluated, episode, model, policy, explanation, candidate, architecture, relation, or action loop.
- Form confusion. A non-metric signal is rewritten as a metric; a bundle is treated as one scalar; an objective is mistaken for a characteristic.
- Substrate confusion. Embodied and preconceptual, latent and distributed, and symbolic-local representations are silently collapsed.
- Plane confusion. Quality of the EntityOfConcern being described, quality of the description, quality of the carrier, and quality of the publication face are silently collapsed across
ReferencePlanevalues and A.7 lanes. - Bridge illusion. Similar wording across traditions is mistaken for sameness.
- Illegal scalarisation. Composite engineering families or explanatory merit are compressed into one number without an admissible scoring method.
- Viewpoint conflict. One stakeholder means architectural attributes, another means usefulness, another means preconceptual fit.
Problem
How can a reader retain the convenience of “quality” while recovering its intended sense, subject and conditions? The difficulty is greatest when a statement connects traditions whose evaluative criteria differ. An action-oriented use may need meaning recovery under A.6.A rather than an evaluative ascription.
Forces
- Breadth vs precision. “Quality” is attractive because it is broad; that breadth can obscure which sense is intended when uses meet.
- Preconceptuality vs auditability. Some uses refer to something real but not yet stably characterised.
- Distributed substrate vs local publication. Some evaluative signals arise in distributed or embodied substrates but must later be published in explicit local forms.
- Comparability vs non-reduction. Engineering and selection settings need comparability, but not every evaluative signal is an admissible metric.
- Cross-tradition dialogue vs false unification. The framework should permit parallels without asserting identity.
- Progressive articulation. A person can notice a mismatch before being able to name a criterion. An account of what they noticed can guide a choice of measure or objective. To claim a measured value, establish how the measure bears on the characteristic and obtain the observation.
Solution
Recover the intended claim first. Use a known endpoint directly; retain a transitional ascription only for unresolved interpretation that the receiving work needs to inspect.
Trigger rule
Inspect quality wording when its sense, bearer, comparison or consequence is unclear, especially in a claimed improvement, admission or cross-tradition use. Familiar phrases such as quality score, model quality and quality requirement are cues to that question. Their appearance alone does not require a new ascription or a repeated repair of an already explicit endpoint.
Operational repair sequence
- Read the whole assertion. Recover what the sentence claims or asks. Retain its wording when an inspectable before/after comparison is needed.
- Recover the subject being described or evaluated. Distinguish the EntityOfConcern being described from its description, publication face or carrier when that distinction changes the claim. If the question could concern a component, the operation of several components together or the encompassing whole, identify the kind of whole and the relation the judgement depends on. A Method can prescribe what its parts must do together. A record of Work can establish which parts were performed together on this occasion. Keep the component result fixed and vary the others or their arrangement. If the requested conclusion can change, restore the relation between the assessed subject and the subject of the original question. An applicable common rule or a direct assessment of the required whole can settle the question. Keep ordinary non-governed wording ordinary.
- Distinguish the plausible interpretations. Use the stated purpose and interpretation scheme, together with any relevant evaluation conditions or comparison, to resolve plausible alternatives. If an existing endpoint already conveys the intended claim, use it and stop this repair. Keep a short Candidate-Set Note only when unresolved alternatives need to remain inspectable for the receiving decision or publication; use an existing explanation when it already does that job.
- Follow the recovered question. For action-oriented wording, a relation, evidence, a gate or another non-quality question, apply its subject rule. Use A.6.A, A.6.P, C.16.P or another recovery pattern only while the corresponding meaning is still unclear. An adequately expressed question can finish without a
QualitySense. - Choose a sense when a transitional ascription is needed. Select the applicable
QualitySenseand retain consequential rival interpretations while they remain unresolved. Explain a rejected alternative only when that distinction helps recover the intended claim. - State the supported interpretation. Use the known endpoint's form, or the transitional form under §4.1 with its required bearer, scheme, probe/model and comparison conditions, evaluator/viewpoint, scope and normal form. Add result, evidence, grounding or cross-local references only for independently established relations used by this claim.
- Apply conditions for a further use. A clarified quality statement can contribute to a decision, evidence claim or other result. Establish that further use under its own rule; the word “quality” and completion of the repair do not establish it.
Transitional repair frame: quality-sense interpretation anchored by qualityTermAscription(...)
A transitional ascription makes an unresolved quality interpretation explicit under one effective ReferenceScheme and a named endpoint pattern or source relation. It records the intended characterization without asserting a result or evidence relation merely through its fields.
qualityTermAscription(...) is the canonical transitional quality-term repair form when the endpoint choice is not yet fixed. The form is not a relation kind by default and does not replace the endpoint source.
Use of qualityTermAscription(...) presupposes enough articulation explicitness to name the bearer, effective scheme, probe/model frame, A.19.CPM profile comparison frame or explicit none, ClaimScope, and at least one candidate quality-sense family. Closure degree may remain low while qualityTermAscription(...) is transitional, but content that is still only a cue pack, forwarded cue, or open explanatory probe stays in A.16.1, B.4.1, or B.5.2.0. If a published record later loses an interpretation-bearing scheme, frame, scope, or direct source relation required for its stated use, retreat via A.16.2; changed witnesses, evidence use, or grounding reopen only the neighboring result or reliance claim they bear on.
The transitional form is:
effectiveReferenceScheme, probeOrModelFrameRef, comparisonFrameRef, and claimScope are explicit. Use the probe or model frame for the domain procedure, including any qualitative comparison it makes. Reserve comparisonFrameRef for a comparison of admitted profiles under A.19.CPM. It identifies the CG-Spec, comparator edition, comparison scope, reference plane and interval. Set it to none when the claim uses no such profile comparison.
The record designates, but does not embed, a viewpoint. A non-none viewpointRef is one U.ViewpointRef whose governed resolution yields an exact viewpoint episteme; the reference, the viewpoint episteme, and the evaluator remain different objects.
qualityResultClaimRef is not assessment work, while witness refs and an A.10 evidence-provenance path establish neither a result nor empirical grounding. Cite empiricalGroundingRelationRef only for a separately obtaining C.2.1 relation between the identified episteme and exact holon under governed observation, intervention, measurement, test, or evaluation relations.
Likewise, cite an F.9 Bridge occurrence and bounded-use claim only when each independently exists. Cite a Card only when that optional package exists. Cite a stance note only when its reference resolves a C.2.1 episteme whose EntityOfConcern is that exact use claim.
At least one of endpointPatternLocator and endpointSourceRelationRef is required. The locator identifies the pattern passage that defines or tests the endpoint. A separate assertion or ClaimGraph is required only when a named later use depends on that rule identity.
For a use whose quality meaning remains consequentially ambiguous, “X has quality” is insufficient. State the supported interpretation, use the known endpoint, or retain the unresolved alternatives in the transitional form when its entry conditions are met.
Discipline note.
QualitySense is a slot value inside the transitional repair form; it is not a replacement for the endpoint FPF pattern or explicit endpoint source reference. The sense token refines the intended quality sense while the endpoint source, applicable pattern, or EntityOfConcern remains explicit.
Separation note.
evaluatorRef and viewpointRef are not synonyms. The evaluator is the observing, criticizing, or selecting party or policy. viewpointRef is a governed reference whose resolution yields one exact U.Viewpoint episteme; selecting or resolving it grants no membership, conformance, authority, or evaluation result.
The checked bearer, any dated assessment work, the resulting claim episteme, witness carriers, an A.10 evidence-provenance path, and an optional EpistemeEmpiricalGroundingRelation remain independently governed. A filled qualityTermAscription(...) may refer to each, but record completion, a result label, or stored witnesses makes none of the neighboring relations obtain.
Polarity discipline (bearer-centred; no silent inverse)
qualityTermAscription is bearer-centred.
A transitional ascription SHALL keep the described or evaluated participant in the bearer position and SHALL publish evaluatorRef and governed viewpointRef separately, using none when either is absent. A known endpoint retains its own form and the same participant distinctions.
- When the evaluating architects are identified as
ArchitectureReviewBoard, “Architects rate the system highly” rewrites toqualityTermAscription(bearerTuple={System}, evaluatorRef=ArchitectureReviewBoard, viewpointRef=none, …). - When the benchmark's evaluation policy is identified as
BenchmarkPolicy, “The benchmark says model quality is high” rewrites toqualityTermAscription(bearerTuple={Model}, evaluatorRef=BenchmarkPolicy, viewpointRef=none, …).
There is no inverse token that silently makes the evaluator the bearer. If inverse wording is used in Plain prose, rewrite it into the bearer-centred form, or use the explicit inverse form supplied by the applicable pattern.
Endpoint-first discipline
When the endpoint pattern or explicit endpoint source relation is already known, publish the form it defines directly. Keep qualityTermAscription(...) only when preserving the transitional ambiguity is itself informative.
Typical direct endpoints are:
- engineering
-ilityheads published as oneCharacteristicor oneQ-Bundle, - selector-context uses published as an
Objectiveheaded byQS.UseValueunless overridden explicitly, - architecture-description uses published under the description-side evaluative head already selected by the viewpoint bundle,
- explanatory-merit uses published under the explicit merit bundle when that bundle head is already known.
Core construct: QualitySense
Every transitional ascription SHALL identify an explicit QualitySense token. A known endpoint can express the intended claim directly without adding this token or record.
A QualitySense token publishes at least:
Where:
articulationMode∈{ preconceptual, exemplar-grounded, proxy-grounded, characteristic-bound, bundle-bound, objective-bound }representationSubstrate∈{ embodied-kinesthetic, latent-distributed, symbolic-local, hybrid }defaultNormalForm∈{ SignalPack, Characteristic, Bundle, Objective }admissibleNormalFormsis the explicitly declared set of admissible normal forms for the sense.defaultNormalFormnames the primary normal form; any additional endpoint forms MUST be declared here rather than inferred ad hoc.probeOrModelFrameKindconstrains only the domain-local probe/model configuration, whilecomparisonFrameRequiredstates whether the endpoint calls for two admitted profiles and their A.19.CPM comparator.bridgePolicycan require F.9 recovery or forbid silent reuse, but it cannot establish a Bridge. If the quality ascription is published, handle publication face, form, unit, carrier, and rendering questions under E.17, E.8, or the applicable publication pattern.
Normative starter set of sense families
A declared local vocabulary under one effective ReferenceScheme MAY add local senses, but the following starter set is normative as the initial disambiguation menu:
Default-form note.
QS.EngineeringQualityFamily and QS.ControlAdequacy default to Bundle.
A declared local use under one effective ReferenceScheme MAY operationalize one explicit head as a Characteristic, but that is a declared operationalization, not a second default normal form.
Normative interpretation rules. Choose the family from the claim; a domain label alone does not determine its sense.
- In NQD, QD or selector use, the evaluative Q head defaults to
QS.UseValueunless another sense is explicitly declared. - An engineering quality claim resolves to one explicit
U.Characteristicwith its CSLC Scale, or an explicit Bundle, preferably a C.25 Q-Bundle when its conditions apply. - Description of experienced phenomenal character uses
QS.PhenomenalCharacter; preconceptual fit or felt rightness usesQS.PreconceptualFit. - A claim about distributed fit in a learned representation uses
QS.LatentFit; a claim about closed-loop adequacy usesQS.ControlAdequacy. A model can also receive a different evaluation under its corresponding sense; the phrase “model quality” alone selects none of them. - Evaluation of an explanation's epistemic merit uses
QS.ExplanatoryMerit. A different claim, such as readability of its presentation, retains its own subject and criteria. - For architecture wording, distinguish the bearer first. Engineering quality of the system and fitness of its architecture description use different subjects and criteria; the latter uses
QS.ArchitecturalDescriptionFitness.
Required slots for a conforming qualityTermAscription
A conforming qualityTermAscription SHALL make explicit:
-
Bearer tuple. Name the exact subject designator or tuple and its arity. A description, carrier, evaluator, or result claim cannot silently replace that bearer.
-
QualitySense. Name the intended quality-sense family; retain the evaluative criteria when the claim evaluates that subject. -
Effective ReferenceScheme. State the effective reference scheme by value in the C.2.1:4.1.2 range
EpistemeInterpretationScheme: supply the designation and interpretation rules with their semantic dependencies so every designator and local sense in the ascription is interpretable. The name of a context or representation language alone leaves those rules unspecified. -
Probe or model frame. Name the exact domain-local exemplar pack, probe pack, test or criticism pack, Q-bundle definition, CG-frame, acceptance specification, control horizon, or other governed probe/model configuration.
-
Comparison frame. For a comparison of profiles under A.19.CPM, name the admitted profiles and the applicable comparator and comparison scope. If the claim needs such a comparison and an input is missing, say which conclusion is still open. For a domain comparison that uses a different operation, describe what was compared and on what grounds in the probe or model frame and in the claim. Use
comparisonFrameRef = nonewhenever the claim uses no CPM comparison. When both operations are used, identify the grounds for each. -
Evaluator and viewpoint reference. State the evaluator or policy and, independently, either one
U.ViewpointRefornone. A non-nonereference SHALL resolve to one exact viewpoint episteme under E.17.0; neither the reference nor its resolution is the evaluator. -
Normal form and result boundary. State whether the ascription uses
SignalPack,Characteristic,Bundle, orObjective. If separately performed assessment work produced a result claim, cite that exact C.2.1 episteme throughqualityResultClaimRef; do not identify the work, result, bearer, or transitional record with one another. -
ClaimScope, selected slices, and time. State one
U.ClaimScopeand its exactU.ContextSlicemembership when the members matter. StateΓ_timewhen omission changes meaning.U.WorkScopeandU.PublicationScoperemain with their own work or publication claims rather than substituting for this claim scope. Keep freshness, qualification, and evidence-decay windows with the evidence, capability, or currentness claims they qualify, instead of treating them as quality. -
Reference plane when relevant. Name the plane when the same trigger phrase could concern the EntityOfConcern being described, its description, a carrier, or a publication face.
-
Representation language and substrate when relevant. When it matters to this use, identify the representation language of the selected expression of the quality-term ascription. If the use depends on reading a source expression or a separately obtained result claim, identify the language of that expression in its own right. Retain the effective interpretation rules of the ascription. A viewpoint-specific decoding convention or an embodied-kinesthetic, latent-distributed, symbolic-local or hybrid substrate may also matter. Name each when omission changes interpretation.
-
Witnesses, evidence use, and empirical grounding when relied on. Name the witnesses relied on for this use. If an evidence-provenance path is relied on, use A.10 for its descriptive account and cite each obtaining direct relation under the pattern that defines it. When empirical grounding is at issue, independently cite an obtaining
EpistemeEmpiricalGroundingRelation, or statenone; witness or record presence does not create that relation. -
Cross-local and endpoint boundaries. Cite an exact F.9 Bridge occurrence and bounded-use claim only when they independently exist. Cite a Card only when that optional package exists, and cite an F.9.1 stance note only when its
EntityOfConcernis that claim. State the endpoint pattern or endpoint source relation, the admissible use, and necessary applicability or stop conditions. Add an explanatory non-admissible-use guard only under F.19:4's independent-ground, plausible-reader, contribution, and smallest-clear-correction test.
Normal-form discipline
A QualitySense SHALL declare one admissible default normal form and MAY declare additional admissible normal forms explicitly.
The normal forms in this section express the recovered quality sense. They are not publication forms by themselves. Publication face, publication form, publication unit, carrier, rendering, export, and front-end questions remain with E.17, E.8, or the applicable endpoint-publication pattern.
QNF-1 - SignalPack.
Use for QS.PhenomenalCharacter, QS.PreconceptualFit, and many cases of QS.LatentFit.
A conforming SignalPack contains:
- exemplar or contrast set or probe set,
- articulation notes,
- source episode, carrier, and observer,
- optional ordinal or thresholded summaries,
- the signal use it supports and the additional declaration needed before treating a proxy as a
Characteristic.
QNF-2 - Characteristic.
Use only when the sense is one measurable characteristic on one declared scale.
This uses A.17, A.18, and C.16 and inherits full scale legality.
QNF-3 - Bundle.
Use when the sense is composite.
Typical for QS.ExplanatoryMerit, many engineering quality families, and QS.ControlAdequacy.
A conforming bundle contains:
- member heads,
- whether each head is Characteristic, status, mechanism, scope, or test,
- aggregation policy if any,
- prohibition on hidden scalarisation.
Engineering note.
For engineering -ility families, the preferred bundle endpoint is Q-Bundle (C.25), because it keeps Measures[CHR] distinct from ClaimScope and WorkScope and from Mechanisms and Status.
Q-Bundle is a C.25-governed bundle endpoint rather than a fifth normal form beside SignalPack | Characteristic | Bundle | Objective.
Do not use a free-floating bundle with hidden metric semantics.
QNF-4 - Objective.
Use for QS.UseValue in selection, generation, or search contexts.
A conforming objective contains:
- CG-frame or objective endpoint source reference,
- admissible comparators,
- acceptance or selector policy,
- reference plane and window,
- relation to novelty, diversity, and constraints.
Functional vs quality-family discipline
C.16.Q SHALL prevent the collapse of function or capability claims into quality-family claims.
- A statement about what a system does uses
A.6.Ffirst when function-like wording hides the FPF kind, relation, or claim, then applies the pattern for the recovered capability, Method, Work, system-role kind or assignment,A.6.Mmodule-interface, architecture, mathematical, evidence, assurance, gate, decision, or release claim. - Choose the sense of how well by the bearer and the evaluation. Maintaining the controlled quantity under a stated class of disturbances over a given horizon uses
QS.ControlAdequacy. Evaluating a system's maintainability under stated service conditions usesQS.EngineeringQualityFamily. - “Quality characteristic” and “functional characteristic” SHALL NOT be used as interchangeable labels.
- In engineering contexts,
-ilitynames are quality-family labels, not automatically Characteristics. They become admissible only as one explicitU.Characteristicor one explicitBundle(preferably expressed throughQ-Bundlewhen composite). - Cross-references are allowed; category collapse is not.
Local repair stances and cross-local Bridge discipline
Within one exact <ReferenceScheme, LocalSenseClaim> interpretation basis, lexical restoration may choose a local sense or rename without asserting an F.9 Bridge. When a use depends on semantic correspondence between two quality senses with different interpretation bases, resolve both exact F.17 SchemeSenseCell values and test the direct F.9 Bridge predicate. Source-local description or separate citation of the two senses needs no such correspondence. Scheme difference, shared spelling, an analogy, a loss note, or a quality record establishes no Bridge.
If the Bridge obtains, cite its exact occurrence and state any proposed comparison, substitution, operationalization, or projection as a separate F.9 bounded-use claim. That claim names the direction, rule, tolerated loss, polarity, and effective ReferenceScheme. Apply A.10 or B.3 only for the reliance branch involved in the proposed use. A Bridge Card remains optional reusable packaging.
Add an F.9.1 stance note only when a short interpretive cue helps a reader understand that exact bounded-use claim. The note is a separate C.2.1 episteme whose EntityOfConcern is the claim. Its optional label may be, for example:
localRename— read this use as near-renaming within its declared local boundary; do not infer cross-local identity.operationalizes— read the receiving expression as a procedural or measurable aid for this use; do not infer work, implementation, permission, or suitability beyond the cited claim.partialAnalogy— read the stated correspondence as partial; do not infer substitution.projection— read this use as a deliberate reduction of the source reading; the F.9 claim still carries its rule and tolerated loss.nonEquivalent— treat this as a warning against equivalence and silent substitution; the label alone asserts neitherDisjoint, negative polarity, nor an evidence score.
These tokens are optional reading labels inside a stance note. They are not Bridge kinds, direct relations, result claims, or substitutes for the Bridge, bounded-use claim, evidence, or loss account.
Examples:
QS.PreconceptualFitandQS.LatentFitare usually only candidates for partial correspondence. If the proposed use depends on correspondence between their exact cross-local F.17 cells, test an F.9 kind such asPartial-overlap; an optionalpartialAnalogynote may help read the resulting bounded-use claim but cannot establish identity.- A progression from
QS.PreconceptualFittoQS.PhenomenalCharacterneeds its exact direct relation or bounded-use account; shared articulation history does not make the senses identical. - A felt cue can prompt a choice of engineering measure without asserting that the measure represents that cue. If the claim instead proposes a cross-local operationalization or projection, establish its Bridge, bounded-use rule and tolerated loss separately from the measurement result. Add a stance note only if it improves the reading.
- An engineering measure can be selected as an objective under a declared CG-frame. That use does not identify the whole engineering family with selector value. When it depends on a cross-local correspondence, state the F.9 use and any demonstrated loss; selecting an objective alone establishes no loss.
- An obtaining F.9 Bridge between
QS.ExplanatoryMeritandQS.UseValuedoes not by itself establish identity. An F.9.1nonEquivalentnote may help read an existing bounded-use claim but cannot replace the Bridge finding or claim polarity. - Pirsig-style dynamic quality may locally cue
QS.PreconceptualFitor sometimesQS.LatentFit. Within one exact interpretation basis this may be a local rename; a proposed correspondence across bases needs exact F.17 cells and F.9 treatment. A source-local account can retain its own meaning. The label alone supplies neither identity nor empirical grounding. - Pirsig-style static quality usually cues a
CharacteristicorBundlepublication under another declared sense; it is not identical with dynamic quality. QS.ArchitecturalDescriptionFitnessandQS.EngineeringQualityFamilyhave different bearer lanes. Any cross-local correspondence must keep the exact description-side and system-side cells, Bridge occurrence, bounded-use claim, and losses separate and must name which description-fitness heads, if any, are proposed to proxy which system-side characteristics.
Change lexicon
Changes to a transitional ascription SHALL preserve the distinctions in this change lexicon. Ordinary wording is sufficient when it makes the changed value or reference clear; the tokens below can support a reusable record:
declareQualityTermAscription(...)— create a new explicit quality-ascription record.withdrawQualityTermAscription(...)— retire a prior record.retargetBearer(...)— retarget the bearer ref or tuple while keeping the repair-form schema.reviseSense(...)— change the value in thequalitySenseslot.reArticulate(...)— changearticulationModewhile preserving the sense family.reProxy(...)— change proxy, probe, or operationalization details.reBundle(...)— change bundle members or aggregation policy.reScale(...)— change characteristic scale or scale type.reProbeOrModelFrame(...)— change the exact domain-local probe or model frame.reComparisonFrame(...)— change the independently governed A.19.CPM comparison configuration.retargetEvaluator(...)— change the evaluator or policy ref without changing the viewpoint by implication.retargetViewpointRef(...)— retarget the governedU.ViewpointRef; resolution yields another exact viewpoint episteme only when the new reference resolves.reReferenceScheme(...)— change the effective ReferenceScheme explicitly; because that changes interpretation, re-check C.2.1 identity for any published claim episteme.rescopeClaim(...)— changeU.ClaimScopeor its exactU.ContextSlicemembers.retime(...)— changeΓ_time.refreshWitnessRefs(...)— refresh witness bindings without silently changing an evidence-provenance path or grounding relation.replaceEvidenceProvenancePath(...)— replace the cited A.10 path of exact direct relations; replacing the path does not by itself constitute a quality result.replaceEmpiricalGroundingRelationRef(...)— cite another independently obtaining C.2.1 grounding occurrence; a record edit cannot make it obtain.retargetBridgeOccurrenceRef(...)— retarget an exact F.9 occurrence ref; it does not retarget a bounded-use claim, optional Bridge Card, or optional stance note by implication.exitQualityAscription(...)— end use of the quality-ascription form and continue with the pattern for the recovered non-quality claim; never silently retype the old record.
A silent sense rewrite is a breaking semantic change.
If the record no longer expresses a quality ascription, close it with exitQualityAscription(...) and publish the recovered claim in the form needed for its use.
A.6.P rewrite note.
retargetBearer(...) is the family-specific form of retargetParticipant(BearerSlot, …). It, retargetEvaluator(...), retargetViewpointRef(...), and retargetBridgeOccurrenceRef(...) are reference-retargeting moves and SHALL preserve the A.6.5 distinction between a reference and the object it resolves. reviseSense(...), reArticulate(...), reProxy(...), reBundle(...), reScale(...), reProbeOrModelFrame(...), and reComparisonFrame(...) refine reviseByValue(...). reReferenceScheme(...) and rescopeClaim(...) change interpretation-bearing values and require an identity check for any published C.2.1 episteme. Witness, evidence-path, result-claim, grounding-relation, Bridge, bounded-use-claim, Card, and stance-note refs change independently; no edit silently rewrites another.
A.6.B boundary classification template for quality-term repair
When a repaired quality statement carries a boundary claim, preserve its A.6.B classification. A local sentence with one recoverable logical job needs no separate Claim Register or publication form. For a reusable boundary account, distinguish:
- L —
qualityTermAscriptionrepair-form skeleton,QualitySensesemantics, normal-form admissibility, cross-local routing, and the rule that any F.9.1 stance note remains a separate optional episteme about an already constituted bounded-use claim; - A — admissibility conditions for using the ascription in selector, gating, and publication lanes (required qualifiers, witnesses, thresholds, qualification windows);
- D — publication requirements (lexical firewall, mandatory rewrites, publication duties);
- E — carrier-anchored evidence and work effects (measurements, traces, critique sheets, probe packs, selector logs).
Where this family is published as a reusable boundary publication, stable L-Q*, A-Q*, D-Q*, and E-Q* claim ids SHOULD be published (or the reused L/A/D/E-classified claim set should be cited by location), and paraphrase drift across quadrants SHALL be avoided.
Do not let the bare word quality carry L/A/D/E claim by itself.
Lexical guardrails
In Tech and normative prose:
-
a consequentially ambiguous quality claim MUST resolve to its known endpoint or, when needed, an explicit transitional
QualitySense; -
high-quality, low-quality, quality metric, quality score, quality requirement, model quality, architecture quality, and solution quality cue inspection of the intended claim;
-
quality characteristic MAY appear only as:
- a source-local term whose cited standard or tradition supplies its meaning, with F.9 used only for a needed semantic correspondence, or
- a family label immediately rewritten into one explicit
U.CharacteristicorQ-Bundle;
-
a claim using quality requirement or quality requirements SHALL make the required head and its requirement or acceptance relation recoverable. Use the applicable rule for that claim; a named Characteristic, Q-Bundle head or objective can supply the evaluated content. The phrase itself establishes no requirement or acceptance result;
-
architecture quality or architectural quality MUST NOT appear without an explicit bearer lane (
EntityOfConcern being described,descriptionor anotherepisteme, publication face, or carrier when the carrier itself is described or evaluated) and, when omission changes meaning, an explicitreferencePlane; -
when quality denotes a QD or NQD selector's Q head, use
QS.UseValueby default. For wording about the reliability of a recorded input, recover what the recording procedure established; -
preconceptual uses MUST NOT be presented as if they were already Characteristics;
-
latent and distributed fit MUST NOT be presented as if it were automatically explanatory merit;
-
if the occurrence is primarily action-invitation talk, the text MUST NOT assign a
QualitySense; use the recovered subject rule directly, orA.6.Awhile the action meaning remains unclear, with source-traditionaffordancewording retaining its cited meaning when used; -
scope words (applicability, envelope, generality, validity) MUST NOT be used as hidden substitutes for
U.ClaimScope,U.WorkScope,U.PublicationScope, or another exact governed scope; -
quoted metalinguistic uses of the token quality are allowed, but SHALL be marked as token-under-discussion, not as a boundary-bearing term.
Progressive elaboration
A transitional ascription can be elaborated as its interpretation becomes clearer. Changed grounds can also require revising or withdrawing earlier content; elaboration is not necessarily monotone:
- Select a
QualitySenseand retain rival candidates while the intended sense remains unresolved. - Name the exact bearer, effective ReferenceScheme,
U.ClaimScope, and any meaning-changingΓ_time, reference plane, language of the selected expression, or substrate. - Name the probe or model frame and record any comparison it makes. Give the separate A.19.CPM frame when comparing admitted profiles under a declared comparator, or
noneotherwise; then name evaluator andU.ViewpointRefindependently. - Choose an admissible normal form and identify any separately constituted quality-result claim.
- Add exemplars, probes, characteristic heads, bundle members, objective pins, witness refs, and exact A.10 evidence-provenance paths as needed. Cite empirical grounding only through an independently obtaining relation.
- If the use depends on cross-local correspondence, resolve exact F.17 cells, the obtaining F.9 Bridge, and the separate bounded-use claim. Add a Card only as optional packaging and an F.9.1 stance note only as optional reader help about that claim.
- Preserve any boundary claim and its material dependencies under A.6.B; add stable hooks only when the receiving use needs them.
- Never move between sense families, frames, schemes, scopes, result claims, or neighboring relations silently.
Archetypal Grounding
Tell
If a draft uses quality for an FPF-governed claim without a recoverable sense, make that sense explicit.
A completed ordinary repair. An architecture team is deciding which edition of a description supplies the dependency information it needs. Its report says “architecture quality improved.” Both describe the same system and serve the same three decision questions. AD1 supplies the required dependency information for two questions; AD2 supplies it for all three. The same coverage rule is applied to both descriptions. The supported statement is: “AD2 supplies the dependency information for all three declared decision questions; AD1 supplied it for two.” This can finish the wording repair. If the commission instead concerns the system's availability or security, report this finding about the description and identify the evidence still needed for the system claim. If the decision questions change, compare both descriptions against the new question set before repeating the improvement claim.
A conforming rewrite publishes either the form for one known endpoint or one explicit qualityTermAscription(...) transitional record with bearer, one QualitySense, effective ReferenceScheme, its probe/model frame and its A.19.CPM frame or none, evaluator and U.ViewpointRef, ClaimScope, admissible normal form, endpointPatternLocator or endpoint source relation, and explicit boundaries among result claim, witnesses, evidence use, empirical grounding, Bridge, bounded-use claim, optional Card, and optional stance note.
Show (Latent fit and control adequacy)
The following pairs make alternative interpretations of each draft inspectable. Use these transitional ascriptions only while selecting the intended reading; once it is known, publish its endpoint form directly.
The examples stipulate their named objects and source relations. Each non-none comparisonFrameRef identifies an A.19.CPM comparison of two admitted profiles under an applicable comparator. Each non-none viewpointRef resolves one E.17.0 viewpoint episteme.
A named result claim is not assessment work, witness refs do not establish an A.10 evidence-provenance path, and neither witnesses nor a result label establish the grounding relation cited beside them. In the model and architecture improvement examples, the cited result claim asserts the improvement under its named comparison frame; the ascription record refers to that claim.
Draft: “The model quality improved.”
Repair A — latent representation line
Here EGR_LatentFitResult_22 denotes a separately established relation between the exact result episteme and exact grounding holon under the governed probe or measurement relations. The run and card alone would not establish it.
Repair B — closed-loop control line
Use this reading if the team's question concerns the policy and model operating together. The same model can produce different control outcomes with a different policy. A supported improvement of the tested pair can settle a choice between the tested deployments. To attribute that improvement to the model, establish its contribution under comparable policies and conditions or use other evidence that distinguishes the contributions.
ClosedLoopCriterionBundleDefinition_5 identifies the control criteria applied over Horizon_H in EnvClass_E. The comparison with the baseline uses the admitted controller profiles and comparator identified by ControlBaselineComparison_CF5.
Show (Preconceptual fit and explanatory merit)
Draft: “Quality matters before definition.”
First recover what remains undefined and how quality is said to affect the work. For one constructed reading, suppose EpisodeNotes_3 states that participants noticed a felt mismatch, used it to revise the problem framing, and only then articulated the criterion for that mismatch. The ordinary repair is: “During PF3, the participants used a felt mismatch to revise the problem framing before they stated the criterion for that mismatch.” Once this is the intended account, the sentence can finish the repair without a transitional record. The records below show partial sense resolutions while the intended claim remains open.
Partial interpretation A — preconceptual fit
Here ExemplarPack_EP3 records the episodes the participants compared and what felt different to them. That is enough to discuss the change in problem framing. The episode notes witness the articulation. No profile comparison or empirical measurement was made.
Partial interpretation B — explanatory merit
In CritiqueSheet_14 the panel compares explanations N5 and N4 against the counterexamples. N5 accounts for the response after the input was removed, which N4 predicts should end immediately. This finding concerns that response under the stated conditions. CriticismBundleDefinitionSource_CB4 cites the supplied definition of the explanatory-merit Bundle. Its criticism criteria determine that endpoint; the shared word “Bundle” does not make C.25 the governor of explanatory merit.
This record identifies an explanatory-merit reading. What is to be defined, how that merit affects the work, and in what sense it precedes the definition remain unresolved until the criticism account supplies them. Preserve those parts of the draft; the sense ascription alone does not complete its repair.
Show (System quality and architecture-description fitness)
Draft: “The architecture quality improved.”
Repair A — quality of the system-side bearer
Repair B — quality of the architecture description
ArchitectureDescriptionProbeFrame_AD12 is one project-local probe frame: it may cite DecisionQuestionSet_DQ7, an architecture-description result under C.30.AD, structural-view adequacy under C.30.ASV, and the retained U.ViewpointRef members resolved from a constituted E.17.1 catalogue. It is neither a viewpoint-family value nor a substitute for the selected viewpoint.
DescriptionFitnessDefinitionSource_AD12 cites the supplied definition of the description-side Bundle, including decision-question coverage and correspondence soundness. C.25 supplies a Q-Bundle only when its engineering-quality conditions apply; the architecture-description and viewpoint patterns supply their own checks.
The shared evaluator does not collapse the two repairs: their bearers, schemes, probe/model frames, scopes, viewpoint references, result claims, and evidence paths differ.
Show (QD or selector lane)
Draft: “Quality in our QD loop.”
Suppose the selector's operator asks how Candidate_7 was valued under its objective. That selects the candidate as the bearer in the example. If the operator instead asks whether the archive covers the required behaviours or the search produced better candidates, identify the archive or search and the coverage or comparison that would answer that question (C.18). One candidate's objective value can contribute to such a conclusion when the relation is established.
Repair
ObjectiveDefinitionSource_CG9 cites the objective definition supplied by CG_Frame_9. C.17 can characterize the candidate's value under that objective; it does not supply the objective's definition merely by being named.
Bias-Annotation
The following biases apply to overloaded evaluative uses of quality in FPF-governed wording.
- Gov bias: this pattern favors explicit evaluative publication and explicit L/A/D/E hooks, which improves auditability but adds drafting overhead.
- Arch bias: a transitional record can make unresolved alternatives reusable, but its fields can burden a question already settled by ordinary prose or a known endpoint. Use the lighter sufficient result first.
- Onto-Epist bias: this pattern distinguishes preconceptual, latent, explanatory, engineering, and selector senses; doing so requires extra lexical work.
- Prag bias: this pattern defaults QD and NQD uses toward
UseValue, which improves selector clarity but can feel narrower than colloquial “quality”. - Did bias: this pattern uses repeated rewrites for teaching; the risk is over-formalizing early exploratory language.
Conformance Checklist (CC-C16Q)
Apply the following conditions to the branch used. A known endpoint follows its own form. Requirements for ascription fields and records apply when the transitional form is used; the semantic distinctions remain relevant wherever they change the claim.
- CC-C16Q-1 - Explicit endpoint classification and explicit sense.
Every repaired claim makes its subject, intended sense and governing conditions recoverable. Use a known endpoint directly. A transitional
qualityTermAscription(...)has an explicitQualitySenseand endpoint classification; an adequate ordinary explanation needs no additional record. - CC-C16Q-2 - Exact bearer and arity. The bearer designator or tuple is explicit; description, carrier, evaluator, viewpoint, work, and result are not substituted for it.
- CC-C16Q-3 - Exact probe/model and comparison frames.
Identify the domain procedure and the basis for any comparison made there. Give the A.19.CPM frame for a comparison of admitted profiles under its comparator, and
nonewhen that operation is absent. A missing profile or comparator leaves a claim that requires the operation unresolved. - CC-C16Q-4 - Effective scheme, evaluator, and viewpoint reference.
The effective reference scheme is explicit. Evaluator and
U.ViewpointRefare separate; a non-nonereference resolves one exact viewpoint episteme and grants no conformance, membership, authority, or result. - CC-C16Q-5 - Substrate and referencePlane are declared when relevant.
Cross-talk across articulation modes (including preconceptual), representation substrates (including latent-distributed and symbolic-local), or
ReferencePlanevaluesworld,concept, andepistemeis not allowed without explicit substrate and, when omission would change meaning, plane declarations. - CC-C16Q-6 - ClaimScope, slices, and
Γ_timeare explicit. OneU.ClaimScope, its meaning-changingU.ContextSlicemembers, and any meaning-changingΓ_timeare stated; work or publication scope does not substitute for claim scope. - CC-C16Q-7 - Admissible normal form and result boundary.
The ascription uses
SignalPack,Characteristic,Bundle, orObjectivewith the corresponding normal-form discipline; any checked object, assessment work, result claim, witnesses, evidence-provenance path, and empirical-grounding relation remain independently identified. - CC-C16Q-8 - No illegal scalarization. Composite senses are not collapsed into one score without an explicit admissible scoring and comparison method.
- CC-C16Q-9 - No silent sense rewrite. Any semantic change preserves the distinctions of the declared change lexicon; changing sense, scheme, frame, scope, or neighboring relation silently is forbidden.
- CC-C16Q-10 - QD default.
When quality denotes the selector's Q head, use
QS.UseValueunless an explicit interpretation applies. Identify the bearer of a claim about the quality of an input, a search procedure or its record. - CC-C16Q-11 - Engineering family discipline.
Engineering
-ilityuses resolve to one explicitU.Characteristicor one explicitBundle, preferably aQ-Bundlewhen composite; they do not remain free-floating adjectives. - CC-C16Q-12 - Functional separation. Function or capability claims remain distinct from quality-family claims.
- CC-C16Q-13 - Bridge accountability.
When cross-local comparison requires semantic correspondence, it resolves exact F.17 cells and cites an obtaining F.9 Bridge plus the exact bounded-use claim when a use is proposed. Source-local description or independent citation of different senses does not require that Bridge. Any optional Card and F.9.1 stance note remain separate; the stance note's
EntityOfConcernis that claim. A stance word,CL, shared label, or loss note establishes none of them. - CC-C16Q-14 - Boundary-claim hook when needed. For a repaired ascription used in a boundary decision, the relevant claims and material dependencies remain recoverable under A.6.B. A separate register or publication form is needed only under its receiving-use conditions.
- CC-C16Q-15 - Lexical firewall. The intended sense is recoverable in Tech and normative claims; already adequate endpoint wording and marked metalinguistic discussion need no further ascription.
- CC-C16Q-16 - Transitional skeleton is complete.
The published skeleton carries bearer position and bearer-kind mismatch repair, sense, effective scheme, exact frames, evaluator,
U.ViewpointRef, ClaimScope, qualifier expectations, normal form, result, witness/evidence/grounding discipline, admissible change classes, and cross-local boundaries without minting universal context, frame, evidence, or grounding kinds. - CC-C16Q-17 - Consequential alternatives remain inspectable.
If sense selection, bearer facet, or the distinction between the EntityOfConcern being described or evaluated, a description or other
episteme, a publication face, and a carrier described or evaluated in its own right is non-obvious, the remaining alternatives are made inspectable before a decision or publication relies on their resolution. An adequate existing explanation can supply this; a separate Candidate-Set Note is optional. - CC-C16Q-18 - Reference resolution is not object substitution. Designators, governed refs, their resolved viewpoint or bearer objects, evaluator, result, frame, scope, grounding holon, and any selected structure remain distinct.
- CC-C16Q-19 - Change verbs dock cleanly with A.6.P and A.6.5.
retargetBearer(...)and the other declared reference moves are used only for ref retargeting; by-value revisions remain distinct from reference changes; a scheme or scope change triggers claim-identity review; edits to witnesses, evidence paths, grounding, Bridge, bounded-use-claim, Card, or stance-note refs do not silently rewrite one another; and silent retyping is forbidden.
Common Anti-Patterns and How to Avoid Them
Consequences
Benefits. The pattern makes a quality statement’s meaning and grounds inspectable across phenomenology, engineering, and search and selection contexts. A transitional ascription keeps consequential alternatives inspectable. Once their meaning is resolved, state the claim in the form supplied by the applicable endpoint.
Trade-offs and mitigations. The pattern adds drafting overhead and can feel heavy in exploratory notes. Mitigation: use an adequate ordinary explanation or known endpoint directly. Restore a missing distinction before relying on it in an evaluation or further claim; use of the Tech register alone does not require the full transitional form.
Rationale
C.16.Q distinguishes the senses of “quality”:
The word “quality” is not treated as one concept. It is treated as a family of quality-sense ascriptions whose members differ by substrate, articulation mode, bearer, effective scheme, probe/model configuration, comparison configuration, ClaimScope, and admissible normal form.
This lets FPF discuss:
- Pirsig-like preconceptual fit,
- representation-learning and neuro-symbolic latent fit,
- explanation quality in criticism-driven inquiry,
- architecture-description fitness under a viewpoint,
- engineering quality families,
- use-value in open-ended evolution,
- control adequacy in action loops,
without forcing them into one universal scalar.
It also makes the distributed-vs-local issue explicit:
- some senses originate in embodied or latent-distributed substrates,
- some are only publishable as symbolic-local CHR, bundle, and objective forms,
- and some require an explicit projection from the first into the second.
It also makes the bearer and plane issue explicit:
- some uses evaluate the EntityOfConcern being described,
- some evaluate its description under a viewpoint,
- some evaluate a carrier or publication face,
- and those uses must not be collapsed without an explicit bearer lane and, when needed, a declared
referencePlane.
Semantic drift usually starts at these bearer and plane distinctions; C.16.Q makes the interpretation choices inspectable.
SoTA-Echoing
The source contributions below address different quality questions. Their common use here is to recover the subject, intended sense and supported conclusion while retaining the distinctions among traditions. Recover the criteria when an evaluation is made.
Scale legality. A numerical quality claim SHALL use the Scale and comparison or aggregation rule required by its meaning. A comparison identifies its comparator set or applicable CG-Spec; a composite score states its admissible aggregation. Neither quality wording nor these source correspondences licenses covert scalarization of bundles, explanations or preconceptual signals.
Source-local and cross-local use. A vocabulary can be described or cited in its own sense. When a proposed use depends on semantic correspondence between different <ReferenceScheme, LocalSenseClaim> bases, resolve the exact F.17 cells and test F.9. The Bridge and bounded-use claim, direction, rule, tolerated loss, polarity, any evidence reliance and any cross-plane representation relation remain separately established. A stance label or loss note supplies none of those relations by itself.
Earlier touchstones such as Pirsig, Popper and Deutsch remain useful for historical lineage and local terminology. Their concepts retain their source meanings. The shared word “quality” supplies a question to clarify, not a general identity across the subjects and practices.
Relations
-
Lives in: C.16 characterization pattern nest as the quality-term realization of E.10.ARCH and C.16.P.
-
Builds on: E.10.ARCH for shared wording-use restoration architecture; C.16.P for characteristic and scale exits; A.2.6 for explicit scope and
Γ_time; A.17, A.18, and C.16 for admissible measurable characteristics; C.25 for engineeringQ-Bundlepublication. -
Coordinates with:
- A.6.P when recovered content is relation construction rather than evaluative characterization;
- A.6.A to recover an ambiguous action-oriented claim or question, then its subject pattern;
- C.2.2a, A.16, A.16.1, A.16.2, B.4.1, and B.5.2.0 for language-state positions, early cues, next-use docking, and retreat or reopen; use A.16.0 only when lineage, branch, loss, or an actual responsibility-handoff history itself must be published as an explicit trajectory account;
- C.2.LS, C.2.4, C.2.5, C.2.6, and C.2.7 for language-state facets;
- C.2.1 for effective ReferenceScheme, exact result-episteme identity, and optional
EpistemeEmpiricalGroundingRelation; - A.2.6 for
U.ClaimScopeandU.ContextSlice; - A.19.CPM for a comparison of admitted profiles under its comparator;
- A.10 for evidence-provenance and bounded reliance;
- C.17, C.18, and C.19 for selector value, novelty, diversity, and policy;
- E.17.0 and E.17.2 for exact viewpoint epistemes and
U.ViewpointRef; C.30.AD and C.30.ASV for architecture-description and structural-view use; - F.9 for exact cross-local Bridge occurrences and bounded-use claims; F.9.1 only for separate optional stance notes about those claims;
- A.6.B when repaired ascriptions become boundary-bearing.
-
Further distinctions: C.3.3 governs any needed comparison or bridge between distinct endpoint kinds; A.19.CN governs questions of comparability; C.2.P recovers a hidden source or publication use. These questions arise from the claim, not from the shared quality label.
-
Publishes vocabulary through: E.10, F.17, and F.18 when the
qualityTermAscriptionrepair-form skeleton, theQualitySensestarter set, and the red-flag rewrites become stable shared vocabulary.
Endpoint-first use
This pattern uses endpoint-first assignment rather than universal governance of all quality language. qualityTermAscription(...) remains useful as a transitional repair form, but it is not the required resting place or durable local record kind for every repaired use of quality.
Explicit endpoint selection
Admissible endpoints after repair include:
- a single
Characteristic, - a
Q-Bundle, - an
E.21 PatternQualityQBundle, when the bearer is one FPF pattern version and the evaluative claim asks whether the pattern is good enough for a declared reader, use, and scope. In this case, do not assign the claim to generalC.25unless the bearer is a non-pattern engineering quality family, - an
Objective, - an explanatory-merit bundle,
- a selector-value endpoint.
Quality wording whose intended claim is already recoverable can use the applicable endpoint directly. A known endpoint needs no additional qualityTermAscription(...) record.
What C.16.Q leaves to other patterns
C.16.Q does not define or test articulation-state characteristics, Bridge truth, bounded-use claims, stance-note identity, evidence-provenance, empirical grounding, profile comparison operations, viewpoint resolution, or representation factors. Use A.16, C.2.LS, C.2.4, C.2.5, C.2.6, C.2.7, F.9, F.9.1, A.10, C.2.1, A.19.CPM, E.17.0, or the applicable representation pattern for those questions.
C.16.Q:End
Characterising Generative Novelty and Value
Type: Evaluation and measurement-use pattern; normative where stated.
Status: Stable
Use this when
Use C.17 to judge the novelty, usefulness or eligibility of a design, code change, theory, policy proposal, dated Work occurrence or finite candidate set. A claim about novelty names what the bearer differs from. A claim about usefulness names what it helps achieve. A claim about eligibility names the conditions it must meet.
If the claim arrives as learning progress, learned novelty, or information gain described as learning, and the bearer or result is still hidden, apply E.10.LRN and the direct result owner first. Enter C.17 only after identifying the bearer and the novelty, use, surprise, creativity, or other characterization question. Stop at the direct result when there is no characterization question to answer.
Begin with the smallest useful answer:
- identify the bearer being discussed;
- for a claim of novelty, identify what the bearer is compared with.
- for a claim of usefulness or eligibility, identify the objective or applicable must-criterion.
- state what the evidence supports for that claim and where its support ends.
Stop when the qualitative answer meets the intended use. Another team can reuse that answer with its evidence, comparison basis or criterion, and limits. Check that those grounds still apply to its question.
Develop a coordinate when the receiver needs its value. A direct qualitative comparison can answer which features differ between two designs. Prepare profiles when the receiving use needs to compare their coordinates under a declared comparator, for example to test dominance across criteria. For later inspection, preserve the grounds for the claim actually made.
What changes in practice. A team can replace an unqualified creativity label with the comparison basis, practical criterion, supported difference, consequence, uncertainty, and stopping point. When numbers matter, it can also reproduce the complete result chain.
Not this pattern when. Use C.18 to generate candidates or maintain an Archive or Front, C.19 to change pool treatment, G.5 to declare selector-facing set results, C.11 to make a choice, and A.13 to characterize agency or autonomy. Use C.17 to report characteristics and bounded conclusions, not to generate, retain, rank, approve, fund, enact, make a service promise, classify a person as creative, measure a person's creative capacity, organize a team, or prescribe a workflow.
Do not infer a person's or System's creative capacity from a C.17 result. Strong agency can still yield a weak result, while useful scaffolding can help produce a strong result; C.17 characterizes the bearer and evidence named in the claim.
The practitioner route
Qualitative first move
Name the bearer and the judgement the reader needs. For novelty, identify a reference set and the supported difference. When the comparison basis is yours to choose, ask whose prior designs, practices or results matter to this judgement and which differences would change the recipient’s use. An inventory can answer a question about its entries. A claim extending beyond it needs grounds that an omitted comparator or a distinction lost by the representation cannot overturn the conclusion. For usefulness, name the objective and the evidence for the practical effect. For eligibility, check the applicable must-criteria. State the limits of each conclusion. The following discussion needs all three judgements.
Compared with the admitted five-year pump-design set, P-22's inspected split-clamp arrangement is a supported difference. The cited assembly evidence supports shorter assembly than the current shop design P-11, and the design must not require new tooling. The inspection does not support claims about other pump families.
This is a C.17 result for an ordinary design discussion. It does not imply a numerical Novelty value or an overall assessment occurrence.
Obtain a coordinate or compare profiles
When the receiver needs a coordinate or a comparison of coordinate profiles under a declared comparator, use the steps that supply its grounds.
- select one A.19 CharacteristicSpace and its
A.19.ECSspecification; - for a corpus-relative claim, fix the finite comparison corpus, inclusion rule, source editions, scope, comparison window and evidence.
- for a usefulness or eligibility claim, identify the objective, acceptance criterion or applicable must-constraints.
- identify how each requested value is obtained and the inputs, assumptions or observations it requires.
- for each coordinate, cite an existing complete
C.16measurement result, perform and constitute the missing C.16 measurement, or state aC.2.1non-measurement ascription under its declared rule; - form only the aggregate conclusion needed by the receiving comparison;
- add an optional profile payload, representation, or record only when a named receiver needs it.
Return only the missing premise that blocks the current coordinate or conclusion. A missing measurement does not invalidate independent qualitative claims or other coordinates.
Dated overall-assessment branch
Use this branch when a statement deliberately asserts dated overall-assessment Work or the receiving use needs that occurrence established. Recover the evaluator System's A.13 core, including its obtaining A.2.1 assignment, and the admitted Method through A.3.1; then use A.15.1 to independently admit the dated Work that enacts that Method. Add F.6 only when the receiving use consumes precise assignment-bound attribution through that same assignment. A missing or failed F.6 attribution alone leaves independently admitted assessment Work intact; recheck any admission premise questioned by the same finding and the conclusions that depend on it. Only the evaluator System performs the Work. A separate MethodDescription may explain the reusable Method; it is not enacted. Coordinate-measurement Work remains its own C.16 Work, and the aggregate result states claims.
Do not infer an A.6.1 operation application from the Method, Work, configuration, or result. If a receiving claim separately asserts an exact operation application or binding, satisfy the current A.6.1 application account and cite that application. Otherwise retain only the C.17-local evaluator System, assignment, Method enactment, dated assessment Work, coordinate results, and aggregate result that the claim needs.
Keep the evaluation objects distinct
Changing a space slot, corpus membership or inclusion rule, model claims or training basis, objective, criterion, constraint, Scale meaning, scope, or window reopens only the coordinate claims that depend on that change and any aggregate conclusion that uses them.
Preserve earlier results
Preserve an earlier result under its original edition and evaluation basis. Relate an earlier and a later result as editions only when both results and the continuity claim are recoverable.
Evaluation configuration
Use the questions that govern the coordinate or comparison being made:
- Which bearer is evaluated, for which use and ClaimScope, over which selected slices and window?
- Which CharacteristicSpace and
A.19.ECSspecification supply the Characteristics, Scales, value meanings, missingness rules, and admissible comparisons? - Which finite corpus or reference set supplies the comparison basis, and how were its members admitted?
- Which Method, model, encoder, distance definition, calibration, and uncertainty basis produced or supports each value?
- If the Method compares a representation or observation instead of the bearer directly, which bearer and representation or observation are related, what describing, projection, measurement, or other stated relation supports the inference, and do the corpus members use a compatible comparison basis? If not, state the mapping and relevant loss.
- Which objective, acceptance criterion, and must-constraints are current, and which source epistemes state them? Their inclusion in the configuration does not make their claims true.
- If the result claims improvement or gain, which baseline and comparison or counterfactual Method make the difference testable?
- Which evidence supports the difference, consequence, coordinate, or comparison conclusion, and what use may rely on it?
For each objective, criterion, or constraint on which the result depends, keep its EntityOfConcern, effective ReferenceScheme, edition or currentness basis, and subject-defined predicate recoverable. Do not use a generic container label to answer several of these questions at once. Add only the source, scheme, scope, model-use, comparison, or evidence relation the current claim needs.
Keep prospective and observed readings distinct. A prospective reading may use a surrogate model and stated assumptions to compare designs before use; an observed reading uses later Work, service, or other outcome evidence. Do not overwrite the prediction as though it had been an observation. Preserve both results when the comparison matters, and reopen only the coordinates and conclusions that relied on the superseded prediction.
Core characteristics
Each selected characteristic has a declared Scale, polarity, admissible operations, missingness rule, and evidence route. An ordinal value is not averaged unless a declared model justifies an interval interpretation.
Novelty: unlike which admitted set?
Novelty describes supported difference from a finite comparison corpus under one declared similarity Method. When that Method returns a calibrated similarity result on [0,1] for each like-for-like comparison, that source result keeps its declared Similarity Scale. The following transformation defines a bounded value on a corresponding declared Novelty Scale, whose meaning is difference from the corpus and whose positive polarity increases as maximum similarity falls:
Novelty = 1 - max similarity(bearer, corpus member)
The maximum requires a nonempty set of comparable members. To report it for the whole declared corpus, establish the largest value, either directly or by bounding the remaining similarities by a value already attained. With partial coverage, report the supported bounds or identify the subset for which a value is known. An empty corpus needs a separately defined initialization convention if a value is required. A declared normalization to [0,1] may be part of the Method; state its source Scale and transformation. If the Method uses another similarity or distance Scale, state the lawful coordinate construction and resulting Scale instead of reusing this formula. Subtracting an unrestricted result from one does not make it bounded.
When the Method compares representations or observations rather than the evaluated bearers themselves, name each bearer and the value compared, together with the describing, projection, measurement, or other stated relation that lets the comparison support the bearer-level claim. Use a compatible basis for corpus members, or state the mapping and relevant loss. A direct like-for-like comparison of epistemes needs no extra representation relation.
A robust top-k variant is allowed when declared. The result identifies the Novelty Characteristic and Scale editions, corpus and inclusion rule, source editions, comparison window, Method, model or encoder edition, distance definition, invariances, calibration, uncertainty, ClaimScope, evidence, and intended use. Changing any load-bearing element creates a different comparison basis or result edition. Those identifiers make the result reproducible; they do not by themselves show that the value is robust. When the value materially affects a comparison or pool treatment, use diagnostics suited to the claim. For example, inspect the nearest corpus members and their distances, repeat the reading with a plausible alternative corpus or similarity Method and report the sensitivity, and remove a claimed invariance to see whether it materially changes the result. If no robustness check was performed, report the supported value and uncertainty without calling it robust.
Novelty is neither timeless originality nor a property detached from its comparison basis. A label such as Novelty@context is not an executable input and must not substitute for the result chain.
Use-Value: useful for which objective?
Use-Value, historically also called ValueGain, reports the bearer's supported usefulness or contribution to one declared objective or acceptance criterion. An ordinary usefulness statement may use an ordinal Scale such as Fail | Partial | Pass; it does not need a counterfactual merely because the bearer is useful for the stated purpose.
When the result claims an improvement or gain, identify both the baseline and the comparison or counterfactual Method that makes the difference meaningful. One bounded measured construction is ValueGain = metric_after - metric_before under a fixed metric and window. An A/B comparison, back-test, or causal-inference Method may provide the comparison when it fits the claim. Cite the before and after measurements and the Work or other evidence on which they rely. A predicted gain identifies its model, error, baseline, and intended later update. Without a baseline and an appropriate comparison Method, say what usefulness is supported; do not report an observed gain.
Use-Value may be one member of a declared Q-set, but it is not the whole Q-set by default. If it stays outside Q, name its use as a side condition or tie-breaker. Do not silently promote Novelty, Surprise, DeltaDiversity_P, or Illumination into dominance.
Surprise: unexpected under which model?
Surprise reports how improbable one declared sample of the bearer is under one generative model. For a discrete probability, a common raw result is -log p(sample) in bits or nats. State the modeled sample unit and encoding and how bearer size is handled. Compare bearers only under a justified common extent, a declared per-unit or code-length normalization, or another calibrated rule suited to the model. For a continuous model, identify the measure as well as the representation; a density value alone is not representation-independent. Otherwise keep the raw model result within its exact basis and do not treat it as a comparable Surprise coordinate. Also identify the model episteme and edition, training basis, preprocessing, fit and out-of-distribution checks, calibration, refresh condition, and limits.
Novelty and Surprise answer different questions. A bearer may be unlike the selected corpus yet unsurprising under a broad model, or close to known examples yet surprising under a narrow model. Keep both results visible when both matter.
ConstraintFit: which must-criteria hold?
ConstraintFit reports satisfaction of the declared must-constraints under their predicates and source epistemes. Use E.5, D.1-D.5, or a service-acceptance pattern only when it supplies the predicate or source for the current must-criterion. A ratio such as passed applicable must-constraints / all applicable must-constraints is allowed when the applicable set is nonempty and the count treats each condition equally. State separately which applicable conditions are unverified and which failed. A weighted result needs its own declared rule and weights. If no must-constraint applies, omit the ratio and state that fact when the receiving use needs it; other eligibility conditions remain to be satisfied.
A failing must-constraint makes the bearer ineligible for the affected use unless the receiving constraint or decision pattern recognizes an independently obtaining exception or waiver effect. A waiver speech act alone does not change eligibility. If no pattern defines the needed effect, return the missing relation rather than treating communication as authorization.
AttributionIntegrity
AttributionIntegrity reports how completely the applicable provenance, authorship, source, and licence duties are met. First identify the duty set and the source that makes each duty applicable. One bounded local construction is satisfied required duties / all applicable required duties on a declared ratio Scale; mark unresolved and inapplicable duties separately rather than treating them as satisfied. Provenance links, licence scans, and acknowledgements are example evidence routes.
When the applicable-duty set is empty, do not compute the ratio. Omit AttributionIntegrity when the receiving use does not need it; when that use needs an explicit disposition, return not applicable under the declared Scale and missingness rule. That disposition is not a pass and cannot by itself pass a filter, break a tie, establish legal adequacy, or satisfy a must-constraint. Keep it distinct from an unresolved duty that does apply.
This reading does not by itself establish legal adequacy. It is measurable but not in the default dominance set; an applicable policy may use it as a filter or tie-breaker. When a duty is a must-constraint, its pass or failure belongs in ConstraintFit and affects eligibility there.
EffortCost
EffortCost reports actual resource outlay through A.15.1 dated Work, B.1.6 resource aggregation, C.16 measurement, and A.10 evidence. Planned effort remains A.15.2 WorkPlan content. Use cost-normalized readings for planning or comparison only under a declared rule; cost is not itself creativity, and a profile does not carry operational actuals.
Retained-set and optional applied readings
Diversity and illumination
Diversity_P describes coverage or dispersion of one declared retained set under a named measurement policy. A local policy may, for example:
- take the average pairwise distance under one declared descriptor map, distance Method and Scale. Identify the pairs the statistic covers. The whole-set mean requires at least one pair and enough information to determine the sum over all eligible pairs. A mean calculated only over observed pairs measures that subset. Give its coverage.
- report how much of a declared feature partition is covered, using a stated covering radius or k-cover rule.
The result identifies the retained set and membership rule, measurement-policy and Scale editions, descriptor or feature source editions, distance or covering definition, comparison window, and evidence. A distance matrix or coverage map can show the calculation. When sensitivity to the kernel, distance definition, covering threshold or admitted-member set can change the receiving decision, use an adequate existing analysis or make the needed comparison and report its consequence. Keep a result on a fixed basis qualified by that basis; claim robustness only with grounds for it.
For a candidate h and retained set S, the same local policy may use the marginal reading DeltaDiversity_P, also written ΔDiversity_P:
DeltaDiversity_P(h | S) = Diversity_P(S plus h) - Diversity_P(S)
This difference requires both readings to be defined by the policy. For a starting set with no eligible pair, report the missing pairwise value or identify the policy’s declared initialization convention.
Illumination is a report over Diversity_P, such as a coverage map or QD-score summary. It is telemetry, not a primitive characteristic and not part of the default dominance set. Use C.18 to maintain an Archive or Front and C.19 to state any pool policy that uses these readings.
Optional retained-set readings include:
FamilyCoverage: coverage of locally defined families under a named policy and Scale;MinInterFamilyDistance: the smallest distance among declared families, with descriptor map, distance definition, Scale, and family-representation rule;AliasRisk: a near-duplicate or alias diagnostic with collision policy, descriptor source edition, and Scale;DescriptorVector: an optional descriptor payload whose dimensions and interpretation the same local policy declares.
These readings characterize the named set. They do not admit sources, select members, or widen applicability. For naming candidate sets, apply F.18's head-term-family anti-inflation rule.
Optional applied characteristics
The following are executable local examples. Use one only when the receiving question needs it and identify its bearer, rule, Scale, and evidence.
ReframeDelta. The bearer is an ordered pair of problem-frame epistemes. One local rule compares the earlier and later frame on an ordinal Scale such asNone | Local | BoundaryShift | Systemic; a boundary or scope diff and a changed causal map support the reading. The frame change does not by itself prove improvement, so state Use-Value separately.Compositionality. The bearer is the design or episteme being assessed. One local rule requires reuse of at least a declared number of components and evidence of at least one new relation among them; it may return a boolean plus a separately defined structure reading. Cite the component graph and component provenance.Transferability. The bearer is the design, result, or episteme whose use is tested in one named receiving setting. One local ordinal Scale isnot supported | supported with stated loss | supported for the stated use. Cite receiving-use pilot evidence and the preserved and lost meaning; use an F.9 Bridge only when a relation between different reference-scheme senses obtains.DiversityOfSearch. The bearer is a finite set of dated Work attempts. Count distinct approach classes under a declared local typology, optionally as a rate over a stated time window, and cite the tagged Work and typology. Cosmetic variants do not create new classes.Time-to-First-Viable. The bearer is one Work episode. Measure elapsed time from a declared start to the first dated result that passes the stated viability criterion; cite the timestamps and passing evidence. If no result passed, reportnot yet obtainedor a right-censored duration rather than the time to the first runnable output.Risk-BudgetedExperimentation. Compare the applicable WorkPlan with the resulting dated Work set. One local rule reports planned exploratory resource use divided by the allowed risk budget and the realized ratio separately, with any overrun visible. Cite the WorkPlan, actual Work, and resource evidence; the reading does not grant the budget or authorize the Work.
Readings about actual attempts, elapsed time, or realized experimentation depend on dated Work; planned experimentation depends on a WorkPlan.
Other domain characteristics
The six applied readings above are examples, not the extension boundary. A configuration may select other Characteristics already established for the current use—for example, time or cost to probe, evidence sufficiency, safety or ethical risk, option value, or regret risk. Keep each selected Characteristic's bearer, Scale, polarity, defining source, Method, and evidence. A safety or ethical must remains an eligibility condition through ConstraintFit; evidence sufficiency does not become creativity; and scope does not become another coordinate merely because every claim needs one.
When one comparison covers several components, attempts, or Work occurrences, identify the lawful aggregation separately for each Characteristic. For example, compatible costs may sum, all declared must-constraints may have to pass, a domain risk rule may use its own conservative combination, and evidence may be combined for a named assurance claim under B.3. If no declared aggregation supports the combined reading, keep the component results separate.
Any prior or default used for Novelty, evidence, risk, or another selected reading remains a separately supported model or policy claim with its source and edition.
Results, profiles, and assessment Work
Coordinate claims
For each selected characteristic, choose one route:
- cite an already constituted
C.16measurement-result episteme and its complete measurand, Characteristic, Scale, Method, dated measurement Work, measurand relation or A.6.1 binding, any further bindings required by the claim, time, uncertainty, and evidence chain; - if the current action measures the coordinate, constitute that complete C.16 chain before using the value;
- for a non-measurement result, state a
C.2.1ascription and show how the declared rule and inputs produce the value. A model calculation and a classification by a criterion can each support such a result.
A numeric model output, criterion label, dashboard cell, or formula alone is not a measurement result.
Aggregate result
CreativityEvaluationResult is one bounded C.2.1 episteme about the bearer. Its ClaimGraph cites the selected coordinate claims, CharacteristicSpace and specification, scope, use, window, evidence, current eligibility consequence, and any frontier or incomparability conclusion. It is not an arithmetic total and does not create a universal creativity kind.
The result may establish eligibility for one comparison or non-domination within one declared set. Two eligible bearers may each perform better on a different criterion. If a missing value could change their ordering, report which comparisons remain open. Choosing, approving, publishing, retaining or enacting a bearer requires the grounds for that action.
Optional payload, record, and rendering
Use CreativityProfile only as the local name for the optional non-arithmetic payload of coordinate-claim references, their arrangement, and current frontier or incomparability annotation. A table, chart, or dashboard is a separate representation. A CreativityEvaluationRecord is a separate optional episteme that packages references for a named receiver.
Do not alternate among result, profile, record, and dashboard as if they were synonyms.
Comparison, gates, and selection boundary
- Use the criterion the recipient actually applies. Among eligible candidates a novelty comparison can settle a preference for the most dissimilar one under the agreed rule. When approval also depends on usefulness or another eligibility condition, establish that condition separately.
- State the selected characteristic subset, polarities, eligibility conditions, and comparability basis of every dominance claim.
- Preserve partial orders and incomparability. A Pareto or constraint-bounded Front follows from the declared rule; a plotted hull does not by itself establish that relation.
- Do not force one scalar creativity score. If a receiving policy uses an index, publish its weights or curves, admissible transformations, uncertainty treatment, sensitivity, and drift rule while keeping the primitive coordinates queryable.
- Do not average ordinal Scales without an accepted model that supports the conversion.
- A result may state a frontier relation over a declared set. Use
C.18to maintain the current Front and Archive,C.19to state pool treatment and tie-break policy,G.5to declare selector-facing results, andC.11to make the choice.
Evidence, uncertainty, and resistance to gaming
Every quantitative result names its derivation or evidence and the calibration, uncertainty and validity window required by its construction and receiving use. Keep aleatory and epistemic uncertainty separate and state any rule that combines them. A claim imported from another source or scale states the preserved meaning, lost meaning, direction, receiving use, and evidence limit; use F.9 only when a Bridge between reference-scheme senses is needed.
Apply these anti-Goodhart guards:
- when a preference or approval uses novelty, establish each further condition that the governing rule requires.
- identify the editions of the corpus, encoder, model or Scale the result actually uses. A change to those grounds can require a new calculation.
- keep Illumination as telemetry unless an explicit policy promotes it for a named use;
- check delayed practical consequences, such as retention, maintenance burden, or cost-to-serve, before treating a proxy improvement as a practical gain;
- connect
DiversityOfSearchto the declared experimentation allowance and report overspend; - retain primitive coordinates and sensitivity when a composite index is used.
When a corpus, inclusion rule, model, encoder, Scale, criterion, window, or cross-source premise changes, leave a compact change account: the previous and new editions, what changed in the basis, which coordinates and aggregate conclusions reopen, whether eligibility or frontier membership changed, and any further calculation or observation needed to settle an unresolved difference. Latest is not a reproducible selector. Use B.4 and G.11 for the refresh; the change account explains the comparison and does not replace the new result.
Goodhart's law is a useful historical warning, not evidence for any result.
Evidence can support a result without establishing assurance. Ordinary use stops with proportionate evidence. Enter B.3 only for a named assurance claim.
When that receiving use requires independent assessment or segregation of duties, name the Systems that performed any bearer-producing, coordinate-measurement, or overall-assessment Work and the assignments relevant to the independence claim. State the relevant conflict or independence evidence. Do not impose this assurance arrangement on an ordinary qualitative discussion that makes no independence claim.
Cross-scale and cross-source limits
When a reading moves across scales, state which Characteristics are preserved, aggregated, projected, or lost and cite the aggregation, projection, transition, temporal cross-scale, or Bridge relation on which the reading depends. Keep a plain distortion note next to the projection. If the receiving use needs the named A.0 qualifier structure, use A.0:QF.2a and its optional OutcomeMapRef, TransitionRelationRef, or BridgeDistortionNote; otherwise the plain loss statement is enough.
Different projections of the same retained set or frontier may preserve different information. A useful projection is not automatically information-preserving.
When claims come from different source schemes or corpora, do not compare by matching labels. State the mapping, Bridge when required, direction, loss, calibration, and receiving use. A target-use pilot may improve evidence for that use; it does not make the source and target bases identical.
Worked cases
Pump design: stop early or open the measurement branch
Identify P-22 as the exact design episteme. Compared with the admitted five-year pump-design set, its inspected split-clamp arrangement is a supported difference. Assembly evidence supports shorter assembly than the current shop design P-11, and the candidate must not require new tooling. The cited comparison uses the same assembly task and conditions. State what was inspected and the limit of that support. A second shop asking the same question can use the supported conclusion when its assembly conditions and comparison basis match.
To use Novelty value 0.42, cite P22-NoveltyResult-4, an already constituted C.16 measurement-result episteme. Its chain identifies P-22 as the measurement subject, the Novelty Characteristic and Scale, exact similarity Method, the calibrated [0,1] CAD-graph similarity result used by the declared Novelty construction, encoder and model edition, uncertainty, dated measurement Work, actual bindings, time, and evidence. The Method compares P22-CADGraph-7, produced from and representing P-22 under CADGraphProjection-2, with graphs produced by the same projection for every corpus design; the result states that this projection omits surface finish and manufacturing tolerances. If the current action measures novelty, constitute that chain first. To resolve the tooling requirement, compare the proposed operations with the available tooling and record what that comparison establishes. It remains open while only the requirement is known.
A CreativityEvaluationResult can now report the supported difference and the assembly comparison with P-11. To establish eligibility, check each applicable must-condition. To place P-22 on a non-dominated front, use the declared comparison over eligible rivals (§7). The two results about P-22 alone leave its front membership open.
If an audit later asserts that an overall assessment occurred, identify the evaluator System, its exact assignment, PumpCreativityAssessment-17 Work, and PumpCreativityAssessmentMethod-2; state that the Work enacts the Method. The MethodDescription explains the Method, the result states claims, and coordinate-measurement Work stays separate. Do not add an A.6.1 operation application merely because those values are recorded. If the audit separately asserts an exact operation application or binding, satisfy the current A.6.1 application account and cite that application.
Software and algorithmic design
Software design. In this worked example, ETL-Parallel-12 is compared with 40 admitted internal pipeline designs through ASTGraphSimilarity-3, whose declared Novelty construction uses calibrated [0,1] similarities. The Method compares AST-graph representations produced by the same declared parser and projection for the evaluated design and every corpus design; the result states that runtime configuration and deployment topology are not preserved by that projection. Its Novelty result is 0.36.
A fixed-workload benchmark reports an 18% lower p95 latency than the serial baseline, but the segregation-of-duties test fails. The benchmark run, corpus edition, nearest-neighbour report, and policy test are the evidence.
The result therefore supports a latency gain but leaves the design ineligible for the stated use; if the candidate is revised for that use, repair the isolation boundary and repeat the affected tests before claiming that the revised design meets this constraint.
Algorithmic search Work. SpikeSet-7 contains nine dated attempts in three declared approach classes over six hours. Tagged Work records support DiversityOfSearch = 3 classes; the first runnable output appeared after 2 h 10 m. The held-out viability test was never run, so Time-to-First-Viable is not established. Run the test when the receiving work needs the viability result. The supported search-diversity reading remains available; time-to-first-runnable cannot substitute for time-to-first-viable.
Health analytics
For Cardio-Readmit-H4, the held-out AUROC is 0.79 against a 0.75 baseline under the frozen test set, clearing the declared uplift threshold of 0.03. The model card, held-out plot, and evaluation Work support that local Use-Value result. No receiving-use pilot has yet been performed at Hospital B, so Transferability there remains unsupported. Use the local result for the applicable pool or choice question, but leave the target-hospital claim open until pilot evidence exists. Use an F.9 Bridge only if the two hospitals' reference schemes require one.
Product reframing
OnboardingFrame-v1 treats onboarding as one completion task; OnboardingFrame-v2 separates job setup from obtaining the first result. The declared ReframeDelta rule returns BoundaryShift, supported by the frame diff and a simpler causal map. Suppose a four-week A/B trial assigns users at random to the existing one-task procedure or to a revised procedure with separate job setup and first-result activities. The revised group’s median time-to-value was 22% lower on the same outcome definition, clearing the 20% objective for this trial. The frame comparison establishes what changed in the model. The trial evaluates the procedures the groups used. Exploratory Work used 9 of the 12 allowed staff-days, so the realized risk-budget ratio is 0.75 with no overrun. The frame diff, A/B report, WorkPlan, and Work records support the result. This evidence can inform the later choice; it does not make the choice.
Scientific and policy proposals
Scientific proposal. ScalingRelation-S4 has Novelty 0.61 from the declared calibrated [0,1] similarity construction relative to the admitted literature corpus. The Novelty Method compares one text embedding that describes the proposal with embeddings produced by the same encoder for every corpus paper; the result names that projection and states that notation and experimental detail may be lost.
Its Surprise is 4.1 bits per token under PriorModel-2, using the same versioned tokenizer and abstract-text sample unit for the proposal and model basis; the per-token normalization handles text length but does not support a claim about equations or full papers.
The corpus, neighbour report, model calibration and derivation evidence support those bounded coordinates. Independent replication is missing; any later empirical-use claim that requires it remains unsupported. Report the available coordinates now and obtain the required replication before making that dependent reliance claim.
Policy proposal. In a municipal permit-triage pilot, Policy-P8 reduces median processing time by 12% against the prior-procedure baseline and passes the declared legal-form test. Subgroup error evidence required by the equity must-criterion is missing, so ConstraintFit and eligibility are not established. Keep the proposal out of an approval-facing comparison until the subgroup test is complete; the time result remains usable for its narrower operational question.
Manager quick start
- Identify the bearer and what the receiver needs to judge. Name the reference for novelty, the objective for usefulness or the applicable condition for eligibility.
- State the smallest supported difference and consequence; stop if that answers the question.
- When numbers matter, select the space and specification and build each coordinate through C.16 or C.2.1.
- For a comparison of profiles, apply the declared eligibility and dominance rules. Preserve open comparisons and established incomparability. When a retained set matters, report its declared
Diversity_Por other needed set reading. - Pass generation or retention questions to C.18, pool-policy questions to C.19, selector-facing declarations to G.5, and choice questions to C.11, with the result references each needs.
Optional one-page comparison brief
When several people must reuse the same comparison, publish one short view containing only the fields they need:
- the working question, bearer, intended use, and stop;
- the finite corpus or reference set, inclusion rule, source editions, and window;
- the selected Characteristics, Scales, polarities, admissible operations, and missingness rules;
- the objective, must-criteria, eligibility consequence, and their sources;
- the Methods, models or priors used, with calibration, uncertainty, and evidence;
- the coordinate-result references and any declared set, frontier, or incomparability statement;
- the applicable C.19 policy, G.5 declaration, or C.11 choice reference when later work relies on one; and
- the basis-change and reopen condition when editions or evidence change.
This brief is a representation of the selected configuration and results. It does not create a result, policy, choice, WorkPlan, prior, or domain default. Put a reusable transform or objective form in the pattern that defines it and cite that definition here; do not make the brief a second source of the rule.
Conformance checklist
Common failures and repairs
Consequences
Benefits. Teams can discuss novelty and value before building a metric stack; stronger results remain reproducible and comparable; trade-offs and incomparability stay visible; and generation, policy, choice, Work, evidence, and publication keep their own boundaries. No particular tool or rendering is required: an implementation is suitable when it preserves the declared configuration, result chain, evidence, and limits.
Costs. Quantitative claims require declared Scales, the relevant methods and inputs, and the evidence and uncertainty needed for their use. A corpus-relative value also requires a fixed comparison set. Comparing results across sources or Scales may require extra evidence or a justified correspondence.
Limits. A C.17 result does not prove that a bearer is good, authorize Work, define universal novelty, supply a selection policy, or establish assurance. It makes the bounded claims and their dependencies inspectable.
SoTA-Echoing and source use
Source-use boundary. The sources below change what a C.17 user inspects or reports. They do not install one creativity theory, automated judge, metric, or search algorithm as the FPF default.
Open questions
The following are research questions, not current requirements:
- Under what evidence can a domain justify a stable distance across several creativity Characteristics without hiding their different Scales?
- How can several agents' partial frontiers be related without importing team-governance assumptions or forcing one scalar objective?
- What evidence is sufficient to treat an ordinal reading as interval-like for one declared frontier-estimation use?
- Which delayed observations best reveal when Novelty, Use-Value, or Illumination has become a gamed proxy?
Relations
-
Builds on:
A.17,A.18,A.19,A.19.ECS,C.16,C.2.1,A.1.1,A.10, andB.3. -
Coordinates with:
E.10.LRNfor ambiguous learning-family wording;A.13for exact evaluator recovery and any separate agency or autonomy claim;F.9for an actual Bridge;F.18for lexical candidate-family diversity;A.0:QF.2afor an optional structured cross-scale qualifier;B.4andG.11for evolution and refresh;A.15.1,A.15.2,B.1.6,A.3.1, andA.3.2for Work, plans, resources, and Method descriptions;A.2.1for the assignment needed by the A.13 core andF.6only for an expressly consumed precise assignment-bound attribution;A.6.1only for a separately claimed application of one exact declared Mechanism operation.
-
Supplies results to:
C.18,C.19, andG.5for their set-side questions,C.11.CRConly when one finite configuration-relative comparison is missing, andC.11for choice, without taking over generation, set stewardship, pool policy, comparison, declaration, or choice.
C.17:End
Open-Ended Search Archive and Front Stewardship
Tech-name:
OpenEndedSearchArchiveAndFrontStewardshipPlain-name: open-ended search archive and front stewardship Type: C-pattern Status: Stable Normativity: Normative unless explicitly marked informative Purpose: retain candidates for exploration purposes and report the non-dominated set under a declared comparison, while preserving the basis needed to interpret either return.
Use This When
Use this pattern when generating or retaining variants raises a practical question: which candidates belong on the current front, which should remain available for exploration, and what changed in the basis for those decisions? A retained stepping stone can matter even when another candidate currently dominates it.
First useful move. Identify the eligible population, the comparison or retention rule and what the recipient wants back. All nondominated candidate identities, the objective values they realize and one candidate per value can give different answers. If missing information affects only some memberships, report the settled part and the decisions still open. For an exploration archive, give the candidates retained and their retention reasons.
When a retention reason is not yet supportable, use B.5:4.1.2 on the candidate, source or failed attempt to develop what later inquiry or action it could enable and under which conditions. Compare that attainable contribution and the retention commitment through C.11.DUA:4.2–4.2.1; use an already adequate basis directly. C.19:4.0 develops this connection when the question concerns treatment of a live pool. Return the supported archive-membership decision, or the particular retention claim still open and the attainable contribution needed to settle it.
An ordinary set description and explanation can finish the work. Use a reusable record when another use needs its references, editions or history. Generation alone can return a variant set without creating an archive or front; use C.40 when constructing or changing the material itself is the missing work.
What Goes Wrong If Missed
The project treats an archive as a shortlist, a front as a decision, illumination telemetry as dominance, a retained stepping stone as current best, or a cultural-style variant as a root cultural kind. Generation looks productive, but the next relation is unclear—for example, whether to retain, compare, declare a selected-set result, choose locally, plan work, measure effects, refresh, or write a cultural-evolution case.
What This Buys
The practitioner can retain a promising but currently dominated variant, report a front without treating it as a choice, and determine which comparison or retention claims must change after an edition or policy changes. Records can preserve those grounds for reuse; they are not the practical result by themselves.
Problem Frame
Open-ended search and quality-diversity work deliberately keep more than one candidate alive. That is useful for engineering, science, design, music, dance, AI-agent frameworks, medical method families, and other evolving practices. The same archive or front label can hide strong candidates, weak but promising stepping stones, coverage-expanding variants, architecture candidates, cultural variants, and telemetry-only signals.
The primary EntityOfConcern in C.18 is the archive or front relation being stewarded: which variants are generated or retained, under which descriptor and characteristic space, with which edition and lineage pins, and with which next relation available. C.18 does not answer a local-choice, selector-result declaration, cultural-evolution, or architecture question.
Problem
An undifferentiated label such as archive, front, Q-front, portfolio, style pool, or candidate set can hide which of these answers the reader needs:
- a front answers current non-domination under a declared comparator or dominance set;
- an archive answers retained exploration value, coverage, stepping-stone value, or future reachability under a declared retention policy;
- telemetry reports search health, coverage, novelty, diversity, or lineage but does not by itself dominate alternatives.
Use G.5 for downstream selected-set result declaration, C.11 for local choice, the applicable C.30 or C.32 pattern for an architecture claim or candidate, C.36 for cultural-evolution case work, A.15.2 for planning, A.15.1 for performed Work, and G.11 for currentness and refresh.
Forces
Solution
Keep the claims about archive membership, front membership, telemetry, generation and later use distinct. Use the following record forms when their persistent account is needed; an adequate existing description need not be recopied.
Archive Record
Use this record when an archive account must remain available for reuse—for example, retained exploration value, coverage, novelty, diversity, stepping-stone value, future reachability, curriculum expansion, lineage, or archive policy. Do not use the archive record as a selected-set result declaration or work permission.
Front Record
Use this record when a front account must remain available for reuse—for example, non-domination, Pareto relation, Q-front membership, comparator currentness, admissibility, or partial-order preservation. The front may feed a later G.5 use, but it is not itself a declared selected-set result; declare that result from the front through G.5 under its own basis.
Filled Archive And Front Micro-Records
For this example's current question, the field names C.36 as the next applicable pattern. If one of the stated changes makes refresh current, G.11 becomes the next applicable pattern instead.
For this example's current question, the field names C.30 as the next applicable pattern. Omit selectedSetResultRef? until an exact G.5 result exists. If declaring a selector outcome later becomes current, start a separate G.5 use; if the front's basis changes, start G.11 refresh.
Generation And Downstream-Use Record
When loop-engineering practice generates many candidates—for example, prompts or harness variants—use C.18 for the generation and any archive/front claims the receiving work needs. Keep the descriptor, retention and lineage basis for those claims; add the record below when their reuse requires it. This does not say that the loop improved. Use E.23 only when one retained object version is changed and re-evaluated through repeated passes under a declared object-under-improvement evaluation; use G.9 for parity between variants and G.5 when a selected-set result must be declared.
When the missing work is to change usable material and examine the resulting variation, C.40 supplies that development and warranted continuation, with or without an already selected contribution. C.40.CD constructs problems and obtaining ways together through target application when those operations are needed. The record below describes generation; filling it does not perform these operations.
Across the archive, front, and generation records, currentnessAsOf names the replay date or window together with the relevant pinned editions; currentStatus states whether the recorded archive/front/generation relation is active, held, closed, or stale for its declared use; and stopOrRefreshReason names the condition that ended it or the trigger that would reopen its currentness. These fields record the boundary but do not perform refresh. When source, descriptor, comparator, policy, evidence, or edition currentness becomes the question, nextGoverningRelation points to G.11 and refreshRef? may cite the separately governed refresh record.
Here @Project is a compatibility and retrieval cue, not a project kind or relation assertion. Fill projectLocality? only after both Work occurrences have been admitted independently and the cited relation obtains under its direct governor. generationWorkOccurrenceRef names the exact dated generation U.Work; compositeProjectWorkOccurrenceRef names the selected composite project U.Work; and exactGenerationToProjectRelationRef cites, rather than creates, the exact work-part, containing-work, decision-use, source-use, or other governed relation. Use work parthood only when the complete A.15.1 basis holds. Otherwise omit projectLocality?: the @Project suffix remains retrieval-only and establishes no project Work, parthood, authority, context, or viewpoint.
For example, a completed harness-generation run may state:
Every optional field whose name ends in Ref? points to a separately identified object, claim, policy profile, or measurement basis. dedupThreshold? is not a reference: it carries one declared scalar threshold value. deduplicationUnit? carries its unit literal.
Fill emitterPolicyRef? and insertionPolicyRef? only when the cited C.19 profile or insertion policy applies to the current pool treatment. When a threshold is inherited, the cited profile supplies dedupThreshold, deduplicationBasisRef, and deduplicationUnit; when it is not inherited, carry the scalar in dedupThreshold? and its basis and unit in deduplicationBasisRef? and deduplicationUnit?. These references and scalars do not give C.19 a generation operation or change the archive and front relations stated through C.18.
In particular, problemCardRef? may cite a C.22.2 problem-side episteme but creates neither an actual Problem nor a ProblematicForRelation under C.22.PFR. Naming a generated variant, archive entry, front membership, telemetry value, or retained-exploration claim alone establishes neither an improvement-result nor a work-result identity and creates no relation from generation Work to a result.
nextGoverningRelation locates the next applicable pattern when another question remains; use none when the current result is sufficient. The locator does not itself make a choice, declare a selected-set result, authorize work, perform refresh or make any relation obtain.
Use this record when a generation account needs the retained references and history. For architecture moves involving candidates named in architectureCandidateRefs, use C.30, C.30.ASV, or C.30.AD as applicable. Use C.36 for cultural-evolution cases involving variants named in culturalVariantRefs. Local choice uses C.11; work planning and performed work use the A.15 family; effect measurement uses its direct measurement and evaluation patterns; refresh uses G.11. P2W carry-through uses E.18.1 when an accepted problem-side distinction must be preserved into the next relation.
Distinguish exploration inside a space from change to the space
Apply this branch only when whether the effective possibility space changed can alter retention, comparison, generation, architecture, or the next decision. Start by declaring the current space through the candidate grammar or type boundary, generator and operators, available building blocks, evaluator or comparator, retention or reproduction rule, goals or actions, and environment that matter for this case. The declaration may be ordinary domain content or a cited description.
Then state the smallest supported mode:
The first result is one ordinary C.2.1 claim naming the earlier and candidate space declarations, changed component, mode wording, counterfactual or trajectory evidence, uncertainty, blocked stronger claim, and next governing relation. Use C.28 for a causal claim that the component change produced the new reachability. Use C.11.CRC when a finite space-changing intervention must be compared with the current configuration, and C.11 for the later choice.
To establish that a change opened a possibility, identify a candidate or action reachable under the new rules and explain what kept it out under the earlier rules. An earlier search may simply have missed it. If the question is whether the whole space expanded, also establish that every previously admitted possibility still has a counterpart under the new rules. A change can open some possibilities and close others.
Stop at ordinary same-space exploration when no action depends on the stronger mode. Reopen only when the candidate grammar, generator, operator set, evaluator, retention/reproduction rule, goals/actions, environment, evidence, or receiving decision changes.
Worked micro-case. A cooling-module search previously admits only fixed rectangular layouts assembled by the same connection operators and evaluated under the same thermal/maintainability comparator. A new layout far from the archive remains exploratory if those rules still admit it. For this comparison, take a validated curved-channel building block and its construction operator as supplied inputs. Adding them is expansive only when the earlier generator could not express the resulting layouts and the comparison still uses a compatible higher-order regime. If the operator itself is missing, C.39.RO and the applicable subject Methods must supply its construction before that addition can be claimed. Replacing candidate admission and retention with a context-adaptive rule that changes which successors count is transformational only when the earlier/candidate rule mapping and observed reachability support that stronger claim. A higher novelty score alone establishes none of these transitions.
Front And Archive Are Different Returns
- Identify the population for the claim. Use
candidateSetReffor the compared candidates and the comparator reference for the characteristics and dominance rule. If the commission ranges beyond evaluated or retained candidates, establish what can still occur in the rest of that domain. - Apply the declared dominance relation to eligible candidates for the same use. With incomplete evidence, consider which candidates could be eligible and what their performance could be, keeping all observations and shared constraints true together. Each such possibility for the candidates together is one admissible joint account. Establish that at least one exists. A candidate is excluded if in each account it is ineligible or an eligible rival dominates it. Name the applicable grounds. Different accounts can identify different rivals. To establish membership, show that the candidate is eligible and undominated in every account. Two accounts that give different membership decisions establish that membership is open. A bound or proof can settle the decision without enumerating accounts.
- Two candidates may each do better on a different criterion. Their established trade-off differs from uncertainty about who does better. The identity of a rival may remain open after the evidence settles a candidate’s exclusion. If the front is computed over recorded comparisons, state which candidates and comparisons those records cover before using it to judge the underlying alternatives.
- For an exact answer, establish both inclusion and completeness for what the recipient counts. Each reported candidate must be eligible and nondominated across the claimed population. Every eligible nondominated candidate must be accounted for. With ordinary Pareto dominance, distinct candidates tied on every objective can all qualify. For the distinct attainable objective values, report each nondominated vector once. For one realization per vector, also supply an eligible candidate realizing each. For all efficient candidates, include every distinct qualifying identity or a representation from which they can be recovered. A bound can settle the value front while leaving tied identities open. Apply the same test across every jointly admissible account when observations are incomplete.
- For an exploration archive, state which candidates the policy retained for coverage, diversity, later use or further search. Occupied cells count retained niches under the specified descriptors. To report regions visited or tested, use evidence of visits or tests, including discarded candidates when they count. Two runs can occupy the same cells after testing different regions.
- Keep tie-breakers and telemetry explicit so diversity, illumination, or popularity signals do not rewrite front semantics.
- Before promoting telemetry or a popularity-like signal into the comparator, dominance set, or selected-set criteria, check which intended archive/front uses or values become worse or more fragile when that signal improves. State the affected uses or values, or the checked loci when none worsened, and cite the policy or decision authority that admits the promotion and any trade-off. If this basis or authority is missing, keep the signal as telemetry. Use it as a bounded tie-breaker only when a separate applicable policy already admits that use.
- Use
RetentionIntent=steppingStonewhen retention exists for frontier expansion or later curriculum value rather than current dominance. - If one source line keeps both returns, say that the front answers current non-domination while the archive answers retained exploration value.
Cultural And Architecture Variant Boundaries
For architecture-candidate generation, C.18 records generation, archive, front, descriptor, telemetry, and retained exploration value. C.30 governs the architecture claim: ArchitectureOf@Context, selected structure or structure kind, affected characteristic, and next architecture move.
For cultural variants, C.18 records the generated or retained variant set and its descriptors, lineage, telemetry, and archive or front relation. Use C.36 for a cultural-evolution case concerning collective-holon or discipline-facing Method, Work, system-role kind or assignment, canon, memory, recognition, selection, mediation, style, tradition, or intervention relations. Use F.17 to identify local senses when needed, F.18 when a durable new name is required, and F.9 when a use depends on semantic correspondence between different local senses. Source-local labels such as style, tradition or genre can retain their cited meanings.
Conformance Checklist
CC-C18-1Before generation, archive update, or front publication, name and pin the descriptor, characteristic, distance, and family-coordinate bases used by that operation.CC-C18-2Archive and front returns are separate from a selected-set result unless one is explicitly declared from them throughG.5.CC-C18-3Telemetry remains telemetry unless a declared policy promotes it into the comparator, dominance set, or selected-set criteria and the governing record names the affected intended archive/front uses or values, or the checked loci when none worsened, together with the authority that admits the promotion and any trade-off.CC-C18-4For archive use, record the applicable retention policy, retained exploration value, lineage, and edition pins. When a stepping-stone use is claimed, record that use and its basis.CC-C18-5Use C.30 family patterns before making an architecture move with a candidate.CC-C18-6Use C.36 for cultural-evolution claims about variants, and term-bridge patterns when durable label or bridge work is needed.CC-C18-7Refresh usesG.11with the smallest affected archive, front, descriptor, edition, or lineage locus.CC-C18-8Agent-loop, harness-loop, workflow-store, or DPF-seed variants retained in an archive name their descriptor, lineage, and next governing relation, together with any telemetry on which the retained-exploration claim relies; archive membership alone establishes no quality improvement. Quality-improvement claims require their own re-evaluation; useE.23only for repeated improvement passes.CC-C18-9A filledprojectLocality?names independently admitted dated generation and composite projectU.Workoccurrences, the subject pattern, and one exact obtaining relation;@Projectalone remains retrieval-only.CC-C18-10Problem-card, result, selected-set, choice, work, and refresh references remain references to separately governed objects or next subject patterns and create none of those identities or relations.CC-C18-11When a claim relies on a source premise, keep its edition and use basis recoverable. A material revision or contrary evidence reopens the dependent archive, front or generation claim through G.11; unaffected claims remain reusable.CC-C18-12When the §4.3a branch applies, exploratory, expansive, or transformational wording names the earlier and candidate effective space declarations, changed component, counterfactual or trajectory evidence, uncertainty, blocked stronger claim, and next relation; distance, rarity, novelty, or local progress alone does not prove space change.
Archetypal Grounding
System-facing case. A robotics team generates gait variants. The front records non-dominated speed and energy relations under declared measures. The archive retains diverse coordination patterns because some are stepping stones for new terrain. Telemetry reports coverage. A selected-set result may later be declared through G.5. If it must be available to an audience, use E.17 for its source-backed publication face and return to source and E.24.PUB for the publication occurrence and availability. Performed test runs use A.15.
Architecture case. A cooling-module project keeps an archive of modular layout variants and a front over maintainability and energy use. C.18 records descriptors, archive policy, front relation, and telemetry. Use C.30 to name the selected structure and affected architecture characteristic, then decide whether to make an architecture move with a retained variant.
Cultural case. A dance-lab project generates movement variants around several source labels. C.18 records generated variants, descriptors, archive membership, front relation, and lineage. Use C.36 to determine whether the lab is deliberately changing the cultural practice; use F.17/F.9 when a label correspondence is needed and F.18 when the work needs a durable new name.
Completed front/archive comparison. Suppose three eligible cooling-module candidates have comparable readings for energy per operating cycle and time for the same service operation. Both quantities are minimized under the declared comparator.
A dominates B because both readings are lower. A and C are incomparable: A uses less energy, while C takes less service time. The front is therefore {A, C}. The project's archive policy also retains one representative of every admitted connector family for later recombination. B is the only available representative of family J, so the archive retains B even though B is absent from the front. That retention is a coverage decision under the stated policy, not evidence that B performs better or will produce a successful later design.
Suppose a fourth eligible module matches A on both comparison coordinates. For these four modules the two nondominated objective vectors stay (8 kWh, 20 min) and (12 kWh, 15 min). A purchaser seeking one design per vector can use A and C. A request for the identities of every efficient module also requires the fourth identity. This conclusion extends to a larger eligible population only when the remaining designs are accounted for under the same comparison.
For the original three modules, return front {A, C} and retain B in the archive as the family-J representative. No selected-set declaration or design choice follows automatically. If another J-family variant arrives, reconsider retention under the same policy; if the comparator or measurement basis changes, recompute the affected front relations.
Bias-Annotation
Lexical and semiotic bias are controlled by keeping archive, front, telemetry, selected-set result declaration, audience publication, local choice, cultural-evolution case, architecture move, work permission, and evidence relations distinct. Mathematical descriptions of descriptor maps, fronts, distances, coverage, or novelty use the mathematical-lens pattern when lens adequacy matters.
Consequences
Positive consequences:
- archives keep exploration value while local choice remains separately governed;
- fronts preserve partial-order and comparator semantics;
- architecture and cultural-variant generation become usable without creating parallel root kinds;
- refresh and source-currentness have clear loci.
Costs:
- keep archive and front accounts distinct when both returns are used; generation alone need not create either return;
- a later selection, choice, plan or measurement needs its own grounds and may need a reusable record;
- descriptor editions and distance definitions require maintenance.
Rationale
Current quality-diversity, illumination search, open-ended engineering, and evolutionary-engineering practice shows that retained diversity, stepping stones, archive lineage, and descriptor currentness often matter before a single choice is justified. FPF keeps that practical gain while preventing archive and front language from replacing comparison, selected-set result declaration, audience publication, architecture, cultural evolution, work, evidence, decision, or refresh patterns.
SoTA-Echoing
The cited editions contribute different parts of archive and front practice. Retention value, current dominance and a claim of open-ended change keep separate grounds.
Relations
Builds on: C.16, A.19.CPM, A.19.SelectorMechanism, and E.18.
Coordinates with:
C.19for current-pool treatment;C.11.CRCfor a finite configuration-relative comparison andC.11for local choice;C.28for causal support of a space-change claim;G.5for selected-set result declaration;E.17for a source-backed publication face and return to source, andE.24.PUBfor the publication occurrence and audience availability;G.9for parity and benchmark comparison;G.11for refresh;E.23when an archived object version enters a declared quality-improvement loop;E.18.1for P2W carry-through;C.30family for architecture candidates;C.32.P2Sfor problem-to-structure carry-through,C.32for candidate palette admission, andC.35for generated or discovered result adequacy before archive or front use of architecture candidates;C.36for cultural-evolution cases;F.17,F.18, andF.9for term and bridge work;- the A.15 family for planning or performed work.
C.18:End
Scaling‑Law Lens Binding (SLL)
Status: Stable Type: Pattern
Use this pattern when. Use C.18.1 when a generator, selector, method family, benchmark, or comparison makes an empirical claim that behavior changes with scale—for example, with resource budget, data, model capacity, iteration budget or freedom of action. First name what is scaled, the window and the outcome being compared. Reuse an applicable probe account; new probing is needed only when the required scale claim remains unsupported. If no scale claim is needed, use the ordinary comparison directly. A conditional scale derivation within a stated mathematical model can use its derivation and assumptions directly, without a scale probe. That result supports the model-conditional claim; a claim about observed behavior or the model’s adequacy needs its own applicable evidence.
What goes wrong if missed. Teams compare unequal budgets, call coverage telemetry an objective, claim a knee without probe evidence, or assume more scale means linear improvement across a window where the behavior has already changed.
What this buys. A scale-law lens: declare the scale variables, ScaleWindow, probe points, elasticity class, parity notes, and policy thresholds before treating an empirical scale claim as usable in selection, parity, refresh, shipping, or mathematical-lens work.
Purpose. Declare which scale variables change, the ScaleWindow and the supported behavior class. Numeric fits remain domain work. G.9 governs parity, and a selector policy must admit any decision use of illumination.
Keywords. scaling law; Scale Variables (S); ScaleWindow; knee; diminishing returns; iso‑scale parity; UNM/NormalizationMethod‑based mapping; scale‑probe; DoE (design‑of‑experiments); segmented regression; knee detection.
Problem frame
Teams often say a method “scales” without disclosing which resources, across what window, and how outcomes respond (convex rise → knee → plateau). Without that, parity is skewed (unequal budgets, unmatched windows), coverage/illumination report-metrics leak into dominance, and “knees” are found late. SLL supplies a notation‑independent lens to make scale behaviour explicit and comparable.
Problem
Omitting Scale Variables and the comparison window causes: (i) unfair parity (compute/data/FoA mismatched), (ii) illumination/coverage report-metric creep into dominance by default, (iii) late detection of knees and budget waste. G.9 already forbids scalarising mixed scales and mandates equal FreshnessWindows/pinned editions; SLL complements this with ScaleWindow & elasticity.
Forces
Notation independence vs useful scaling heuristics; local context vs cross‑context generality; telemetry vs objectives (illumination stays report‑only telemetry unless policy promotes it); early exploration vs reproducible policy.
Solution — binding lens for generator/selector profiles (normative)
Profile annotation
SLL.Profile annotates a MethodFamily/Generator or Selector profile with the scale claim and its basis.
Fields (conceptual descriptors).
- S — Scale Variables. Minimal set of adjustable variables with a declared order for the Context:
compute(steps/tokens/FLOPs/time/energy),data(size/quality),model capacity(params/branches),iteration budget,freedom‑of‑action (FoA)/environment richness, etc. Declare each variable’s Characteristic and Scale, with a unit when applicable under C.16, and bindSto a ScaleWindow. Ordering the input does not assume a monotone improvement in the outcome; do not treat qualitative data quality or freedom of action as a ratio quantity without that measurement basis. Keep planned budget values with A.15.2; bind dated resource-use accounts to A.15.1, B.1.6, and A.10. Where training/inference trade, name the phase the claim concerns. - ScaleWindow. The declared range of resource settings under investigation. Identify the settings actually tested and the response, statistic, task family and conditions to which the result applies. To extend a conclusion to untested settings, state the model, bound or other grounds for that extension.
FreshnessWindowgoverns the age of the evidence. - Scale‑Probe. At least two (preferably ≥ 3) parity-respecting points in
Swithin the ScaleWindow, with uncertainty appropriate to the empirical claim. Use replicates, seeds and confidence intervals when the trial design and variability require them; deterministic results retain their conditions and any applicable measurement or numerical error. Pick points via a small factorial or Latin-hypercube when that design answers the multi-knob question. - ElasticityClass
χ ∈ {rising, knee, flat, declining}— a qualitative class supported by the probe at the settings supported within the declared ScaleWindow. Leaveχunassigned when the observations and their uncertainty do not support a class, and state what remains unresolved. Numeric exponents/fits live in domain annexes, not Core. - ParityNotes.
iso‑scale parity?flag and loss notes if not achieved, plus applicable Bridge, Φ, and Ψ IDs for actual crossings under G.9; any numerical assurance penalty follows its applicable B.3 model.
Norms (SLL).
- SLL‑1 (Declaration). Any profile claiming scale behaviour SHALL declare
Sand a ScaleWindow for the Context. - SLL‑2 (Probe). An empirical conclusion about a change with scale SHALL use an applicable scale‑probe (≥ 2 points in
S, with the applicable uncertainty treatment in §4.2) and record the supported χ, or the unresolved classification. An existing probe may supply this basis while its window, definitions and conditions remain applicable. Multi‑knob probes SHALL hold unspecified knobs fixed or pinned, and disclose invariants. A result at one setting can establish what happened there. Repeat trials at that setting can improve the estimate for that setting. A change with scale remains open until the evidence covers distinct settings. - SLL‑3 (Parity). A comparison claiming matched scale SHALL establish iso‑scale parity. When heterogeneous knobs such as FLOPs and tokens require a mapping, apply a justified UNM/NormalizationMethod mapping before comparing outcomes under G.9; their names alone supply no conversion. Keep FreshnessWindows/editions equal/pinned as required by G.9. If iso-scale parity cannot be established, state the loss and restrict the comparison to what its applicable rule admits. Loss notes alone do not make an otherwise unsupported comparison admissible. Record any seeds/replicates used, the uncertainty basis, ComparatorSet and applicable policy ids in telemetry/SCR.
- SLL‑4 (Selection lens). Within the same Context and ScaleWindow, if novelty, use value and constraint fit are tied, selectors MAY use illumination as a tie-breaker when an applicable pinned policy admits that bounded use. This SHALL NOT change default dominance. Promotion into dominance needs its own explicit CAL policy; otherwise illumination remains report-only telemetry.
- SLL‑5 (Knee test). A knee is claimed only when a monotone rise is followed by a slope drop across adjacent intervals formed by at least three distinct probe points within the ScaleWindow. Compare slopes along one ordered resource coordinate with other influential conditions held comparable. If several resources change together, define the path and its unit of advance, and attribute the knee to that path. The slope comparison must be admissible for the declared Scales and establish a drop exceeding the policy-defined threshold under the uncertainty or error treatment appropriate to the probe. For stochastic probes, assess the slope drop under the governing uncertainty rule using a method whose assumptions fit the trial design. If that rule requires a significance test, report its result and significance level. If it specifies a probability or bound on the drop, give the probability or bound and its supporting assumptions. Report the level of any interval used. For deterministic probes, establish the threshold exceedance throughout the applicable error bounds. Cite the threshold policy (E/E-LOG), comparison method and uncertainty basis. Failure to establish a knee leaves that claim unsupported. Report rising, flat or declining only when the probe supports that class; otherwise leave χ unassigned. Absence of a confirmed knee is not evidence of rising performance.
- SLL‑6 (Telemetry invariants). Probes SHALL export the uncertainty basis, any seeds/replicates used, edition pins, applicable policy ids and resource-account units governed by C.16 and B.1.6, with dated-work and provenance links under A.15.1 and A.10, to G.11.
Method — minimal SoTA probe recipe (notation‑agnostic; informative).
- Choose ordered scale variables
Sfrom the decision the result will inform. Identify whose resource is varied, which system or work produces the response, and the range of settings the decision must cover. When using component readings to judge a whole service, establish how their allocation, timing and interaction determine the service result. A direct service measurement or sufficient bound can settle the question. Declare the measurement basis for the selected variables. - Choose attainable settings that can distinguish the effects in question. Three to five settings can reveal curvature along a single ordered resource. A two-state variable permits a two-state comparison. For several resources, choose a design that can distinguish the relevant main effects and interactions.
- For stochastic trials, choose replication and an uncertainty method adequate to the claimed trend; bootstrap intervals are one option when their assumptions fit the data. Retain the seeds used. For deterministic trials, retain the conditions and applicable error bounds.
- Estimate the contrast needed for the claim. For a knee, use the adjacent slopes in SLL‑5. Logarithmic slopes require positive values and operations admitted by the Scales under
A.18. A segmented fit or a knee detector can assist that comparison. - Record the response and statistic, task family, settings tested and conditions held fixed. Make the evidence available through the profile or
SLL.Card@Context. An observed run demonstrates that its outcome occurred under those conditions. A feasible solution with a verified objective value supplies a lower bound when maximizing that objective. Meeting a target with a verified amount of an ordered resource supplies an upper bound on the minimum needed under the same conditions. If several resources are tracked separately, report the feasible combination. A claim about expected performance, population coverage or a required success probability needs evidence for that statistic. - If χ changes across the window, split the ScaleWindow and re‑classify per segment.
Consumer relation fields - minimal inputs and outputs (conceptual)
G.9 parity planning uses S and ScaleWindow to align budgets, pin editions, and declare any required UNM or NormalizationMethod mapping; run evidence records the applied mapping and pins. G.11 carries policy-id, PathSliceId, edition pins, the uncertainty or error basis, and any seeds, replicates or CI level used, under the applicable parity conditions.
Archetypal Grounding (post-2015; informative)
- LLM scaling. Historical Kaplan-style and Chinchilla-optimal regimes illustrate different compute/data allocations. Mixture-of-Experts and retrieval-augmented families require their own capacity and inference-budget account.
- RL/Planning. Model-based optimization & general agents vs hand-tuned controllers; slopes reported wrt budget/FoA under safety envelopes.
- QD/OEE. MAP-Elites, CMA-ME, DQD, QDax; POET/Enhanced-POET families: coverage/illumination as telemetry metrics; parity uses fixed grids/spaces and edition pins.
Constructed classification case. A matched-probe account supports unchanged performance across the declared resource window, with uncertainty small enough for the comparison's stated tolerance. Report flat; the absence of a knee does not turn it into rising. In a second account, the uncertainty still permits both an increase and a decrease that would matter to the choice. Leave χ unassigned and retain that limitation. A supported monotone increase without a confirmed slope drop can instead support rising. These supplied accounts illustrate classification and its limit; they are not empirical scaling results.
Constructed bounded-error knee. At resource counts 1, 2, 3, deterministic throughputs are 10, 18, 19, each with absolute error at most 0.1, under the same tested conditions and admissible linear scales. The first adjacent slope is in [7.8, 8.2]; the second is in [0.8, 1.2]. Both rises are positive and the slope drop is in [6.6, 7.4]. A declared drop threshold of 5 is therefore exceeded throughout these bounds, supporting a knee over these three probe counts. No interpolation beyond the admitted counts is claimed. If wider error bounds admit a drop of 5 or less, this knee conclusion is no longer established; use only the behavior class still supported by those observations and bounds.
Bias-Annotation
Conformance Checklist (CC-SLL)
Sdeclared orS = N/Awith rationale.- For an empirical scale comparison, give the probe settings, response, task family, invariants and applicable uncertainty. Name the supported class or the remaining uncertainty. With observations at only one setting, report what is established there and leave the scale comparison open.
- A matched-scale claim has established iso-scale parity, with the editions, seeds when used, and comparator recoverable. If the resource regimes cannot be matched, report the individual results or a narrower comparison whose grounds hold. Any numerical assurance penalty has a basis in the receiving B.3 model.
- An illumination tie-break cites the policy admitting it and its lens id in E/E-LOG provenance. When that policy depends on χ, include the supported class or leave the dependent preference open.
- Knee claims cite the policy threshold, slope-comparison method and applicable uncertainty or error basis; include the statistical significance level and any CI level when those are used.
Common Anti-Patterns and How to Avoid Them
Hidden budget mismatches; averaging ordinals across families; illumination in dominance by default; unpinned editions; slope claims without an applicable uncertainty basis; training/inference phase mixing → cure with G.9 parity (equal windows/editions; normalize‑then‑compare; return sets), phase‑label the claim, and record slope uncertainty per Scale‑Audit discipline.
Payload — exports
SLL.Card@Context (UTS row; editioned):
⟨S{knobs, applicable units, phase when relevant}, ScaleWindow, Scale‑Probe{points≥2, design, uncertainty basis, seeds?, CI?}, ElasticityClass χ?, ParityNotes{iso‑scale?|loss, invariants}, BridgeIds?/Φ/Ψ, PolicyIds? (E/E‑LOG), PathSliceId?⟩.
UTS row template (conceptual).
SLL.Card@Context := S=(ordered variables; applicable units=…; phase when relevant=…), ScaleWindow=[LOW…HIGH], Probe=(points=…, design=…, uncertainty basis=…, seeds?=…, CI?=…), χ?=rising|knee|flat|declining, ParityNotes=(iso=true|false; invariants=…), Bridge/Φ/Ψ=(…), PolicyIds=(…), PathSliceId=(…).
Here χ? means that the class is omitted when unresolved; uncertainty is not a fifth behavior class. Keep the probe account and its limitation available to the receiving comparison. An unassigned class cannot support a class-dependent preference. Use the linked profile or the probe account to recover the response statistic, task family, tested settings and conditions, and the grounds for any conclusion between or beyond those settings.
Consequences
Benefits. A comparison can show which knobs were scaled, what window is covered, how much probe evidence supports the slope class, and what parity or normalization losses are declared, with any numerical assurance penalties justified by the receiving B.3 model. Dominance changes require their own admitting policy.
Trade-offs. An empirical comparison across resource settings needs observations at at least two scale points and recoverable invariants, applicable phase and uncertainty information, any seeds used, and applicable policy thresholds. Reuse of an applicable account can avoid new probing. The gain is that selectors, parity harnesses, refresh telemetry, and mathematical-lens uses can cite one bounded scale claim instead of guessing whether the observed behavior transfers.
Stop condition. Stop at C.18.1 when the scale variable, ScaleWindow, probe basis, supported class or explicit classification limit, and parity notes suffice for the receiving question. An unresolved class can finish this inquiry while leaving a scale-dependent preference unsupported. Select further probing only when its attainable contribution to the decision warrants its burden; use C.11.DUA if that choice needs resolution. Move to G.9, C.19, G.11, C.29, or a domain annex when the next question concerns parity, selector policy, telemetry refresh, a mathematical lens, or a numeric fit.
Rationale
C.18.1 exists because scale claims are easy to overread as universal improvement claims. The pattern keeps scale behavior bounded by a declared scale variable, scale window, probe basis, uncertainty, elasticity class, and parity notes before the claim is reused.
SoTA-Echoing
Which evidence is sufficient for a scale claim? Retain the distinction in C.29 between a consequence inside a declared mathematical model and a claim about the modelled phenomenon. The conditional cost equations in C.19.2:5 establish their break-even horizon from the supplied assumptions; empirical probes would answer whether those assumptions describe actual handoffs. Requiring a probe for the derivation adds work without strengthening that conditional result, while using the derivation as observed performance omits the needed evidence. This pattern therefore limits its probe requirements to empirical scale claims and adapts C.19.1's claim-matched evidence rule: keep parity and uncertainty conditions, and reuse an adequate account. Reopen when the assertion extends beyond the model conditions, the empirical window or the supported uncertainty.
Scaling-law fits, knee detectors, segmented regressions and experimental-design methods support the scale claim under their own conditions. Keep the measurement, parity and selector-policy grounds required by the use. Apply C.29 only when a mathematical representation choice, transfer or reliance question remains unresolved.
Relations
C.27 temporal-claim relation.
- C.27 may flag: a claim that more review capacity, tool calls, tokens, data, model capacity, parallelism, freedom of action, sprints, or another declared scale variable changes rate, learning, recovery, throughput, stabilization, or improvement.
- This pattern keeps: scale variable, scale window, scale probes, and elasticity value.
- Non-admissible use: more scale does not by itself establish linear improvement, and a scale word does not create a C.27 rate-change claim by itself.
- Neighboring-pattern use: for a comparison or benchmark, cite G.9 for parity; if linear improvement is assumed, name the scale variable and scale window or state the improvement claim as an assumption.
Builds on: C.16/17/18.
Coordinates with: C.19 (lenses/policies), G.5 (set‑returning selector), G.9 (parity; ParetoOnly default; UNM/NormalizationMethod‑based mapping), G.10 (shipping).
Use C.24 only when a fixed action or option needs tool or service call planning; a scale claim about budgets or freedom of action alone does not require that plan.
Pedagogical cue. Say what is scaled, use the probe basis and steer only by the behavior it supports.
C.29 mathematical-lens use relation
C.18.1supplies empirical scale-window and scaling-law evidence forC.29when a mathematical-lens use claims observed scale behavior, universality, knees, exponents, coarse-graining validity or diminishing returns. Apply that evidence requirement to the empirical claim; an adequate conditional derivation within a declared model retains its mathematical grounds.C.19.1governs a separately claimed method scale preference. If no such claim is live, keep the local result and its stop condition.
C.18.1:End
Explore-Exploit Live-Pool Governor
Type: C-pattern Status: Stable Normativity: Normative
Plain-name. Explore-exploit governor.
Intent. State and test exploration and exploitation policy over live candidate pools so frontier treatment, graduation, narrowing, and sunset treatment stay explicit and auditable. A C.19 result governs pool treatment only. C.19 does not generate candidates; use C.19:4.4 when the next result requires another operation.
Use this when
Use C.19 when several candidate lines, family regions or frontier segments remain live and the question is how to treat that pool: widen exploration, keep the frontier, narrow to a subset or sunset a line. Name the governing policy and what would justify changing the treatment. Use C.11 when the remaining question is a local choice among fixed options.
If a proposed pool-policy premise is expressed as learning progress, information gain, novelty, or an articulated former cue, recover its result owner first. Use E.10.LRN only while learning wording hides that result, A.10 when the needed evidence or source basis requires a replayable account, C.17 or C.18 only when the question concerns characterization or possibility-space change, C.11.CRC only when a finite configuration-relative comparison is missing, and C.11 for local option or probe choice. Stop before C.19 unless the remaining question is policy over a live pool.
What goes wrong if missed
- scalarized top-1 picks are mislabeled as "the frontier", so it becomes unclear whether the result names one lens-ranked winner or the admissible live set
- exploration continues without one named pool, one named governing lens, or one explicit next treatment
- local option choice, pool policy, enactment planning, selector-result declaration, and publication availability collapse into one blurred result
What this buys
- one explicit pool-governance result for exploration, graduation, narrowing, and sunset treatment
- one explicit link from lens or policy state to the next pool-side treatment
- one repeatable way to preserve heterogeneity and frontier discipline without forcing inadmissible totalization
First-minute questions
- Which live pool, frontier segment, or family region is under governance?
- Which lens or policy state is governing it?
- Is the next admissible pool treatment to widen, keep the frontier, narrow to a subset, or sunset a line?
- If the question concerns none of those treatments, which subject pattern applies, and why is it no longer pool policy?
- What justifies active exploration or cheaper retention now, and what change would end that justification? Readiness for exploitation is a separate question.
First output
For loop-engineering practice, use this first output only when the question is pool policy over live candidates such as loops, harnesses, workflows, method families, or framework seeds. A C.19 record may say that the pool should widen, keep its frontier, narrow to an internal subset, or sunset a line under a declared lens. If the question leaves pool policy, retain the established result or qualified open question and use the handoff in C.19:4.4.
When a treatment is supported, state the live pool, governing policy, treatment and the event that would justify changing it. A sufficient ordinary explanation can complete that result. To record it for reuse, use governingLens and one currentTreatment token from the closed set widen | keep_frontier | narrow_to_subset | sunset_line. When a missing fact or commitment can change the treatment, report what is established and identify what would settle the choice. Use nextQuestionPatternLocator from C.19:4.4 when another question has become current.
A claim-bearing account of the pool treatment is an episteme under C.2.1; identify its EntityOfConcern and claim content when that identity matters. An existing adequate account can supply it.
If the account cannot identify the pool, policy, treatment and relevant change trigger, state the missing basis; that pool-policy question remains open.
Problem frame
C.19 describes policies and lenses for treating an identified live pool. Use the applicable C.18 generation, archive or front account when the treatment depends on it; the existence of a record is not an entry condition.
C.11 concerns local choice among a fixed OptionSet; C.19 concerns treatment of several live candidate lines, family regions or frontier segments. Either question can arise without first completing the other.
Immediate failure indicators for this pattern:
- the pool-policy result cannot name the live candidate pool whose treatment it states
- the governing lens or policy state is missing
- the next pool-side treatment exists only as one vague promise to continue exploration later
Use C.19 while the question concerns treatment of a live pool; otherwise use the applicable exit in C.19:4.4.
Problem
Ad-hoc exploration mixes ordinal and interval claims, silently scalarizes partial orders, and loses lens or policy provenance, undermining admissibility and reproducibility.
Forces
- Readiness vs. continuation — exploitation needs its direct qualification, while active exploration and cheaper retention need their own prospective contribution and affordable commitments. None of these judgements follows from the others.
- Heterogeneity vs. focus — fairness quotas by family vs. depth on proven lines.
- Lens expressiveness vs. audit — scalarised choices must not be called 'the frontier' and MUST record lens ids.
Solution
Decide which lines remain worth exploring, which remain worth retaining without new probing, and which no longer warrant a place in the live pool. Use the pool's intended contribution and horizon, the resources and opportunity window still available, and what keeping each line displaces. State the resulting treatment and the change that would make it worth reconsidering.
Judge readiness for exploitation separately. A line may warrant further research or cheap retention before it supports deployment or transfer. Conversely, neither missing readiness nor past expenditure justifies keeping it indefinitely. Retaining an archive record under C.18 need not keep the line active or fund another experiment.
Develop the continuation basis when needed
Use an adequate policy, analysis or professional subject Method directly. When the prospective contribution or feasible continuation is not yet clear, begin with the candidate artifact, source account, failed attempt or possible receiving use that is available. B.5:4.1.2 supplies the constructive examination; the connection below returns its result to the pool-treatment question.
- Follow what the candidate could enable. Inspect how an existing artifact or source example takes an input to a result, then how that result could be used. For a failed attempt, find where the observed trace first departs from the expected operation. Use the subject Method, a worked application or a competent contributor to explain what that difference could change. A research line may instead aim to explain an obstruction, develop a model or make a new question answerable; name that attainable kind of contribution at its research horizon.
- Derive a condition or question from that relation. Follow the proposed effect forward and work back to the inputs and contributions it requires. Inspect the corresponding interface, assumption, observation or arrangement in the available material. Where the relation suggests a revealing contrast, examine a case with and without that premise, or make a small trial that can distinguish them. Keep the established consequence separate from an untested interpretation. If the relation itself cannot yet be explained, use B.5:4.1.2 to construct a smaller case or obtain the specific subject assistance needed; the useful first result may be a better question.
- Make the proposed continuation attainable. Walk the remaining performance from available input to needed output. For an operation that can change feasibility, use its Method to determine the preparation, material, competence and duration it requires. Place those demands in the available time windows and compare overlapping demands on the same people, setup or access with ongoing work and other live lines. Read the allocations or agreements supporting those commitments, or obtain the responsible participant's answer. This identifies displaced work and unresolved dependencies. An earlier account can supply these facts.
- Compare the commitment with the contribution. Apply C.11.DUA:4.2 and 4.2.1 to active inquiry, continuing on the present basis, lower-burden retention and other available continuations. Include the work of obtaining and adapting the Method, the execution and retention demands recovered above, and consequential waiting or checking. Use qualitative or ordinal comparison when it suffices. An exploratory attempt can earn a bounded allocation through the understanding or possibility it could develop, before a particular discovery is known.
- Return to the treatment decision. Bring the supported contribution, its conditions, the feasible commitment and the comparison back to the pool policy. Use them to justify a treatment or to preserve the dependent judgement still open. When an unknown condition matters, identify the observation, source, agreement or construction that could settle it and whether obtaining it is worthwhile. When no useful continuation can yet be described, identify the missing relation or subject contribution and compare a bounded attempt to obtain it with retention, deferral or ending the live commitment.
Causal data and causal-policy exploration hook
When exploration collects data for a causal claim, learns or evaluates a causal policy, or uses counterfactual replay as a reason to treat a live line, the pool-policy result stays within C.19 and cites C.28 for the causal-support conclusion.
Optional PoolPolicyResult.causalUseSpec?:
Omit this tail when the pool treatment makes no causal claim and consumes no causal-support result. Include it when effect, counterfactual replay, causal-policy support, or causal evidence changes the treatment. The C.28 result remains evidence support; it does not authorize ranking, retirement, deployment, or graduation. C.19 makes the pool-treatment decision under its own policy.
Policy fields. EmitterPolicy is a context-local, versioned policy with canonical fields:
{ emitterPolicyId, name?, regimeKey ∈ {UCB, Thompson, BO-EI, GP-UCB, PES, InformationGain, …}, params, explore_share∈[0,1], temperature τ≥0, rebalance_period, wild_bet_quota≥0, graduationConditionRef?, assuranceResultRef?, epsilon_dominance ε, cell_capacity K, insertionPolicyRef, dedupThreshold, deduplicationBasisRef, deduplicationUnit }.
graduationConditionRef cites the direct domain or policy condition for moving a line into exploitation or extending an already supported use. assuranceResultRef is present only when satisfying that condition relies on one exact B.3 result for a named assurance use and bounded scope. Neither field is an assurance level.
The pool policy separately states why active exploration or retention is worthwhile, what continuing commitments it needs, and what would defeat that basis. An exploration horizon may extend beyond the next local decision; it still needs a defensible prospective contribution and obtainable resources.
emitterPolicyId is cited as emitterPolicyRef; the profile is not a U-kind, generation operator, staffing instruction, budget approval, or Work record.
Decision-subject clarification. Attribute any later choice to one declared DecisionSubject at explicit DecisionSubjectGranularity. Record measurement spaces and admissible policies in the semantic-frame epistemes that state them. Use LOG to describe lenses and policies; that description does not enact a choice.
EmitterPolicy use. A C.18 generation or archive account cites this profile only when pool treatment, insertion or deduplication uses it.
Use the ordinary default tokens defined in G.Core and G.5. The rules below explain their pool-policy consequences.
Decision-theory bridge. Use C.11 for theory-side choice among already-available options and for the meaning of ProbeBudget, ValueOfInformation, and ValueOfComputation. A pool-policy record may use those outputs as inputs to its treatment judgement. A probe's lack of value for the next local choice does not by itself settle the value of longer-horizon exploration; that prospective contribution and its opportunity cost belong to the pool policy. A worthwhile research line need not become a prerequisite for the present decision.
Ordinary default references (if policy is unspecified):
- Dominance: consume
DefaultId.DominanceRegimefromG.CoreandG.5; in ordinary Q-front use this means{Q components}withConstraintFit=passas eligibility gate. - Tie-breakers: the current policy may use a Novelty coordinate,
DeltaDiversity_P/ΔDiversity_P, Surprise, or Illumination only when it names that tie-breaker. Unused optional tie-breakers need no result.- For Novelty, cite each bearer's coordinate-result episteme: a complete C.16 measurement result for a measured value, or a C.2.1 ascription when the declared rule permits that reading. Before comparing bearers, confirm compatible Novelty Characteristic and Scale editions, corpus/reference set and inclusion rule, similarity Method and encoder/model editions, ClaimScope, and the window, uncertainty and support required by the construction and comparison.
- For Surprise, cite the coordinate result and its generative-model and training-basis editions, Scale, ClaimScope, and the window, uncertainty and support required by its construction and comparison.
- For
DeltaDiversity_P, cite the retained set, candidate, calculation or measurement rule and Scale editions, descriptor or distance basis, resulting marginal reading, and the window and support required by the use. - Illumination remains telemetry over
Diversity_Punless the named policy explicitly promotes it. A promoted use still cites the report and its measurement basis. The words Novelty, Surprise, and diversity alone are not executable policy inputs.
- Archive:
K=1,ε=0, deduplication inCharacteristicSpace. - Policy family: one uncertainty-aware explore policy family with one declared regime key and explicit change triggers;
UCB-class with moderate temperature andexplore_share ≈ 0.3–0.5is one didactic starter profile, not the semantic default family. - Provenance (minimum for the current use): record
DescriptorMapRef.editionandDistanceDefRef.editionwhen their definitions are used. A consumed DHC coordinate carries the active fields of C.21'sDHCReplayBasis, includingDHCMethodRef.edition. CiteemitterPolicyRefandinsertionPolicyReffor the policies used. An applied deduplication threshold keeps scalardedupThreshold,deduplicationBasisRef, anddeduplicationUnit, whether supplied by a cited policy or stated here. RetaintimeWindowandseedswhen they affect replay.
Use-value and declared-Q boundary. C.16.Q is the pattern for the selector-context meaning of use-value and its Objective form. When use-value participates in the current Q, declare QS.UseValue as an objective head in that exact Q and cite the current Q/comparator basis. When it does not participate in the current Q, keep the use-value criterion explicitly outside Q as a declared side condition or tie-breaker. A pool-policy record may use either declared position but cannot silently promote use-value into Q or construct the Q model.
Scalarization lenses (policy‑level). A lens J_ℓ declares: (a) hard eligibility conditions (e.g., ConstraintFit=pass), (b) soft aggregation (weights or curves), (c) trust policy (how any applicable assurance result and any declared CL discount enter).
Conformance. A pool-policy record MUST name the lens used to pick from a frontier; scalarized rankings MUST NOT be presented as “the frontier”; the lens id MUST be recorded in the provenance of each selection.
Promotion rules (policy).
- Tie-breaks. Use only the constituted and compatible results named by the current policy. Promotion of Surprise or Illumination into the dominance set MUST be declared by lens or policy id and captured in provenance.
- Graduation. A candidate line or pool member moves from Explore to Exploit only when eligibility holds and the direct condition cited by
graduationConditionRefis satisfied. When that condition relies on assurance,assuranceResultRefcites the exact B.3 result whose named use and bounded scope support the judgement. An optional profile may supply evidence; neither the profile nor a label graduates the line. - Continue, retain, narrow or sunset. At
rebalance_period, judge the line's prospective exploration or stepping-stone contribution against the remaining opportunity, obtainable resources, retention burden and displaced lines. Keep active exploration only while that commitment is warranted; cheaper retention may remain worthwhile without new probing. Narrow, pivot or sunset when the applicable continuation basis no longer warrants the current treatment. An unsatisfied graduation condition blocks the use it governs, not continuation by itself. A defeated continuation basis can justify retirement even if much has already been spent. The optional profile remains evidence, not the treated object. Pool-policy operation. Determine the treatment of the identified live pool. Obtain a needed generation, archive, front or implementation result through the corresponding operation in §4.4.
Pool-policy pass (per rebalance_period).
- Identify the live pool. Read any C.18 archive/front account on which the treatment depends and retain its replay boundary; front or archive recomputation remains a C.18 question.
- Recover the governing lens and the continuation basis; use §4.0 when that basis still needs construction. Record the policy values used, such as
explore_share, emitter-profile preference,wild_bet_quota, or an admitted heterogeneity constraint. - Apply the conditions for the proposed treatment. For exploitation or a wider supported use, apply eligibility and
graduationConditionRef, citing the bounded B.3 result when assurance is needed. For exploration or retention, assess the continuation basis and the whole pool's competing commitments. Return the supportedcurrentTreatmentfromwiden | keep_frontier | narrow_to_subset | sunset_lineand state whether a retained line warrants active probing or only lower-burden retention. If a needed premise is unresolved, retain the supported account and identify the dependent treatment judgement that remains open. - If that judgement requires fresh candidates, a changed emitter mix or temperature, archive insertion, or front recomputation, set
nextQuestionPatternLocator = C.18and pass only the desired emitter profile, quota or constraint, and the generation/archive/front reason. Use §4.4 to obtain the needed generation, archive or front result. - If carrying out the treatment requires dated implementation, planning, staffing, or budget use, pass the policy record to the A.15 family; the policy record itself grants none of them.
- State the supported conclusion or the dependent question left open. When a reusable account of a settled treatment is needed, use
PoolPolicyResultwithlivePool,governingLens,currentTreatment,changeTrigger, and the inputs required by the next subject pattern. A qualified open question is not recorded as a settled treatment.
Named lenses (policy‑level) The following lens profiles are illustrative heuristics, not norms. Practitioners MAY reuse or modify them.
- Frontier‑sweeper — maintain attention on the full front; promote only when the direct graduation condition holds.
- Barbell — enforce
explore_share ≥ θwith awild_bet_quota; otherwise exploit top‑trust region. - Spike‑first — pick highest Use‑Value subject to
ConstraintFit=passand a small Cost‑to‑Probe cap. - Safety‑first — minimize SafetyRisk subject to
Use‑Value ≥ θandConstraintFit=pass. - Platform‑option — maximize Option‑Value under probe cost bounds.
- Pilot-then-scale — optimize Use-Value on the declared pilot scope. Set
currentTreatment = widenonly whenassuranceResultRefcites the exact B.3 assurance result whose supported scope includes the proposed wider pool, andchangeTriggernames the satisfied assurance condition and that newly supported scope; otherwise keep the pilot scope. - Heterogeneity-first (illustrative profile). Use only when the applicable policy already admits a heterogeneity constraint or sampler policy. The applicable policy may declare a
FamilyCoverageorMinInterFamilyDistancegate, a family or subfamily quota, or a diversity-promoting sampler. Use that policy's thresholds, quotas and sampler rules, and record only the policy values and IDs used.
Conformance (lens recording). A pool-policy record that uses a lens MUST record its lens id. Include emitterPolicyRef when the current treatment uses that profile.
Explicit pool-policy result
Canonical record vocabulary. A serialized PoolPolicyResult uses the field governingLens and exactly one currentTreatment token from widen | keep_frontier | narrow_to_subset | sunset_line. Reader prose may say widen, keep the frontier, narrow to a subset, or sunset a line, but those phrases are labels, not alternate serialized values. Do not use lens as a second field name.
When the policy and available evidence support a treatment, state the treatment and its grounds. A record preserves that result for later inspection or reuse. An inquiry can also establish why the choice of treatment remains open (§4.3).
That result should state:
- the live pool, frontier, or family scope under governance;
- the governing lens id or policy state;
currentTreatment, chosen fromwiden | keep_frontier | narrow_to_subset | sunset_line;- the event or threshold that would justify changing that treatment next.
A result may therefore state, for example:
livePool = frontier_FgoverningLens = barbell_policy_v2currentTreatment = keep_frontierchangeTrigger = graduation_condition_v3 is satisfied for one retained line
or, for one narrower family region:
livePool = family_region_betagoverningLens = heterogeneity_firstcurrentTreatment = narrow_to_subsetchangeTrigger = quota satisfaction plus a compatible cited C.17 Novelty coordinate result clearing novelty_floor_policy_v2
Both examples identify the pool, policy, treatment and change trigger; their different scopes require different continuation grounds.
Closure rule over the live pool
A supported pool-treatment conclusion identifies the pool and its next treatment. If a needed premise is unresolved, the inquiry may finish with the supported account and the dependent judgement left open, as in §4.3.
- Close as
widenwhen wider exploration is warranted under the pool's contribution, horizon and resource conditions. If widening also asserts a wider supported use, satisfy that use's graduation and assurance conditions. - Close as
keep_frontierwhen keeping the live lines is warranted under the current policy. A line retained for later reconsideration need not receive a new probe. - Close as
narrow_to_subsetwhen the continuation basis warrants a smaller internal live set, without implying that one scalar winner has already been chosen. - Close as
sunset_linewhen a line's prospective contribution, feasible continuation or retention no longer warrants its burden under the pool policy. Failure to graduate is not enough. Its archived result may still be retained under C.18 when that separate retention remains useful.
For a decision about the present pool, apply both policies to the same candidate lines and evidence. Compare how the policies would treat these lines. With the same permitted actions they can still give different commitments, resource allocations or sampling probabilities. Retain the differences that matter to the present decision. For a comparison starting from this pool, hold its initial candidates and the external opportunity and resource conditions comparable. For performance across pools, establish a comparable basis across the population in question. Trace how each policy’s treatments would affect the candidates explored, the evidence obtained and the later choices available. Identify which differences follow from the policy and which depend on an external change. A prospective comparison can remain conditional on an outcome still to be observed.
When the question has stopped being pool policy, carry the established result or qualified open question into the applicable handoff in C.19:4.4; the next pattern is recorded outside currentTreatment.
One internal retained subset here is still one pool-treatment result. It is not yet a declared Shortlist or RankedShortlist, and it has no ShortlistId merely by being retained. When a downstream use needs declaration or audience availability, use C.19:4.4.
If the account cannot identify which pool remains live, which lens and policy apply, or which event would justify changing the treatment, it has not established a settled pool-policy result. Preserve the supported part and the unresolved question.
Minimal pool-policy record
The smallest useful C.19 record usually states:
livePool = ...governingLens = ...currentTreatment = widen | keep_frontier | narrow_to_subset | sunset_linechangeTrigger = ...nextQuestionPatternLocator? = ...only when the question is no longer pool policy- one or more native direct-owner reference fields only when an already constituted result or claim supports the treatment; retain the field name, kind, identity, claim episteme when applicable, and subject-pattern locator supplied by that owner—for example the A.2.2
capabilityHolderRefandcapabilityStatementRef; an information-gain or articulated-endpoint use keeps the ref name and kind defined by its own direct owner—rather than replacing them with one C.19 signal or cue a10RelianceRef? = ...when the needed evidence or source basis requires a replayable A.10 account; retain its relied-on claim, bounded pool-treatment use, evidence-provenance path, applicable window andRelianceDispositioncompetenceModelRef? = ...only when it cites one exact model episteme used by the pool policy; that model is neither the capability, the owner-defined result, nor proof that the treatment may rely on eithergoalSpaceExpansionPolicyRef? = ...only when one independently declared archive or curriculum expansion policy governs goal- or task-space growthassuranceResultRef? = ...when graduation, scaling, or widening relies on one exact B.3 assurance result and its bounded supported scopewhyNotLocalChoice = ...when the result might otherwise be mistaken forC.11
A record may therefore read:
When currentTreatment = narrow_to_subset, livePool still names one internal retained subset or one live pool subset. Retaining the subset does not declare a public Shortlist or RankedShortlist or assign a ShortlistId. If selector-facing result declaration is now required, the C.19 record leaves currentTreatment as the last pool treatment and fills nextQuestionPatternLocator = G.5, with the reason that the next question concerns result declaration rather than pool policy.
Goal and task space growth is one pool-policy doctrine over the archive or curriculum side. When autotelic or capability-discovery pressure is active, cite goalSpaceExpansionPolicyRef only for an independently declared policy.
Retain every supporting information-acquisition, capability, novelty, objective, articulated-endpoint, or other result or claim through its native direct-owner reference and kind. Add a10RelianceRef when the needed evidence or source basis requires a replayable A.10 account. Use competenceModelRef only for one exact model episteme, never as an alternative name for the capability, result, or reliance account.
These inputs may support widen, keep_frontier, narrow_to_subset, or sunset_line; none becomes a generic signal or cue, default Q, dominance coordinate, probe choice, or selector-facing shortlist by entering the record.
When the available evidence leaves the treatment open, report the pool, applicable policy, remaining treatments and the premise that would decide among them. Reserve PoolPolicyResult and its currentTreatment field for a treatment that can already be justified.
Worked closure slice
Four short contrasts keep the closure law practical.
Several family regions remain live. When the point is to keep several lines active under one declared lens, the pool-policy result must not imply that one local choice has already been made:
A capability claim supports pool treatment.
An A.2.2 statement about holder diagnostic_agent_v4 supports its qualified ability in task region alpha; it does not establish transfer into region beta. In this constructed case, a declared curriculum-expansion policy warrants keeping both regions live while beta's prospective contribution and retention burden remain acceptable. That policy does not establish transfer. In this case, the curriculum review needs a replayable account of the capability statement's support and bounded use, so the record cites its A.10 reliance account:
capabilityHolderRef identifies the holder System; capabilityStatementRef identifies the C.2.1 episteme asserting that System's bounded ability; and a10RelianceRef qualifies only the stated bounded reliance. None is renamed as a signal or cue, entered into the declared dominance set, or emitted as a ChoiceResult.
The same missing beta qualification permits different continuation decisions.
Hold the alpha-only capability evidence fixed in these three hypothetical policy conditions. None supports deployment in beta.
A compatible Novelty floor or quota may affect these judgements when the policy justifies it for this continuation use. Its failure is not a universal retirement rule, and the unsatisfied transfer condition is unchanged across all three cases.
For the third condition, the record can state:
Obtaining an initially missing beta basis. Start with the alpha and beta configuration artifacts, their code and input definitions, a beta trial log, a saved input capture and one shared test window. Read the beta output and the receiving analysis: the output has elapsed time and scores, while the receiving analysis joins scores to equipment events by event identifier. The failed trial reports unmatched rows.
Following the producer code shows that its projection after sorting omits the identifier. This yields a question: can the identifier be preserved through that operation so the receiving analysis can use the scores? The capture still contains the identifiers, and the replay Method works on that saved capture. A small replay can examine the mapping before using the shared setup.
Ask the receiving team which comparison would consume the joined data and read the corresponding protocol. That use supplies the reason to investigate the mapping; the diagnostic value of beta's scores remains to be established by the comparison.
Recover the remaining replay work from its Method and compare its demands with the actual allocations. If replay fits available analysis time while the shared setup is committed to alpha, compare the limited beta inquiry with cheap retention; a full beta trial remains conditional on obtaining a suitable window. If no receiving comparison is yet established, retain that specific question and use the bounded examination in §4.0 when it is worth its burden.
The result may be a supported use for further comparison or a reason to end that attempt. These obtained facts and remaining conditions supply the pool-treatment comparison above; a changed allocation or receiving operation reopens the affected judgement.
The pool has already been narrowed and the next question is selector-facing result declaration. When one internal retained subset is already explicit and the next question is to declare it for downstream use, close the pool-policy question by naming the applicable pattern instead of presenting that subset as though it were already one selector result:
Cultural and style live pools
For a cultural or style live pool, state how the variants remain live under one policy. The pool might contain style, tradition, Method-family, Work-family, canon, scene or technique variants; use the local kind. The same minimal record can preserve that account.
When source labels need disambiguation, use F.17 to identify their local senses. Use F.18 when a durable new name is needed. Source-local labels can retain their cited meanings.
Include termBridgeRefs only for an independently established F.9 relation between identified sense cells. That reference identifies the relation; a pool-treatment claim relying on it must also establish the named use, direction, correspondence rule and tolerated loss, with A.10/B.3 reliance when required. An existing adequate claim episteme can carry this account.
If the question becomes the cultural-evolution case, use C.36. Result declaration, publication availability and currentness use §4.4.
Exit from pool treatment
When fresh candidates, a changed emitter mix or temperature, archive insertion or front recomputation are needed, use C.18 with the desired policy values and the pool-policy reason. C.18 keeps the generation, archive and front claims distinct. If obtaining the variants requires construction or change of usable material, use the generation Method, with C.40 when that development is the missing contribution; a generation record does not perform it.
When one evaluation suffices to answer whether one bearer or version should improve, use the applicable object evaluation, with E.22 when its framing is needed. When the object version will be improved through repeated passes under a declared evaluation, use E.23. For either exit, pass the exact bearer or version, objective or criterion, evidence, and pool-policy reason that made improvement current. A C.19 treatment is neither a generation operation nor an improvement result.
An internal subset retained by narrow_to_subset is still the live pool named by one C.19 policy record. It is not a public Shortlist, RankedShortlist, or ShortlistId-bearing selector artefact, and no public selector artefact is emitted by this pool-policy use. Front and Archive retain their C.18 meanings; a scalarized pick does not rename either one.
When the retained set must be declared for downstream comparison, registry use, or another selector-facing use, pass G.5 the exact declared source set, lens or policy id, eligibility conditions, dominance set, tie-breakers, promotion policy, and provenance pins. Use G.5 to declare the selected-set result and any stable public shortlist identity required by a named use. The C.19 account retains the preceding treatment or qualified open question and the reason for declaring the result. If the question also concerns audience availability, use E.17 for a source-backed publication face and return to source and E.24.PUB for the publication occurrence and availability.
When the question is which option to choose, pass the fixed option set and comparison basis to C.11. For planning or performed work, use the applicable A.15 pattern. Use C.24 after the action or option is fixed and tool or service calls must be planned. When the question concerns edition, source, descriptor, policy or evidence currentness, use G.11. Retain the established pool-policy result or the qualified open question in each handoff.
For the handoff, identify the pool question and preserve the evidence the receiver can use. For an established treatment, give the treatment and its change trigger. For an open choice, give the possible treatments and the fact or agreement that would decide between them. Include an archive or front reference when the account relies on one. The receiving pattern governs declaration, publication, choice, work or refresh.
Archetypal Grounding
System grounding
A product-search or architecture-search team can keep several family regions alive even after one line looks best locally. keep_frontier under frontier_sweeper_v3 is warranted while those regions' prospective contribution and commitments fit the pool policy. Graduation is reconsidered when graduation_condition_v3 is satisfied; continued exploration or retention is reconsidered when its own basis changes. The team need not choose between premature exploitation and indefinite funded search.
Episteme grounding
A source collection may retain one tradition for the range of its evidence and another because it makes a different family of questions accessible. Decide whether those contributions warrant their continuing cost under the current pool policy. To call a tradition nondominated, also establish under the declared comparator that no eligible rival dominates it. A line kept for later exploration may already have a dominating rival.
Collective and contextual grounding
A regional or stakeholder-diverse pool may have to sunset one line while keeping others alive to preserve coverage, fairness quotas, or contextual fit. Use C.19 to state and test that pool-treatment decision only while the question is still about the live set. Once the result must become one local choice, one enactment plan, or one declared selected set, apply the pattern that defines and tests that result.
Bias-Annotation
Keep partial orders unless the applicable policy justifies scalarization. Apply only operations supported by each Scale; an ordinal value supplies no interval or ratio arithmetic. Preserve lens and policy identity for selections, independently of notation or tool.
Conformance Checklist
-
C19-1 When a C.18 generation or archive record relies on a named C.19
EmitterPolicy, it SHALL cite that profile inemitterPolicyRef?. If the active insertion policy is not inherited, record it ininsertionPolicyRef?. If the deduplication threshold is not inherited, record scalardedupThreshold?together with itsdeduplicationBasisRef?anddeduplicationUnit?; never encode that scalar as a reference. These fields are not required without that policy dependence. -
C19-2 The characteristic set and indicators used for dominance MUST be declared and eligibility conditions applied first. If use-value participates in current
Q, the record cites the C.16.QQS.UseValueobjective head in that Q; otherwise it states that the criterion remains outside Q. (References to C.18 generator operators are descriptive only; LOG exports no Γ.) -
C19-3 If a lens is used, its id MUST be recorded; do not label scalarized top-1 as "frontier".
-
C19-4 Promotion of
SurpriseorIlluminationinto dominance MUST be explicit in policy. -
C19-5 A pool-policy record creates no
SystemRoleAssignmentStateRelation, system-role assignment, permission, plan, budget or Work occurrence. For each such claim needed when implementation follows, establish its context, scope and independent basis under the applicable subject rule. Precise role, assignment or assignment-state claims need their respective grounds; ordinary implementation wording does not require asserting all of them. -
C19-6 When a pool-treatment lens uses dominance or tie-breakers, it MUST document or cite the applicable
Eligibility → Dominance → Tie-breakerscomparison. Eligibility is for the affected use. Continuation and retention without that comparison follow §4's contribution, commitment, and opportunity-cost basis; exploitation readiness is separate. For every tie-breaker used, cite a constituted result with the compatible basis required above; unused optional tie-breakers need no result. Any promotion of Surprise or Illumination into the dominance set MUST be named by lens or policy id and recorded in provenance. -
C19-7 (pattern-change boundary). A project-local choice or revision of an
EmitterPolicy,DescriptorMap,DistanceDef, sampler, quota, orδ_familythreshold stays under C.19 and the decision or result that consumes it; it does not invoke E.15 merely because a profile changed. When the definition is changed in an existing FPF pattern edition, use E.15 to compare the exact predecessor and candidate, classify the effect, and repair dependent consumers. Compare materially plausible candidate definitions when the question requires it. Use C.18 for their generation/archive account, C.40 for missing construction and C.19 only when their live-pool policy is the question. No default heterogeneity quota or sampler is defined here. Keep the policy and card ids in the existing decision, change, or SCR result that needs them; create no separate authoring trace. -
C19-8 When a heterogeneity-first profile is used, provenance MUST name each admitted heterogeneity constraint and its governing policy id. If a family or subfamily quota applies, record the exact quota vector and family-definition id; if sampling applies, record the sampler class, seed when relevant, and sampler-policy id.
-
C19-9 A
PoolPolicyResultMUST identifylivePool,governingLens,changeTrigger, and exactly onecurrentTreatmenttoken fromwiden | keep_frontier | narrow_to_subset | sunset_line;lensand space-separated treatment spellings are not alternate record fields or values. -
C19-10 If the question under repair is local option choice, an enactment-facing plan, selector-facing result declaration, or publication availability,
C.19MUST name the applicable pattern rather than restate it:C.11, the applicable A.15 pattern (C.24only for tool or service call planning after the action is fixed),G.5,E.17, orE.24.PUB. -
C19-11 If goal- or task-space expansion, autotelic pressure, or capability-discovery support is used, the record MUST cite
goalSpaceExpansionPolicyRefonly when one independently declared policy governs the treatment; retain each supporting result or claim under its native direct-owner reference, kind, identity, claim episteme when applicable, and subject-pattern locator; adda10RelianceRefwhen the needed evidence or source basis requires a replayable A.10 account; and usecompetenceModelRefonly for one exact model episteme. None of these inputs becomes a generic signal or cue, default dominance coordinate, probe choice, or selector result merely by supporting the treatment; any dominance promotion still requires the explicit lens or policy rule and provenance required above. -
C19-12 If exploration collects data for a causal claim, learns or evaluates a causal policy, or treats counterfactual replay as support,
PoolPolicyResult.causalUseSpec?MUST carry the target rung, claim kind, available support-component refs, supported use, unsupported use, and the C.28 support-result ref when one is consumed. -
C19-13 A pool-policy record for live loop-engineering candidates—for example, loops, agent harnesses, workflows, or DPF seeds—names the pool, governing lens, current treatment, and change trigger. Fresh generation, archive work, or front recomputation uses
C.18as the pool-policy pass specifies. Any other next-result question uses the transfer inC.19:4.4; C.19 does not absorb improvement, declaration or publication, choice, Work, or refresh. -
C19-15 Graduation, scaling, or widening an already supported use MUST cite its direct
graduationConditionRef. If that judgement relies on assurance,assuranceResultRef?cites the exact B.3 result andchangeTriggernames the satisfied condition and bounded supported scope. Widening exploration, continuing a line, retaining it without new probing, or sunsetting it MUST follow the stated continuation basis, including its contribution, feasible commitments and opportunity cost. Failure to graduate alone supplies neither a retirement decision nor a reason for indefinite continuation. A policy threshold or label does not create an assurance result.
Common Anti-Patterns and How to Avoid Them
-
Using deployment readiness to settle research continuation. Judge active exploration and cheaper retention by their prospective contribution and remaining cost. Keep the qualification for deployment or transfer; past expenditure is not a reason to keep funding the line.
-
Treating one scalarized top-1 as the frontier. Avoid by naming the governing lens and keeping the live frontier distinct from any lens-ranked pick.
-
Treating an unresolved policy question as a settled treatment. Return the supported
currentTreatmentwhen its conditions hold. Otherwise identify the unresolved premise and dependent judgement; if another question has become current, name its subject pattern. -
Letting
SurpriseorIlluminationquietly become dominance criteria. Avoid by promoting them only through one declared lens or policy id and recording that promotion in provenance. -
Absorbing neighboring questions. Avoid by using the exact handoff values in
C.19:4.4instead of adding a neighboring result's fields or claims toPoolPolicyResult.
Consequences
- the result states whether the pool is being widened, kept live, narrowed, or sunset; if the question leaves pool policy, the record names the next subject pattern separately
- active exploration, cheap retention and exploitation can receive different justified decisions; heterogeneity need not collapse into one scalar winner
- the cost is stricter provenance and the need to name lenses, policies, and change triggers explicitly
Rationale
Pool governance decides how to treat several still-live lines over the relevant horizon. A local choice among fixed options can select an action without deciding which other lines to continue or retain; a pool-policy result makes those commitments and their change conditions explicit.
When an applicable policy independently admits coverage or heterogeneity pressure, keep that pressure explicit until one declared reason justifies retirement or use of a different subject pattern. Sunset a line only when the current pool-policy result states the policy reason for retiring that line; name the next subject pattern only when the next question no longer concerns pool policy under C.19.
SoTA-Echoing
The cited editions inform different policy pressures; none supplies a default algorithm, quota or sampler. If a changed source premise affects a pool-policy claim, reopen that dependent use through G.11 while retaining unaffected results.
Relations
C.27 temporal-claim relation.
- C.27 may flag: a temporal claim that changes exploration, exploitation, narrowing, widening, convergence speed, or search cadence in a way that changes admissible use.
- This pattern keeps: pool-policy result and explore and exploit governance, including
keep_frontier,narrow_to_subset, andsunset_line. - Non-admissible use: faster narrowing is not automatically a positive result; it may collapse exploration health, diversity, archive coverage, or frontier discovery.
- Exit: use C.19 for the pool-policy result; use C.27 only for the temporal-claim adequacy question when speed or change affects admissible use.
Builds on: C.18, C.16, A.19.CPM, A.19.SelectorMechanism, and B.3.
Coordinates with:
E.10.LRNonly for unresolved learning-family wording;A.10for a replayable account of the needed evidence or source basis;C.22.PFRfor actual Problem identity;C.18for generation, Archive, Front, and possibility-space change;C.32.P2S,C.32, andC.35for architecture-alternative carry-through and candidate admission;C.28for causal-use support;C.17andG.9for evaluation and parity inputs;C.11.CRConly for a missing finite configuration-relative comparison;C.11for local option or probe choice;- the other next-result patterns and transfer values named in
C.19:4.4.
C.19:End
Bitter‑Lesson Preference (BLP)
Status: Stable
Purpose.
State the empirical Bitter Lesson: in search, learning, planning, and related computational work, general methods able to use increasing compute or data have often displaced hand-engineered special cases. Treat that history as a comparison pressure, not as proof about every bearer. Here bearer names the candidate being compared under the scale claim or generality policy. A project may declare an analogous preference for a module, platform, organization design, evidence arrangement, or other bearer only as a separate local policy, with a scale predicate, objective vector, comparison basis, and evidence form appropriate to that bearer. Safety, cost, admissibility, uncertainty, and non-dominance remain visible; the word general creates no preference by itself.
Builds on. C.19 (E and E‑LOG), C.24 (Agent‑Tools‑CAL; ATC‑2), B.3 (Assurance), E.3 (Precedence), E.5 (Guard‑Rails).
Coordinates with. G.5 (Selector), G.8 (SoS‑LOG Bundles), G.9 (Parity), G.11 (Refresh‑Telemetry), A.0 (On‑Ramp).
Keywords. general-solution preference; scale‑amenability; BLP‑waiver; iso‑scale parity; Scale‑Audit; slope vector; alpha and delta tolerances.
Use This When
Use C.19.1 when a current choice or policy makes a scale claim or invokes a declared generality preference: a narrower special-purpose approach is preferred over a general alternative, or a general approach is preferred because its measured performance is expected to improve across a declared scale window. For search, learning, planning, and agent substrates, the empirical Bitter Lesson can supply the motivating line. For a module relation, platform, organization design, evidence-bearing episteme or work arrangement, or selected structure, state that the move is a local analogy or policy rather than an empirical Bitter-Lesson result.
The pattern governs only that scale-based comparison, preference, or waiver. It neither proves architecture adequacy nor turns a bearer label into a holon kind. If the project is using a bounded specialization and makes no scale advantage or durable generality claim, keep the use local under the bearer's direct pattern and stop here.
When comparing a general adaptive loop with a specialized cycle or direct repair for an E.23 use, use C.19.1 only if the decision relies on scale advantage or a declared generality policy. The E.23 loop still names the object under improvement, evaluation, cost and risk account, protected trade-offs, and stop or switch condition.
What Goes Wrong If Missed
A team treats "more agentic", "more automated", "more specialized", or "works on this benchmark" as proof that one bearer should displace a more general scale-amenable bearer. Another team repeats the opposite error: it invokes the Bitter Lesson as permission to ignore safety, cost, task-family fit, or a narrow heuristic that wins inside the declared scale window. In both cases, the selector loses parity, waiver, and scale-window discipline.
What This Buys
The practitioner gets a cheap first probe before an expensive audit. It distinguishes a supported scale comparison, a declared local analogy or policy, a bounded use with no scale claim yet, and a high-stakes claim that justifies a fuller Scale-Audit. When comparison proceeds, task family, scale window, parity, uncertainty, cost, safety, and waiver remain explicit.
Not This Pattern When
Apply the pattern that defines and tests the current question: C.30 or C.32 for architecture adequacy and synthesis, G.5 for selected-set result declaration, E.17 for a source-backed publication face and return to source, E.24.PUB for the publication occurrence and audience availability, E.23 for repeated object-version improvement, the A.15 family for work, and A.21 for gate decisions.
First Output
First examine the scale claim and its available basis before selecting a Scale-Audit. This cheap scale-claim probe may use an existing adequate account; it does not by itself require a new experiment or note. Identify the two candidates and their direct patterns, task family or receiving use, proposed scale predicate, objective vector, comparison basis, feasible evidence, safety boundary and stakes. Return one of four results:
no scale claim yet: no supported scale claim is available. State whether no such claim is needed or which claimed basis remains unresolved. A bounded use may proceed under its direct pattern only if it needs no missing scale conclusion or override of a declared preference; this result supplies neither preference nor waiver;local analogy or policy: for an analogy, identify the bearer family, the proposed transfer and its grounds. For a local selection policy, identify the actual candidates, the policy edition and the condition under which it decides their use. This can include a policy governing two computational methods.bounded scale comparison: state the smallest parity and uncertainty method adequate for this use;full Scale-Audit selected: state why the claim, stakes, feasible evidence, and receiving use justify the added work.
A BLP-waiver is needed only when a declared generality preference would otherwise decide the use.
Problem frame
Bespoke computational heuristics can win locally while failing to exploit larger compute, data, or search budgets. General methods can improve across a declared scale window, but the empirical record is strongest for computational search, learning, and planning. A module organization, institution, work arrangement, or episteme does not inherit that empirical result by analogy. A project may adopt a broader policy, but it must state the bearer-specific scale relation and evidence rather than treating all growth, reuse, or generality as one phenomenon.
Without this separation, teams either repeat the Bitter Lesson as a slogan or impose a costly machine-learning experiment recipe on bearers for which seeds, FLOPs, compute slopes, or data sweeps have no meaning.
Policy clauses (normative)
BLP-1 — Probe first; select audit depth by claim and risk.
Every BLP use starts with the scale-claim examination above. A claim ready for comparison needs the bearer kind and direct pattern, a recoverable scale predicate, objective vector, comparison basis, evidence form and named receiving use. If any is absent, identify the missing basis and return no scale claim yet: the asserted claim remains unsupported, rather than disappearing. A useful local analogy or policy may be retained as such, but it supplies no empirical conclusion and does not bypass an applicable preference or waiver condition.
When the scale claim is current, choose evidence proportional to the bearer, stakes, feasible observations, and receiving use. A selector-facing superiority claim, durable reusable-bearer policy, safety-material override, or expensive irreversible choice normally justifies a fuller audit. A reversible local probe may need only a small matched comparison. The method shall preserve:
- (a) Parity and admissibility: comparable task family or use, safety boundary, budget basis, current editions, and set-returning Pareto comparison unless a declared policy lawfully selects another operation.
- (b) Bearer-appropriate scale dimensions: compute, data, model capacity, or freedom of action for computational methods when they vary; for another bearer, state its capacity, resource, reuse, throughput, coordination, or other scale predicate and explain why it provides a comparable scale basis for the two candidates.
- (c) Uncertainty appropriate to the evidence: repeated seeds or bootstrap intervals for repeatable stochastic trials when useful; measurement error, interval estimates, case comparison, or another justified form for other bearers. Do not demand seeds from an organization design or FLOPs from an episteme.
- (d) Cost, resource, and safety visibility: report the accounts material to the decision through their direct patterns. Add B.3 only for a named assurance claim and receiving use, including when a material-reliance threshold requires such a claim.
- (e) Objective honesty: keep quality, risk, cost, and any policy-promoted coverage or illumination coordinates separate unless their Scale permits the declared operation.
- (f) Risk-selected design: choose a design that can answer this claim. Fractional factorial, Latin-hypercube, three-level sweeps, heteroscedasticity treatment, and repeated-seed designs are options for suitable multi-knob experiments, not a universal minimum. Record why the selected design is adequate and what it cannot establish.
- (g) Claim-matched tests: use knee or budget-constrained regret tests only when the asserted advantage depends on a knee or dominance inside the audited window.
BLP-2 — Preference rule with alpha and delta tolerances.
Among admissible options at comparable budgets, evaluate the performance contrast at the points and on the objectives that matter to the receiving choice. Apply alpha when a tolerance is needed and delta when the use depends on an assurance comparison. Account for uncertainty in the contrast, including dependence when the runs are paired. A scale-based preference over the audited range requires established Pareto dominance there. If the responses are equivalent under the comparison rule, each wins on a different point or objective, or the evidence still admits materially different orderings, report no scale-based preference and say which of these conclusions is supported. A separately declared local policy can decide only the cases covered by its terms. Cite the policy and its E/E-LOG edition for that choice. For an agentic use, keep the applicable alpha and delta in the comparison, local-policy or waiver result consumed by the C.24 CallPlan.
BLP‑2.1 — Valid waiver grounds (override transparency).
Establish which preference is in force and who may authorize a departure. A course the governing policy already permits can be chosen under its terms. Use local exception provisions where they govern this decision and the departure meets the other binding commitments. If it requires changing a commitment, obtain the decision of whoever holds that authority. To claim a waiver under BLP‑2.1 itself, establish at least one of the following grounds:
- Admissibility override: guard rails, ethics, or precedence make the general bearer inadmissible (
E.5,E.3).- Scale‑probe overturn: under iso‑scale parity in the declared ScaleWindow, the heuristic sustainedly outperforms with uncertainty accounted for.
- Complementary bias: the heuristic is an inductive bias that improves the general method without blocking scale (graceful degradation as
Sgrows).For a waiver under BLP‑2.1, record a BLP-waiver with rationale, admitted review System, waiver-review responsibility relation or A.6.RCD missing governor, and expiry or review in the DRR. Any system-role kind or assignment needed by the review Work is cited separately.
BLP-2.2 — Task-family specialization compatibility.
A bounded specialization that makes no scale-advantage, selector-facing superiority, durable generality, or override claim may be used under its direct pattern with an explicit task family, work target, budget guard rails, and evidence locus. It needs neither a full Scale-Audit nor a BLP waiver merely because it is specialized.
If a scale claim becomes current, run the cheap probe first. Select a full audit only when the claim, stakes, feasible evidence, and receiving use justify it. A specialization produced by a general substrate or described as a complementary bias is not automatically compatible: state the non-blocking scale relation and evidence when that fact is relied on. Low-human-overlap or newly discovered approaches remain admissible under the same bounded-use, parity, safety, cost, and evidence rules; novelty neither proves nor defeats scale amenability.
BLP-3 — Prescription architecture is a separate question.
The Bitter Lesson establishes no universal preference for prohibitions over positive instructions and no general right to autonomous sequencing. For tool-call planning, use C.24 with its budget, stop, replan, and Guard-Rail rules. For another WorkPlan or normative constraint, use the A.15 planning family, E.3, E.5, and the direct policy or commitment pattern. A project may adopt a minimal-prescription policy only with its own trigger, safety boundary, evidence, and review condition; it is not a consequence of a BLP comparison.
BLP‑4 — Heuristic‑Debt register (mandatory).
Record Heuristic Debt only when an admitted heuristic functions as reusable solution-family policy, selector-facing preference, durable override of a general scale-amenable alternative, DRR-backed scale waiver, or project-side choice that claims scale advantage or BLP override. Ordinary local bounded tactics that make no reusable-bearer, scale-advantage, selector-facing, or override claim may remain local and bounded without Heuristic Debt publication. BLP.HeuristicDebtEntry is a C.19.1-local or G.11-linked policy and debt entry; it is not a universal U.* record kind unless separately admitted under E.24.UK, with F.18 for naming, C.3 for Kind semantics, and E.9 for the decision record. For a live debt entry, record scope, admitted review System, debt-review responsibility relation or A.6.RCD missing governor, expiry or review window, and a de-hardening plan; any system-role kind or assignment needed by review Work remains separate. Track the entry in CalibrationLedger or BCT and cite it in SCR.
BLP-5 — Adaptation policy is a separate question.
The Bitter Lesson does not by itself require feedback-driven adaptation or make disabling adaptation a waiver. When adaptation is current, use C.22.1 for the task-family adaptation claim, E.23 for repeated object-version improvement, and C.24 for tool-call planning and replanning, together with the applicable privacy and Guard-Rail patterns. A product policy may require or prohibit adaptation, but that result needs an objective, evidence, a risk boundary, and a review date.
BLP‑6 — Precedence & safeguards. BLP draws on P‑10, P‑11, P‑7 and P‑1. Apply E.3 to the principles being compared; the combined policy has no single precedence class. It does not supersede Guard‑Rails (E.5) or precedence rulings (E.3). Where a declared NQD or C.19 E‑LOG policy promotes illumination into dominance, use that lens for the audited window.
BLP‑7 — Publication discipline.
When a durable Scale-Audit is performed, its artifacts SHALL be exported to G.11 with the used edition pins, uncertainty method, alpha and delta tolerances when current, ComparatorSet, policy reference, and qualification window. A no scale claim yet or local bounded-use exit creates no audit package.
Conformance Checklist (CC-BLP)
- The cheap scale-claim probe names the bearer kind and direct pattern, task family or receiving use, scale predicate, objective vector, comparison basis, feasible evidence, safety boundary, and stakes.
- The result says
no scale claim yet,local analogy or policy,bounded scale comparison, orfull Scale-Audit selected; it does not hide the choice of audit depth. - A performed comparison uses parity, current editions, explicit uncertainty, material resource and safety accounts, and lawful Pareto or declared policy operations.
- The evidence method fits the bearer. Compute/data sweeps, seeds, bootstraps, and factorial or Latin-hypercube designs appear only when they answer the claim.
- A comparison that establishes equivalence or a trade-off, and a comparison whose evidence cannot yet settle dominance, each records its own conclusion. A subsequent choice by local policy names the condition that the policy covers.
- A BLP‑2.1 waiver identifies the policy it overrides, rationale, admitted review System, responsibility relation or missing governor, and expiry or review window.
- A live Heuristic Debt entry meets the bounded
BLP-4trigger; ordinary local tactics create no debt record. - Prescription architecture and adaptation policy are routed through their direct patterns rather than reported as consequences of BLP.
- An actual durable audit is exported to
G.11; a bounded-use or no-claim exit is not padded into an audit artifact. - A narrower specialist bearer names its task family, work target, and the scale, waiver, or bounded-use ground relied on.
Anti-patterns & remedies
- Slogan as evidence.
General,agentic, orBitter Lessonis treated as proof. Repair with the cheap probe and a comparison basis. - Analogy as empirical result. A module relation, organization, or episteme is assigned compute/data scaling semantics. State a local analogy or policy and use a bearer-appropriate predicate and evidence form.
- Universal experiment recipe. Every claim receives seeds, FLOPs, and a multi-factor design. Select the smallest method that can answer the risk-bearing claim.
- General wins by default. Compare the actual performance contrast with its uncertainty. Marginal error bars may overlap even when paired runs establish a consistent advantage. If the contrast leaves the order open, record that uncertainty. If two options each win somewhere in the required range, record the trade-off. A generality preference requires a policy that covers the established case.
- Single-winner leaderboard. Hidden budget mixing or scalarization replaces Pareto comparison. Restore comparable windows, objective coordinates, uncertainty, and policy identifiers.
- Debt without trigger. A local bounded tactic is entered into Heuristic Debt. Apply the
BLP-4trigger before creating the entry. - Elegant-math override. A specialized mathematical lens is selected because of elegance or prestige while scale advantage is live. Use the proportionate BLP comparison; otherwise keep the lens local under the
C.29stop condition.
Archetypal grounding (post-2015; informative)
These computational examples illustrate where a scale comparison may be useful; they supply no empirical conclusion for another bearer family. Each comparison needs its task family, comparator, budget and safety basis, uncertainty, applicable alpha/delta tolerances and any source-currentness condition required by its receiving use.
- LLMs: prompt programs, retrieval-augmented policies, and MoE policies compared with narrow task-specific pipelines; set-returning selection across editions and budgets.
- RL and planning: model-based optimization and general agents compared with hand-coded controllers, subject to alpha and delta tolerances and safety.
- Preference learning: comparisons between RLHF and DPO families.
- QD and OEE: MAP-Elites, CMA-ME, DQD, and QDax; POET and Enhanced-POET; illumination remains report-only telemetry unless policy promotes it.
Completed bounded comparison. Suppose a specialist and a general search method are tested on the same fixed task set at three matched budgets on the same processor. The account includes the full computational cost, keeps the other material conditions fixed and supplies the following uncertainty bounds for the percentage of tasks solved. Both methods satisfy the same applicable safety and eligibility conditions; no separate assurance result is consumed. The budget tolerance is alpha = 0 CPU-seconds at each matched point.
The specialist wins at the first probed budget, while the general method wins at the other two even after allowing for the supplied uncertainty. Neither response dominates across all three probed budgets. Return no scale-based preference for that whole comparison; the word general does not reverse it. The result supports the stated comparisons at those points, not an interpolated law between them or a claim outside the window.
If the receiving choice is confined to 100 CPU-seconds, the specialist's bounded advantage can be used under that question's comparison rule. A separately declared prefer-general policy, if applicable, needs its own tie-break or waiver treatment. No such policy is stipulated here, so no waiver or debt entry is created merely to report these results. If a tariff changes, measured CPU time can remain usable while monetary costs are recalculated. If the task mix changes, compatible results for the affected groups can sometimes be reweighted. A previously untested group needs its own evidence or justified transfer. Reuse only the conclusions supported for the new choice. This is a constructed application, not a new benchmark finding.
SoTA-Echoing
Payload - exports
BLP.Policy@Context is an editioned local policy row, not a universal kind. It records:
<scopeBranch={empirical-computational | declared-local-analogy}, PreferenceDefault={neutral | declared-prefer-general}, alpha?, delta?, scaleProbeResult?, proportionateComparisonMethod?, fullScaleAuditRef?, WaiverRegister?, E-LOG policyIds?, G.11 telemetryPins?>.
The row omits fields that are not current. PreferenceDefault=declared-prefer-general identifies a local tie-break policy, not an empirical conclusion. A full audit reference appears only after the risk-selected audit exists.
Relations
Depends on: G.5 and G.9 (selector and parity), G.11 (refresh telemetry), A.15.1 for dated work, A.15.2 for work plans, B.1.6 for resource and cost aggregation, C.16 for measurement, and A.10 for provenance, C.18 (NQD‑CAL), C.19 (E and E‑LOG), F.7 and F.9 (bridges, CL, Φ, and Ψ). Constrained by: E.5 Guard‑Rails and E.3 precedence.
C.32 architecture-synthesis use relation
When using C.32 to generate candidate architectures, C.19.1 applies only if a claim about a candidate asserts scale advantage or invokes a declared local generality policy. For a universal module relation, platform, organization design, evidence arrangement, or selected structure, that branch is a project analogy or policy: the empirical machine-learning literature is not proof. Name the bearer and its direct pattern, the holon under change when one is current, the bearer-specific scale predicate, objective vector, comparison basis, feasible evidence, admissibility boundary, and intended receiving use.
BLP neither selects the architecture nor turns method-family, practice, role-side, or culture wording into a holon kind. If the candidate removes parts, carries several functions, or is described as reusable, with no recoverable scale claim or declared generality-policy claim, keep the question in C.32 and C.31; return no scale claim yet rather than manufacturing an audit. A TRIZ-style ideality move enters BLP only when the declared comparison relies on scale amenability inside a named window.
C.29 mathematical-lens use relation
When a mathematical lens is chosen over a general, scale-amenable bearer because it is elegant, specialized, or theoretically prestigious, and a scale-advantage or declared generality-policy claim is current, C.19.1 governs the scale-advantage and preference claim. A C.29 application may state CandidateMathObject, LensMappingMode, PreservedStructure, LostStructure, LensUseBoundaryValue, declaredLensUse, and StopCondition, with optional blockedLensOverread? for an overread that passes F.19's plausible-reader test; it does not supply BLP compatibility, scale dominance, or waiver evidence.
If scale advantage or a declared generality policy is live, start with the cheap probe and cite the resulting bounded comparison, risk-selected Scale-Audit, or applicable BLP-waiver. If neither is live, keep the mathematical lens local and bounded by its C.29 stop condition.
Memory hook. Test what scales; label policy when evidence does not decide.
When choosing between a general adaptive loop, a specialized object-family cycle, or a mixed operation-family set for an E.23 use, C.19.1 applies only when the decision relies on scale advantage or a declared generality policy. Start with the cheap probe. Compare the declared characteristic values for material resources, tools and instruments, adaptation attempts, skilled attention, rework or delay, risk exposure, and avoided loss on their admitted scales. Keep the components separate. Enforce the binding conditions for this result before excluding an option on grounds of dominance. Its rival must be eligible for the same work. Use the declared project policy to choose or hold among the nondominated alternatives. Net-cost arithmetic is permitted only after every term has been converted to one declared unit through an admissible conversion whose basis, uncertainty, and scope remain visible. Repeated automation alone does not satisfy BLP; the record names the object under improvement, evaluation, protected trade-offs, bounded cost and risk condition, and stop or switch condition. If neither a scale claim nor a declared generality policy is current, E.23 proceeds without a BLP audit or waiver.
C.19.1:End
Use-Bounded Apparatus Application
Type: Architectural (A) Status: Stable Normativity: Normative
Use this when
Use this pattern when one practical result matters and a relevant method, model, formalism, assurance technique, ontology, or other direct-kind apparatus is available, but the work needed to configure and apply it may cost more than the result warrants. Start here whether one apparatus is already selected or a real choice among available alternatives has become current.
The first useful move is to name the practical use, result kind, claimed guarantee, constraints, and reuse horizon, then ask whether the next adaptation and application work can reach a useful result within the available budget. This keeps a small, adequate path small while letting repeated or high-consequence use justify richer configuration.
Not this pattern when. If an adequate way to obtain the result is still unexplained, use C.39. If usable material needs feasible variation and examination, use C.40. Use C.18 when the question concerns the generation account or archive/front stewardship. If the live question is a local choice over an existing option set, use C.11. If the blocker is an ontology conflation, use A.7.1; if it is a material conflict among FPF premises, use A.7.2.
The primary reader is an engineer, method or model selector, or technical lead. Use the instructions below to guide bounded application; the problem-facing result remains with the pattern that defines or tests it. Reader position, guidance and performed Work have different meanings. Section 4.1 gives the identity-bearing conditions when a claim about a Method, MethodDescription or dated Work is needed.
Problem frame
A team can have a rich apparatus and still lack an economical way to use it. A maintenance group may need only one typed relation distinction for a one-off repair, an architecture team may need a configured model across hundreds of handoffs, and an assurance team may need a formal technique only after the required guarantee becomes stronger. In each case, displaying more apparatus is easier than proving that its setup work changes the result.
The concern is one bounded apparatus application for one declared use and horizon. The apparatus retains its direct kind; this pattern does not introduce a generic U.Apparatus kind. The application question is distinct from candidate generation, local choice, planning, performed work, and the domain result.
Problem
Without a use-bounded application method, users make two symmetrical mistakes. They configure an entire rich basis because it is available, or they keep a lightweight path after recurrence, automation, transfer, or consequence would repay a more capable configuration. They may also call candidate generation “choice”, call a choice record a work plan, or treat an application note as the practical result.
The failure is not merely excess documentation. It obscures who performs the work, what result the work must produce, which guarantee is being claimed, and which neighboring pattern contains the defining content for the next question.
Forces
Solution
Identify the use, apparatus, method, performer and result
Before choosing how to proceed, identify what the application involves.
- Declared use: the practical question, direct result kind, claimed guarantee, non-negotiable constraints, and horizon.
- Selected or candidate direct-kind object: the method description, model, ontology module, formal technique, or other governed object being considered.
- Application guidance and Method: use this pattern’s instructions to guide the application. When identifying a particular Method, apply A.3.1. Treat a claim-bearing episteme as its
U.MethodDescriptiononly when A.3.2 establishes that the already admitted Method is its EntityOfConcern and it makes substantive claims about that way of doing. Using the guidance does not itself require that classification. - Performer and work: when a claim identifies particular dated configuration or application
U.Work, recover the admitted performingU.Systemthrough A.13 and independently admit the Work under A.15.1, including its enacted Method under A.3.1. Add F.6 afterward only when the present use needs precise assignment-bound attribution. In a short account, expose assignment identity, species, participants, or attribution detail only when the use relies on them, attribution is ambiguous, or source wording must be repaired. - Problem-facing result: the domain, engineering, assurance, architecture, or other subject-pattern result inspected after the work.
A person or an AI system can read these instructions and also perform the work. To attribute a particular performance, establish who actually carried it out.
Select the application branch
One current apparatus. For an already selected method, recover its required operations, the inputs and settings it accepts, and the result and guarantee it supports. Check the conditions of this application and supply its inputs. The existing configuration may already suffice. If extra setup is needed, establish what it will enable and whether its cost fits the budget and the expected use. The single-apparatus path needs no OptionSet or ChoiceResult.
Candidate generation or reframing. If no adequate current object is available, use C.39 to develop a missing explanation of how to obtain the result, or C.40 to vary usable material and examine the difference. Use C.18 when generation-account or archive/front claims are needed. The declared use and eligibility basis can guide that work.
Local choice. Only when two or more already-available eligible alternatives, or another genuine local-choice question over a live set, are current, use C.11 for OptionSet, ChoiceRule, probing, and ChoiceResult.
Post-choice enactment. For a singular selected direct-kind object, use A.15.2 to plan its application when a plan is needed and A.15.1 for the dated application Work. C.24 is the pattern for sequencing, budgeting, checkpointing, and replanning only when the application involves tool-call work.
Admit candidates by one use-bounded predicate
UseBoundedApparatusCandidateEligibilityPredicate@Context is a local eligibility predicate, not a U-kind, relation kind, or candidate-generation method. A candidate is eligible only when it has a credible adaptation path to the same declared use, direct result kind, claimed guarantee, scope and horizon, and non-negotiable constraints. A candidate that cannot meet one of those values stays outside the current option set rather than becoming a “weaker” member of it.
When the claimed guarantee or requirement is disputed, use C.11.DUA to appraise its purpose, contribution, burden and current force. Keep apparatus eligibility under the current declaration separate from that appraisal. Resume application under a revised declaration only when the change is available through the authority or agreement that governs it.
Apply C.11 to the current comparison. If the evidence or the holder’s priorities leave a requested choice open, return the supported comparison, the outstanding question and who can settle it. A completed ChoiceResult has one of four forms:
choose nownames one selected option or a retained tie-set;reject current setreturns to a named candidate-generation pattern or closes with no current application;probe againretains one probe and its epistemic budget because it can still change the choice;reroutenames the neighboring question and its subject pattern.
These four dispositions form the complete current C.11 result set. “Configure the rich basis”, “adapt again”, and “use a lighter method” are option or plan contents, not extra choice-result values. A tie is not a hidden winner.
Perform the bounded application
- State the practical use, direct result kind, claimed guarantee, constraints, horizon, and current apparatus state.
- For an already selected apparatus, check its configuration and conditions of use (§4.2). Under
A.3.1, determine whether the original Method admits the proposed inputs, parameters, operations, required participants, result and acceptance conditions. If a defining condition changes, identify the revised way of doing the work or state what remains unresolved. Use the subject method to establish what guarantee the revised procedure supports. If candidates are missing, useC.39for an unexplained way to obtain the result orC.40for feasible variation and examination of usable material. AddC.18when a generation account or archive/front claim is needed. - When alternatives are current, identify them by their direct kinds and apply the shared eligibility predicate.
- For each current path, identify what must be learned, supplied, integrated or checked to meet the required guarantee. Estimate the work still needed. An existing configuration that meets the requirements can be used immediately.
- Compare only the dimensions that can change the current path for the declared use and horizon. These may include time and budget, prior exposure, post-threshold efficiency, transfer, retention, interoperability, downside and reversibility; use their subject-pattern values when needed. Do not compress them into an undeclared scalar.
- Use the current comparison at the strength it supports (§4.3). If the choice remains open, give the holder the supported alternatives and the premise or priority needed to decide. A completed
C.11 ChoiceResultmay identify a worthwhile probe or another decision. A rejected set leaves no selected application. When a set is accepted, distinguish ties from incomparability and establish who may choose a member for use. - When one apparatus is selected for use, carry out its application. Use A.15.2 when a plan is needed and C.24 when tool-call planning is the question. A precise dated-Work claim uses the performer and admission conditions in §4.1.
- Inspect the result against the criterion for this use. For another application, check that the accepted method, inputs and assumptions cover its conditions. To establish that the recipient can apply the method with a stated degree of help, examine that recipient’s own performance under those conditions. Keep an application or configuration note when reuse, dispute, automation or consequence makes it useful.
Stop and reopen
Stop when the direct result is usable at the claimed guarantee and no further distinction or setup is expected to improve the declared use enough to justify its burden over the stated horizon. This is a positive result, not a claim that the unused apparatus is inferior.
Reopen when a consequential counterexample, failed result, changed use or guarantee, changed recurrence horizon, new candidate path, automation need, or changed adaptation cost alters the lawful path. Reopen only the affected application, candidate, choice, plan, or work question under its subject pattern.
Archetypal Grounding
Repeated audited handoffs. A fleet has 400 maintenance handoffs each year. The required result is an audited maintenance-decision episteme with a fixed safety and interoperability guarantee. Candidate generation is complete: an ontology-backed Method and a lighter local decision Method are eligible; their descriptions are available. A spreadsheet macro is excluded because it cannot preserve required relation-occurrence identity.
In this constructed planning account, both eligible Methods satisfy the same guarantee and material non-time conditions. The complete expected setup and per-handoff effort are comparable person-hours, including the required learning, review and integration. Forty setup hours are available. The ontology-backed Method needs 40 hours initially and 6 minutes per handoff; the lighter Method needs 2 hours initially and 14 minutes per handoff.
The setup difference is 38 hours and the per-handoff saving is 8 minutes: the methods tie at 285 handoffs. At the declared 400-handoff horizon, C.11 returns choose now for the ontology-backed Method on the supplied equal-guarantee basis, saving an expected 15 hours 20 minutes. If the feasible horizon falls to 100, the same basis favors the lighter Method. These are conditional estimates, not observed maintenance results; changed costs or guarantees reopen their respective questions.
An admitted maintenance-information System performs the dated configuration and application Work using the selected ontology-backed Method. The account recovers the System's A.13 core and independently admits the Work under A.15.1. Because the current question is whether setup cost is repaid across the handoffs and consumes no assignment-bound attribution, this example does not open F.6 or expose assignment identity, species, participants, or attribution detail. A later attribution-bearing claim adds F.6 through the same obtaining A.13 assignment.
A separately admitted maintenance-decision System then performs decision Work using the selected Method. Its A.13 core and independent A.15.1 Work admission are recoverable; this setup-cost example consumes no assignment-bound attribution, so F.6 remains unopened. Under A.6.1, application MaintenanceDecisionApplication-1 of operation DecideMaintenance, declared by MaintenanceDecisionMechanism-E1, returns AuditedMaintenanceDecision-1 under result declaration DecisionResult; no Work-to-result or production claim is made. To check the forecast, compare actual setup and effort per audited handoff on comparable work and verify that the required guarantee still holds. Reconsider the application if the revised costs, guarantee or expected number of handoffs change which path is worth using.
One-off naming repair. A team already has the typed-relation method for one local, reversible naming decision. No rival is live and the current method has a credible small path, so no option set or choice result is created. The team performs the minimal wording and typing work and returns a scoped terminology decision. Recurrence, integration, a failed result, or a stronger guarantee may later open candidate generation and choice.
Non-use. If the blocked result comes from missing telemetry while the method, state kinds, and action distinctions are already clear, return to measurement and evidence work. Apparatus selection cannot manufacture the missing observation.
Bias-Annotation
The main bias is prestige-by-apparatus: richer form, newer tooling, or familiar terminology is treated as practical superiority. The mitigation is one declared result and guarantee, direct-kind candidates, adaptation/work cost, and a positive one-apparatus path. A second bias is automation optimism; cheaper computation does not erase evidence, human attention, integration, or consequence review.
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Consequences
The pattern makes cheap use a first-class result. Teams can apply one current method without ceremony, yet repeated or consequential uses can justify configuration or a real choice. The cost is naming the result, guarantee, applicable condition or predicate, and enough work economics to support the path. That bounded burden prevents apparatus display from replacing practical return.
Rationale
Application is the common job; selection is conditional. Starting from application preserves the frequent case in which a subject pattern is already selected and only the next useful adaptation matters. When alternatives are real, existing generation and choice patterns provide better result semantics than a fifth local decision vocabulary. Direct-kind discipline prevents an umbrella comparison word from becoming a new ontology.
The aphorism is: make the apparatus earn its setup work.
SoTA-Echoing
These sources change the positive method: the user starts from a result, may obtain value before total configuration, opens choice only for a live set, and reopens when use economics or guarantee changes.
Relations
- Coordinates with:
C.39for a missing explanation of how to obtain the result;C.40for feasible variation and examination;C.18for the generation account and archive/front stewardship;C.19for explore/exploit policy over still-live candidate pools;C.19.1for scale-amenable bearer preference;C.22.1for adaptation signatures;E.23for repeated improvement; andC.31.ASAPfor architecture-scale preference. - Uses conditionally:
C.11only when a local-choice question over a live eligible set exists. It consumes, but does not extend, the fourChoiceResultdispositions. - Hands off enactment to:
A.15.2for work plans,A.15.1for dated work, andC.24only for tool-call enactment planning. The domain pattern contains the defining content for the practical result. - Description-level specialization:
A.7.1narrows the method claims stated here for consequence-guided ontology analysis. It retains the declared use, problem-facing result, claimed guarantee, horizon, useful threshold, separation among plan, Work, and result, stop, and reopen. It retains the candidate and choice branch only when that branch is triggered. This wording adds no relation occurrence between the described Methods and asserts neitherU.SubkindOfnor a world relation.
C.19.2:End
Composition of U.Discipline (Discipline-CAL)
Status: Stable Type: Pattern
Use This When
Use this pattern to establish what makes an existing field a discipline, to design the connection between knowledge and practice in a proposed field, or to maintain that connection as the field changes. Begin with the work the discipline is meant to sustain. Practitioners may be producing something, investigating a phenomenon, or keeping an activity effective over time. What must they understand to do that work, and how can the results of their work test or improve that understanding?
Typical moments include:
- comparing rival traditions without letting a shared label, Bridge or comparison table merge them;
- reusing a practice across semantic localities while preserving local meanings, direction, tolerated loss and field boundaries;
- testing whether a Method, theory, standard or subdiscipline is a constituent rather than merely used, cited, registered or taught in the field;
- determining continuity while a canon, standard or accepted practice changes, and identifying another whole when the reidentification rule fails;
- using an applied, multidisciplinary or transdisciplinary classification under its stated criterion.
Primary EntityOfConcern. One exact U.Entity candidate being tested for recognition as U.Discipline. The candidate is recoverable before the classification result.
First useful move. What does the proposed field enable practitioners to do or explain? Start with a problem the field is meant to address. Trace the knowledge that guides how practitioners address it and the reusable practice through which results can challenge that knowledge. This gives you candidates for both kinds of contribution. A field may also use a result supplied from outside. To distinguish the two, establish which contribution the field's own practice requires or currently realizes as an admitted alternative and how it connects to the others. For recognition of the discipline, test the exact candidate and its independently identified parts under section 4.1. A narrower question about one use can be settled by the grounds for that use.
What goes wrong if missed. A field name starts doing the work of a canon, method repertoire, institution, bridge, comparison policy, and maturity score at once. The reader may then treat a card or graph as establishing parts, or treat changes to a journal, standard, method list, or dashboard as establishing a new discipline. Cross-locality reuse can also look valid after local meanings, evidence lanes, or comparison rules have drifted.
What this buys. A practitioner can say what the discipline is made of, how the parts form one practice-and-knowledge whole, what remains the same through change, and which surrounding objects are merely used by discipline work. That makes comparison, transport, stewardship, and edition work inspectable without reducing the field to a document list or domain label.
Not this pattern when.
- Use domain or catalogue work for a subject-area label.
- Use the effective
ReferenceScheme,ClaimScope, and F.17 local-sense row for one local meaning. - Use F.9 only when a semantic relation is claimed between two exact F.17
SchemeSenseCellvalues from different semantic contexts and its predicate obtains. - Use C.2.1 and its publication or source-use patterns for one theory, standard, canon item, or their exact episteme edition.
- Use C.3, C.2.1, and direct historical or provenance relations for a school, variant, lineage, or classification.
- Use A.3.1 and B.1.5 for a Method or composite Method.
- Use A.15.1 for dated Work.
- Use A.22 for a selected relation organization.
- Use A.19.CPM for an actual comparison.
- Use C.21 for field-health characteristics.
- Use E.24.PUB for publication.
Use C.20 when determining, explaining, or relying on whether one exact field-level candidate has the required construction and continuity. The result may be true, false or unknown. Missing constructive facts leave the discipline claim unknown; they do not establish that the candidate is only a collection. Continue with the separately supported objects or collection when those suffice.
Problem Frame
Disciplines can persist while theories, standards, methods, organizations, publications, and vocabulary change. In practice that persistence is carried and refreshed through knowledge canons, codified practices and standards, and institutional carriers such as journals, professional bodies, curricula, committees, and laboratories, while none of those carrier families is identical to the discipline. FPF therefore needs a typed, provenance-preserving field-level composition account. That persistence cannot be explained by a stable name or by one fixed five-position card. It needs exact constituents, direct constructive part relations, whole-forming facts, a field assembly, a reidentification rule, and a whole characteristic that the composition produces or sustains.
The construction requires both reusable practice and knowledge. A bibliography without reusable ways of investigating or intervening is a knowledge collection. A method catalogue without field-level claim organization is a repertoire. An institution may sustain both through work, but the institution is a separately identified system. C.20 asks when exact parts and their couplings warrant one further whole.
Problem
Without a direct construction test:
- labels, organizations, document sets, curricula, or registries are mistaken for disciplines;
- a canon item or method is called a part merely because a community cites or uses it;
- rival traditions are either flattened into false consensus or treated as separate disciplines without an identity test;
- a changed standard, publication, or health score is mistaken for a new discipline edition;
- bridge loss, comparison admissibility, evidence strength, and selector policy leak into discipline identity;
- an optional selected structure is treated as a context holon, subdiscipline, or breadth count.
Forces
Solution - recover the whole from direct construction
Run the complete recognition test
U.Discipline is the admitted durable holon kind for a field-level practice-and-knowledge whole. It is neither a U.System nor a U.Episteme by default. C.20 supplies its construction criterion; A.1 recognizes a candidate under that kind when all six components below and the practice-and-knowledge condition hold.
Recover six constructive components for the same candidate:
- Exact candidate. Identify one exact
U.Entityand its proposed field boundary. A public name or description is only a designator or episteme about that candidate. - Exact constituents. Identify every claimed part under its direct kind and identity pattern. The construction must contain at least one exact knowledge-bearing contribution and at least one exact reusable-practice contribution; these are assembly contributions, not fixed part kinds or heterogeneous card positions.
- Constructive part relations and assembly. Recover each obtaining
disciplinePartOf(part, candidate)occurrence and every other exact whole-forming claim needed by the assembly. A list, co-use, adjacency, or common publisher supplies none of them. - Identity and reidentification rule. State which constituent, relation, boundary, or coupling changes preserve this candidate and which identify another whole or end the current one.
- Composition-grounded whole characteristic. State at least one exact A.17-governed Characteristic whose value or state is produced or sustained by the practice-and-knowledge composition and is not attributable to one constituent alone. Use A.18 for Scale legality and C.16 only when a value is measured.
- Larger-assembly compatibility. State the candidate boundary, exposed practice-and-knowledge interfaces, relevant whole characteristics, and identity-preservation conditions that make it admissible as a possible constituent under at least one governed larger field assembly. This establishes possibility, not an actual larger-discipline part relation.
The candidate is recognized as U.Discipline only when all six components and the C.20-specific practice-and-knowledge condition hold. A materialized classification assertion is a separate C.2.1 episteme about the candidate. Evidence can support that assertion and receiving work can rely on it, but neither creates the discipline.
Direct DisciplinePartOfRelation
C.20 directly governs DisciplinePartOfRelation, expressed in Plain register as disciplinePartOf(partEntity, candidateDiscipline). The first participant is one exact U.Entity already identified under its subject pattern. The second is the exact candidate U.Entity under the C.20 test; this participant meaning does not presuppose that the candidate has already passed the test.
The predicate obtains throughout an interval exactly when all of the following are true:
- the candidate's actual field assembly includes the exact part as a required contributor or a currently realized admitted alternative for one required knowledge-bearing or reusable-practice contribution;
- at least one exact whole-forming claim is true: that contribution connects to the candidate's field boundary, to another required contribution, and to the declared composition-grounded whole characteristic;
- the candidate's direct reidentification rule treats the current part, its contribution, and any allowed replacement as belonging to this continuing assembly.
Eligibility, citation or participation alone does not establish the required contribution and whole-forming parthood.
One occurrence is identified by <exact part entity, exact candidate, maximal continuous obtaining interval>. If the same part leaves the assembly and later returns, the two episodes are distinct occurrences. A changed observation or evidence window does not split an ongoing occurrence; actual cessation and resumption under the direct predicate do. Reidentifying either participant also identifies another occurrence.
Multiple contribution claims for the same part do not create several part occurrences during the same continuous interval. If typed declaration reuse is needed, the direct relation has only the two participant meanings above; other contribution claims and qualifications do not add participant slots. Qualifiers and whole-forming claims keep their subject patterns.
State whole-forming claims before choosing notation
Work from the field question to the contributions that can answer it. Which distinctions in the knowledge change a practitioner's investigation or intervention? What result of doing that work can count against a claim or lead to its revision, and how would practitioners recognize it? A trace through one investigation can identify a coupling. For an account of the whole field, explain how the coupling holds over the intended work or recover the further contributions and connections that work needs. Where practitioners use an outside result, establish separately whether the knowledge or method that supplied it belongs to their field. For a proposed field, these connections can guide the design. To describe an existing field, establish which connections operate. Describe the resulting assembly in the language of the field.
- which exact claim-bearing contributions supply the field's knowledge commitments, distinctions, explanatory resources, or admissible questions;
- which exact methods or other independently governed parts supply reusable ways of investigating, designing, intervening, evaluating, or learning;
- which claims state how knowledge contributions constrain, explain, or qualify practice contributions;
- which claims state why separately governed results of actual practice are relevant to evaluation, revision, or replacement of knowledge contributions;
- how incompatible or rival contributions coexist without being silently equated;
- which boundary, stop conditions, exposed interfaces, and substitution conditions keep the assembly one field-level whole.
Each whole-forming statement stays at the lightest truthful disposition supplied by its direct relation pattern or A.6.RCD: an existing direct predicate, a local compound claim, or a reusable predicate-definition episteme. A readable arrow such as uses, supports, tests, belongs to, standardizes, or aligns does not admit a relation kind and does not make a part relation obtain.
The assembly rule names the exact current part relations, contribution meanings, whole-forming claims, permitted alternatives, incompatibilities, boundary conditions, failure conditions, and substitution conditions. A C.13 Gamma_m.sum construction trace may report those facts when a named use needs an inspectable account. The trace is a C.2.1 episteme and creates none of the parts, relations, assembly, identity, or characteristic.
A construction account lets the reader review the field as a whole, trace the provenance of its claims, and use it in separately governed publication and admissible comparison. A Discipline Card or another readable form can present this account. Its completeness depends on the parts, obtaining relations, assembly and identity grounds above.
Identify canon contributions as exact epistemes and claims, and distinguish standard epistemes, Methods and practice claims. Identify the organizations supporting the discipline as systems, and recover their relevant system-role kinds, assignments and Work separately. Identify semantic crossings through exact F.9 occurrences and separate bounded-use propositions, and comparison, evidence, assurance and acceptance policies through their declarations. Section 4.6 supplies intake prompts and subject patterns for these objects. Their place in a card or argument list establishes neither discipline parthood nor identity.
Identity, continuity, and change
The discipline's identity is not the extensional part list. It is the exact candidate with its field boundary, practice-and-knowledge assembly principle, obtaining constructive relations, whole-forming architecture, and declared whole characteristic under one direct reidentification rule.
The same discipline may continue through a permitted canon revision, method replacement, institutional change, publication change, or temporary part-relation change when the rule admits that variation and the field boundary, required contribution meanings, essential couplings, and whole characteristic remain within their declared continuity conditions. The rule must say which substitutions are permitted and how the replacement contribution reconnects to the assembly.
To maintain a discipline, start with the work its practitioners must continue to do. Follow how relevant knowledge guides that work and how results reach the people who can test or revise the knowledge. Find which connections still work under present conditions and where the connection is at risk or has broken. Continuing a sound practice can be sufficient. If results have stopped reaching practitioners, restore that exchange. If the knowledge or method itself needs revision, establish how the revised contribution will work with the others. For a commission to sustain the field in operation, carry out the needed work and establish that the connections still serve it. For an inquiry about continuity, evidence of existing practice can already answer the question.
A change outside those conditions identifies another candidate or leaves the stronger claim unresolved. When a receiving use asks whether the old whole can still explain the case, run B.2's existing-whole explanation check after the direct C.20 facts are recovered. Loss of evidence, another description edition, a renamed field, or a stale registry entry does not itself end or create a discipline.
"Discipline edition" can concern different objects. Recover the changed subject:
- revised canon, standard, method description, classification assertion, or discipline description is another C.2.1 episteme when its identity discriminator changes, with historical continuation tested separately;
- a newly available edition has its own E.24.PUB publication occurrence, particular expression, carrier, audience, use, and availability interval;
- an unchanged discipline during a proper interval may be described through the direct temporal and A.14 phase route;
- a changed field assembly outside the reidentification rule requires another discipline candidate, not a renamed description.
Identify a reusable publication form and its conformance separately when the current use relies on its rules. Keep the change rationale as claim content of one exact C.2.1-identified revision or decision episteme.
When historical continuation is asserted, the assertion about the exact EpistemeEditionRelation cites the exact source use, applicable continuity policy or rule, and the preserved and deliberately changed claim, EntityOfConcern, and scheme features that the rule consumes. Revision Work, Method, provenance, change facts, and evidence are case facts; no label establishes continuity.
When availability changes, trace the publication transition through the ending and beginning E.24.PUB occurrences with their selected edition, audience, bounded use, particular expression, carrier, and availability intervals.
The rationale, edition-relation assertion, publication occurrences, and discipline reidentification result remain distinct; none decides another merely by being recorded.
Require a composition-grounded whole characteristic
Name at least one exact Characteristic under A.17, its Scale under A.18, and its direct measurement or evaluation route only when a current use needs a value. A useful local choice is the degree to which the assembly sustains a replayable practice-knowledge coupling: independently identified knowledge contributions constrain field practice, and separately governed practice results can be used to test, revise, or replace those knowledge contributions under declared rules. This is a local characteristic choice, not a new universal kind or scalar score; C.16 governs a measurement chain only when dated measurement Work attributes a value.
The value must depend on the composition. Citation count from one episteme, frequency of one method, size of one organization, bridge count, publication count, or a single maturity label is not a composition-grounded whole characteristic. Rival traditions can coexist when their boundaries, incompatibilities, permitted uses, and contribution relations are explicit; cohesion does not require false consensus.
C.21 is the pattern for discipline-health characteristics such as reproducibility, standardization, diversity, and disruption balance. A health reading can reveal pressure to inspect construction or reidentification, but it is not a part, assembly fact, identity rule, or classification result.
Keep neighboring objects with their subject patterns
Keep the three boundaries explicit. A domain is a subject-area or catalogue designation. A semantic locality is the bounded use of exact local meanings under its effective ReferenceScheme, ClaimScope, and direct F.17 pattern; it is not another holon or a discipline constituent by locality alone. A discipline is the exact field-level practice-and-knowledge whole recognized by the C.20 construction test.
The same or similar word across domains or semantic localities establishes none of shared meaning, one discipline, parthood, or identity. When a use relies on a semantic crossing, state the two exact local senses and cite an F.9 Bridge only if its predicate obtains; field or locality difference alone creates no Bridge. Then test each discipline candidate independently; even an obtaining high-congruence Bridge does not merge its endpoints.
Use these intake prompts to recover candidate contributions and surrounding dependencies, keeping each item under its subject pattern:
- canon candidates: theories, models, reference works, definitions, proof traditions, benchmark descriptions, and other epistemes treated as canonical;
- practice and standard candidates: accepted Methods, norms, standard procedures, measurement conventions, and admissible comparison rules;
- institutional and organizational candidates: journals, committees, curricula, professional bodies, laboratories, and other institutional arrangements that may carry or refresh field work;
- cross-locality candidates: exact F.9 Bridges, F.17 term or local-sense rows, Bridge descriptions, and loss observations used across semantic localities or source traditions;
- comparison candidates: exact Characteristics, Scales, Units, comparators, evidence policies, and CG-Spec declarations used to make a named comparison admissible.
The palette prevents those reader questions from disappearing; it is neither a required five-part decomposition nor a constructor. Filling every category is not required. Every item remains under the subject pattern below and becomes a C.20 part only when disciplinePartOf independently obtains.
Any entity in this table can become an actual part only when it is independently identified and the C.20 predicate separately obtains. Its ordinary association with the field is never enough.
Optional bounded-model-use structure
Discipline work can use an independently selected BoundedModelUseStructure when the selected organization of exact relations changes how a named model is interpreted or used in that work. State the exact model, direct relation occurrences, selection basis, receiving work or claim, and the organization that matters. A.22 is the pattern for the structure and A.15.1 is the pattern for the dated work.
Selection establishes neither discipline parthood nor discipline identity; use the C.20 construction test for those claims.
Breadth, tradition, lineage, and school claims
Applied, multidisciplinary, transdisciplinary, tradition, lineage, and school labels do not carry a universal C.20 classification rule.
- For a project-local kind, use C.3 to state exact intent, membership predicate, scope, and the candidate disciplines that satisfy it.
- A multidisciplinary claim normally needs several independently identified disciplines and exact contribution or use claims; co-listing fields or counting viewpoints is insufficient.
- In an interdisciplinary project, identify the result that depends on combining contributions from distinct disciplines. For a transdisciplinary principle or Method, identify the relation or operation that practitioners can interpret in different eligible disciplines and the conditions each interpretation must preserve (C.39.WA:4.3). A project can use such a principle within one discipline or while combining several. If the claim concerns the formation of a new discipline, test that candidate under section 4.1.
- A tradition, lineage, school, variant, edition, or provenance claim identifies its exact subject and the exact historical-continuation, source-use, method, similarity, edition, derivation, or provenance relation under that relation's subject pattern. Tradition or lineage can organize variants or editions within or across disciplines as an ordinary auxiliary value or a C.3 project-local kind; none is a discipline part, subkind, or public U-kind by label. A public kind requires its own direct governor, identity and use rule, and E.24.UK admission.
- An applied classification states what application-facing criterion is satisfied. Method use, one project, one organization, or one selected structure does not establish it by itself.
The classification can remain a C.2.1 claim without admitting another public U-kind. A changed label or classification result does not reidentify the discipline unless the direct C.20 rule independently says the field assembly changed.
Crossing, comparison, evidence, health, and publication are conditional branches
Open these branches only when the named discipline use needs them.
Cross-field or cross-sense use. A cross-field use may use independently governed claims without a semantic crossing. When an exact cross-context semantic relation is claimed, first resolve two exact F.17 SchemeSenseCell values; F.9 identifies an obtaining Bridge only when their semantic-context projections differ and the direct predicate and dependencies are satisfied. A field, scheme, cell, or plane difference alone creates no Bridge. For an obtaining Bridge, keep a separate current C.2.1 bounded-use proposition whose EntityOfConcern is that Bridge and whose ClaimGraph names the proposed action u, exact direction d, use-specific correspondence rule r, tolerated semantic loss t, and affirmative or negative polarity. Observed loss and counterexamples remain evidence; the permitted-loss tolerance remains claim content. Neither one reidentifies the Bridge.
For ordinary evidence reliance, recover the descriptive A.10 evidence-provenance path and the independently established direct relations it cites for the same bounded use. Only RelianceDisposition=pass supports that affirmative use; degrade supports only the named narrower use, while abstain, reopen, evidence-needed, assurance-needed, or blocked-current-use does not pass the attempt. Use B.3 only when a named assurance claim is current, and require its result for that same bounded assurance use. Authorization and any actual comparison, substitution, translation, publication, or Work occurrence remain with their subject patterns.
Crossing visibility and presentation checks. Keep an exact Bridge and separate bounded-use proposition discoverable in an account that relies on that obtaining semantic relation. For a ReferencePlane-only crossing, keep the applicable plane relation and policy discoverable instead; if both facts are current, keep both under their own predicates. Materialize an E.18 CrossingBundle only when an independently current E.18 structural GateCrossing exists and a named downstream use relies on durable evidence for that structural crossing; if it also relies on cross-semantic correspondence, add E.18's separate F.9 block. E.17 governs publication packaging and A.21 governs any actual GateCheck or GateDecision.
For lane purity, inspect the current B.3 result by value: every characteristic keeps its bearer and scale, and any aggregation cites its domain model and assumptions; there is no universal F-G-R-CL fold. Apply E.10 lexical checks to the published labels, keep normative prose notation-neutral, and treat every discipline column as didactic only.
Comparison or aggregation. Identify the exact Characteristics under A.17 and their Scales, applicable Units and legal operations under A.18; C.16 supplies measurement results when the inputs are measured. A mean over an ordinal Scale and an aggregation that mixes unconverted or incommensurable Units are inadmissible: stop that operation, establish a lawful transformation or comparison scheme, or retain the separate values. When using A.19.CPM, bind its required profile pair, comparator, claim scope and selected slices, predicate when used, reference scheme and plane, evaluation point or interval, and evidence policy; retain its set-valued result. CPM does not aggregate. A numeric or profile aggregation uses its own justified combining rule and recoverable inputs and contributors; a selected A.19.ULSAM or B.1 Γ-fold retains that interface's arguments and result binding.
Ordinary comparison or folding mathematics needs no asserted dated U.Work. When the account asserts dated comparison or aggregation Work, or the receiving use consumes it, recover its A.13 core and independent A.15.1 admission; use F.6 only for precise assignment-bound attribution. A calculation application and its qualification window do not by themselves establish a dated Work occurrence.
When a comparison or aggregation is governed by a G.0 CG-Spec, its pre-operation check names the Characteristic ids, ScaleComplianceProfile with applicable Unit, polarity and legal-operation conditions, MinimalEvidence, and the admitted comparator or declared Γ-fold and contributors. Missing or unknown required declarations or evidence block that governed operation. Ordinary comparisons and aggregates still need compatible inputs and a lawful operation; they need no CG-Spec unless the receiving use selects that interface.
If the operation relies on an obtaining F.9 Bridge between exact F.17 SchemeSenseCell values, use the separate Bridge, bounded-use-claim and reliance branch above. A ReferencePlane crossing instead uses its applicable relation and policy; when both relations matter, retain both. These relations do not supply the operation's scope, comparator, inputs or result.
Evidence, currentness, and assurance. Use A.10 for source use and the evidence-provenance path, and G.11 for selected-edition currentness. When an imported source claim supports a construction, classification, comparison, publication, or assurance use, keep the exact source and edition plus only the lane tags, freshness window, or validity conditions that the receiving use consumes. Use this apparatus only for claims that need such reliance.
Open B.3 only for a named assurance claim. Its AssuranceResult identifies the target claim, assurance use, basis, disposition, limits, and reopen condition. If the argument consumes a characteristic or calculation, name its bearer, property, scale, unit, interpretation, basis, dependency model, assumptions, rule, and rival as applicable. C.20 supplies no default F/G/R/CL lanes, weakest-link fold, SpanUnion, congruence penalty, plane penalty, or assurance arithmetic. A Bridge-related penalty is current only when a named B.3 assurance claim uses an applicable declared domain model whose current facts and rule yield it; that value belongs only to the calculated AssuranceResult for that named use.
State ReferencePlane only when it changes an exact input or interpretation; when an exact plane crossing is current, cite the applicable plane relation and policy. The selected plane and its relation identify neither a discipline part nor a part relation.
Insufficient or unknown support narrows, abstains, requests evidence, reopens, or blocks the attempted assurance use under B.3; it does not make a part relation or whole obtain or cease.
Health or state view. C.21 can evaluate typed discipline-health characteristics. Local values such as emerging, consolidating, codified, or fragmenting need their defining criteria and support, with Scale, observation period and currentness conditions as required by the claim. A decision threshold belongs to the applicable decision rule; represent it as a G.4 AcceptanceClause when that formal acceptance interface is used. A state label, health vector, score, or dashboard transition does not itself reidentify the discipline.
Description and publication. The claim-bearing whole expressed by a discipline description, construction trace, comparison report, health series, card, or registry row is identified as an episteme under C.2.1 when that pattern's constitution test holds. E.24.PUB governs any actual publication occurrence and keeps selected episteme edition, audience, bounded-use declaration, particular expression, carrier, and availability interval distinct. Updating or publishing the account changes neither the field assembly nor its past.
Archetypal Grounding
System/Episteme contrast
The examples distinguish the field candidate from a system, episteme and the relations used to describe its construction. Each cell retains its own subject pattern; the table is not an assembly rule.
Safety engineering as a positive construction
Suppose SafetyEngineering-SE is one exact field candidate. C.2.1 independently identifies HazardCausalityCanon-v5 as a claim-bearing episteme. A.3.1 independently identifies HazardAnalysisMethod-v4 and IncidentLearningMethod-v2 as reusable Methods. None is a C.20 part yet.
The field assembly makes these exact contributions current:
HazardCausalityCanon-v5supplies the hazard, causal, and acceptable-argument meanings used by the two Methods;HazardAnalysisMethod-v4supplies the reusable analysis practice by which those meanings constrain identification and treatment of hazards;IncidentLearningMethod-v2supplies the reusable practice by which separately governed incident and evaluation results can challenge or refine the canon claims.
The three exact relation occurrences are disciplinePartOf(HazardCausalityCanon-v5, SafetyEngineering-SE)@I-SE-core-1, disciplinePartOf(HazardAnalysisMethod-v4, SafetyEngineering-SE)@I-SE-core-1, and disciplinePartOf(IncidentLearningMethod-v2, SafetyEngineering-SE)@I-SE-core-1. They share the maximal current continuous obtaining interval I-SE-core-1 = [2024-04-01T09:00Z, open): its left boundary is the dated field-assembly activation at which all three contributions first became jointly required.
At the continuation check 2026-07-31T23:59Z, each contribution remains required, the canon meanings remain connected to method preconditions and result meanings as stated by the exact whole-forming claims, the same candidate and reidentification rule continue, and no cessation boundary has occurred; therefore the right boundary remains open. If any contribution ceases and later resumes, that part's interval ends at cessation and a new disciplinePartOf occurrence begins at resumption. The field assembly requires those contributions, and the reidentification rule admits them as current parts.
The local A.17 Characteristic PracticeKnowledgeCouplingReplayability has the exact subject SafetyEngineering-SE and the ordinal A.18 Scale broken < partial < replayable: its criterion asks whether the named canon-to-Method constraints and result-to-canon revision routes remain executable across the declared safety-work classes. No constituent can have that whole value alone.
If a current use needs a measured value, C.16 requires the exact measurement subject, method, model, Scale, dated measurement Work, uncertainty, and result episteme; the label alone supplies no reading.
The reidentification rule permits a compatible canon revision or replacement Method only when the field boundary, causal and argument meanings, practice-knowledge feedback, and whole characteristic remain within declared continuity conditions. Removing the feedback route, changing the governing safety concern and field boundary, or replacing the assembly by unrelated document and method lists falls outside the rule.
Safety Journal, a standards committee, a university curriculum, a laboratory, a bridge to resilience-engineering terminology, a comparison CG-Spec, and a published discipline card remain separately governed. This example establishes no C.20 part predicate for them, so their parthood remains unresolved. A positive or negative part claim needs the corresponding C.20 grounds.
For A.1's larger-assembly test, the separately identified rule episteme EngineeringFieldAssemblyRule-v3 describes a governed larger-field construction whose applicability requires one constituent field to expose stable hazard-constraint meanings, analysis-result meanings, and identity-preserving interfaces to design and verification practices. SafetyEngineering-SE has those interfaces and its reidentification rule preserves them, so it is compatible with that possible assembly while remaining the same candidate. The rule episteme does not create the compatibility facts, and this result asserts no actual disciplinePartOf(SafetyEngineering-SE, Engineering-E) occurrence.
A C.13 Gamma_m.sum trace may report the candidate, three parts, three part-relation occurrences, whole-forming claims, assembly, reidentification rule, and whole characteristic. Writing or publishing that trace does not construct SafetyEngineering-SE.
A department, corpus, and method registry are not yet a discipline
Safety Department A performs research and teaching Work. Its repository contains standards and papers, and its registry lists analysis Methods. Those facts identify the organization, its research and teaching Work, the standards and papers, and registry membership under their subject patterns. Any publication occurrence additionally needs its own audience-availability basis under E.24.PUB.
They leave the field constituents, disciplinePartOf occurrences, practice-knowledge assembly, whole characteristic and reidentification rule unresolved. The discipline classification is therefore unknown on these facts. The recovered collection and work organization may suffice for the current question; do not treat filled card positions as grounds for a positive classification.
Cross-field reuse does not create a transdiscipline
A safety-engineering team proposes to reuse one resilience-engineering term. F.17 resolves the two exact SchemeSenseCell values, and their semantic-context projections differ. The stated directed BridgePredicateProfile applies to those endpoint readings at its stated as-of basis; the current endpoint meanings satisfy its relation-kind-specific semantic condition and Boolean truth condition; and all required dependencies are present. F.9 therefore identifies an obtaining directed Bridge. A separate bounded-use proposition states the direction, mapping rule, tolerated loss, and polarity for this hazard-review use. Review Work may rely on that proposition through A.10, and a comparison may apply A.19.CPM.
The directed correspondence relates the two meanings of the term. To rely on that correspondence in the hazard review, the bounded-use proposition must support this use in the stated direction, and the A.10 reliance check must permit it. If the teams combine their methods to produce this review, the joint result can establish interdisciplinary work. A resilience principle has transdisciplinary reach when the same relation admits warranted interpretations in different disciplines under stated conditions. To show that the safety team used it here, identify which distinction, inference or choice in its hazard review depended on the principle. Evidence from the second discipline may already establish the wider reach. To establish a new discipline, identify the candidate whole and examine all six construction components in section 4.1.
Canon revision, same field, and another field
HazardCausalityCanon-v6 changes claim content and is another C.2.1 episteme. Historical continuation from v5 is tested separately. The same SafetyEngineering-SE can continue if its C.20 rule permits that replacement and the required contribution meanings, whole-forming couplings, boundary, and whole characteristic persist. The old part occurrence ends and the new one begins; a description or evidence window does not decide those intervals.
If the replacement rejects the field's former causal object, changes the practice-knowledge feedback rule, and changes the field boundary beyond permitted continuity, the same-discipline claim fails. The project then tests another candidate and uses B.2 only if a receiving use needs a whole-reidentification conclusion. A new canon edition, renamed department, or republished card alone decides neither branch.
Bias-Annotation
These questions expose different ways a field account can acquire unsupported meaning.
Scope. Every use is bounded by exact field candidates, semantic localities, scopes, editions, and receiving work. There is no context-free or global discipline established by name, popularity, institutional reach, or a shared vocabulary.
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Consequences
Benefits. The construction account distinguishes constituent replacement from reidentification without freezing a universal canon or Method repertoire. It preserves rival traditions through their explicit contributions and admissible directed uses. Comparison, selection and publication can use that account and their own characteristic, evidence or authority grounds; they do not supply missing construction facts.
Costs and trade-offs. A positive discipline claim needs exact parts, direct relation intervals, whole-forming claims, an assembly, a reidentification rule, one whole characteristic, and a larger-assembly compatibility account. Ordinary use can state these in readable sentences. A comparison adds its comparison basis and lawful operation; a selected CG-Spec adds that interface's declarations. Cross-local use that relies on an obtaining semantic Bridge adds its bounded-use claim and reliance grounds. Trace, publication, health, selector or assurance apparatus is added for the receiving use that needs it.
Risks avoided. The pattern prevents a field from being inferred from a label, identity from an institution, parts from publication, breadth from context counts, merger from a Bridge, ontology from a health score, and method or discipline membership from a registry.
Rationale
A discipline is durable because a governed practice-and-knowledge assembly can continue through permitted replacement, not because all surrounding artifacts remain fixed. Direct part relations establish which exact entities are constitutive during which intervals. Whole-forming claims explain their contributions and coupling. The assembly and reidentification rule then distinguish continuity from a different field, while the composition-grounded characteristic shows why the candidate is a whole rather than a collection.
Keeping organizations, Work, methods, epistemes, structures, Bridges, comparisons, publications, evidence, and health readings separate lets each object affect the discipline claim through its actual relation, without making it an automatic part or identity discriminator.
This separation keeps every discipline-related claim local to an exact EntityOfConcern and effective ReferenceScheme, comparison admissible only under its current Scale and comparator declarations, and ordinary evidence-bearing reliance explicit under A.10 and any actual named assurance claim explicit under B.3. It also lets plural traditions, obtaining directed Bridges with use-specific tolerated loss and any material observed loss kept visible, typed health readings, and the grounds for verification, validation and correspondence of meanings remain inspectable together.
Scale compliance, B.3's requirement for an applicable domain model and stated assumptions, and Bridge hygiene stay with their subject patterns: C.20 adopts their constraints for a named use but defines no universal field score or reliability fold. Charisma, prestige, institutional reach, or a field name cannot substitute for the required construction, Characteristics, and source evidence.
SoTA-Echoing
Currentness and reopen. Reopen only the affected construction, identity, crossing, or health decision when its direct source interface changes. A newer discipline study can change the local breadth criterion or worked case without weakening the exact-part, direct-relation, assembly, and reidentification boundary.
Relations
Builds on. A.1 for constructive holon recognition; A.6.REL for relation-occurrence discipline; A.14 and direct part patterns for mereological meaning; C.13 for an optional construction trace; A.17 and A.18 for the composition-grounded Characteristic and Scale, with C.16 only for an actual measurement chain; E.24.UK for public-kind admission.
Coordinates with.
- C.2.1 for canon, standard, classification, description, trace, and result epistemes.
- A.3.1 and B.1.5 for Methods and method composition.
- A.15.1 for dated discipline work.
- A.22 for optional selected structures.
- B.2 for a separately current whole-reidentification question.
- C.3 for project-local breadth and tradition classifications.
- C.22 for selector-facing problem typing and TaskSignature assignment.
- C.23 for MethodFamily evidence and maturity.
- E.10 for lexical precision.
- F.17/F.18 for local-sense term publication and durable naming.
Conditional branches. F.9 for exact Bridges; A.10 for a recoverable evidence account and B.3 for a named assurance claim; C.21 for discipline health; A.19.CPM and G.0 when their comparison and CG-Spec interfaces are used; G.4 for formal acceptance clauses; E.24.PUB for publication; G.5 for selector use of independently grounded discipline or Method claims.
G.2 source-pack boundary. When one current G.2 harvest or synthesis informs a discipline account, preserve the anti-monoculture intent by keeping plausible rival lineages and the semantic localities relevant to that named synthesis visible, with an explicit coverage or omission rationale and explicit crossings where needed. C.20 sets no minimum number of Traditions, local frames, cards, contexts, or matrix rows; G.2 alone governs conformance of its pack. A plural or conforming pack, TraditionPalette, or BridgeMatrix supplies neither a discipline construction nor a part relation.
Constrains. A C.13 discipline trace consumes C.20 facts and creates none. C.21 evaluates an already identified discipline.
G.5 may consume an independently identified Method together with exact discipline-relevant construction or classification, Characteristic, and evidence or assurance claims only through its own declared predicate. An ungrounded discipline name supplies none of field belonging, comparison value, evidence strength, selection, or truth.
A registry or family row, selector request, policy, result or shortlist, CG-Spec reference, and EvidenceGraph row can make an actual G.5 use auditable; none lets G.5 infer Method membership, discipline parthood, actual selector application or selection, or claim truth.
A selected structure, Bridge, comparison, publication, evidence graph, health reading, or label never supplies a missing C.20 part occurrence.
C.20:End
Field Health & Structure (Discipline-CHR)
Status: Stable Type: Pattern
Purpose. Characterize the health, maturity, and structure of a scientific or engineering discipline in a typed, reviewable way, without collapsing the result into taste, anecdotes, a dashboard view, an audit label, or one score. C.21 defines discipline-health Characteristics and the conditions under which their readings may be compared. It does not make a dashboard, publication, record, or Work occurrence part of the discipline.
Use This When
Use this pattern when a team must say something practical about the health, maturity, or structure of an already identified discipline. Typical questions concern reproducibility, formal standard status, actual adoption, cross-tradition alignment, disruption balance, evidence resolution, diversity, engineering-claim recoverability, or pressure to mistake representations for their subjects.
What goes wrong if missed. Field-health claims become labels: incompatible readings are compared, ordinals are averaged, formal approval is treated as adoption, entropy and concentration are read in the same direction, stale evidence looks current, and a dashboard or standards list is treated as the discipline or as proof of health.
What this buys. The practitioner can state and interpret a bounded health claim, compare compatible readings and identify what additional measurement or correspondence a proposed use needs.
First useful move. Start with the question about the field and what the recipient will do with the answer. Which difference in practice would change that answer or its explanation? Choose a Characteristic that can reveal it, the population or corpus it concerns and the relevant period. If the recipient needs several conditions to hold together, establish which must hold for the same units or at the same time. Report the supported answer. A single reading or a bound can finish the work when it decides the question.
Not this pattern when. Use C.20 to decide whether the candidate is a discipline, C.16 to construct a measurement, A.19 for comparison or aggregation work, F.9 only when the use relates distinct local senses, E.24.PUB for audience availability, G.12 for a reusable dashboard view, and G.4 for a selected formal acceptance interface. C.21 supplies none of those results merely by naming a health Characteristic.
Problem Frame
A discipline can persist as its constituents and associated epistemes, practices, standards, institutions, and Work change. Teams routinely say “replication is improving,” “the field is fragmented,” or “standards are converging.” Such a sentence can be useful before it becomes a dashboard row, but any relied-on comparison needs exact Characteristic, Scale, scope and comparison-basis semantics, with measurement conditions when measured readings are used.
C.21 treats health as separately typed coordinate claims, not one scalar health value. A threshold or preferred band belongs to the receiving decision rule rather than the Characteristic; use G.4 when its formal acceptance interface is selected. A dashboard represents already constituted claims.
Problem
Five recurrent failures make discipline-health claims unreliable:
- Object collapse. A definition set, Method, MethodDescription, measurement Work, result episteme, series episteme, publication occurrence, form, and carrier are called one “DHC artefact.”
- Scope slippage.
ClaimScopeand a selectedTargetSliceare treated as interchangeable, although the scope states where the claim holds and the slice is only an optional computation or publication input. - False crossing. Different sources or editions are assumed to require a Bridge even when C.16 direct-comparability conditions hold; conversely, distinct local senses are compared without their obtaining F.9 relation and loss account.
- Scale collapse. Formal recognition is ordered with adoption, alternative ratios are called one Characteristic, or entropy and HHI are placed in one field despite opposite directions.
- Assurance inflation. A cheap readable claim is forced through evidence graphs, registries, dashboard pins, and publication machinery that its receiving use does not consume.
Forces
Solution — Discipline Health Characterisation (DHC)
The objects used by DHC
“DHC” names this vocabulary and method of use.
Rendering, measuring, assembling a series, uploading, and maintaining availability may each be Work when performed. A work record or carrier does not make that Work occur.
One replay basis for every persisted coordinate
Every persisted, compared, aggregated or published C.16 measurement coordinate makes the following values recoverable. This is a field group, not another public kind. An ordinary non-measurement health claim retains its own support; writing or publishing it does not create a measurement chain:
DHCReplayBasis := <DisciplineRef, IntendedUse, ClaimScopeRef, ComparisonBasis, CharacteristicRef.edition, ScaleRef.edition, UnitRef.edition?, DHCMethodRef.edition, MethodRef, MethodDescriptionRef.edition?, MeasurementModelRef.edition?, CalibrationBasisRef?, TimeOrPopulationBasis, DHCDefinitionSetRef.edition?, TargetSliceRef?, DistanceDefRef.edition?>
DHCMethodRef.editionresolves the same Characteristic, Scale, Method, MethodDescription, model, calibration, and uncertainty semantics named by the active fields. A mismatch is not repaired by choosing one field as “primary.”DHCDefinitionSetRef.editionappears only when a named reusable definition selection exists.TargetSliceRefappears only when the named computation or publication consumes an A.2.6 selection. Every selected slice must be shown to belong to, or otherwise be covered by, the authoritativeClaimScope; the slice never substitutes for that scope.DistanceDefRef.editionappears only when the Scale comparison or target-distance rule uses a separately declared distance.- Evidence paths, lane tags, currentness, assurance, acceptance, public names, and publication refs are added only when the receiving use consumes those separate results.
Portable Characteristics
Each bullet below names one exact Characteristic and Scale family. A DHC use selects only the coordinates needed by its question.
-
ReproducibilityRate — ratio in
[0,1], Unitreplicated_claims/tested_claims, for a nonempty tested population. Identify which claims were actually tested and how they entered the denominator. Declare the protocol, independent-team condition, source corpus or cohort and time window. A higher ratio means a greater share replicated among the claims tested. An estimate for a wider population also needs grounds for extending the result to the untested claims. -
FormalRecognitionStatus — nominal by default. Values such as
none,draft,approved,withdrawn, or another lifecycle vocabulary belong to one named standards body and exact status scheme. Use an ordinal only when that scheme itself supplies a lawful order. There is no generalde facto < de jureladder and no default health polarity. -
PracticeAdoptionRate — ratio in
[0,1], Unitadopting_units/eligible_units, for a nonempty eligible population. Identify the eligible units and state which units enter each reported denominator. Give the adoption criterion, window and treatment of partial adoption. With incomplete observations, distinguish confirmed adopters, confirmed non-adopters and units whose status is unknown. A rate calculated over respondents answers a question about those respondents. To estimate adoption across the eligible population, account for how units were observed and what is known about the others. -
AlignmentDensity — ratio; Unit
obtaining_relations/100_compared_cells. Count only exact obtaining directed F.9 relations in the declared F.17 cell set. For each counted relation, state its orientation and admitted-use qualifier; keep observed loss in its evidence account, separate from any bounded-use claim under section 4.2. A higher value means denser declared alignment for that set; any preferred health band needs a receiving criterion, using G.4 when formal acceptance is selected. -
DisruptionBalance — interval reading over one exact disruption/consolidation method and corpus. A receiving use may select target-is-best polarity with an explicit target band and distance rule. The raw reading supplies no preferred band; use G.4 when formal acceptance is required.
-
EvidenceUnitResolution — ordinal, compare-only, under one exact segmentation scheme whose levels are nested, for example
artifact < section < claim < subclaim. Higher means a finer addressable unit under that scheme. It does not say how many claims an artifact contains or how densely claims are supported. -
ClaimsPerArtifact — ratio; Unit
claims/artifact, with exact claim segmentation and artifact population. It measures claim breadth or packing, not support density. Declare a target band when the use needs one; no universal monotone health polarity applies. -
SupportAnchorsPerClaim — ratio; Unit
anchors/claim, with exact anchor admissibility and claim segmentation. It measures support-anchor density, not claim size. It has no universal monotone health polarity. -
TraditionShareEntropy — one exact entropy Characteristic and Scale, with log base, normalization, category set, and population fixed. Higher entropy means greater dispersion on that scale. Any desired band is separate.
-
TraditionShareConcentration — HHI or another exact concentration Characteristic, normally ratio in
[0,1]; higher HHI means greater concentration and therefore lower dispersion. Do not place it in the entropy field.1 - HHImay be introduced only as an explicit transformation to a separately declared receiving Scale. Comparing that result with normalized entropy still requires an explicit common comparison rule.
Engineering-grade extension Characteristics
A discipline-health use may add these coordinates when its question needs them. They do not become evidence, assurance, gate, release, Work, or project-authority results.
-
EngineeringClaimJustificationRecoverability — ordinal, polarity higher-is-more-recoverable. It asks whether the exact construction, source, model, lens, or relation carrying an engineering claim's force can be recovered for the intended use. The reading cites the direct pattern and rule that define or constrain that force.
-
SemioSubstitutionPressure — ordinal or ratio as separately declared, polarity lower-is-less-substitution-pressure. It asks how often a representation, fluent wording, record, dashboard, view, or source chain is mistaken for its engineering subject, relation, or claim.
When either extension is active, explain the claim kind or use boundary and the direct pattern or rule that supports the reading. Include a limit, prohibited use or reopen condition when it changes how the result may be used.
Comparison and legality rules
- Direct same-semantics comparison. Compare readings under the same Characteristic and compatible Scale semantics and scope. Measured readings also retain C.16's applicable measurement-definition, model, calibration, Unit and time or population conditions. State the basis used; different source labels or editions alone require no Bridge. For a claim about the whole field based on part of it, give the sampling or measurement assumptions that support the extension, or report what holds for every admissible status of the unobserved units. Restricting a report to the observed population changes the population it describes. A new observation may settle an answer under the existing model. For strict model refinement under B.5.MR, the new model must preserve the earlier assertions under the shared interpretation, admit at least one possible case and exclude an answer the earlier model allowed for the same inquiry.
- Cross-local comparison. When a comparison relies on a semantic relation between distinct F.17 local senses, cite the exact obtaining F.9 Bridge only after its predicate and dependencies hold. Keep its separate C.2.1 bounded-use claim, with the use, direction, correspondence rule, permitted-loss tolerance, and polarity, distinct from observed loss and current reliance. Follow F.9's A.10 reliance branch or, for a named assurance claim, its B.3 branch for the same bounded use. If the assurance use requires an R adjustment, cite the applicable domain model and keep the adjustment in that use's AssuranceResult. The relation supplies none of ClaimScope, measurement, comparison, or acceptance semantics.
- Reference-plane crossing. When a reading is used across distinct world, concept, or episteme planes, cite the exact crossing basis. Any R adjustment follows the named assurance-use and domain-model condition in rule 2. A dashboard row or source label does not establish the crossing.
- Cross-scale transformation. A conversion, normalization, distance, or aggregate names its exact Method, Scale, legal operation, and loss or uncertainty. No common scale is inferred from similar labels.
- Freshness. A persisted or reused coordinate carries its observation window and applicable currentness rule. Staleness leads to the receiving pattern's degrade, abstain, or reopen result; it does not rewrite the historical measurement.
- Target bands. “Target-is-best” is not “higher-is-better.” A comparison to a band uses an explicit distance-to-band rule and keeps the receiving acceptance criterion separate from the raw reading.
Three progressive uses
Minimal readable health claim
State the discipline, intended use, one Characteristic, the ClaimScope, comparison or observation basis, time stance, and ordinary result. Name the support relied on. Stop when this answers the working question. Do not require an EvidenceGraph, UTS row, registry, dashboard, definition set, series, or publication occurrence.
Measurement, comparison, or aggregation
Open C.16 when an actual measurement is claimed. Make the DHC replay basis recoverable, identify measurement Work and result separately, and apply A.18 legality. For direct comparison use the same-semantics branch in section 4.2. For a comparison that relies on a cross-local semantic relation, add the exact F.9 branch. Open G.0, A.19, normalization, distance, evidence-reliance, or assurance only when the operation or receiver consumes it.
Reusable series, dashboard, or publication
Create a DHCDefinitionSet only when a reusable selection of definitions is needed. Create a DHCSeries episteme only when the receiving use needs ordered coordinate-result refs across windows. Use G.12 for a dashboard representation and refresh wiring. When an audience must be able to obtain a selected edition, use E.24.PUB and keep the availability relation, form, carrier, and any publishing Work distinct. Public naming and registry behavior remain conditional on a named use.
Archetypal Grounding (five domains)
Computer vision: direct comparison without a Bridge
Two ReproducibilityRate results concern the same named set of tested benchmark claims in matching 24-month windows. They cite the same Characteristic and ratio Scale editions, the same DHCMethodRef.edition, and compatible model, calibration and independent-team conditions. The team can compare the rates for that set. If two reports tested different subsets of a common benchmark, state what was tested in each. An inference about the common benchmark population also depends on how the claims were selected. The source reports have different publication editions, but no distinct F.17 local senses are being related, so no F.9 relation is required.
Formal benchmark approval and actual benchmark adoption have separate results, FormalRecognitionStatus and PracticeAdoptionRate. Ten teams are eligible this quarter. Count a team as an adopter if it used the benchmark in a qualifying test during that quarter. The records establish use for four teams and no qualifying use for two. Either status remains possible for each of the other four, with no further joint restriction. Between four and eight of the ten could be adopters. A decision that requires at least six adopters is still open. Confirming qualifying use by two distinct teams among the four with unknown status would settle it even if the last two statuses remain unknown. The rate among the six teams with known status answers a different question.
Biomedicine: evidence resolution without ratio substitution
One claim reports EvidenceUnitResolution = claim under ClinicalClaimSegmentation-3. A separate result reports SupportAnchorsPerClaim = 2.4 anchors/claim for the declared corpus. Neither value is substituted for ClaimsPerArtifact. Replication is a separate ReproducibilityRate over independent cohorts and a 36-month window.
Software performance engineering: explicit cross-local use
The compared cells are OpenTelemetry:SLO/latency-objective@E4 and VendorB:SLO/service-level-target@E7. For this example, take F9-SPE-SLO-12 as an independently established directed F.9 Bridge from the OpenTelemetry cell to the VendorB cell under a satisfied predicate profile.
A separate C.2.1 bounded-use claim states suitability for “compare service-latency objective coverage in the 2026 survey” in that direction, using only rows with the aligned 30-day window and tolerating no rolling-window mismatch. The evidence loss note records that the target cell permits a different rolling-window convention; the comparison still requires current A.10 reliance on the bounded-use claim.
That directed relation is one counted member of the declared AlignmentDensity cell set; it does not make all tracing-ecosystem readings comparable.
Decision-making: entropy and concentration stay separate
TraditionShareEntropy uses normalized Shannon entropy with base and category set fixed. TraditionShareConcentration uses HHI over the same population and has the opposite dispersion direction. A view may show both. If a receiving comparison wants one orientation, it declares 1-HHI as a transformation and still does not equate that Scale with normalized entropy.
Evolutionary architecture: banded disruption
DisruptionBalance is computed over one declared corpus and method edition. The result is interpreted against an explicit target band and distance rule; a higher raw value is not automatically healthier. Architecture decision records and fitness tests remain inputs or neighboring objects, not evidence that the field itself is healthy.
Making a health claim
- State the question about the field and what result the recipient needs.
- Identify the discipline under C.20. Select the Characteristics and Scales that distinguish the answers that matter for this use.
- State the ClaimScope, observation or comparison basis and period for each reading. If several conditions must hold for the same cases, identify the relation among the observations or a bound sufficient for that conclusion. Counts over separate groups can be accurate while leaving their overlap open.
- Give the resulting health claim with its grounds. Stop when it answers the question.
- If measuring, recover the C.16 chain and DHC replay basis. If comparing or aggregating, retain the compatible inputs and legal operation, adding the measurement, crossing or formal-interface conditions used.
- If repeated windows matter, construct a series episteme from exact result refs.
- Use G.12 when the needed dashboard represents measured DHC coordinate results.
- If audience availability matters, establish E.24.PUB publication separately.
Bias-Annotation
C.21 counters dashboard, standard-status, popularity, and pseudo-precision bias. A public standard can have little adoption; a widespread practice can lack formal recognition; a dense relation map can preserve important losses; and a polished dashboard can represent weak or stale claims. The progressive path keeps the remedy proportional.
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
- Treating discipline health as one scalar before separately typed coordinates exist.
- Treating
ClaimScopeandTargetSliceas two spellings for one object. - Requiring a Bridge merely because sources, schemes, or editions differ.
- Making a comparison that relies on semantic correspondence between distinct local senses without the obtaining F.9 relation and its use/loss basis.
- Ordering formal status and adoption on one universal maturity ladder.
- Calling
claims/artifactandanchors/claimalternative Units of one Characteristic. - Writing
<entropy/HHI>as if either formula produced the same reading. - Treating a definition set, method description, Work record, series, dashboard row, or publication carrier as the measurement result.
- Requiring publication and assurance machinery before one useful health claim exists.
Consequences
Benefits. Field-health claims remain readable, scale-admissible, comparable when justified, and reusable without turning a dashboard or standard into authority.
Costs. A numerical or reused claim must expose its definition and comparison basis. A comparison that depends on semantic correspondence between distinct local senses additionally needs the obtaining relation, bounded-use claim, loss account and applicable reliance basis.
Risks avoided. The pattern prevents false maturity ladders, hidden polarity reversal, scope substitution, unnecessary Bridges, stale trends, treating a publication as a result, and treating a record as assurance.
Rationale
Discipline health has multiple dimensions. Reproducibility, formal recognition, adoption, alignment, disruption, evidence resolution, and diversity answer different questions and can move independently. Their definitions, measurement results, series, representations, and publications also change for different reasons. Keeping those distinctions explicit makes the result both more precise and easier to use.
SoTA-Echoing
Relations
- Builds on: C.20 for the discipline, A.17-A.18 for Characteristic and Scale, C.16 for measurement and comparability of measured readings, A.2.6 for ClaimScope and optional selected slices, and C.2.1 for result and series epistemes.
- Coordinates with: F.9 for actual cross-local sense relations; G.0 and A.19 for numerical operation, comparison, or aggregation; G.4 for a selected formal target-band or acceptance interface; G.9 for parity; G.11 for currentness and refresh; G.12 for dashboard representations; A.10/G.6 and B.3 only for named evidence-reliance or assurance uses; E.24.PUB for publication availability.
- Constrains: G.5 and G.10 when they consume a DHC coordinate: they carry the same active DHC replay basis rather than a generic method-spec or metric-edition pin.
Practitioner Quick Template
C.21:End
Task Typing and TaskSignature Assignment (Problem-CHR)
Status: Stable Type: Calculus (C)
Purpose. Declare a minimal, portable TaskSignature for selector-facing use after the problem-side episteme is stable enough for Principles-to-Work, eligibility, acceptance, or policy-governed choice. C.22.2 carries the first problem-framing episteme for a messy signal.
Use C.22 to constitute one CHR-grounded U.Signature and, when a receiving use is current, relate the exact problem-side episteme to that signature through TaskSignatureAssignmentRelation. Typed characteristics and unknowns stay visible. The declaration includes scope and only those basis, currentness, evidence-use, or crossing relations on which the receiving use relies; none adds a generic setting, carrier, or organization as a participant.
Body-level kind boundary. TaskSignature is a C.2.1 episteme and a species of existing U.Signature, conformant to A.6.0 direct declaration fields, Vocabulary, Laws, and Applicability. Its declaration may use ordinary language when the required content is recoverable; it is not a record format and introduces no new root U-kind. TaskSignatureAssignmentRelation is a separate obtaining relation among one exact problem-side episteme, one exact TaskSignature episteme, and one exact use episteme. ProblemCard is the C.22.2 problem-side episteme used before that assignment. KindSet contains C.3 U.Kind values for selected entities. Descriptor maps, telemetry hooks, policy ids, and selector fields remain signature vocabulary or projections unless an exact admission predicate and current subject assertion establish another kind.
Primary EntityOfConcern. This pattern defines or constrains one TaskSignature episteme. Inside it, EntityOfConcernRef identifies the exact task or work target declared for the receiving use; it does not identify the signature, TaskKind, carrier, organization, or publication. TaskKind, optional TaskFamilyRef, KindSet, characteristic bindings, and scope relations are declaration content. A later SelectorOutcome remains a downstream result.
Use This When
Use this pattern when one stabilized problem-side episteme must be related to a selector-facing TaskSignature for eligibility, acceptance, or policy-governed selection. Typical cases include solver choice, method-family eligibility, QD archive selection, open-ended generator selection, or specialization claims that need a declared task family or work target.
The working moment often sounds like this: "We are about to compare possible ways of doing, but which facts about this problem make a method family eligible, comparable, or unacceptable for this use?" Construct the smallest A.6.0 TaskSignature that a later selector can consume without selecting a method in advance, then assign it to the exact problem-side episteme and receiving use.
If problem framing remains contested or stale, use C.22.2. If a sufficient signature and assignment already exist and the current question is selection, use G.5. If a method is selected and dated enactment is being prepared, use A.15.2.
What goes wrong if missed. Unstated selector inputs let eligibility drift; incompatible scales are combined, unknowns are coerced and acceptance criteria are mistaken for characteristic definitions. A reuse that depends on semantic correspondence can also proceed without its local senses, obtaining F.9 relation and separate bounded-use claim.
What this buys. The downstream selection question gets one separately constituted TaskSignature with typed vocabulary, laws, applicability, unknown handling, and ClaimScope. Evidence, freshness, and cross-semantic relations are added only when that receiving use needs them. Its assignment is replayable while publication and serialization can vary without changing the signature.
Intent
Make the task distinctions relevant to a selector decision explicit in a typed TaskSignature. A problem reaches C.22 when its problem-side episteme is stable enough to constitute and assign that declaration without selecting a method in advance. The signature is the smallest CHR-typed A.6.0 declaration sufficient for eligibility, acceptance, and policy-governed selection without inadmissible arithmetic or silent coercions.
Term split used in this pattern
TaskSignatureassignment means one obtainingTaskSignatureAssignmentRelationamong an exact problem-side episteme, exact TaskSignature, and exact receiving-use episteme; it does not pre-bind a method.ScopeSlice(G)means the exact A.2.6U.ClaimScopevalue used by this declaration; it is not an evidence-path slice, baseline-set slice, container, or assignment participant.thresholdhas three distinct uses here:- articulation and closure thresholds stay with cue or prompt subject patterns such as
B.4.1andB.5.2.0; - acceptance-gate thresholds stay with
G.4; - a work-measure threshold target used in a specialization claim is only the declared success mark for that task family or work target.
- articulation and closure thresholds stay with cue or prompt subject patterns such as
Name and kind map for code-shaped heads. The names below identify different structural positions; capitalization does not make them peer kinds.
ProblemCard relation
ProblemCard is the C.22.2 C.2.1 episteme used to stabilize one problem-side representation before downstream Principles-to-Work.
A ProblemCard can prepare TaskKind, scope, and characteristic bindings for a candidate TaskSignature. Identify which adequate signature was adopted for this problem-side episteme and receiving use under §5.2. Other adequate descriptions may be available. If adequacy or adoption is unresolved, retain the candidate signatures and identify what would settle it.
Assigning a TaskSignature does not transfer the card's claims into it. The signature keeps only its A.6.0 declaration content; the card remains the reviewable problem-side episteme that explains why this problem can proceed to characterization, comparison, search, refresh, retirement, or another subject pattern.
The corresponding claims remain with their named subject patterns.
Problem Frame
Selector-facing problem case
For selector-facing C.22 use, a problem case applies when the problem-side episteme is stable enough to construct a minimal TaskSignature and assert its TaskSignatureAssignmentRelation for eligibility, acceptance, or policy-governed selection. Method absence or contestability is a common reason for this downstream use; it does not define what counts as a problem.
Use C.22.2 when a symptom, opportunity or proposed work item still leaves open whose work is affected or what would count as addressing it. Return there also when a changed ReferenceScheme or ClaimScope undermines the existing formulation. An already agreed assignment with adequate grounds can proceed directly to the work it requires. When selection becomes current, cite the A.19.SelectorMechanism relation and exact G.5 policy refs rather than moving selection policy into the signature.
Unknown-first discipline. In the TaskSignature, preserve a live unrecovered trait as unknown when its value rule permits it. Name the exact downstream policy that interprets a live unknown for the receiving use. C.22 introduces no universal outcome set; C.23, G.4, G.5, or another direct pattern defines or constrains the resulting eligibility, acceptance, or selection disposition.
When the receiving selector cannot recover the relevant task distinctions, it cannot filter or abstain on a justified basis. Acceptance criteria may leak into scoring, or cross-semantic reuse may proceed by a shared name instead of an obtaining relation.
The receiving use needs a use-bounded TaskSignature that:
- establishes MM-CHR admissibility for Scale and applicable Unit and polarity before aggregation;
- carries tri-state unknowns explicitly; and
- states the exact scheme, scope, and only the basis and currentness relations needed by the current use.
Keep the direct support required for the claim; open A.10 when its evidence, provenance or bounded reliance needs a replayable account.
Open B.3 only when a named assurance claim about an exact target claim is current for this receiving use, including when a material-reliance rule requires that assurance claim; its assurance result is not a TaskSignature field.
Problem
Without recoverable descriptor meanings, Eligibility and Acceptance can depend on inadmissible operations, such as ordinal means or unconverted mixed units. Without the F.9 split, a cross-scheme comparison can also equate unlike local meanings or import an assurance penalty even though no current assurance policy selected it.
Forces
Solution — Problem CHR, TaskSignature, and assignment relation
Local TaskSignature mantra. Stabilize the problem; name the receiving selection question and task kind; keep only traits that can change eligibility, acceptance, or selection; type each live trait; preserve unknowns, scope, and any basis or currentness relation the use relies on; declare the TaskSignature, assign it to the problem-side episteme for that use, and stop before selecting a method.
Apply that formula as follows:
- Confirm that the problem-side representation is stable enough for selector-facing use; otherwise use
C.22.2. - Name the work for which a method is needed and the decision that will use its description. Use the kind of that work as
TaskKind. When the practitioner needs a method for making a comparison, the comparison itself is the work to type. Identify the particular target or task family when the receiving decision depends on it. - Read the eligibility conditions, acceptance clauses or comparison criteria of the receiving use. For each condition, identify the task-side fact or characteristic whose value the condition consumes, and recover it from the problem-side account or its source. Include that trait and its interpretation; leave unrelated optional extensions absent. If two feasible tasks fit the same description but require different eligibility or acceptance decisions, identify the characteristic that accounts for the difference. When a method must be chosen before work starts, establish what the chooser can learn by then. A supported conditional choice or a declared rule for an unknown value can suffice. Keep an unresolved interpretation visible.
- Identify each live Characteristic and Scale, its Unit and polarity when applicable, and the admitted comparison relation. Declare ReferencePlane when interpretation or the receiving comparison contract requires it. Apply these conditions before aggregation or comparison.
- Preserve a live but unknown value as
unknown; include or reference the exact scope, evidence relation, freshness or edition condition on which later use relies. When cross-semantic reuse is current, resolve the two local senses and test the F.9 Bridge separately; a shared label, scheme, or plane does not establish it. - Close with one minimal
TaskSignature. Pass later eligibility and acceptance claims toC.23andG.4, and method-family selection toG.5; do not put their outcomes back into the signature as if they were problem traits.
Positive closure, bounded non-use, and local return
Close the C.22 use positively when the direct TaskSignature fields, Vocabulary, Laws, and Applicability are complete and the assignment obtains under C.22:5.2. Its assertion identifies the exact problem-side episteme, TaskSignature, receiving-use episteme, effective ReferenceScheme, ClaimScope, and current qualification conditions.
Each live characteristic has its Scale, applicable Unit and polarity, required ReferencePlane, admitted comparison relation, and value or admitted unknown; every relied-on evidence, freshness or edition relation is named. When cross-semantic reuse is current, its exact local senses, obtaining F.9 Bridge, and separate bounded-use claim are recoverable.
A downstream selector can now consume the assigned signature without guessing, but no eligibility verdict, acceptance result, method recommendation, selector outcome, WorkPlan, or dated Work is claimed.
Close by bounded non-use when problem framing is not stable enough for a TaskSignature declaration, when no selector-facing receiving use is current, or when the current question has already become eligibility, acceptance, selection, planning, or performed work. A non-current optional extension remains absent. Report a particular assignment when the evidence establishes adoption of that adequate signature under §5.2. Keep unresolved alternatives as candidates.
Return to the smallest affected TaskSignature position when its receiving question, exact target, effective ReferenceScheme, ClaimScope, TaskKind, task-family reference, characteristic meaning, scale, unit, polarity, reference plane, unknown status, evidence-use relation, freshness condition, or edition changes. Recheck the F.9 endpoint senses, Bridge predicate, bounded-use claim, or reliance result only when that exact dependency changed. Keep the upstream ProblemCard and downstream selection history unchanged unless that exact change invalidates them under their own subject patterns.
Worked local repair. A machining TaskSignature originally records surface finish as an ordinal visual grade. The named use later adopts measured roughness Ra on a ratio scale in micrometres with a named measurement and evidence relation. Repair the affected characteristic head, scale, unit, admitted comparisons, and evidence relation. Keep the machining TaskKind, unaffected constraints, scope, and prior Work history. Reopen eligibility, acceptance, or Method-family selection only when its earlier result relied on the replaced finish head; state the new result under the exact downstream predicate with its subject-pattern locator.
Apparatus proportionality
Use the lightest signature declaration and assignment relation that the named receiving use can consume:
- Minimal selector-facing use. Materialize one TaskSignature with only the live fields needed by the current eligibility, acceptance, or selection question. This is the ordinary positive result of C.22.
- Reliance-bearing use. Add an addressable
ProblemProfileepisteme only when delayed feedback, audit, transfer, automation, expensive reversal, or another named use relies on replay beyond the local assignment relation. Pin the exact problem-side episteme and edition, TaskSignature edition, receiving use, every relied-on field-basis relation with its subject pattern, qualification window, review trigger, and any current evidence or currentness relation. When this use crosses local meanings, test F.9. Add a Bridge and separate bounded-use claim only if its predicate is true; otherwise keep the local values separate and stop that reuse. - Extension-bearing use. Add QD, OEE, archive, generator, parity, or specialization positions only when that exact downstream relation is current and its direct pattern requires those values.
If no selector-facing receiving use needs a TaskSignature, finish without creating one.
Minimal CHR fields (tri‑state aware).
Selector-side field boundary. The fields below are live only after problem framing has been stabilized enough to ask eligibility, acceptance, selection, Method-family, or policy-constrained choice questions. They are not a universal problem-framing checklist and do not replace the C.22.2 Thin ProblemCard pass for a messy signal. Each live characteristic field is CHR-typed by its Characteristic and Scale, with Unit and polarity where applicable under MM-CHR discipline. A live predicate may preserve unknown only when its exact value rule permits it; the cited downstream policy states what follows. This aligns G.4 and G.6 without making their results C.22 values.
Optional extension absence rule. If QD, OEE, archive, generator, parity, specialization, or another optional relation is not live for the current case, the corresponding optional fields are absent, not unknown. Use unknown only for a live field whose value is currently unknown. An absent non-live extension triggers no downstream disposition.
-
DataShape— data regime and admissible transforms (e.g., tabular, sequence, graph; density; stationarity claims). -
NoiseModel— uncertainty class and robustness envelope (e.g., iid Gaussian; heavy‑tailed; adversarial budget). -
ObjectiveProfile— objective heads (Scale, applicable Unit and polarity, and ReferencePlane when interpretation or the receiving contract requires it), admissible order relations (lexicographic, Pareto), and medoid or median operations where admissible. Weighted sums across mixed scale types are inadmissible; ordinal heads use order-only guards. For QD tasks, enumerate quality heads, diversity or descriptor-space heads, and any policy-authorized QD contribution heads; see DominanceRegime below. Do not introduce a default QD score. If a scalar or set-scalarization policy is live, cite the governing CAL policy and keep the uses of dominance and telemetry explicit. -
RegularityTraits— method-relevant structure (convexity, differentiability, separability, monotonicity) as CHR-typed predicates with guard macros (for example,ORD_COMPARE_ONLY,UNIT_CHECK,POLARITY_CHECK). IncludeConditionClasssuch as stiffness or kappa proxies where applicable. -
Constraints— explicit hard and soft constraint classes (feasibility predicates; ResourceEnvelope and RiskEnvelope). Declare acceptance-gate thresholds in G.4 AcceptanceClauses. A signature or implementation may reference those clauses for its use; it does not redefine them as Characteristic properties or hidden constants. -
ShiftClassand stationarity — CHR‑typed claims about regime stability (iid | covariate‑shift | concept‑drift | adversarial). When this field is live but unrecovered, its value isunknown. The cited acceptance or selector policy governs the consequence; omit the field when the receiving use does not need it. -
Evidence and assurance (conditional). Cite the direct evidence-use, provenance or subject relation that supports the receiving claim. Use A.2.4 when an episteme's evidence-use or status-use relation needs classification, and A.10 when evidence, provenance or bounded reliance needs a replayable account. State source edition, currentness, or freshness only to the degree that reliance requires.
Open B.3 only for a named assurance claim about an exact target claim and receiving use, including when a material-reliance rule requires it, and use only the assurance lanes and fold that its declared policy requires. A TaskSignature by itself requires neither all TA/VA/LA lanes nor a Gamma-fold.
-
ScopeSlice(G)— the A.2.6U.ClaimScopevalue that bounds this declaration's claims aboutEntityOfConcernRef(discipline governance in CG‑Spec; Domain is a catalog mark only). -
SizeAndConditionProfile— size and condition proxies (n, m, kappa, sparsity) with units declared where applicable; a unit mismatch makes the current comparison unsupported until the direct acceptance or selector policy supplies its governed result. -
Freshness(conditional) — the validity window for a descriptor only when the receiving use relies on its currentness. -
Missingness— MCAR, MAR, or MNAR (or mapped equivalents) per CHR.Missingness; Acceptance and flow use preserve the declared missingness semantics. -
KindSet— selected C.3U.Kindvalues for the entities addressed by the TaskKind; separates EntityOfConcern kind from Scope (USM).
QD and Illumination extensions (normative; ties to C.18 and C.19).
Select these fields from the receiving use. A QD search constructs and maintains diverse candidates under its descriptor, archive and emitter rules; that use activates the QD fields below. Retaining exploration evidence under PortfolioMode=Archive, comparing a front, or returning several alternatives does not by itself require QD. An open-ended generator use activates GeneratorIntent and its own required fields; add the QD fields when that use also performs QD. Keep the source collection, its retention policy, the comparison and the emitted selector outcome distinct under G.5:4.4b–4.4e.
CharacteristicSpaceRef— reference to the exactU.CharacteristicSpace, with its dimension and CHR-typed coordinates as required by the declared descriptor and comparison rules. Declare ReferencePlane where coordinate interpretation or the receiving contract requires it. Pin its defining episteme separately withcharacteristicSpaceDescriptionRef: U.EpistemeRefand resolve both to the same space under A.19:5.1.1.1.ArchiveConfig— archive topology (grid, CVT, or graph), resolution (bins or centroids), K‑capacity,InsertionPolicyRef(elite replacement, dedup, or novelty), andDistanceDefRef.edition(declare metric or pseudometric status and invariances; normalisation is admissible only when the applied scale transform is admitted by CG-Spec); admissibility follows CG‑Spec.EmitterPolicyRef— reference to the emitter policy governed by C.19 and applicable to this TaskSignature; edition id recorded.DominanceRegime—{ParetoOnly | ParetoPlusIllumination}. Default =ParetoOnly(illumination remains report‑only telemetry unless CAL explicitly authorisesParetoPlusIllumination, policy‑id cited).IlluminationSummary— a telemetry summary overDiversity_P; reported by default; excluded from dominance unless a CAL enablesParetoPlusIllumination(policy‑id cited).IlluminationMap(parity-run) — parity-run publication is complete when an IlluminationMap publication (grid, CVT, or graph perArchiveConfig) records coverage per niche or cell withDescriptorMapRefandDistanceDefRef.edition. A single-score leaderboard does not satisfy this comparison use; compare under the declared CG-frame.PortfolioMode— resolve its values and default under G.5CC-G5.23.Archivepreserves exploration evidence. It neither activates a QD search nor names the emitted selector result; G.5:4.4e distinguishes retention, source set and outcome.Budgeting— evaluation, time, and batch budgets, including E/E‑LOG exploration budget id; units declared (CG‑Spec).TelemetryHooks—PathSliceIdonly when an E.18 path-slice reference is current, plus decay and refresh policy ids, edition counters, descriptor-map updates, and policy-id updates upon illumination gains.GeneratorIntent(OEE) — optional intent to use a registeredGeneratorFamily(G.5), with pointers toEnvironmentValidityRegion,TransferRulesRef, and coverage and regret reporting expectations.
Admissibility. Before any numeric comparison or aggregation, establish CSLC admissibility for Scale and the applicable Unit and polarity, and cite CG-Spec.Characteristics. Record ReferencePlane when its interpretation or the receiving contract requires it. Preserve unknown for the downstream policy; do not coerce it to 0 or false, and do not invent a C.22-local disposition.
TaskSignature declaration and assignment
TaskSignature is a C.2.1 episteme and a species of A.6.0 U.Signature. It uses A.6.0 identity and declaration content directly rather than a flat record schema. Add an optional SignatureManifest only when dependency replay requires imports and provided names; SignatureId and edition are designators and currentness handles, not substitutes for identity.
The field families in C.22:5.1 are projections of Vocabulary and Applicability. They are not extra conceptual rows and do not redefine A.6.0.
The assignment is a separate relation with exactly three direct participants:
The relation obtains while that exact TaskSignature is adopted as the task-typing declaration for the exact problem-side episteme and the use stated in the exact receiving-use episteme, under the stated scheme, scope, and qualification conditions. Co-publication, a card field, a shared label, or one record row does not make it obtain.
One occurrence is identified by the three participants plus its maximal continuous actual assignment extent. A participant change yields another occurrence; actual withdrawal and later readoption yield distinct occurrences even when the same three participants return.
TaskSignature identity and publication. The tuple <declaration content, EntityOfConcernRef, effectiveReferenceScheme> determines TaskSignature episteme identity under A.6.0 and C.2.1. SignatureId and edition designate and track that episteme. A semantic change to direct declaration fields, Vocabulary, Laws, Applicability, the exact target, or the effective scheme identifies another episteme. Admit it as a TaskSignature only if it satisfies A.6.0; relate it as a revised signature edition only when C.2.1:4.5's source-use and continuation conditions obtain.
Two E.17 publications, database rows, cards, or files may present the same edition when they resolve to the same tuple and add no new claim. A ProblemProfile may cite the signature and assignment or include the declaration content. Test TaskSignature membership and C.2.1 identity from that content; the assignment relation remains separate.
Minimality rule. Include only declaration positions needed to determine eligibility, acceptance, or admissible selection for the named use. Additional traits remain outside Vocabulary until a later use makes them current.
Values are CHR-typed and tied to the exact measurement, evidence-use, source-use, representation, or scope relation that justifies their use when such a relation is current. Each reliance-bearing field basis names that relation and its subject pattern; generic provenance or support wording is not a replay basis. Unknowns preserve their direct missingness semantics.
TaskSignature invariants. A positive assignment satisfies all six conditions:
- The TaskSignature exposes its exact
EntityOfConcernRef, effective reference scheme, direct declaration fields, Vocabulary, Laws, and Applicability. - The exact problem-side episteme, exact TaskSignature, exact receiving-use episteme, obtaining conditions, and occurrence extent of the assignment relation are recoverable.
- Every live field has an admitted filler kind or scale discipline and, under reliance, an exact basis relation with a subject pattern.
- A live but unrecovered value is
unknownonly where the field's exact value rule permits it and a downstream policy states how the named use handles it. - A non-current optional extension is absent; absence and unknown are not interchangeable.
- Eligibility verdicts, acceptance results, selected methods, selector outcomes, WorkPlans, and Work occurrences are absent from the TaskSignature and remain with their direct patterns.
Lowering and withdrawal conditions
Stop relying on the assignment for the current use when its problem-side episteme, TaskSignature, receiving-use episteme, scheme, scope or qualification conditions cannot be recovered. Actual withdrawal of adoption or failure of an obtaining condition ends the occurrence under §5.2; missing access to its account does not by itself rewrite the assignment history. The TaskSignature may remain valid for another assignment. Use C.22.2 only when the problem-side representation itself is no longer stable enough.
Identify the resulting episteme and apply C.22:5.2's signature-membership and edition-continuity checks when a direct declaration field, Vocabulary, Law, Applicability claim, EntityOfConcernRef, or effective reference scheme changes. Lower or remove one vocabulary position when its filler kind, scale, unit, polarity, reference plane, direct basis relation, or subject pattern cannot support the claimed use. Preserve unknown only when the position remains live and admitted. Split any selected method, selector outcome, acceptance result, plan, or Work occurrence into its subject pattern.
A changed or invalid signature position reopens an earlier downstream result only when that result relied on the changed position. The downstream pattern repairs or supersedes its own result. A revised signature does not imply that the actual Problem disappeared or that prior Work did not occur.
Evolution and currentness boundaries
Use C.22 to revise the smallest affected identity or declaration-content component; identify the resulting episteme and claim a new TaskSignature edition only under the identity, membership, and continuity conditions in C.22:5.2. A changed problem formulation requires C.22.2 before a replacement assignment is made.
G.11governs relied-on source edition, freshness, decay, telemetry, and currentness relations; its result may trigger signature review but does not rewrite the signature by itself.C.18andC.19govern archive, front, lineage, and live-pool evolution.G.5governs selected-set and method-family selector results.E.23governs repeated object-version improvement.
C.22 introduces no local refresh object and does not rewrite earlier selector results or dated Work without an explicit dependency.
TaskKind fills SubjectKind. TaskFamilyRef? names one comparison-relevant family in Vocabulary when specialization, transfer, or parity is live. KindSet and A.2.6 scope slices determine the ranged extent. They remain declaration content, distinct from a selected method or selector result.
DesignRunTag hygiene. Do not mix DesignRunTag positions in one signature edition. If design-side information is reused in run-time Work, identify the actual receiving Work and its relations independently. Cite an E.18 structural GateCrossing only when a selected transformation-flow use contains that occurrence; it does not require a package and does not establish an F.9 Bridge.
Specialization-claim reference discipline (normative)
A specialization claim identifies the holder, dyad, team, or scoped portfolio, the task family or TaskSignature, the work target, and the criterion and grounds for usable performance. Identify what the holder did and what assistance was available during the performance. A claim of acquisition needs evidence of change from the relevant prior state under the support conditions of the intended work. When the receiving use compares acquisition or adaptation, C.22.1 supplies the adaptation budget, prior-exposure and threshold-crossing conditions; retain freshness and provenance to the extent reuse depends on them. A method may be selected, refined, or retired in this process, while the specialization claim concerns the stated holder or portfolio. Preserve the task-family and work-target references required by the C.22.1, G.5 or G.9 receiving use.
Low-human-overlap or newly discovered task families remain admissible when those task-family or signature references are explicit by value.
When a low-human-overlap comparison matters to the receiving use, identify the target task family and a bounded reference set of documented human practice. Define the task correspondence through the inputs, objectives, constraints and conditions relevant to that use. State the direction, weighting or coverage rule and the local threshold meant by "low". These choices make the comparison reproducible; an ordinary task-family declaration needs no such comparison.
For example, suppose one of five equally weighted target task classes has a match in a completely examined finite reference set. Directional coverage is 1/5, below a declared threshold of 1/4. If one class is matched, two are known not to match and two remain unclassified, possible coverage instead ranges from 1/5 to 3/5, so that evidence cannot settle the same low-coverage claim. The result concerns this reference set and task correspondence. Task admissibility and the holder's specialization retain their own tests.
Provenance, schemes, and planes
Record the effective reference scheme, U.ClaimScope, and any ReferencePlane needed to interpret a relied-on value. A difference in scheme or plane opens a comparison question; it does not by itself establish a Bridge, forbid use, or impose a penalty.
For cross-semantic reuse, recover the two exact F.17 local senses and test the direct F.9 predicate. Cite a Bridge only when that predicate is true. Then state the proposed use separately: the action, direction, correspondence rule, tolerated loss, and polarity. If no Bridge obtains, keep the local values separate and return the missing comparison or translation question instead of asserting correspondence.
Keep the grounds required for that bounded use, and open A.10 when its evidence, provenance or reliance needs a replayable account.
Open B.3 only for a named assurance claim about an exact target claim and bounded use, including when a material-reliance rule requires it; only that B.3 result may use edge-scoped CL and the policy it declares. An optional local F.9 CL note is evidence shorthand, not a use threshold or automatic penalty.
A card, gate check, reusable package, or publication is required only when its own receiving pattern independently needs it. The assignment receives no generic setting participant, and a domain, organization, location label, or shared carrier supplies none of these relations.
Attachment & use.
The bullets below state which TaskSignature fields and relations each downstream use reads. C.22 does not execute eligibility, acceptance, selection, archive treatment, or generator-family choice. Their verdicts and returned sets remain results of the named direct patterns.
- Eligibility gates read TaskSignature against each MethodFamily.Eligibility (C.23) and CG‑Spec.MinimalEvidence for referenced characteristics.
- Acceptance clauses (G.4) use these fields for acceptance-gate threshold predicates (acceptance-gate thresholds live in Acceptance only).
- Selection kernel (G.5.S3) applies an admissible order (often partial); weighted sums across mixed scale types are inadmissible. A non-dominated front or retained archive supplies candidates. G.5:4.4b–4.4e governs the emitted outcome: retained alternatives use
ShortlistorRankedShortlist, while joint inclusion and other outcomes keep their own conditions. Preserve the source reference, comparison basis and any unresolved ties.PortfolioMode=Archivechanges evidence retention, not that result family. In an active QD use, illumination enters dominance only when CAL authorisesDominanceRegime=ParetoPlusIlluminationwith its policy id; otherwise the telemetry is reported separately. - When
GeneratorIntentis present, G.5-governed selection may use a registeredGeneratorFamily(POET-class); the selection domain becomes pairs{environment, method}, with Environment guarded byEnvironmentValidityRegionandTransferRulesRef(C.23 wiring). When the active branch or receiving use requiresIlluminationSummary, obtain that retained-set report under C.17:5.1, including itsDiversity_Pmeasurement policy, Scale, descriptor or feature basis, and distance or covering definition. Generation alone requires no such report. Illumination remains report-only unless its governing policy changes dominance participation as above.
Unknowns.
An identity position needed for positive closure cannot be replaced by unknown. A live characteristic or predicate may preserve unknown only when its exact value rule permits it. The TaskSignature cites the downstream policy that defines or constrains the consequence; C.22 performs no implicit coercion and declares no universal outcome set.
Publication.
When a named receiving use needs a reader-facing publication of the ProblemProfile specified in C.22:5.0b, publish that accepted C.2.1-conformant episteme through E.17. Use E.24.PUB when that use depends on publication availability, its declared boundary, or publication-occurrence identity. The profile references the bound TaskSignature and only the evidence, currentness, F.9 bounded-use, and representation relations on which that use relies.
Apply F.18 and F.17 Name Cards when a durable new name is being admitted; do not create a card merely because a local field or Bridge is present. Keep vendor or tool examples in Plain explanatory use rather than letting them become normative selector inputs.
When no replay beyond the local assignment relation is needed, the TaskSignature closes without a separate ProblemProfile.
Open‑Ended tasks (GeneratorFamily) (normative).
When open-ended generation of tasks or environments is current, the TaskSignature supports that use only when it includes GeneratorIntent with pointers to EnvironmentValidityRegion (admissible region for generated environments), TransferRulesRef (cross-environment transfer constraints), and coverage and regret telemetry expectations. G.5 declares its outcome over {environment, method} pairs under G.5:4.4b–4.4e; the generating archive remains a source.
Report coverage and regret under their declared policies. Add IlluminationSummary when the active branch or receiving use requires that retained-set report, with the measurement basis specified in C.17:5.1. These telemetry results remain excluded from dominance unless a CAL policy promotes them (policy id recorded in SCR; see DominanceRegime).
The SCR change record retains current editions and actual edition increments of the definitions on which the active generator or telemetry use depends. These include TransferRulesRef.edition and any characteristic-space description, descriptor map, distance or covering definition, and emitter policy used by that branch. An illumination increase retains its governing policy id; a generator use without illumination requires no illumination record.
Archetypal Grounding (Tell–Show–Show)
The examples show which TaskSignature fields supply eligibility or acceptance inputs. Keep that dependency recoverable when a downstream decision relies on it—for example, an ordinal service level permits order comparison, while budget arithmetic requires aligned units. The compact field projections below illustrate declaration content; §5.2 governs the complete signature and assignment.
A. Differential equations (continuous systems, solver choice).
TaskSignature fields. DataShape=ODE, stiff?=unknown, SizeAndConditionProfile={n≈10^3}, ObjectiveProfile={↓error@ratio, ↑throughput@ratio}, ConstraintRefs={budget-envelope relation, safety-predicate relation}, RegularityTraits={Lipschitz known?=unknown, Jacobian sparsity=high}, Missingness=MAR.
Attachment. Selector consumes TaskSignature; eligibility filters MethodFamilies whose eligibility conditions require known stiffness or differentiability, with unknown yielding degrade or abstain per family. Acceptance tests its safety_gate condition on the declared ordinal safety level using order comparison (ORD_COMPARE_ONLY), and uses unit-aligned sums for ratio-scale budgets. The selector declares a Shortlist retaining the non-dominated candidates and its comparison basis; no cross-ordinal weighting.
B. Mixed‑integer optimisation (planning and scheduling).
TaskSignature fields. DataShape=MIP, NoiseModel=deterministic, ObjectiveProfile={↓cost@ratio, ↑service_level@ordinal}, Constraints={SLA hard, workforce soft}, RegularityTraits={convex_relaxation=available}, SizeAndConditionProfile={vars~10^5}, Missingness=MCAR.
Attachment. CG‑Spec forbids means over service_level (ordinal); Acceptance holds acceptance-gate thresholds; Eligibility checks convex-relaxation availability; Selection applies the lexicographic guard (assumption-fit before evidence-fit before resource). If this receiving use makes a named assurance claim about an exact target claim, including when a material-reliance rule requires it, use B.3 to obtain its separate assurance result from the relied-on evidence relations and the declared policy; otherwise this example adds no assurance fold. If the admissible comparison remains partial, the selector can declare an unordered Shortlist from the non-dominated candidates, with the comparison basis and unresolved ties.
Practice example: the cited SciML Problem Interface constructs an immutable problem value before solver use and supports explicit
remakewhen problem fields change. C.22 adapts only that problem-before-selector separation; it does not import Julia types as FPF ontology.
C. Quality-Diversity archive and declared set (illumination).
TaskSignature fields. DataShape=policy‑search; ObjectiveProfile={↑reward@ratio, ↑coverage@ratio (report‑only)}, DominanceRegime=ParetoOnly, PortfolioMode=Archive, CharacteristicSpaceRef(d=3, characteristics=CHR‑typed), characteristicSpaceDescriptionRef=CS-description-1, ArchiveConfig(grid, res=32×32×16, K=1, InsertionPolicyRef=elite‑replace, DistanceDefRef.edition=v1), EmitterPolicyRef=v2, Budgeting{eval=1e6}, TelemetryHooks{PathSliceId=…}. Include PathSliceId only when an E.18 path slice is current.
Selection result. The QD archive remains the declared source. G.5 returns a Shortlist of retained alternatives from it, with exact members, ordering and basis pins under G.5:4.4b–4.4e. Coverage and illumination are reported separately and excluded from dominance under the declared ParetoOnly policy. Changes to DistanceDefRef.edition or the emitter policy are editioned and logged in SCR.
D. Open‑ended environment generation (POET‑class).
TaskSignature fields. GeneratorIntent{GeneratorFamilyRef=…, EnvironmentValidityRegion=… (CHR‑typed), TransferRulesRef=…, CoverageMetric=…}, PortfolioMode=Archive.
Selection result. G.5 declares a Shortlist over the eligible {environment, method} pairs and cites the source, ordering and selection basis. TransferRules govern cross-environment policy reuse. This use reports its declared coverage and regret with their edition and policy basis, separately from the selector outcome. It requires neither a QD search nor an illumination report; a later receiving use that needs illumination obtains its measurement basis under C.17:5.1.
E. Physical manufacturing method-family eligibility.
Problem-side episteme. PartFamilyFinishingProblemCard-E2 : U.Episteme is the exact C.22.2 ProblemCard for a shop that must finish AlloyPartFamily-17 on one machine under ShopInspectionScheme-E4 and a production-window ClaimScope. The receiving question is which available finishing-method families can be compared without presuming one of them.
TaskSignature. SurfaceFinishingEligibilitySignature-E1 declares EntityOfConcernRef=AlloyPartFamilyFinishingTarget-17, effectiveReferenceScheme=ShopFinishing-Scheme-A, TaskKind=surface-finishing work, ScopeSlice(G)=AlloyPartFamily-17 during [2026-09-01T00:00Z, 2026-10-01T00:00Z), ObjectiveProfile={surface roughness Ra@ratio in micrometres with downward polarity, throughput@ratio}, ConstraintRefs={geometric-tolerance relation, heat-distortion relation, resource-envelope relation}, and material hardness as a live unknown, with the rule for obtaining the needed measurement and the policy for handling that unknown in eligibility.
The TaskSignature makes eligibility reviewable; it does not select grinding, honing, polishing, or another method and does not establish that any part was finished.
Assignment. FinishingMethodEligibilityUse-E1 : U.Episteme states the exact receiving eligibility-comparison use. TaskSignatureAssignmentRelation(PartFamilyFinishingProblemCard-E2, SurfaceFinishingEligibilitySignature-E1, FinishingMethodEligibilityUse-E1) has exactly the problem-side episteme, signature, and receiving-use episteme as participants.
It obtains only while that exact signature is adopted as the task-typing declaration for that exact card and the use stated in FinishingMethodEligibilityUse-E1, under ShopFinishing-Scheme-A, the declared part-family scope, ShopInspectionScheme-E4, and the production window above. Withdrawal of that adoption or change of a participant or qualification ends this assignment occurrence; a shared row, carrier, or publication does not make it obtain.
F. Clinical rehabilitation method-family eligibility.
Problem-side episteme. CohortRehabilitationProblemCard-E3 : U.Episteme is the exact C.22.2 ProblemCard for a rehabilitation service with Cohort-2026-Q3 and a stated capability-change question under clinical safety constraints.
TaskSignature. RehabilitationFamilyComparisonSignature-E1 declares EntityOfConcernRef=RehabilitationCapabilityChangeTarget-4, effectiveReferenceScheme=ClinicalRehabilitation-Scheme-C, TaskKind=rehabilitation work to develop the specified capability, and ScopeSlice(G)=Cohort-2026-Q3 in the declared care setting during [2026-08-01T00:00Z, 2026-11-01T00:00Z). It also declares outcome characteristics with their scale kinds, contraindication and resource constraints, and unknown tolerance or comorbidity values preserved as unknown. Include the evidence relations and follow-up windows only because this comparison use relies on them.
C.22 makes the comparison inputs explicit. Diagnosis and treatment recommendations remain claims governed by the applicable clinical patterns. For a separately current evidence or benefit claim, apply its evidence or clinical evaluation pattern; for a named gate decision, use A.21; for a care-permission claim, use A.2.8.PER and its GrantedPermissionRelation@Context predicate; for decision authority, recover the applicable authority predicate. Return missing-governor under A.6.RCD only when no governing rule can state or test that exact needed relation claim.
If clinical Work occurs, establish its actual performer System and obtaining assignment under A.13, then independently admit the dated Work and enacted Method under A.15.1. Expose the assignment identifier and declared species, and apply F.6, when this account or its receiving use consumes precise assignment-bound attribution through that assignment.
Missing or failed F.6 attribution alone leaves independently admitted Work intact; recheck an admission premise and dependent conclusions if the same finding questions that premise. A local system-role classification or assignment supplies none of the separate permission or authority grounds above.
Assignment. RehabilitationInterventionFamilyComparisonUse-E1 : U.Episteme states the exact receiving comparison use. TaskSignatureAssignmentRelation(CohortRehabilitationProblemCard-E3, RehabilitationFamilyComparisonSignature-E1, RehabilitationInterventionFamilyComparisonUse-E1) has exactly the problem-side episteme, signature, and receiving-use episteme as participants.
It obtains only while that exact signature is adopted for that exact card and the use stated in RehabilitationInterventionFamilyComparisonUse-E1, under ClinicalRehabilitation-Scheme-C, the declared cohort and care ClaimScope, the qualification window above, and the stated evidence-use conditions. Withdrawal of that adoption or loss of a participant or qualification ends this assignment occurrence; cohort labels, records, carriers, and organizations add no signature field or fourth participant.
Bias-Annotation (lexical and discipline guards)
- Selector and policy references. For selection behavior called a strategy, cite the A.19.SelectorMechanism relation and the G.5 criteria, policy or
SelectorOutcomeused in the claim. - Transdiscipline vs domain. Comparability uses the exact ReferenceScheme, ClaimScope, characteristic meanings, and
U.Disciplinerelation when current. A Domain label is only a catalog mark; it supplies no norm, Bridge, scope, or comparison rule. - Descriptions and specification uses. Use E.10.D2 when the wording must distinguish a description of an entity from the use of a description as a specification.
Conformance Checklist (normative)
-
Minimal A.6.0 declaration.
TaskSignatureexposes exactEntityOfConcernRef, effective reference scheme,SubjectKind,RangedValueKind, optionalResultKind,SliceSet, andExtentRule, plus Vocabulary, Laws, and Applicability. AddSignatureManifestonly when dependency replay needs imports and provided names; it does not supply signature identity. -
Signature and assignment present. Every exported selector-facing case names one TaskSignature identity and edition plus one
TaskSignatureAssignmentRelationwhose exact problem-side episteme, TaskSignature, receiving-use episteme, obtaining conditions, and occurrence extent are recoverable. No setting, carrier, or organization is added as a fourth participant. Current characteristic bindings are CHR-typed; a live unknown preservesunknown, while a non-current optional vocabulary item remains absent.1a. Publication and designators do not define identity. Two E.17 publications or serialized records that resolve to the same
<declaration content, EntityOfConcernRef, effectiveReferenceScheme>identify the same TaskSignature episteme. Carrier, layout, serialization,SignatureId, or edition label alone does not create a new identity component. -
CHR admissibility proven. Any numeric comparison or aggregation cites CG-Spec by Characteristic id and proves CSLC admissibility; no mean on ordinals; no unit mixing.
-
Unknowns remain typed. A live unknown remains
unknown, cites the direct downstream policy, and is not coerced. The acceptance, eligibility, or selector pattern records its own governed result. -
Evidence and assurance are conditional. For relied-on evidence or provenance, retain the supporting claim and its direct relation, with only the source edition, currentness and freshness conditions the use needs. Apply A.2.4 to an episteme's evidence-use or status-use question and A.10 when replay of evidence, provenance or bounded reliance needs its account. When a named assurance claim about an exact target claim is current for this receiving use, including when a material-reliance rule requires it, cite its separate B.3 result, applicable assurance lanes, and declared fold. Otherwise the TaskSignature needs neither an evidence dossier nor an assurance fold.
-
Cross-semantic use is separated. Declare
ReferencePlanefor a value or objective head when its interpretation or comparison needs it. A scheme or plane difference alone creates neither a Bridge nor a penalty. Resolve two exact F.17 local senses, test the F.9 predicate, and state any proposed use separately. -
Acceptance thresholds live in CAL. G.4 AcceptanceClauses govern acceptance-gate thresholds. Signatures and implementations reference those clauses instead of redefining them as CHR properties or hidden constants.
-
Selector-use support. The TaskSignature exposes the scales, units, polarities, and admitted order relations needed by
G.5; it carries no mixed-scale scalarization or local selector verdict.G.5governs the emitted outcome and its source under G.5:4.4b–4.4e, including retained alternatives drawn from a non-dominated front. -
Bridge and use claims remain distinct. For current cross-semantic reuse, cite the two exact local senses and an F.9 Bridge only when its predicate is true. State the action, direction, correspondence rule, tolerated loss, and polarity in a separate bounded-use claim; if the predicate is false, keep the local values separate.
-
Packaging is conditional. Create an F.9 card, terminology row, Name Card, or reusable package only when a named receiving pattern independently needs that artifact. Packaging describes an already tested relation and separate use claim; it establishes neither.
-
Structure and gates are conditional. Apply E.18 crossing checks only when a selected transformation-flow structure and exact
GateCrossingare current. Apply A.21 only when a named gate decision is current. Each pattern supplies its own result; neither a structure nor a gate makes F.9 obtain, and C.22 requires no crossing or gate package otherwise. -
QD fields (when QD is in scope). A TaskSignature for an active QD search is complete only when it carries CHR-typed CharacteristicSpaceRef with the dimension required by its declared descriptor and comparison rules, the separate characteristicSpaceDescriptionRef to its exact defining episteme, ArchiveConfig (topology, resolution, K,
InsertionPolicyRef,DistanceDefRef.edition) and EmitterPolicyRef; each characteristic declares its ReferencePlane when interpretation or the receiving contract requires it.PortfolioMode=Archiveor a set-valued selector outcome alone activates none of these fields. -
DominanceRegime default.
DominanceRegimedefaults toParetoOnly. Illumination enters dominance only through a cited CAL.Acceptance policy enabling that relation; the SCR records the policy id. -
Telemetry (when used). The active telemetry record carries PathSliceId when an E.18 path slice is current, the applicable decay and refresh policy ids, and the exact space-description episteme and the editions of the distance and emitter definitions on which that telemetry depends. An illumination increase is traceable to its governing policy id. Ordinary evidence retention does not require an illumination record.
-
GeneratorIntent (when OEE is in scope). A TaskSignature supports the claimed OEE generator-family use only when
GeneratorIntentcitesEnvironmentValidityRegionandTransferRulesRefwith ids resolvable in G.5 and C.23. Any downstream abstention is their result, not a C.22 output. -
Budgets. When
Budgetingis live, its evaluation, time, and batch values carry declared units and the applicable E/E-LOG exploration-budget id. -
Archive-comparison support. The TaskSignature supplies the measurement and comparison basis required by the claimed comparison. When that comparison uses distance or novelty, its distance definition and novelty measures must be CSLC-admissible and editioned, including
DistanceDefRef.editionwhen used. Retaining the alternatives in an archive does not by itself require distance. The archive or selector pattern defines or constrains any downstream abstention or returned-set result. -
Planes. QD heads and characteristics declare a
ReferencePlanewhen their meaning or comparison depends on it. A plane difference alone creates neither correspondence nor an assurance adjustment. Use item 8 for a current cross-semantic claim and item 4 for any separately current B.3 assurance result. -
Unknown QD values. A live unknown QD field remains
unknown, cites the policy governing its downstream use, and is not coerced or mapped by C.22 itself. -
Specialization claims referenced. Name the holder or scoped portfolio, task family, work target, performance criterion and supporting basis. For an acquisition claim, support the relevant change; for a C.22.1 adaptation comparison, retain its budget, prior-exposure and threshold-crossing conditions. Keep the TaskSignature edition and assignment, freshness and provenance required by the receiving use.
Common Anti-Patterns and How to Avoid Them
Selector Fields And Evidence Relations
Inputs.
- One stabilized problem-side episteme and CG-Spec ids;
- a
ProblemProfile(...Description) only when C.22:5.0b's replay need is current; - the supporting evidence-use or provenance relation when relied on, with an A.10 account only when replay of that basis needs it;
- D.CTX only when that separate context relation is current;
- the QD values or references named by CharacteristicSpaceRef, ArchiveConfig, and EmitterPolicyRef only when QD is live;
- GeneratorIntent only when OEE is live.
Produces. Declare one TaskSignature episteme as the U.Signature species specified in C.22:5.2. When that signature is adopted for the problem-side episteme and receiving use, one separate TaskSignatureAssignmentRelation obtains among that signature, the exact problem-side episteme, and exact receiving-use episteme under C.22:5.2.
TaskSignature is neither a new root U-kind nor a record kind: its A.6.0/C.2.1 identity tuple and declaration content determine episteme identity; signature membership and edition continuity follow C.22:5.2, while publication, carrier, and serialization remain outside identity. Optional QD, archive, generator, PortfolioMode, and telemetry vocabulary appears only when current.
Used by. G.5 (Eligibility and Selection kernel), G.4 (Acceptance and Evidence), C.23 (admit, degrade, and abstain rules and method-family maturity checks).
Consequences (informative)
- Admissible selection. Selection is explainable and inspectable: every admission or rejection reason cites the TaskSignature positions and CHR rules on which it relied. When a separate B.3 assurance result is current, that result cites its own evidence relations, applicable assurance lanes, and declared fold.
- Use-bounded first, cross-semantic use explicit. The exact EntityOfConcern, effective ReferenceScheme, ClaimScope, and receiving use are primary. When local meanings differ, an obtaining F.9 Bridge and a separate bounded-use claim make the proposed reuse inspectable; evidence reliance, assurance, gates, and packaging remain optional neighboring questions.
- Common downstream inputs. G.1-G.5 use one typed TaskSignature; acceptance-gate thresholds are separated into Acceptance; unknowns are not guessed.
- QD and OEE-ready. Typed QD and GeneratorIntent fields make declared returned-set structure and open-ended generation contexts explicit, with admissible dominance, editioned distances, and policy-aware illumination.
Rationale
Selecting a method before the relevant eligibility, acceptance, unknown-handling, and admissible-comparison relations are explicit invalidates the selector-facing problem record. When the selection use relies on evidence, provenance, currentness, or assurance, those separate relations and results must also be explicit; C.22 does not require them otherwise.
SoTA-Echoing
Wolpert and Macready's "No Free Lunch Theorems for Optimization", 1997, remains historical lineage for the warning that method superiority is distribution-dependent. It does not by itself supply the current C.22 field set, a selector policy, or evidence that one TaskSignature is adequate. The following sources inform different parts of this task-typing use.
Relations
Builds on: C.16 MM-CHR, G.0 CG-Spec. Coordinates with: G.4 Acceptance, G.5 Selector, C.18 NQD-CAL, C.19 E/E-LOG, C.23 Method-SoS-LOG, and C.32.P2S when typed problem pressure continues into architecture selected structures and synthesis. Constrained by: E.10 for selected EntityOfConcern, Description-episteme, specification-use, and publication-lane wording. Coordinate with E.18 only when a named receiving use depends on a selected transformation-flow crossing or gate position; E.18 supplies neither the F.9 Bridge nor a mandatory publication package.
Practical Use Checks
- If two candidate approaches are answering different
TaskKinds or differentScopeSlice(G)cuts, a direct comparison is not admissible yet. - For a specialization claim, recover the holder or portfolio, task family, performance criterion and grounds. Recover acquisition history and the C.22.1 adaptation budget and threshold-crossing basis when those are the claim or comparison being made.
- If crossing, normalization or missingness changes the comparison, state the relevant meaning and rule in the signature or its cited basis. Ordinary explanatory prose is sufficient when it supplies that declared content; an undocumented assumption is not.
- If
QDorOEEheads are in scope, archive and generator fields belong in the same typed signature rather than in a detached explanatory appendix.
Generator-specific difficulty targets
When a declared search or learning policy seeks feasible but challenging tasks, state the feasibility and difficulty criteria for that task family in the policy. Use C.19 when the question is treatment of a still-live pool, and C.18 for the generation or archive account. Difficulty alone does not establish that a TaskSignature is informative. An ordinary signature includes only traits needed by its receiving use, with no general requirement to make the task harder. G.1 applies when authoring a reusable generator/selector kit; it supplies no universal task-difficulty target.
C.22:End
Task-family adaptation signature
Status: Stable
Purpose. Compare how a holder, dyad or bounded specialist portfolio acquires usable specialization on a declared task family. An adaptation signature connects the threshold, acquisition time and cost, and prior exposure with any claimed transfer, retention, downside or entry into a new solution corridor. An explicit account in ordinary prose is sufficient when these distinctions and their grounds are recoverable; a shared record can support later selector or parity use.
Keywords. adaptation signature; task-family specialization; time-to-threshold; budget-to-threshold; prior exposure; corridor entry; stepping stone; transfer; retention; downside field.
Problem frame
Final task score alone does not tell whether a holder, dyad, or bounded specialist portfolio acquired usable specialization quickly, under what budget, with what prior exposure, whether the resulting competence transferred, or whether the adaptation path entered a new solution corridor. If those elements are not published together, the adaptation claim splinters across task typing, probe notes, selector prose, and parity notes, and later readers can no longer tell what was being compared.
Problem
A comparison of specialization acquisition needs a common task family and work target, with enough acquisition history to distinguish equal final scores reached under different conditions.
Use this when
- the live task-family adaptation claim is not only that a holder or dyad solved a task, but how fast it acquired usable specialization on a declared task family
- comparison must account for the work-measure threshold target, prior exposure, adaptation budget, transfer field, and reuse window
- movement into a new solution corridor or stepping-stone family is part of the novelty claim
What goes wrong if missed
- adaptation claims collapse into vague
got betterlanguage with no declared work-measure threshold target or budget-to-threshold account - parity later compares outcomes that were reached under different prior exposure, different work-measure threshold targets, or different reuse windows
- nonhuman or unfamiliar solution corridors are either treated as novelty or dismissed as noise because the corridor entry was never typed
What this buys
- adaptation speed becomes reviewable by value on the same declared
TaskFamilyand work target - later
G.5 / G.9portfolio and parity work can compare specialization claims on the same declared basis - stepping-stone or solution-corridor movement becomes visible as one typed part of the adaptation claim
Forces
Solution — one adaptation signature over the C.22 anchor
- Start from the work the holder is expected to perform under the stated support conditions. What must the holder produce, which decisions belong to them and which results or instructions may someone else supply? Choose a threshold that distinguishes doing that work. An accepted product can answer a commission to deliver it. To assess the ability to construct a method, look for the holder's choices about its operations and how they work together. If the claim covers a task family, establish which variations in the work the available performances or other grounds cover. Use the shared field meanings below when passing that comparison to G.5 or G.9.
- Bind the adaptation signature to the declared
TaskFamilyReforTaskSignatureand work target. Name whose capability is being judged and what tools, references or help are part of the intended work. Record who performed the operation on which the threshold depends and what help was supplied during practice and the check. A joint performance can establish what the participants can do together. For a claim about one participant, use a performance that shows what that participant can do under the intended support conditions. Reuse an existing account when it supplies those grounds. - The signature should expose at least:
thresholdTargettimeToThresholdbudgetToThresholdpostThresholdEfficiency?priorExposureDeclarationtransferTarget?transferGain?retentionWindow?downsideEffect?corridorEntryBaseline?corridorEntryEvidence?steppingStoneEvidence?
- Anchor these fields to the same work target and work-measure threshold semantics declared by
C.22. State the time origin, budget boundary and threshold-observation rule. If the threshold was not reached or the available history cannot establish its first crossing, say so and retain the observed limit; do not invent a crossing time or cost. C.22carries the declared task-family anchor, task typing, and baselineTaskSignature.C.22.1narrows the adaptation-signature question to threshold timing, reuse, downside, and corridor-entry disclosure over that existing anchor.
Corridor, transfer, and durability discipline
- If the adaptation claim depends on entering a new solution corridor, publish the
corridorEntryBaselinefirst: the prior repertoire, baseline set, or comparison family relative to which corridor entry is being claimed. - State
corridorEntryEvidencethat supports the claimed difference from that baseline, such as a reproducible solution class, a stable descriptor shift or a supported stepping-stone sequence. Name what distinguishes the corridor; an unfamiliar sample alone does not establish entry. - If a stepping stone mattered, publish the stepping-stone evidence as part of the adaptation signature.
- Corridor or stepping-stone notes do not replace the work-measure threshold account; they explain why the adaptation path matters, not whether the threshold was reached.
- A fast threshold result is not yet enough to claim durable specialization.
- For transfer, identify the target task family, baseline and observed change there. To attribute that change to the earlier practice, give grounds that distinguish its effect from repeat testing, practice on the target or changed assistance when those explanations are plausible (C.28). State how far the evidence reaches across the target family.
- For retention, report when and under what support conditions performance was demonstrated again. Use a comparable criterion over the stated delay. An unobserved interval can remain open.
- If specialization harms neighboring task families, narrows reusable competence, or creates de-specialization cost, publish that in
downsideEffect?. - If post-threshold performance matters to later exploitation, publish
postThresholdEfficiency?to show performance after threshold crossing.
Worked moment
Consider two research setups, A and B, adapting to the same catalyst-search family. This is an illustrative comparison. The declared threshold is at least 80 accepted predictions out of the same 100 held-out cases under a fixed validation rule. Both start with the same training corpus and no prior exposure to those held-out cases. Time starts when adaptation begins; the budget counts all adaptation and evaluation calls. The rule checks threshold attainment at the end of each round.
A first passes a scheduled check after 2 hours and 120 calls, B after 6 hours and 480 calls. Both finish at 85/100. The histories show what each setup spent by its first observed pass. To conclude that either acquired the capability during adaptation, establish how it performed beforehand under comparable conditions. A failed check followed by a pass brackets the observed change at the declared checkpoints.
On a named adjacent family, a before/after test gives A 60/100 then 70/100 and B 60/100 then 60/100. A improved by ten percentage points on this test. B's observed score stayed the same. These observations leave open how much change the original practice caused. A comparison that accounts for repeat testing, other training and assistance can support a transfer claim for the conditions it covers. Retention beyond the measured run is untested. A corridor-entry claim would need the prior repertoire and evidence of a repeatable solution class absent from it.
The comparison can finish with these bounded claims. If A's prior exposure later proves different, reopen the equal-exposure comparison while retaining the observed histories. A later retention claim needs its own window and observations. A selector may use the signature under its applicable rule; the signature alone chooses neither setup.
Consequences
- Acquisition time and cost can be compared with prior exposure and any claimed post-threshold efficiency on one recoverable basis.
- Selector and parity pattern applications can consume a stable upstream adaptation signature without silently redefining its fields.
- Claimed corridor entry and known downside remain visible alongside the specialization gain.
Rationale
The reader needs one place where the adaptation claim stays whole. C.22 keeps the task family and work target explicit.
A.15 distinguishes probe plans from performed Work. A claim about a particular dated Work occurrence needs A.13 and independent A.15.1 admission; precise assignment-bound attribution additionally uses F.6. C.24 defines call-planning CheckpointReturn records after an action or option is fixed; E.16 governs budget, guard, and ledger enforcement for autonomy claims. Probe results, checkpoint records, and budget records may supply evidence for the adaptation claim.
G.5 and G.9 later compare several candidates or parity runs. C.22.1 keeps the specialization story readable across those neighbouring pattern applications by connecting threshold timing to the claimed reuse, downside and corridor-entry evidence.
SoTA-Echoing
Claim 1. A comparison of specialization acquisition keeps the declared work-measure threshold and prior exposure explicit. It reports the observed time and cost to threshold. If the first crossing is unresolved, give the time or cost bounds the observations support. Distinguish a successful first check at the end of a month from evidence that acquisition took a month. Report non-attainment for the monitored interval only when the observation rule supports it.
Practice source, local alignment, and adoption decision. FactorMiner v1 (2026-02-16), §§4.1–4.4, separates factor quality and redundancy, the effect of experience memory on screening yield, and factor-evaluation computation time. SkillOpt v2 (2026-05-25), §§3.1, 3.4–3.5 and 4.3–4.4, changes an external skill text while holding the task-execution model fixed, selects text revisions through held-out validation, and evaluates transfer and training tokens per test-score gain. This pattern adopts these distinctions among the changed object, attained result, evaluation or training cost, and tested reuse.
The shared field set is an original FPF construction. For the question in §§1–6, §7 adapts those narrower comparisons by connecting the acquisition history to one TaskFamilyRef or TaskSignature and work-measure threshold. A terminal score and total cost per score gain can describe the attained result and average improvement cost, yet leave the resources needed to reach a particular threshold unresolved.
Recording threshold timing and prior exposure adds observation effort but makes histories with the same final score distinguishable; §8 then separates claimed transfer, retention, and downside. The pattern rejects substituting either paper's reported metrics for this shared acquisition account.
Reopen the comparison if a rival protocol supports the same specialization judgement with less observation, or exposes a consequential difference that these fields cannot express.
Claim 2. A corridor-entry claim needs a baseline and evidence for the claimed difference.
C.17 separates novelty and diversity relative to a declared comparison basis; C.18 distinguishes generated candidates, retained archives and fronts. Here that distinction supplies a conditional question: when acquisition includes a claimed new corridor, which prior repertoire and observations establish it? The fields in §8 make that evidence recoverable without treating one unfamiliar sample as reproducible entry.
Claim 3. Selector and parity comparisons depend on compatible interpretations of the fields they consume.
G.5 and G.9 consume this shared field set. Their users must retain its threshold, prior-exposure and conditional reuse meanings when comparing candidates; a silent downstream redefinition would change the comparison basis.
Evidence-source note. Use each source result with its measured object and evaluation conditions. Factor admission and skill-edition selection concern different objects from a holder's acquisition of usable specialization. For timeToThreshold and budgetToThreshold, relate the acquisition evidence named in §11 to the declared threshold and prior exposure; faster evaluation or tokens per score-point gain alone cannot supply that history. Transfer, retention, and corridor-entry claims keep their separately stated evidence requirements.
Relations
C.27 temporal-claim relation.
-
C.27 may flag: a claim that a holder, dyad, team, explicitly scoped specialist portfolio, or agent acquires usable specialization faster on one declared
TaskFamilyReforTaskSignature. A method may be selected, refined, or retired as part of that story. -
This pattern keeps: threshold target, time-to-threshold, budget-to-threshold, prior exposure, transfer, retention, downside, corridor-entry evidence, and adaptation-signature fields.
-
"learns faster" wording without a task-family anchor leaves the adaptation claim incomplete. Faster threshold crossing alone does not establish durable specialization; retention needs its own window and evidence, and any additional transfer, downside or corridor-entry claim needs its stated grounds.
-
Next-question boundary: classify the claim as Dyn1 when only a trend is live; use C.24 for call planning after the action or option is fixed; use the C.22.1 adaptation signature when specialization is the live adaptation question.
Builds on: C.22 TaskSignature anchoring, C.19.1 BLP compatibility, A.15 system-role-kind and assignment, Method, WorkPlan, and Work-occurrence separation, C.24 call planning and CheckpointReturn semantics after the action or option is fixed, E.16 budget enforcement.
Coordinates with: G.5 selector specialization profiles, G.9 adaptation parity, G.11 later telemetry/refresh reuse.
Coordinates with: E.23 when repeated object improvement includes a claim of durable task-family specialization. C.22.1 supplies the adaptation account for that claim; E.23 governs the improvement loop.
Constrained by: E.10 lexical discipline.
Not this pattern when
- the claim only needs to name the task family and work-measure threshold target, with no adaptation-speed or transfer claim; ordinary
C.22anchoring is enough - the question under repair is already selector or parity law across candidate selected sets; that belongs to
G.5 / G.9 - the threshold or prior-exposure basis cannot yet be established: retain a narrower observation or an unresolved adaptation claim. Missing corridor evidence leaves that corridor claim unresolved; it need not prevent a separately supported threshold account
Conformance checklist
CC-C22.1-1An adaptation signature SHALL bind to one declaredTaskFamilyorTaskSignature, one work target, and one work-measure threshold target.CC-C22.1-2An adaptation signature SHALL report prior exposure and the observed time and budget to threshold, or the supported limits on each. Distinguish a known crossing, a threshold met by the check with the first crossing undetermined, and a threshold known to have remained unmet. If the observations establish none of these, report what remains unknown.CC-C22.1-3Any declared transfer, retention, post-threshold-efficiency, downside, corridor-entry, or stepping-stone claim SHALL identify what is claimed and for which target. Give the grounds for that claim under §8, including the baseline or observation window when the claim depends on it.CC-C22.1-4This pattern may refine specialization timing and reuse claims over the declaredC.22anchor, but it SHALL NOT redefine acceptance-gate thresholds, task-family attachment, or selector/parity law governed by another FPF pattern.CC-C22.1-5Downstream selector/parity pattern applications SHALL cite or consume the same published adaptation-signature field set rather than silently redefining threshold, prior-exposure, transfer, retention, downside, or corridor-entry terms.
C.22.1:End
Problematic-For Relation
Type: Conceptual (C) Status: Stable Normativity: Normative unless marked informative
Plain name. Actual problem.
Problem frame
Use this when. Use this pattern when an actual condition may be adverse for one entity and use, and a receiving claim needs to distinguish the actual Problem from a signal, criterion description, evaluation, assessment claim, ProblemCard, or claim that no suitable method is currently known.
First useful move. Name the actual-condition relation. Then name the predicate, entity, scope, and interval for which that predicate applies. If the condition falls on the adverse side of that applicable predicate, say plainly: "This condition is a problem for this entity in this scope." Expose a PFR occurrence only when another claim needs that Problem identity.
What goes wrong if missed. A card or evaluation result is allowed to create a Problem; the same criterion is applied to the wrong entity or scope; a new description edition creates a false new Problem; or one continuously adverse episode is split every time evidence is sampled. Conversely, two adverse episodes separated by actual non-adverse behavior collapse into one occurrence.
What this buys. Actual Problems can exist before discovery, can be referenced while still ongoing, and can be distinguished across repeated adverse episodes. One applicability relation supplies the predicate, problem-for entity, claim scope, and declared criterion-applicability window used by PFR; its actual occurrence extent is separately derived from uninterrupted obtaining. Measurements, evaluations, evidence, claims, cards, and method search remain available without becoming Problem identity.
Battery example. A technician needs to know whether a vehicle may be released for its intended start. Recover the release rule, its battery and load conditions, and what the available reading establishes under those conditions. A bound wholly below the permitted voltage can settle that decision. A different question arises if a service history must identify the particular low-voltage episode. That use needs an actual voltage-state relation with defined participants, conditions of occurrence and temporal identity. The reading, alarm and report can support the claim about the battery. The formal PFR occurrence remains unresolved while those grounds are missing.
Once such a governor exists, the applicability relation can connect its selected voltage predicate to the vehicle, intended-start U.ClaimScope, and declared criterion-applicability window; discovering a method still changes solvability rather than PFR actuality.
Use another pattern when. Use C.22.2 when the current object is a problem-side card, signal, hypothesis, forecast, scenario, anticipated-condition claim, or reviewable formulation rather than an actual PFR.
Use C.27 when a rate, rhythm, recovery, or regime claim is being used to change action, C.28 when causal support is at issue, or the direct forecast, scenario, counterfactual, or anticipated-condition governor when that claim is current.
Use the selected A.19 comparison, G.4 acceptance, state, gate, or measurement pattern when the current question is how to evaluate or support the adverse predicate.
Use E.18.1, E.23, and the direct NQD or OEE patterns for repeated problematization, search, work, evaluation, and continuation.
Problem
An actual Problem depends on an actual condition and on a criterion that applies to one entity, scope, and interval. It is neither a record nor a free-standing quality label. The relation obtains when the condition is on the adverse side of that applicable predicate.
FPF needs one identity for this dependent evaluative relation without copying values defined by ProblemCriterionApplicabilityRelation into additional writable PFR slots. It also needs to distinguish continuous adverse episodes from repeated episodes without inferring a recovery from missing observations or evidence.
Forces
Solution
Model an actual Problem as one obtaining ProblematicForRelation, a dependent evaluative U.Relation between exactly two relation occurrences: the actual condition and the applicability of a characteristic-space predicate.
Use one criterion-applicability relation
CharacteristicSpacePredicate is a by-value predicate used by an A.19 comparison, acceptance, state, gate, or other direct consumer. It is not a new U-kind, publication record, description edition, or comparison result. Its meaning is recoverable from the declared characteristic-space coordinates, scales, normalization or bridge values, operator, cut or band, polarity, and the selected direct consumer's governed comparator, admissibility, and predicate-use semantics.
A performed evaluation remains U.Work; its result episteme and evidential-support relations remain separately governed, and none of these constitutes predicate meaning.
Before testing adversity, identify the characteristic the predicate compares, how the actual condition supplies its value and why it bears on this entity and use. The predicate carries one ConditionToPredicateInputRule as part of its semantics:
- Direct input: the actual-condition participant is already a governed characteristic-assignment or state relation whose subject pattern exposes the exact characteristic-space coordinate, scale, and value used by the predicate.
- Projected input: when that relation does not itself expose the needed point, the rule cites one exact governed projection or bridge, its source relation kind and participant positions, target characteristic space, coordinate and scale, and the direct relation or predicate connecting that input to the problem-for entity and receiving use.
Establish which input the rule takes from the actual condition and how it relates to the affected entity. A reference to the condition alone leaves that connection open. The evidence may bound the input closely enough to decide the comparison (§4.3). When two projections are plausible, the rule names the selected one; identify the nearest inadmissible alternative only when one fails the declared input or problem-for-link conditions. ConditionToPredicateInputRule is a pattern-local by-value rule inside CharacteristicSpacePredicate; it is not a U-kind, relation occurrence, evaluation result, or copied PFR field.
The F.18 NameCard identifies the applicability relation kind.
Use one individuable dependent U.Relation for applicability:
This relation states that one exact predicate currently governs one exact entity and claim scope under one declared criterion-applicability window. Its direct applicability predicate is satisfied while that criterion remains selected and governing for that entity, scope, and window; it does not ask whether any actual condition is presently on the adverse side. The applicability occurrence can therefore continue across adverse, non-adverse, and later adverse condition intervals.
It ceases when the criterion is withdrawn or replaced for that use, the entity or claim scope changes, the declared window no longer covers the use, or another direct applicability condition fails. One occurrence is identified by the four fixed participants plus the maximal continuous period of actual applicability. Changing a participant yields another occurrence; actual loss and later restoration of applicability yields distinct occurrences even when all four participants stay fixed.
A coextensional description edition or carrier change does not change either the predicate participant or the occurrence. Assessment windows, evidence-relevance intervals, description editions, claim-currentness windows, and adverse-truth intervals remain with their own claims and relations.
A semantic predicate change selects a different predicate participant; it is not an edition-only repair. If the new predicate replaces the old predicate for the same use and the old applicability therefore ceases, the old applicability occurrence ends and a new occurrence begins only if the new predicate applies. The PFR dependent on the old occurrence can then end, and another PFR can begin under the new occurrence when adverse truth obtains.
If both predicates remain applicable, two applicability occurrences may coexist; a new criterion description alone does not prove replacement or cessation.
Keep the PFR signature reduced to two participants
The F.18 NameCard identifies the dependent evaluative relation kind.
The obtaining ProblematicForRelation occurrence is the actual Problem. A ProblemCard, criterion description, assessment claim or local Plain label may describe or designate it without supplying a second world-side problem identity.
The complete non-derived participant set is:
The first reference resolves to the exact obtaining relation that constitutes the actual condition under its direct pattern. The second resolves to the exact obtaining applicability relation from C.22.PFR:4.1.
PFR has no separately writable condition-bearer, predicate, problem-for-entity, claim-scope, applicability-window, assessment-window, or description-edition slot. The actual-condition relation fixes its condition participants, and the applicability relation fixes the predicate, problem-for entity, claim scope, and declared applicability window. Assessment windows and description editions remain with their own claims. A readable claim projects the applicability values from that participant:
This is a derivation from one participant, not a consistency check between copies.
Use predicate truth as the obtaining condition
ProblematicForRelation obtains exactly when all three conditions hold:
- The exact
ActualConditionRelationobtains under its direct pattern. - The exact
ProblemCriterionApplicabilityRelationobtains because its criterion remains governing for the exact entity and claim scope under the declared criterion-applicability window, independently of whether the actual condition is adverse. - For the actual-condition participant, the predicate's exact
ConditionToPredicateInputRulemust yield the declared characteristic-space point or value and the direct link to the problem-for entity and use; that point falls on the adverse side under the selected scale, comparator, cut or band, polarity, and admissibility semantics.
The selected direct consumer identifies the input through the declared assignment or projection and supplies the comparator and admissibility rules. The actual value determines which side of the criterion obtains. An evaluation supplies grounds for a claim about that value or about the comparison.
To establish the side of the criterion from incomplete evidence, first establish the condition, the applicability of the criterion to this entity and use at the time in question, and at least one input value compatible with the evidence. If all compatible values fall on the adverse side, the evidence establishes adversity at that time. If all fall outside it, the evidence rules it out there. One feasible value on each side leaves the comparison open. A bound wholly on one side can settle the question. A bound extending to both sides may include unattainable values. When no value fits the evidence, revisit the observations and their interpretation (C.16.IR:4.3). The time at which adversity is established can lie inside an episode whose start or end is still unknown. The two obtaining relations in §4.2 remain the PFR participants. Measurements and evaluation outcomes can support assertions about them and about adversity.
A Problem can therefore obtain unnoticed. Later detection can involve work, evidence, and claims about the already obtaining relation; it does not create retroactive actuality.
When an account asserts a particular dated evaluation U.Work, recover its actual performer U.System through A.13, independently admit the Work under A.15.1, and name the U.Method actually followed. Name any declared A.6.1 operation application separately when its returned-value binding is claimed.
Add F.6 only when the evaluation account or a receiving use expressly consumes precise assignment-bound attribution through the same obtaining A.13 assignment; missing or failed F.6 attribution alone leaves the evaluation Work intact while its independent A.13/A.15.1 grounds still hold. A short PFR explanation may omit an assignment identifier that no later claim uses.
That Work may return the separate evaluation result true, false, or unknown defined by the selected evaluation pattern; a C.2.1 assertion may state the result.
Establish support for reliance under the assertion's direct subject rule; use A.10 when the evidence, provenance or bounded reliance needs a replayable account. B.3 governs any named assurance claim, and G.11 may qualify the assertion's current edition. The practitioner may rely on the assertion for the receiving Work, decline or defer that reliance, or reopen its basis. unknown is an evaluation result, never a world-side PFR value; test PFR obtaining through the three conditions in §4.3.
Identify repeated adverse episodes from world-side continuity
The direct occurrence rule is ontic. With the two participant occurrences fixed, one PFR occurrence is the maximal continuous episode during which both participants obtain and the selected condition point is actually on the adverse side. Actual cessation of either participant or actual movement to the non-adverse side ends that occurrence. Later renewed adverse truth starts a later PFR occurrence. Measurement, evaluation, demonstration, assessment, and evidence availability neither start nor end either occurrence.
Use this world-side identity basis:
adverseEpisodeStart is the episode's actual inception under the declared temporal reference, not the first observation or report time.
When admitted time grain cannot distinguish co-inceptions, or when a receiving history must keep one reference while a claim about inception remains revisable, individuate the occurrence under A.6.REL and assign a stable pfrOccurrenceId that designates it. The resolution record may carry the current boundary claim, but neither its asserted start nor its asserted end becomes a mutable identifier field.
The PFR's actualAdverseExtent is derived from the world-side episode: its end becomes fixed when the episode actually ceases, and a later claim may recover or correct that boundary. The recovered end and completed extent describe the occurrence; they do not replace it or change its assigned reference. This applies A.6.REL's participant-plus-episode rule without making a currently known boundary constitutive.
Participant references alone remain insufficient because the same participants can enter adverse, non-adverse, and later adverse episodes. A universal evaluation reference is also insufficient because an unnoticed PFR can obtain and several assessments can support one occurrence.
Keep world-side occurrence and current boundary claim separate
Use [adverseEpisodeStart, open] only in a current claim whose evidence supports that this same PFR occurrence still obtains at the claim's stated reference time. open is a claim-side endpoint sentinel, not the clock time, an identity field, or a substitute for missing evidence. The stable occurrence reference remains unchanged when a later claim records the recovered actual end.
A current assertion or description distinguishes three cases in ordinary words:
- Current adversity. Report adversity at the reference time when the observations and criterion establish it. Publish
[adverseEpisodeStart, open]when the start and continuity of this occurrence through the reference time are also supported. - Established cessation. Report cessation and whatever bounds the evidence establishes for its time. If the condition was adverse at 09:00 and non-adverse at 11:00 while the criterion continued to apply, the 09:00 episode had ended by 11:00. Publish
[adverseEpisodeStart, adverseEpisodeEnd]when both endpoints are established for that occurrence. - Unresolved continuity. When a gap could contain cessation and restart, retain the known occurrence reference and supported segments. State which claims about the gap remain open.
Later evidence may revise the assertion, including its claimed start, end, or continuity, without creating or changing the world-side episode. If it supports an earlier actual cessation, complete the extent of the earlier occurrence at that boundary. If it also supports later renewed adverse truth, identify the later occurrence from its own actual inception. Until that distinction is supported, do not attach the later adverse segment to either the old or a new occurrence by default.
Consider three successive condition intervals A, B and C while holding one continuously obtaining applicability occurrence fixed. The criterion remains governing for the same entity, scope, and declared window through A, B, and C; only world-side adverse truth changes:
- During A, the condition is actually adverse, so the first PFR obtains from
A.startuntil its actual cessation atA.end. - During B, the condition is actually non-adverse, so the first PFR does not obtain.
- During C, the condition is actually adverse again, so a later PFR begins at
C.startwith the same two participant references and a different stable occurrence reference.
Evidence that supports A, B, and C warrants the corresponding two-occurrence assertion. Return continuity unresolved when the available evidence cannot decide whether cessation or restart occurred. A missing assessment, unavailable measurement or stale item alone settles neither continuity nor recovery; other adequate evidence may settle the claim. Adjacent or overlapping assessment windows likewise do not split or join world-side episodes by themselves.
Keep assertions, reliance, anticipated conditions, solvability, and cards separate
An assertion about the PFR obtaining predicate has affirmative or negative claim polarity. An affirmative assertion may designate an independently established occurrence; a negative assertion denies predicate satisfaction for the named participants and qualification but does not erase or reidentify an earlier occurrence. Judge support for that assertion under the PFR obtaining predicate and the applicable subject rule.
Use A.2.4 for a question about an episteme's evidence-use or status-use relation, and A.10 when evidence, provenance or bounded reliance needs a replayable account; B.3 applies only when a named assurance claim is current. Assertion polarity, support, and reliance therefore answer different questions.
A possible or anticipated problem remains a forecast, scenario, counterfactual, or anticipated-condition claim in ProblemCard or another episteme until an actual-condition relation, an applicability relation, and adverse predicate truth all obtain. C.2.1 governs its assertion identity and polarity. Use C.27 when a rate, rhythm, recovery, or regime claim is being used to change action and C.28 when causal support is at issue. The direct claim pattern defines or constrains assumptions, horizon, and non-actual semantics. None of those claim-side facts establishes a current PFR.
A ProblemCard is one C.2.1 episteme with one exact ClaimGraph, one independently identified EntityOfConcern, and one effective reference scheme. It may carry claims designating several PFR occurrences only when those claims are jointly about that one EntityOfConcern under the direct pattern that identifies it. When two PFR references lack such a joint concern, split the ClaimGraph and card.
Conversely, several cards may designate the same PFR through different ClaimGraphs, schemes, viewpoints, or receiving uses. Card count, merge or split, currentness, assessment window, publication, carrier, and edition change neither PFR actuality nor identity.
C.22.2:20.1b illustrates all three branches:
- Two Robot-7 PFR episodes share one A.1-identified
Robot-7and one card; - Robot-7 and Robot-8 PFRs have no direct joint EntityOfConcern and force two ClaimGraphs and cards;
- Two differently qualified cards retain the unchanged
PFR-InspectionAssignment-17reference.
A claim that no supported method is currently available concerns the admitted method set, evidence, constraints, and intended use. Selecting or discovering a method changes current solvability; it does not end PFR while the actual condition remains adverse. Performed repair work can end PFR only when an independently recovered actual change makes a participant cease or moves the selected condition point to the non-adverse side.
Repeated problematization, method search, work, evaluation, and continuation occur in work and transformation flows governed by E.18.1 and E.23. A claim or plan may carry a reference to the same PFR while work and transformation occurrences participate in a selected transformation-flow structure. Neither that reference use nor any flow-structure relation enters PFR identity. A later PFR is a later occurrence because a participant changes or because actual adverse truth begins again after actual cessation, not because another card, assessment, or flow visit exists.
Preserve the lightweight path
For a first use, name only what decides the case:
- the actual condition, when it is established, and what the evidence shows about the input from that condition;
- the criterion that makes that point adverse, including the selected input path and cut or band;
- the entity for which it is a Problem, the use or claim scope, and the applicability window.
Then write one ordinary sentence:
At
<time or supported interval>,<condition>falls on the adverse side of<criterion>for<entity and use>. This criterion applies during<applicability window>.
Stop there when the next work only needs to recognize the Problem. Cite the direct condition and criterion patterns, but do not fill a PFR record, repeat either relation signature, or assign a PFR identifier. The NameCards and signatures above define reusable semantics; they are not a mandatory user form.
Add explicit identity only when another claim must compare, qualify, change, nest, plan from, or refer back to this particular Problem occurrence. That receiving claim then names the exact actual-condition occurrence, exact applicability occurrence, actual adverse inception or stable PFR identifier when needed, claimed extent, and any evidence or assessment claim on which its reliance depends. The evidence remains separate from the world-side Problem.
Archetypal Grounding
Worked first use — an assignment that violates a release criterion. InspectionReleaseAssignment is a declared U.SystemRoleAssignment species defined under A.2.1. It defines the holder and assigned-kind participant meanings, assignment predicate, applicability, and occurrence identity. Occurrence InspectionAssignment-17 has admitted System Robot-7 as holder and local kind InspectorSystemRole as assigned-kind value. It obtains without interruption during [2026-07-13T09:00, 2026-07-13T17:00].
MaintenanceRoles-2026, Maintenance-Scheme-A, and the interval description may interpret or describe the assertion; they are not extra world-side participants. A roster row and InspectionAssignmentAssertion-17 may describe the assignment and interval; neither makes the relation obtain.
A.2.1 identifies InspectionAssignment-17 through the complete real participant set and the maximal uninterrupted period in which the direct species predicate obtains. Demonstrated non-assignment after the stated 17:00 boundary ends it. If the same participant set again satisfies the predicate on the next day after that actual gap, InspectionAssignment-18 is a later occurrence; an evidence gap by itself neither ends nor splits the first occurrence. The subject pattern therefore supplies the participant meanings, obtaining rule, temporal extent, recurrence, and same-versus-new-occurrence rule required of ActualConditionRelationSlot.
The by-value predicate NoInspectorSystemRoleBeforeValidation-v2 uses one direct ConditionToPredicateInputRule: from InspectionAssignment-17 it reads the assigned-kind participant as coordinate assignedSystemRoleKind = InspectorSystemRole on a nominal system-role-kind scale and the holder participant as the direct link to problem-for entity Robot-7. Its adverse region is the declared set of system-role kinds prohibited for that System's autonomous-inspection release before validation, which contains InspectorSystemRole. The same kind assigned to another System is the nearest inadmissible input because its holder participant does not link the condition to Robot-7; a roster row is inadmissible because it is not the obtaining assignment occurrence.
Robot7ReleaseCriterionApplicability-4 is the separate applicability occurrence. Its four participants are that predicate, Robot-7, autonomous inspection release before validation as the exact U.ClaimScope, and declared window [2026-07-13T08:30, 2026-07-15T18:00]. Because both participant relations obtain and the selected point is adverse, PFR-InspectionAssignment-17 obtains on [2026-07-13T09:00, 2026-07-13T17:00]. The ordinary first-use sentence is:
Robot-7 is assigned as inspector through
InspectionAssignment-17, whileInspectorSystemRoleis prohibited for this System's autonomous-inspection release before validation during the declared 13–15 July window. This assignment is an actual Problem for Robot-7's release during its 13 July 09:00–17:00 assignment episode.
Evaluation and reliance remain separate. Admitted SafetyReviewSystem-2 : U.System performs independently admitted, dated InspectionReleaseCheckWork-21 under independently obtaining SafetyReviewAssignment-9 : SafetyReviewWorkAssignment <: U.SystemRoleAssignment; F.6 uses the assignment's holder projection, and the Work enacts InspectionReleaseCheckMethod-3. The evaluation result is true for the adverse predicate. InspectionReleaseCheckAssertion-21 states that result; its evidence path and assurance tuple can warrant one release decision, G.11 can qualify the assertion edition, and the release decision-maker can rely on the assertion or decline that reliance. None is a third PFR participant.
A ProblemCard may later designate PFR-InspectionAssignment-17; creating, revising, splitting, or publishing that card changes neither PFR participant nor actuality.
If evaluation never occurs, the PFR still obtains. If the assertion is stale, the basis for that reliance may become unresolved, but the world-side relation is not thereby created, ended, or split.
Actuality and recurrence. Selecting another staffing Method and writing a Work order do not end the PFR. Only actual cessation of the A.2.1 assignment predicate at the stated 17:00 boundary ends it. When the same direct species predicate resumes on the next day as InspectionAssignment-18 while the applicability occurrence still obtains, A.2.1 identifies that later system-role-assignment occurrence and its later adverse inception grounds PFR-InspectionAssignment-18. An evidence gap alone establishes neither continuous assignment nor withdrawal and reassignment.
Unestablished condition — installed component. Fuse-R17 : U.System and Panel-7 : U.System may be named in a parts claim, but current A.14 supplies no installed-part relation kind, installed-part participant meanings, obtaining predicate, temporal extent, recurrence, or same-versus-new-occurrence rule. Under A.6.REL:5.2, ComponentOf-FuseR17-Panel7 is therefore not introduced as an individuated installed-part occurrence here. A parts list, inspection note, removal report, or reinstallation wording cannot fill ActualConditionRelationSlot; the fuse case remains non-conforming until an accepted direct installed-part pattern supplies the required relation definition.
Unestablished condition — battery voltage. Current A.18 and C.16 can govern the voltage characteristic, scale, measurement work, result, uncertainty, and assertion. They do not supply a direct voltage-state relation with participant meanings, obtaining, temporal extent, recurrence, and occurrence identity. Therefore TerminalVoltageState-12 is not introduced as a relation occurrence here, and the low-voltage case remains non-conforming until an A.6.RCD decision selects or assigns that direct governor. MeterReport-88, an alarm, or a maintenance card may support a claim but cannot fill ActualConditionRelationSlot or backdate PFR.
Unestablished condition and use — proof gap. Current proof and assurance patterns define or constrain proof epistemes, obligations, evidence, and relying decisions, but this case lacks direct patterns supplying both (a) an individuable unresolved-consequence relation with its obtaining and episode law and (b) the exact proof-use or acceptance object and applicability relation needed by the case.
Keep the condition gate and problem-for/applicability gate separate. UnresolvedConsequence-17, ProofUseEntity, and an omnibus proof-gap object are not admitted by this wording; the case remains non-conforming until both direct governors close.
Unestablished condition — clinical condition. A diagnosis, assessment, measurement report, and patient label are epistemes or context-dependent cues, not the clinical-condition occurrence or patient identity. This case lacks a clinical-condition pattern supplying the needed participants, obtaining, recurrence, identity, and temporal extent. Until one does, keep the case non-conforming.
If patient also names a work-facing classification, recover the classified System and local system-role kind. If an assignment is claimed, recover its holder System, assigned kind, occurrence, and declared U.SystemRoleAssignment species; none supplies the missing clinical-condition relation.
Unestablished condition and use — missed transfer. First distinguish transfer Work, a world-side transfer or delivery relation, a commitment or acceptance relation, and a package or record episteme. An E.18 connection position resolves to a selected relation; neither that position nor the phrase hand-off failure establishes a missed-transfer condition. This case lacks a direct pattern supplying the exact missed-transfer condition relation, and the phrase receiving work does not decide whether the problem-for entity is intended U.WorkPlan, dated U.Work, or another governed entity. The case therefore remains non-conforming until separate direct governors settle both questions.
Unestablished condition with two proposed uses — hot surface. This case lacks a direct pattern supplying an individuated hot-surface condition relation; a temperature reading cannot substitute for it. If such a governor later exists, hold its one occurrence fixed and use two distinct applicability occurrences when an exact receiving U.Work and an exact U.System have different problem-for fillers, scopes, declared windows, or applicability continuity. Adverse truth would then yield two PFR occurrences sharing only the condition reference. Until the condition gate closes, this remains a multiplicity test, not an asserted example.
Repair outcomes. Keep four outcomes distinct:
- A repair method is selected;
- Repair work is planned;
- Dated repair work occurs without a demonstrated condition change; or
- Dated repair work is connected through its direct change/result governors to actual cessation or non-adversity of the condition.
While the criterion continues to apply, this PFR ends when the actual condition ceases or becomes non-adverse. It may end before anyone observes the change. To credit the repair with that outcome, establish the change and how the performed work brought it about. A record can support the particular claim its content and provenance warrant.
Bias-Annotation
This pattern has an actuality bias: Plain problem names an obtaining dependent relation. Forecasts, scenarios, counterfactuals, hazards, and problem formulations remain useful epistemes under their claim governors without backdating actuality.
It has a predicate-centered bias because adverse truth is load-bearing. The applicability relation prevents a criterion from becoming globally adverse by label; exact entity, scope, and interval stay explicit.
It also has a continuous-time bias for occurrence identity. Discrete-state and event-based domains can supply equivalent actual episode boundaries under their temporal reference. An unresolved gap in the evidence needed to decide continuity leaves that assertion open in either representation; it does not decide the world-side boundary.
Conformance Checklist
- The actual condition is an explicitly individuated obtaining
U.Relationunder its direct pattern. CharacteristicSpacePredicateis given by value and includes one exactConditionToPredicateInputRule: either a direct governed characteristic assignment/state relation or a named governed projection/bridge to the exact coordinate, scale, and problem-for link. A criterion-description identifier, arbitrary relation reference, or plausible alternative projection is insufficient.ProblemCriterionApplicabilityRelationhas the predicate, exact problem-for entity, claim scope, and declared criterion-applicability window as its four participants. It obtains while that criterion governs those participants, independently of adverse-condition satisfaction; its occurrence extent is the maximal continuous period of actual applicability.- PFR has exactly two non-derived participant slots: actual condition and criterion applicability.
- Predicate, entity, scope, declared criterion-applicability window, and actual applicability-occurrence extent are projected from applicability rather than copied into PFR.
- The selected direct consumer identifies the input from the actual condition, its link to the entity and use, and the applicable comparison. Evidence supporting the adverse side at a stated time may establish an exact value or constrain the input sufficiently to settle the comparison under §4.3.
- Evaluation work, outcomes, measurements, evidence, assessment claims, descriptions, and cards do not create or identify PFR by default.
- Repeated PFR occurrences use the two participant references plus actual adverse inception. Add a stable occurrence identifier only for co-inceptions or a receiving history that must survive revision of the claimed inception boundary; the currently asserted start, derived end, and completed extent are not mutable key fields.
[adverseEpisodeStart, open]is used only by a claim that supports current continuous obtaining; a later supported end completes the claimed extent on the same stable occurrence reference.- Actual non-adverse B between actual adverse A and C yields two world-side PFR occurrences. Evidence can support, refute, or leave that boundary unresolved; a gap alone proves neither one occurrence nor two.
- Method availability and solvability claims remain separate from PFR actuality and identity.
- Possible conditions remain forecast, scenario, counterfactual, or anticipated-condition claims under their direct governors until both participant relations and adverse predicate truth obtain; assertion polarity and reliance posture do not substitute for those obtaining conditions.
- Ordinary readable use can stop before explicit PFR materialization when no receiving claim needs Problem identity.
- Installed-component, battery-voltage, proof-gap, clinical-condition, missed-transfer and hot-surface cases remain non-conforming until their named direct condition and, where applicable, problem-for/applicability governors exist; a phrase, measurement, diagnosis, record, card, or position in a flow structure does not supply those relations.
Common Anti-Patterns and How to Avoid Them
Consequences
Benefits. PFR gives actual Problems stable identity without turning cards or evaluations into world-side constituents. Two entities or scopes can use one predicate without collision, repeated adverse episodes remain distinguishable, and ongoing episodes can be referenced before closure. Claims state what the evidence leaves unresolved.
Costs. A load-bearing use must recover a by-value predicate, its exact condition-to-input rule and problem-for link, and one exact applicability relation. Direct consumer patterns must state how the selected point is obtained and how adverse truth is evaluated, including comparator and polarity semantics. Temporal identity requires a real recovery boundary rather than assessment timestamps.
Limits. C.22.PFR does not define domain criteria, measurement methods, comparator semantics, acceptance policy, evidence sufficiency, method search, repair work, or problem-card publication. It governs only actual Problem obtaining, dependence, projection, and occurrence identity.
Rationale
Applicability is independently useful: it states which predicate applies to which entity and use under which declared criterion-applicability window even when no adverse condition currently exists. Keeping those four participants canonical there prevents disagreements between duplicated fields, while maximal continuous actual obtaining distinguishes repeated applicability occurrences without adding a fifth participant. PFR adds the missing fact: the named actual condition is adverse for that applicability occurrence.
The maximal continuous adverse episode resolves an identity collision, but its completed interval value is not a stable reference key. Participant references plus actual adverse inception retain one occurrence reference before and after closure; the derived end completes its extent. Actual non-adverse behavior ends the occurrence and later renewed adverse truth starts another. Assessments can support, refute, or leave those boundary claims unresolved, but they neither supply the world-side boundary nor create the occurrence.
SoTA-Echoing
External-source qualification and reopen. The three non-FPF rows are qualified to the cited 2026 material and the current FPF direct relation and temporal interfaces used here.
The selected seminar material is only a practice-pressure comparator for separating problematization, criteria, method search, performed work, and results; it supplies no ontology or occurrence law.
The cited 2026 gUFO preprint is only a research comparator for dependent relations, reification, and occurrence identity; its category hierarchy is not imported.
The TypeDB Core Concepts documentation is only implementation evidence that relation instances can participate in relations; database semantics do not establish FPF truth or identity.
Reopen only the affected row if later seminar material changes the separation being tested, a later gUFO edition changes the relevant relation or occurrence treatment, TypeDB changes the cited relation-participation semantics, or a current FPF subject pattern changes so that the comparison no longer tests this Solution. A carrier, hyperlink, or layout change with unchanged source content does not reopen the decision.
Relations
A.6.RELgoverns explicit individuation of both PFR participants and PFR itself when a receiving use needs identity.A.6.RCDgoverns each missing direct-condition or applicability relation decision; C.22.PFR keeps the affected case non-conforming until an existing direct predicate closes it or a separately admitted direct subject pattern supplies obtaining, recurrence, and occurrence identity.A.6.5governs the two PFR participant SlotSpecs and the four applicability SlotSpecs.A.19governs the characteristic space used byCharacteristicSpacePredicate; the selected direct consumer governs its condition-to-input rule and comparator semantics,A.19.CPMgoverns comparison when that is the consumer, andG.4governs typed acceptance clauses when acceptance is the consumer.C.16,A.18, and direct condition or measurement patterns define or constrain characteristics, scales, actual characteristic assignments or state relations, and measurements. F.9 governs any semantic Bridge named by the input rule; none of these adds a PFR participant.C.22governs selector-facing task typing and TaskSignature assignment after a problem-side episteme is usable.C.22.2governs ProblemCard claims, signals, forecasts, scenarios, anticipated-condition cues, descriptions, next use, and publication without creating PFR; the direct claim pattern defines or constrains each claim carried there.A.15.1andA.3.4govern repair work and changes to the actual-condition relation.E.18.1,E.23, and direct NQD and OEE patterns define or constrain repeated problematization, method search, work, evaluation, and continuation; relations locating or ordering those occurrences in a transformation-flow structure do not enter PFR identity.C.27.TAgoverns temporal aspect statements when interval publication or temporal adequacy is current.A.2.4governs an episteme's evidence-use or status-use relation;A.10supplies a replayable evidence, provenance or bounded-reliance account when needed.B.3governs a named assurance claim, andG.11source or claim currentness.
C.22.PFR:End
ProblemCard
Type: Calculus (C) Status: Stable Normativity: Normative
Plain name. Problem-side card.
Intent. Give a practitioner one compact C.2.1 episteme that turns a messy signal into reviewable problem-side claim content before Principles-to-Work or selector use. The card preserves practical formulation functions while every world-side relation, evidence use, decision, method, plan, and Work occurrence remains with its direct governor.
Use this when. Use this pattern when work starts from a signal, anomaly, drift, risk, hypothesis, stakeholder demand or concern, set-derived candidate, underused capability, new constraint, new environment, opportunity-like cue, or solution-shaped request, and a downstream task-typing, method-family, planning, evidence, gate, autonomy, or P2W use needs one reviewable problem-side episteme.
Do not use this when. Use another pattern directly when the current question is already actual PFR obtaining, task typing, method selection, Work planning, evidence, provenance, assurance, gate decision, autonomy, archive or selected-set governance, mathematical-lens use, or ordinary discussion with no receiving use. C.22.PFR governs the actual Problem; C.22 governs TaskSignature constitution and assignment.
Boundary summary. C.22.2 yields one reviewable ProblemCard, a P2W-ready problem-side input for downstream C.22, or a stop naming the direct pattern for the claim, relation, or boundary now current. It yields no world-side Problem, task assignment, selected method, WorkPlan, Work occurrence, gate result, or evidence verdict.
C.2.1 constitution. One ProblemCard is one U.Episteme constituted by one ClaimGraph, one independently identified joint EntityOfConcern, and one effective reference scheme. U.ClaimScope, assumptions, horizons, qualification windows, viewpoints, and receiving uses qualify only the claims and direct relations that need them. No setting, carrier, organization, card id, or publication is a fourth constitution component or identifies an actual Problem.
A composite or component U.Work under A.15.6 may be referenced by a claim or receiving use; it neither constitutes the card nor identifies the Problem.
Claim-family, actuality, and solvability boundary. A card may carry several claims only while their ClaimGraph has that one joint EntityOfConcern. An actual-problem assertion states affirmative or negative polarity for the exact ProblematicForRelation obtaining predicate. An affirmative assertion may designate an independently established occurrence only when both participants and its adverse episode are established under C.22.PFR; negative polarity creates, ends, erases, or reidentifies no occurrence.
Judge support under the assertion's direct subject rule. A.2.4 governs a question about an episteme's evidence-use or status-use relation, A.10 supplies a replayable evidence, provenance or bounded-reliance account when needed, and B.3 applies only when a named assurance claim is made. G.11 separately governs assertion-edition currentness.
Forecast, scenario, counterfactual, and anticipated-condition claims retain their assumptions, horizon, and direct governor and do not assert a current PFR.
A method-availability or solvability claim concerns the admitted method set, evidence, constraints, and intended use; selecting a method changes that claim, not an independently obtaining PFR. Only non-adversity or cessation of the actual-condition or criterion-applicability participant can end its maximal continuous adverse episode; missing, stale, refuted, or absent evidence cannot.
Problem Frame
A working team can begin with symptoms, anomalies, stakeholder signals, constraints, risks, old solution evidence, comparison ideas, solution temptations, underused capabilities, new environments, opportunity-like cues, or members of a retained set. Opportunity-like signals still need an exact EntityOfConcern, effective ReferenceScheme, ClaimScope, claim family, not-wish reason, improvement or acceptance probe when the next use needs one, and next use.
Problematization becomes useful here when it produces claim content a practitioner can inspect. The card distinguishes:
- the observed signal from the claim family being considered;
- the one joint EntityOfConcern from nearby affected entities;
- the effective scheme and scope from generic setting words;
- actual-PFR assertions from forecasts and solvability claims; and
- problem-side readiness from method, Work, gate, or evidence conclusions.
The first-minute question is:
Can I state one concern, claim family, scheme, scope, improvement or acceptance probe when the next use needs one, and next use without turning a signal, card field, setting label, evidence result, or proposed method into the actual Problem?
Thin First Use and Output Kind
Thin First-Use Form
The first substantive use is the Thin form. It is a practitioner-facing prompt for the smallest reviewable card, not a demand to complete a schema.
Develop the content from accessible material. If the concern is still a rough signal, use the constructive examination in B.5:4.1.2 at the scale needed here:
- Follow an actual episode, inspect the artifact or recover the source statement that prompted attention. Separate what it shows from a proposed cause or solution. Trace what result is used, by whom or by what next operation, and what happens there.
- Recover why that consequence matters. Read the applicable purpose, acceptance condition, commitment or constraint, or ask the participant responsible for that use. Identify the entity affected in that use and trace how the observed condition bears on its criterion. The person reporting the signal and the subject of the card can each differ from the affected entity. If the criterion or connection is unknown, preserve that specific inquiry rather than asserting an actual Problem.
- Form the claim at the extent the material supports. Identify the participants and relation, the cases or conditions covered, and the meaning of the terms used. Distinguish a reported current condition from a forecast, an explanation to test or a claim about available means. Inspect a contrasting case when it can distinguish materially different formulations; retain an adequate existing formulation directly.
- When the next use needs a check, derive it from the claim the recipient must decide. A clarification can be tested by asking which recorded event satisfies the applicable rule. An improvement in service needs a comparison of service outcomes across the relevant work under comparable conditions. Identify the observation or argument that can settle the claim. Use an adequate existing result or method; obtain the particular missing observation, interpretation or subject assistance when it prevents that check. A source inquiry can still be the sufficient next use before such a check is available.
- Carry the supported formulation and its consequential uncertainty into the Thin form below. Choose the next operation because its available inputs and attainable result answer the remaining question. If no useful operation can yet be described, return to the episode or request help on that missing connection.
A ProblemCard is complete for its current use when it states:
- what signal or cue made the practitioner stop;
- the one exact joint
EntityOfConcern, effective reference scheme, andU.ClaimScopefor the claims being carried; - which claim family is current: actual-PFR assertion, anticipated-condition claim, method-availability or solvability claim, or another directly governed problem-side claim;
- what observation, need or constraint makes the concern worth addressing, independently of a proposed solution;
- what would count as improvement or as an acceptance probe when the next use needs one, or what remains missing; and
- one next use.
The next use may be P2W-ready, characterize, compare, search, refresh, retire, archive, abstainOrNoChange, or apply the direct pattern governing the named claim, relation, or boundary. If an improvement check or acceptance probe is absent, the card may preserve the signal and choose one of the other uses but cannot claim P2W-ready.
C.22.2 - ProblemCard is the pattern heading. A ProblemCard instance is the C.2.1 episteme described above. ProblemCardRef is only a reference value designating one such episteme; it is not another durable kind. Plain-register glosses and local labels do not replace the Tech name.
The local not-wish reason label means the positive grounds for addressing the concern; a stakeholder need can supply such grounds, while a proposed solution alone cannot.
Local labels include problem-formulation follow-up reason, validation boundary, risk condition, solvability band, P2W-ready, reviewable, stale, refreshed, retired, archived, abstainOrNoChange, and firstPrinciplesCue. They describe claim content or a local disposition; they create no FPF kind, state-machine kind, gate result, relation occurrence, or mathematical-lens object.
Use E.10.MOVE only when move-like wording no longer denotes one of these local dispositions and instead hides a pattern-use recommendation, Work-entry readiness, performed Work, gate, transformation, source relation, architecture move, call-planning move, or another directly governed value.
Reference labels ending in Ref are reference positions, not kind names. This includes ProblemCardRef, sourceSetRef, rivalProblemFormulationRef, and representationOrWordingUseRelationRef.
Semantic locality comes from the exact EntityOfConcern, effective ReferenceScheme, ClaimScope, declared assumptions or window, and receiving use carried by the relevant claims. A broad domain, organization, practice, or location label is only recognition material until those exact values are recoverable. None becomes a container participant or establishes global Problem identity.
Plain gloss for P2W-ready: problem-side input ready. It means ready as input to downstream P2W or selector reasoning, not ready for Work execution, gate passage, method selection, evidence reliance, or autonomy.
When a downstream reader asks whether intended Work can enter a work boundary, the card may supply problem-side cues, acceptance probes, constraints, and freshness conditions, but A.15.5 governs the readiness relation. C.22.2 decides no full-kit preparation, commitment, launch gate, or performed Work.
Required Solution Use
The Solution turns observed signal material into one C.2.1 episteme and one governed next use, not a completed form.
- Capture the symptom, anomaly, risk, stakeholder cue, drift, hypothesis, or other observed signal before naming an actual Problem. If the concern is not yet clear, put available source entries side by side and compare the events or meanings they record. Use what this comparison reveals to develop the question, as in the support example.
- Trace the signal to the work affected and what its recipients need to do. If their difficulty arises where two contributions meet, identify the hand-off, whole or other relation that gives the claims a common subject. An account of each contribution can leave their interaction unresolved. State the joint EntityOfConcern, effective ReferenceScheme, ClaimScope and claim family. Where the claims have independently identified subjects and no common one for this work, make separate cards.
- Separate the signal detector, actual-PFR assertion if independently grounded, anticipated-condition claim, improvement check, candidate acceptance criterion, method-availability claim, monitored risk signal, and proxy-distortion risk. These are not one card status or one PFR participant set.
- Pay only for current complexity. Add conditional content only when it changes the current next use; otherwise stop at the lighter card or name the direct pattern for the claim now current.
- If a reformulation changes the represented concern, meaning, scope or intended use, establish the actual relation on which reuse of an earlier disposition depends. Use the applicable representation-transition, retargeting or structural-reinterpretation rule; F.9 is needed only when a relied-on semantic correspondence meets its Bridge conditions. Ordinary wording cleanup can preserve the disposition without a new relation.
- Close by the next use. A filled card without a truthful next use is not a successful result.
Cheap stop. The smallest card that gives a truthful next use is sufficient. A conforming use does not require heavier fields merely because the full inventory exists.
Open the relation-boundary aid only when another claim or relation changes the next use.
Relation Boundary Aid
Once the Thin card is legible, use this aid for an unresolved claim or relation that changes its next use:
Which claim being made, relation, or boundary changes the problem-card use, and which FPF pattern defines or constrains that claim, relation, or boundary?
If the claim, relation, or boundary does not change the current problem-card use, leave it out of the card. If it does change the move, keep only the local cue or reference that makes the card reviewable, then apply the subject pattern for that claim, relation, or boundary.
Over-capture symptom: the practitioner spends the pattern use classifying FPF patterns while one or more core Thin items in C.22.2:2.1 remain unstable.
Repair: return to the Thin form in C.22.2:2.1 and complete only its core items. Reopen this aid only for the claim, relation, or boundary that changes the next use.
Use Boundaries and Record Budgets
Use C.22.2 when a signal is not yet a problem-side record and downstream task typing, P2W, method-family selection, work planning, evidence use, gate passage, autonomy control, or selected-set use depends on such a record. A known method does not close this pattern when the problem signal, scope, acceptance probe, or EntityOfConcern remains unstable.
Use another pattern directly when the question under repair is already that pattern's EntityOfConcern or governed relation:
A.15for work planning or performed work;A.2.4for an episteme's evidence-use or status-use relation,A.10for a needed replayable evidence/provenance/reliance account,G.6for a needed evidence-graph representation, orB.3for a named assurance claim;A.21for gate decision;E.16for autonomy;C.11for a local choice among explicit options;C.18,C.19, orG.5for archive, pool, front, or selected-set governance.
C.22.2 may still preserve the problem-side cue or reference that explains why that pattern is now current.
Use record budgets:
- Thin record budget: signal, joint EntityOfConcern, effective ReferenceScheme, ClaimScope, claim family, not-wish, not-slogan, not-ticket, or not-preselected-work reason, provisional improvement check or acceptance probe when needed, and one next use.
- Standard record budget: Thin fields plus the current comparison, acceptance, risk, validation, freshness, unknown-handling, or P2W-readiness fields needed for downstream use.
- High-relation record budget: Standard fields plus only the relation references needed when public, disputed, high-risk, set-derived, cross-scheme, cross-use-boundary, evidence-adjacent, autonomy-adjacent, gate-adjacent, agentic, temporal, causal, representation, or Part-G relations are current.
Stop at Thin when it gives a truthful next use. Stop at Standard when it supplies enough problem-side input for a minimal TaskSignature, TaskKind, or ProblemProfile under C.22. Apply the governing FPF pattern for the claim being made, relation, or boundary instead of enlarging the card when the issue under repair is no longer the problem-side record itself.
Field Labels and Current-Use Conditions
Use these readable labels only when current for the case:
-
problem signal and exact signal reference;
-
one ClaimGraph, joint EntityOfConcern, effective ReferenceScheme, and ClaimScope;
-
actual-PFR assertion with polarity and PFR reference only when C.22.PFR independently establishes the occurrence;
-
anticipated-condition, forecast, scenario, or counterfactual claim with assumptions and horizon;
-
method-availability or solvability claim with admitted method set, evidence qualification, constraints, and intended use;
-
primary viewpoint or an explicitly recovered system-role-kind, assignment, relation-participation, responsibility, or other direct concern as claim qualification, never as card or PFR identity;
-
symptom detector, problem hypothesis or cause-theory cue, and rival formulation when current;
-
improvement check, acceptance probe, characterization relation, characteristic or Q-bundle relation, indicator selection, and parity or comparison relation when current;
-
mandatory constraints, risk condition, validation boundary, freshness or expiry condition, and unknown handling;
-
problem-formulation follow-up reason when it changes review, discrimination, or receiving use;
-
representation-transition, retargeting, Bridge, structural-reinterpretation, or wording-use reference when an earlier disposition may no longer transfer;
-
sourceSetRef, source-set kind, selection or retention criterion, budget or window, and declared next use when the signal comes from a set, pool, front, archive, shortlist, or selected set; -
firstPrinciplesCuewhen a mathematical structure changes the formulation; and -
subject-pattern cue naming the direct pattern, claim kind, exact receiving-use reference, and stop condition when an outside claim changes the card use.
If a conditional value is not current, omit it rather than writing unknown. Use unknown only where the current value rule permits that result.
If a required current value is unavailable, state whether the next use is blocked, degraded, sandboxed, or must be reconsidered under a named condition. A stale value needs the applicable G.11 currentness qualification.
An irrelevant field can be omitted silently; explain an omission only when the receiving reader could otherwise rely on a missing value or mistake it for a completed check.
When a comparison of options, retained candidates or rival formulations is part of the claim, state its basis and applicable comparison or parity relation. Merely preserving alternatives does not assert comparability or require a parity result. Any claimed parity or selected-set result remains under its direct governor.
problemCardWitnessRefs may be used as a local recoverability group, not as a new kind or evidence graph:
Evaluate generated problem descriptions and evaluator feedback by their content and the support required for the receiving use. They may supply problem-side claims or supporting grounds; a generated description may itself serve as the card when it satisfies the C.2.1 constitution and current card use. Apply the relevant validation, currentness and action-entry conditions to the claims and proposed action.
Anti-Pattern Checks and Worked Slices
Anti-pattern checks start from the local card use:
- card-as-executable-work request: the card is treated as executable work while method, plan, and work occurrence remain undecided;
- form-completion: every field is filled because the template exists, even though the Thin next use would be truthful;
- readiness shortcut:
P2W-readyis declared from signal and scope alone, without improvement check or acceptance probe; - source-claim shortcut: a preselected solution, work request, proof-looking reference, gate-looking cue, or authority-looking cue replaces one or more core Thin items in
C.22.2:2.1; - scalar shortcut: archive, set-return, Goldilocks, NQD, OEE, partial-order, stepping-stone, or indicator material collapses into one readiness score;
- prestige shortcut: first-principles or mathematical wording is kept without practical payoff, preserved and lost structure when current, problem-formulation follow-up reason, and stop condition.
Local stop rule: if the encountered material tries to carry a claim outside C.22.2, the card keeps only the cue or reference that changes problem formulation or the next use, then names the governing FPF pattern and claim kind named by value to use before that claim is relied on.
The worked cases in §20 show both a Thin card and set-derived recognition. Use the applicable case to recover a needed distinction; an ordinary card need not reproduce both examples or acquire their conditional fields.
Conformance Checklist Requirements
The checklist protects a completed or reviewed card from overread; the writing order remains the Thin form and next use.
Problem-Kind Recovery
Problem remains an ordinary word when no FPF-governed claim is being made. Recover it only when wording changes a governed kind, relation, selector use, evidence claim, causal-use claim, assurance claim, decision, or use boundary.
The card may reference candidate ProblemProfile, TaskSignature, source set, PFR, forecast, solvability claim, or first-principles cue only when that reference changes its use. No reference is promoted into a local kind or card constitution component.
Problem, Task, Method, Work, and Result Split
ProblemCard remains usable while a method is unknown, contested, or not yet selected. A known method does not make the card ready while any core Thin item in C.22.2:2.1 remains unstable. If the problem-side episteme and method are accepted and only planned execution remains, apply A.15.
Use C.22 to prepare the TaskSignature from the exact problem-side episteme or ProblemProfile needed by one receiving use. Report the adequate signature actually adopted for the stated problem-side episteme and use under C.22:5.2. Another adequate description can remain available. If adequacy or adoption is unresolved, identify the candidates and the missing grounds. When the question becomes selection, planning, performed Work, result, gate, evidence, or reliance, apply its governor instead of expanding the card.
Relation to C.22
C.22 governs TaskSignature constitution and TaskSignatureAssignmentRelation. ProblemCard comes earlier: it states which problem-side claims are ready for P2W, search, comparison, characterization, refresh, retirement, or another direct pattern.
A card may prepare TaskKind, characteristic bindings, scope, and candidate TaskSignatures. C.22 assigns one only when the exact problem-side episteme, TaskSignature, and receiving-use episteme satisfy its obtaining rule. A plausible signature remains a candidate until its adequacy and actual adoption for that use are established.
No card detail is copied wholesale into TaskSignature. Downstream method, Work, result, evidence, gate, autonomy, archive, portfolio, and selected-set claims remain with their direct patterns.
Characterization, Indicators, and Comparability
State the characterization and comparison basis when the formulation or next use depends on it. A card that only preserves a signal or proposes an inquiry can finish without a measurement or parity account.
These FPF patterns supply the relevant subject content:
C.16carries measurement characterization, backing, and comparability discipline;A.19carries characteristic, scale, unit, polarity, and indicator-use discipline;C.25governs engineering-quality bundles;C.16.Qhelps recover other quality-like descriptions and the Characteristics their use needs;G.9carries parity, comparison-window, comparator, budget, unit, repeatability, and reproducibility pins;G.0carries comparison-frame and CG-Spec governance;G.4carries acceptance clauses and threshold predicates;- use
G.5for selector-facing selected-set result declaration when the problem enters a selected set; when actual audience availability is separately current, useE.17for a source-backed publication face and return to source andE.24.PUBfor the publication occurrence, form, carrier, audience, bounded use, and availability.
Missing characterization or parity relation is a current disposition. When that relation is current, apply the characterization, parity, search, or pool pattern before treating the problem as ready for P2W.
The C.22.2 candidate acceptance criterion separates functional check, constraint compliance, risk or safety boundary, parity or comparison relation, and freshness window when those relations are current. Apply G.0 for the comparison frame, CG-Spec, or comparability governance. Acceptance clauses and acceptance threshold predicates apply G.4; C.22.2 may name only the needed relation, cue, or reference. Passing a test, improving one observed indicator value, or naming an acceptance phrase is not by itself sufficient for P2W use.
Source Record-Form Recovery
Source record names are recovered by use, not by label shape. This section prevents source prose from becoming local C.22.2 subobjects.
If the source supplies problem-side material, make the needed content available in the card, directly or through an accessible return to the exact source content. The receiving reader must be able to interpret what that return supplies; when the card travels without access to the source, include the content and interpretation needed for its next use. If the source supplies another FPF-governed claim, keep only the local cue and apply the pattern that defines or constrains that claim.
Portfolio, Archive, and Set-Return Treatment
Archive, portfolio, pool, front, shortlist, selected-set, and set-return material remain source and set cues for the current problem-side record. ProblemCard preserves sourceSetRef, source-set kind, selection or retention criterion, and the declared next use when current; portfolio and archive governance stays with the named subject patterns and does not become a local problem-card kind.
A singleton problem card is the degenerate case. If it came from a portfolio, front, archive, or pool, the selected problem remains traceable through sourceSetRef: the lightweight reference to the source-set kind, source reference, selection or retention criterion, budget or window, review cadence, and direct pattern when current. sourceSetRef is a reference field, not a new kind and not a downstream claim carrier.
sourceSetRef preserves the recoverable set-source relation when current: Palette, Front, Archive, ExplorationArchive, Shortlist, RankedShortlist, SelectedSet, LivePool, or another accepted source-set form. If the set-source relation is not recoverable, the card may keep a set-finding cue, but it does not claim selected-set readiness or archive-derived readiness.
When multiple plausible problem formulations remain current, C.22.2 does not bind one TaskSignature prematurely. Each optional rivalProblemFormulationRef states the rival formulation, EntityOfConcern, effective ReferenceScheme, ClaimScope, preserved concern, lost concern, reason not selected yet, and next discrimination action. It is not a CG-Frame, not the E.8 Problem Frame, and not a representation-frame kind.
The next discrimination action may be to characterize, compare, retarget, reopen the source material or source relation, choose a local problem formulation, or apply the relation-bearing pattern. Reframing is triggered when EntityOfConcern, effective ReferenceScheme, ClaimScope, viewpoint qualification, or cause-theory cue changes the problem representation enough that readiness or use-boundary cannot be inherited by wording continuity.
Goldilocks and Set-Return Docking
Use a Goldilocks interpretation only when a declared search or learning policy seeks a problem of suitable difficulty for current capability. Explain that fit through the applicable characteristics, constraints and comparison or retention rule. It supplies neither a universal readiness score nor a new problem kind.
A Goldilocks, stepping-stone, or archive-derived problem is represented by its source-set reference, selection or retention criterion, and current next use, not by one difficulty, priority, or readiness score.
ProblemCard carries only the problem-side readiness fields and relation references:
- source set kind when archive, pool, front, shortlist, selected set, or portfolio language is current;
- solvability band;
- characteristic-space, declared problem-side characteristic descriptor, or Q-bundle relation;
- declared difference criterion when novelty or diversity is claimed, or apply the governing characterization or comparison pattern;
- non-scalar trade-off, dominance, partial-order, or set-return relation when that relation is being made;
- measurability or explicit unknown handling;
- reversibility or containment for safe probing when current;
- stepping-stone option value when retention matters;
- expiry or refresh condition;
- selected-set, archive, pool, front, or parity relation when that relation is being made.
The local solvability band label means a scheme-and-scope-qualified interpretation of feasible-but-not-trivial fit under current capability, constraints, validation boundary, and optional set-return relation. The declared search or learning policy supplies its difficulty and feasibility criteria; a scalar criterion or ranking remains usable with its measurement and policy basis. No universal difficulty or readiness scale is introduced.
If the band cannot be tied to a characteristic, Q-bundle, comparison, retention, or capability relation and its qualification, treat Goldilocks wording as informal recognition only and bind any selection, set-return, or parity claim to its direct governor.
The governing family is C.18, C.19, G.5, G.9, G.11, A.6.P:7a and C.16.Q, selected for the relation needed. C.22.2 can state this qualified problem-side fit before P2W. C.22:15 addresses difficulty targets only when a declared search or learning policy uses them; a TaskSignature includes the resulting trait only when it can change the receiving eligibility, acceptance or selection question.
Structure Cue That Improves Formulation
Use an adequate domain equation, model or calculation directly when it answers the formulation question. Enter C.29 when choosing or transferring a representation could supply a missing inference, or when reliance on a representation exceeds its supported correspondence.
firstPrinciplesCue is a local label for a structure that may change the formulation. It is neither a mathematical-lens kind nor a requirement to produce a C.29 record.
The problem card may ask whether a first-principles or mathematical structure helps find or improve the problem formulation, not only whether an already-mentioned mathematical expression helps the problem formulation. Useful cues include state space, graph, boundary, topology, symmetry, invariant, variational or constrained-optimization structure, probability or information structure, resource bound, obstruction, scale window, composition, or coarse-graining choice.
The practitioner-facing use is:
Identify the structure, obtain its relevant consequence, and say how that consequence changes the formulation. When the inference relies on a representation or transfer, explain the correspondence, assumptions and omitted distinctions that bound its use.
Distribution by principles:
When no useful mathematical construction is needed, proceed with the ordinary formulation. Keep an unresolved candidate cue only when it can guide the next observation or construction.
Validation, Reliance, AI-Agent Cues, and Safe Probing
ProblemCard exposes three local fields for downstream use:
- problem-formulation follow-up reason: why this formulation is worth keeping, reviewing, discriminating, or moving onward now;
- validation boundary: what has been checked for the current next use, what may be used now, and which use is governed by another pattern;
- risk condition: the monitored risk, cost-of-error concern, or containment concern that may change the safe next action.
Use these fields to state a local reliance disposition, not to authorize downstream action.
Cause-theory cues may focus problem formulation inside ProblemCard. Association, intervention, counterfactual, responsibility, expected-effect, or causal-evidence claims are governed by C.28 plus evidence, provenance, or assurance patterns when those claims are being made.
Environment design and safe probing may appear as problem signal reference, validation boundary, risk condition, or subject-pattern cue. For a probe affecting a controlled EntityOfConcern, identify the applicable safety or authorization condition. Existing adequate grounds may permit the action; the card itself supplies no missing permission, release authorization or gate result.
Freshness, Expiry, and Unknown Handling
C.22.2 includes a section-local state and disposition vocabulary for ProblemCard; this vocabulary is not a new FPF kind. These labels describe the card's current governed use; they are not required states in a transition sequence, event kinds, or gate records. The local labels are:
Freshness names the affected locus:
- Problem signal. Ask whether it is still present, recurring, solved, absorbed, duplicate, unnecessary, or no longer worth downstream work.
- Effective ReferenceScheme or ClaimScope. Ask whether the applicable meaning, cut, assumptions, window, or receiving use changed enough to alter the formulation.
- Characterization or parity relation. Ask whether measurement, comparison, and parity relations are current enough for the intended use.
- Problem-formulation reason, source material, or source relation. Ask whether cited sources, provenance, reason references, and source references remain current.
- Source-set reference. Ask whether archive, front, pool, shortlist, or selected-set membership and its selection or retention criterion remain current.
- Representation or wording-use relation. Ask whether wording, diagram, functional description, transformation-flow path, Bridge, retargeting, or other representation change alters the EntityOfConcern, effective ReferenceScheme, ClaimScope, viewpoint qualification, comparison relation, governed next use, or relation needed for inheritance.
A stale source material, source relation, or evidence reference does not always retire the problem; it may require refresh while the problem remains reviewable. A stale problem signal may lead to refresh, retire, archive, abstain or no-change, or a subject-pattern cue for the claim, relation, or boundary that is checked.
Freshness or expiry failure is a current disposition. A stale or unknown-bearing problem card may remain reviewable as a problem-side record, but it does not become P2W-ready unless freshness and unknown handling permit the intended downstream use. A stale problem card does not silently remain usable as P2W input.
When freshness, expiry, or unknown handling fails, choose one of these current dispositions:
- refresh the problem card or its characterization or comparison relation under
G.11,C.16,A.19,C.25, orG.9; - retire or deprecate the problem-side record under the relevant archive, pool, selected-set, or refresh pattern;
- continue only as explicitly governed bounded-risk use under the subject pattern for the claim being made, relation, or boundary.
Unknown-handling fields state whether they permit use, require degraded use, abstention, or sandbox treatment, or make the current problem formulation blocked. No P2W, no change, or abstain-for-now may be a successful next use when the signal is stale, duplicate, already solved, already absorbed, unnecessary, or not currently worth downstream work. Before using the ProblemCard as input to TaskSignature preparation or assignment under C.22, check whether the problem signal is still present and whether prior work has already solved or removed the problem.
Representation and Wording-Use Relation Continuity
C.22.2 names A.6.3.RT, A.6.4, E.17, F.9, E.18, and E.10 only when changed problem formulations, diagrams, functional descriptions, transformation-flow paths, wording, or PathSlice examples carry a current representation, bridge, retargeting, structural-reinterpretation, or wording-use claim. The card may preserve the local cue, reference, or problem-formulation follow-up reason, but it does not prove continuity or use-boundary inheritance by wording similarity.
Framing is not wording repair. A framing change applies when EntityOfConcern, effective ReferenceScheme, ClaimScope, viewpoint qualification, comparison relation, use-boundary inheritance, or next use changes.
Wording-use repair is current only when wording, diagram, functional description, transformation-flow path, Bridge, retargeting, or another representation change alters the carried problem-side representation, EntityOfConcern, effective ReferenceScheme, ClaimScope, viewpoint qualification, comparison relation, governed next use, or subject-pattern cue.
Ordinary wording cleanup triggers no representation-continuity relation and does not block a Thin ProblemCard.
C.22.2 may not treat changed-problem examples as governed relations unless the appropriate accepted FPF relation is named.
Source and P2W Carry-Forward
Apply the following source details as constraints on C.22.2 use and its P2W-facing input.
C.22.2 remains the problem-side output. P2W carries its problem-side cues and, when current, receiving-use and Work references into method-family selection, work planning, performed-work interpretation, and result measurement.
SoTA Decision-Use Source Material
The following external sources are adopted or adapted only where they change this pattern's answer.
For a claim extending beyond these uses, obtain its support under the pattern that defines or tests that claim; the dispositions above supply no additional warrant.
The following questions connect the source contributions to decisions in the card. Use the row whose distinction is current:
Rationale
ProblemCard gives the practitioner one compact problem-side record between vague problem talk and downstream P2W. The card is useful because it is light enough for ordinary use and specific enough to show when comparison, characterization, evidence, selection, mathematical-lens use, method, work, gate, autonomy, bridge, representation transition, or refresh requires another FPF pattern.
The card gives the practitioner one thing to write, inspect, and challenge. A practitioner can see whether a problem is ready without first assembling the problem-side record from TaskSignature, Q-bundle, parity report, evidence note, selected-set output, and refresh record. Claims beyond the problem-side record stay with their subject patterns.
The archive and portfolio distinctions remain current when they matter because the card preserves sourceSetRef and names the subject pattern for any current set, archive, or portfolio claim. Changed problem formulations, diagrams, functional descriptions, or transformation-flow path interpretations require the accepted representation or retargeting relations before a local cue or readiness disposition is reused. Current SoTA and first-principles cues matter only when they change fields, relation references, boundaries, or the problem formulation itself.
Problem-Card Use Invariants
Misuse Modes and Repairs
Use-Quality Checks
These checks protect the card's practical use.
Worked Slices and Anti-Cases
Five-Case Worked Slices
These rows are recognition slices, not complete Thin cards. Before using one, complete the Thin contract in C.22.2:2.1; add the relation cues shown here only when they change that card's next use. The compact filled Thin example follows in 20.1a.
Compact P2W-ready Disposition Slice
Consider a support team seeing repeated reopened hand-offs after a new interface policy. Its source episteme SupportHandoffPolicy-E7 contains the two clauses compared below. A colleague proposes rewriting the escalation workflow. The support team has been asked to establish what the policy requires to happen to a pending case. Clarification of that rule can complete this inquiry. If an authorized requester has also commissioned a revised workflow, the same clarification guides its development and the revised workflow is still due.
The first-line entry uses "complete" when a case is sent. The specialist entry uses it when the case is accepted. Both events can be recorded correctly. The shared queue must decide when to remove an item from the sender's list. For a case that has been sent but is still awaiting acceptance, the two readings give different decisions. The team can now ask which event authorizes that change. If the queue already distinguishes dispatch from acceptance and applies a single unambiguous rule, the two labels alone give no reason to revise it. The reopened cases still merit inquiry.
A claim that the service improved would need different evidence. To compare reopening rates, identify all eligible hand-offs in comparable periods, or use a sample that supports an estimate for them. For each period, determine which hand-offs reopened within the same follow-up time and compare their proportions, accounting for changes in the case mix that affect the comparison. Check the unresolved safety, compliance and customer-impact exceptions. Looking only at reopened cases can reveal how a hand-off went wrong. It leaves the reopening rate unknown.
The P2W export is narrow: signal, joint EntityOfConcern, effective ReferenceScheme, ClaimScope, claim family, not-wish or not-preselected-Work reason, improvement check or acceptance probe, and next use. In this case it also carries the qualification window, validation boundary, follow-up reason, and subject-pattern cues because the stated next use relies on them.
Add freshness, unknown handling, source-set, representation, evidence, or other conditional content only when it is current. If the improvement check or acceptance probe is missing, the card stays reviewable-only or source-finding and cannot claim P2W-ready.
If the next user wants evidence sufficiency, a gate decision, Work authorization, or selected method, the card preserves the cue and its direct governor carries that downstream claim.
Card/PFR Cardinality Replay
Every PFR reference below designates a world-side occurrence established under C.22.PFR; no card-side fact supplies its participants, adverse extent, or identity. PFR-InspectionAssignment-17 and later PFR-InspectionAssignment-18 are the two Robot-7 occurrences described in C.22.PFR:5.
For the unrelated branch, InspectionReleaseAssignment is a declared U.SystemRoleAssignment species. Occurrence InspectionAssignment-27 has admitted System Robot-8 as holder and local kind InspectorSystemRole as assigned-kind value. It obtains without interruption on [2026-07-13T10:00, 2026-07-13T10:30]; MaintenanceRoles-2026, Maintenance-Scheme-A, and the interval description interpret or describe the assertion but are not extra assignment participants.
Robot8ReleaseCriterionApplicability-4 uses predicate NoInspectorSystemRoleBeforeValidation-v2, maps the occurrence's assigned-kind participant to the adverse nominal coordinate and its holder to Robot-8, and names Robot-8, its release ClaimScope, and [2026-07-13T09:30, 2026-07-13T12:00] as the other applicability participants. Those facts establish PFR-InspectionAssignment-27 on [2026-07-13T10:00, 2026-07-13T10:30] under C.22.PFR.
Anti-Cases
Machine-Assisted Drafting Boundary
Machine-assisted ProblemCard drafting is only a drafting aid. Before the draft is used for P2W or selector-facing work, a practitioner checks the card's local fields and any subject-pattern cues for claims outside C.22.2.
Check the Thin content in §2.1 against the available source material. A machine-generated field value is a proposal until its claim is supported; the same record budget applies as for a human draft.
Check problem-formulation follow-up reason, validation boundary, freshness or expiry, unknown handling, source-set reference, representation or retargeting relation, and outside-pattern cues only when that content is current for the chosen next use. Omit a non-current field; do not invent a value or write unknown merely because the drafting aid exposes that field.
First Practical Entry Aid
This section is a discoverability aid only. It helps a practitioner or assistant find a candidate pattern; it does not prescribe a transition sequence and does not require opening C.22.2 for every problem-sounding text.
These likely practitioner entry phrases point to C.22.2:
- "We have a problem, but it is not yet clear what work should be done."
- "This looks like a ticket, but I am not sure the problem is stated."
- "A signal or anomaly keeps recurring before method selection."
- "We selected this candidate from a front, archive, pool, or selected set, but need to state why it is a problem now."
- "P2W would otherwise use 'implement X' as if it were reviewable problem-side output."
- "There is a symptom, but we do not yet know what to solve."
- "We need to know whether this problem is ready for P2W or should apply another pattern."
Direct-entry cues that are not C.22.2:
- accepted method or work planning: use
A.15; - proof, provenance, reliability or assurance claim: use its direct subject rule; add
A.10for a needed replayable account,G.6for a needed graph representation, orB.3for a named assurance claim; - local choice among explicit options: use
C.11; useG.5when selector-facing set-result declaration is current; when that result already exists and actual audience availability is current, useE.17for its source-backed publication face and return to source andE.24.PUBfor the publication occurrence and availability; - agent tool-call, gate, or autonomy claim: use
C.24,E.16, orA.21;ProblemCardmay only name the problem-side cue or relation named by value; - ordinary discussion with no downstream receiving use: no
C.22.2use.
First-use Thin-card test:
Aim for a Thin card that fits on one page when only the core content is needed. Use 2.1 to recover: signal; one joint EntityOfConcern, effective ReferenceScheme, and ClaimScope; claim family; not-wish or not-preselected-Work reason; improvement check or acceptance probe when needed; and one governed next use.
That use may be P2W-ready, characterize, compare or parity, search or pool, refresh, retire, abstainOrNoChange, or apply the FPF pattern that defines or constrains the claim, relation, or boundary outside the card.
Downstream Cue Export
ProblemCard exports problem-side claim content, not authority over downstream use.
The compact export contains:
- problem signal and exact signal reference;
- one joint EntityOfConcern, effective ReferenceScheme, and ClaimScope;
- claim family and polarity: actual-PFR assertion, anticipated-condition claim, method-availability or solvability claim, or another named direct claim;
- the observation, need or constraint that warrants addressing the concern independently of a proposed solution;
- improvement check or acceptance probe when needed, or the gap that prevents P2W-ready use;
- one next use and its disposition: reviewable-only,
P2W-ready,abstainOrNoChange, refresh, retire, archive, or subject-pattern cue; - current qualification window, exact PFR, source-set, A.15.6 composite or component Work, or representation reference only when independently governed and relied on; and
- problem-formulation follow-up reason, validation boundary, freshness condition, and stop only when the receiving use relies on them.
Choose the next pattern by the receiving question:
- For P2W carry-through, use E.18.1 with the accepted problem-side distinctions.
- For TaskSignature constitution and assignment, use C.22.
- For selector-facing selected-set use, apply G.5 when that relation is current.
- When the missing contribution is development of reusable material through feasible variations, use C.40; use C.18 to preserve the generation or archive account when needed.
- For intended or performed Work, use A.15 only after its object is current; when the question is whether intended Work may enter its boundary, A.15.5 governs
WorkEntryReadiness@Context. - For evidence, gate, autonomy, or any other claim, apply the direct pattern; the whole card never carries that claim by itself.
Consequences
Benefits
- FPF gains a clear problem-side output for problematization as the input P2W can use.
- P2W uses a typed problem-side record rather than a slogan, ticket-shaped wish, or preselected method.
- A source-grounded formulation can expose that a requested rewrite, selected candidate or indicator change does not yet answer the concern. The next use can then seek the missing distinction, retain alternatives or stop before unnecessary work.
- Current archive, pool, front, shortlist, set-return, parity, refresh, evidence, and
C.29patterns are reused instead of duplicated. - The positive use of mathematical and first-principles thinking is preserved: it can find missing structure, not only check already-written mathematics.
- Characterization and parity are required when problem reviewability depends on them.
- Representation-change relations are handled through named relation references rather than local proof inside the problem card.
Costs of Use
- A
ProblemCardadds a small writing step before P2W. The cost is justified only when the signal is not yet reviewable before downstream use. - The practitioner keeps the card small: preserve the split between problem, task, method, work, and result; keep
TaskSignatureminimal; and add conditional fields only when their relation is current. - For problems emitted from archives, pools, fronts, selected sets, or portfolios, the practitioner preserves
sourceSetRefor the set-source relation without turning the card into a portfolio, archive, selected-set, or work-planning carrier. - Keep source-local or ordinary domain wording when it supplies the needed meaning. Recover its governed object or relation only when that distinction changes the card or its next use.
Relations
- Builds on:
C.2.1,E.2,E.9,E.10,C.2.P,A.6.P,C.16.Q,C.16,A.19,C.22,C.25,C.29,G.5,G.9,A.6.3.RT, andA.6.4. - Coordinates with:
C.11,C.18,C.19,C.22.1,C.22.PFR,C.24,C.27,C.28,A.15,A.15.5,A.21,E.16,G.6,G.11,A.10,B.3,E.17,E.17.ID.CR,A.6.3,F.9,E.18,E.18.1,C.32.P2S, andE.10.MOVE. UseC.22.PFRwhen actual Problem obtaining or identity is current. UseC.32.P2Sonly when a problem-side record exposes architecture pressure that must continue into selected structures, candidate synthesis, decision, realization, actual-structure feedback, or refresh.
C.22.2:End
MethodFamily Evidence & Maturity (Method‑SoS‑LOG)
Status: Stable
LOG (logic) for deductive shells for admissibility
SoS‑LOG means Science‑of‑Science LOG.
Registration boundary. A MethodFamily is registered by one exact G.5 registry row and registry edition. That record names the family, its admitted A.3.1 Method members and grouping basis, and intended selector use. Establish evidence, claim scope, validity, and the decision result under their respective rules.
Problem frame
Use this when a G.5 selector must admit, restrict or refuse a registered MethodFamily for a declared TaskSignature and use. Apply executable rules to typed descriptors and the applicable evidence, retaining the branch and its grounds. Family maturity is evidence about that bounded admission use; an admission threshold belongs to a G.4 AcceptanceClause.
An ordinary application of a sufficient domain Method needs no MethodFamily admission record. Use C.11 for a local choice when no selector-facing family admission is required.
Problem
Unstructured “readiness” stories and undisciplined evidence lead to:
- (i) Illicit scalarisation across mixed scale types,
- (ii) ambiguous gating conditions that a dispatcher cannot apply,
- (iii) reuse after the family, evidence profile, claim scope, qualification window, or comparison basis changed, or reliance on an unstated source-local, kind, or plane relation, and
- (iv) Immature families entering production.
We need a notation‑independent LOG layer whose executable rules use TaskSignature (S2) + EvidenceProfiles to return admit / degrade / abstain, retaining relation losses and any justified assurance consequences under R4.
Forces
- Pluralism vs. dispatchability. Competing Traditions expose different invariants; selection must compare without semantic flattening.
- Maturity vs. opportunity. Open‑ended exploration (E/E‑LOG) must coexist with run‑safe exploitation; immature ≠ forbidden → provide safe degrade paths.
- Unknowns (tri‑state). Missing or
unknownvalues in live S2 fields must propagate to Degrade(mode) (including sandbox) or Abstain; no silent coercions. - Meaning and implementation. Keep the family, its description, rule and evaluation result distinct. A particular notation or software implementation must preserve the same admission conditions.
Solution — Method‑SoS‑LOG: deductive shells over Eligibility & Evidence
Objects & heads (LEX/I‑D‑S)
MethodFamily(registered in G.5) carries Eligibility and artefact identity.MaturityCard(this pattern) carries evidence‑aware maturity.SoS‑LOG.Rule(this pattern) is an executable rule schema.
Evaluating a rule returns one of {Admit | Degrade(mode) | Abstain} for a (TaskSignature, MethodFamily) pair.
A qualifying description episteme uses …Description; …Spec names that same episteme only after the E.10.D2 specification-use gate grants the named use.
Rule schema (normative)
Apply the existing rule set for the registered family when it covers the task, intended use and current conditions. When you must write or revise a rule, start from the work the family is being admitted to do. Identify the circumstances that would make that use acceptable, restricted or unavailable. Derive the task facts, evidence and governing acceptance clauses needed to distinguish those results. A permitted bound may settle a threshold while the exact value is unknown. If the applicable rule requires the exact value or a particular item of evidence, retain that requirement. Express the resulting conditions as an executable rule set for MethodFamily f.
with the following branch obligations:
What the family claim covers. State which tasks the claim covers and what it promises about the registered members. Evidence that one member works on one task can support that use. A claim that every task has a suitable member allows the member to vary with the task. Promising one fixed member throughout requires evidence for that member throughout. Permitting arbitrary substitution requires the promised performance from each permitted member on each task where it may be used. When the work requires several methods together, identify what each supplies to the joint result and the conditions under which their contributions fit. B.1.5.CM supplies the construction of a composite Method. A registered member that already provides this whole can be evaluated directly for that use. Membership or separate admission of its constituents leaves their compatibility to be established.
Identify the member, task conditions and performance criterion behind each result. Cover the combinations the claim needs through applicable trials, a warranted account of why the result carries over, or both. For a fixed set of tasks, report the observed proportion or the bound the outcomes establish. Eight confirmed successes among ten eligible tasks establish at least 80% for that set even while two outcomes are unknown. To estimate the rate for a future population, identify the population and explain why the available results carry over. For a sample estimate, give the selection method and its uncertainty. Even if every task in a sample succeeded, a guarantee for every future task needs grounds covering those future tasks. If successful members differ by task, an agent can be assigned to choose among them when the information and a supported way to choose will be available in time. It can still be known that a suitable member exists before it can be identified.
R0 — CG-Spec gate (precondition). For the exact G.5 registry row and MethodFamily, verify the cited CG-Spec.MinimalEvidence and EvidenceProfile for every CHR characteristic used by the family's acceptance clauses and flows, under the declared claim scope and selected slices, qualification window, and intended selector use. Failure ⇒ Abstain with reasons. Publish the consulted CG-Spec, EvidenceProfile, registry, and policy editions.
Rationale: selector legality requires an explicit CG‑Spec gate. Publish associated ReferencePlane notes alongside the consulted ids.
R0.QD — QD pre-gates (when QD is used). Apply these gates when the TaskSignature enables a QD search or QD illumination use, including an open-ended generator that uses QD. Resolve that use and its fields under C.22:5.1. Retaining evidence under PortfolioMode=Archive or emitting a set of alternatives does not activate this branch. For the active QD use, verify:
- (i) CharacteristicSpaceRef resolves to the CHR-typed coordinates and dimension required by the declared descriptor and comparison rules, with ReferencePlane wherever interpretation or the receiving contract requires it;
- (ii) ArchiveConfig is lawful (topology, resolution, K>0,
InsertionPolicyRef,DistanceDefwith edition id and declared metric/pseudometric status); - (iii) EmitterPolicyRef present (with edition id);
- (iv) resolve DominanceRegime; if absent, use default= ParetoOnly.
Failure of any ⇒ Abstain with reasons.
R1 — Admit. Admit IFF
- (a) S2 satisfies Eligibility predicates of f (tri‑state aware),
- (b) the exact EvidenceProfile minima referenced by Acceptance/Flows for f are met for the declared claim scope and selected slices, qualification window, and intended selector use (post R0),
- (c) all relevant CAL.AcceptanceClauses (G.4) evaluate to true under lawful CHR comparisons,
- (d) any maturity gating (e.g., a floor on Maturity rungs) is expressed as an AcceptanceClause and referenced here by id (no acceptance thresholds inside LOG).
LOG never sets acceptance thresholds; its rules use and cite Acceptance verdicts.
R2 — Degrade. Apply R0 and, when applicable, R0.QD before any degrade branch, including U2, U3, and R7. Failure returns Abstain for the attempted use.
After those gates pass, return Degrade(mode) only through a declared family branch when an admitted unknown or an unmet acceptance condition permits a narrower scope or execution mode under the applicable CAL failure behavior; mode ∈ {scope-narrow | sandbox | probe-only}. An eligibility violation still requires R3 abstention.
Record the exact S2 unknowns or unmet conditions, narrowed claim scope or execution mode, qualification window, governing policy edition, and result. If the branch changes the intended use, re-evaluate R0 and eligibility for that bounded use before relying on it. LOG-Degrade never changes CHR scales or planes or turns a failed CAL verdict into a pass.
Note (CAL vs LOG). CAL‑level degrade.order (fall‑back to order‑only comparisons) is governed by G.4/CG‑Spec and is not a LOG mode. SoS‑LOG never overrides CAL outcomes; a LOG branch only narrows Scope(G) or execution mode (e.g., sandbox, probe‑only), it does not alter CHR scales or admissible orders.
probe-only MUST cite the applicable probe-budget and Acceptance-bound guards. Cite an E/E-LOG policy when that policy governs exploration over a still-live pool.
R3 — Abstain. If S2 violates Eligibility or R0 fails, return Abstain with the failed rule, reasons, and available policy, evidence-profile, scope, and qualification-window basis. Report missing references and unperformed judgements under §4.2.1. Abstain is mandatory for illegal CHR operations and when a conclusion depends on an F.9 Bridge, kind relation, or plane relation that has not been established.
R4 — Relation and loss routing. Cite an F.9 Bridge, kind relation, or plane relation only when the admission decision relies on that obtaining relation. Record its participants, direction, what meaning is preserved and what is lost, receiving use, and applicable policy edition.
When the admission use makes a separate named assurance claim, identify its exact target claim and receiving use under B.3.
Assess loss under the receiving-use rule. A numerical assurance penalty requires an applicable domain model under B.3:4.3–4.4 and affects that model's declared assurance result. Expression properties, claim scope and admission conditions retain their own governing rules.
A changed registry row, evidence profile, claim scope, qualification window, or intended use is not by itself a crossing.
R5 — Proof hooks. Every branch MUST retain its A.10 evidence/source basis for the conclusion it makes. For an early Abstain, cite the available basis for the failed prerequisite and name the required evidence that could not be recovered.
Evidence used by a branch retains the lane tags (TA/VA/LA) and freshness windows required by its CG-Spec.MinimalEvidence and EvidenceProfile; missing requirements remain explicit gaps under §4.2.1.
Cite Bridge ids + loss notes when the branch relies on a Bridge; the decision is SCR‑visible. When G.6 EvidenceGraph is present, also publish EvidenceGraph path id(s) for the branch (admit/degrade/abstain). A branch verdict is not its own evidence basis.
R6 — Retention, source and selector outcome. Resolve PortfolioMode under G.5 CC-G5.23; it governs retention of exploration evidence. After Admit, a QD archive may supply candidates to G.5, but the archive remains a source. The emitted outcome follows G.5:4.4b–4.4e, with its result family when applicable, exact members, ordering and basis. Retaining an archive supplies neither a new result family nor a QD requirement.
When illumination telemetry is used, it remains report-only unless CAL authorises DominanceRegime=ParetoPlusIllumination and the policy id is recorded in SCR; coverage and regret likewise remain reported telemetry unless their governing policy promotes them.
R7 — GeneratorFamily branches (open‑ended). If S2 includes GeneratorIntent, SoS‑LOG MUST:
- (i) verify
EnvironmentValidityRegionis declared and lawful; - (ii) verify
TransferRulesRefexists; ifunknown⇒Degrade(scope‑narrow)orAbstainper family policy; - (iii) treat the selection surface as pairs
{environment, method}; publish coverage and regret under their declared telemetry policies. When the active branch or receiving use requires IlluminationSummary, obtain that retained-set report under C.17:5.1 with itsDiversity_Pmeasurement policy, Scale, descriptor or feature basis, and distance or covering definition. Generation alone requires no illumination report. These results remain report-only unless their governing policy promotes them under R6.
R8 — Telemetry and refresh hooks. When an active QD or open-ended generator branch records an illumination increase or changes its search archive, publish the current editions and actual edition increments of the definitions on which the changed use depends.
For QD, these include its characteristic space, descriptor map, distance, insertion and emitter policies under R0.QD and C.22.
For open-ended generation, retain TransferRulesRef.edition and the applicable validity and telemetry policy ids; add QD definitions only when that branch uses QD.
A separate illumination report retains the measurement basis required by C.17:5.1.
Ordinary retention of selection evidence follows its own retention and source references; it does not require QD definitions. Expose PathSliceId for refresh or decay in SCR only when an E.18 path slice is current.
Admit on the complete R1 basis. Degrade only through a declared narrower-use branch after its prerequisites pass. Abstain when a prerequisite fails or the policy supplies no permitted branch.
Report the branch actually reached
An early Abstain completes the admission result for the attempted use. Keep the failed rule, reasons, known family and registry edition, TaskSignature, intended use, claim scope, qualification window, and available source and policy references. Name any required reference that could not be resolved and the available basis for that finding. The report can stop there without completing later judgements.
Retain each premise or result that was established or validly reused for this branch, with its source edition and the scope, window, evidence profile, and use that make it applicable. Reuse does not require recomputing the result. An earlier result whose applicability is unresolved remains unavailable as a premise for this use.
Distinguish these situations for each affected entry:
These descriptions qualify report entries; they do not extend the S2 value sets, the closed maturity rungs, or the Acceptance verdict domain. An applicable earlier judgement used by the branch is reported as reused, not as unevaluated merely because it was not recomputed. A known result not used by this branch may be cited separately with that limited purpose.
First use. A registered family resolves, but the evidence profile required by R0 cannot be recovered. Report Abstain, the missing profile and available source basis, maturity not evaluated when no applicable judgement is available, and Acceptance not reached after R0. Do not insert L0 or false to fill those result positions. If R0 passes but Eligibility evaluates to false, report that predicate result and R3 Abstain; later Acceptance may remain not reached after R3. Keep a previously established maturity result if it is applicable and used.
Admit still requires the complete R1 evidence, Eligibility, and Acceptance basis. A declared Degrade(mode) retains the premises that selected its branch and any consumed maturity results. An unmet Acceptance condition stays unmet when its failure behavior permits a narrower use. R2 still requires R0 and eligibility to be checked for that changed use before reliance; a failed R0 or an Eligibility violation for the original attempted use remains Abstain.
Maturity ladder (poset, not a scalar; Description, not Spec)
When a maturity judgement is established for an admission use, publish or cite its editioned MaturityCardDescription for the exact evaluated MethodFamily, G.5 registry edition, evidence profile, claim scope and selected slices, qualification window, and intended admission use (UTS enum ids; scale kind = ordinal; reference plane declared).
Here L0–L4 name maturity rungs for the evidence about a registered family. For example, maturity L2 (Replicated) identifies independent replication under the stated conditions. A model abstraction level L2 identifies what a disciplinary model specifies (B.5.MR:4.3). A replication can test a claim at any model abstraction level.
Cite an existing card when its judgement remains applicable. If no applicable judgement is available at an early stop, report that state under §4.2.1 without creating a card or assigning a rung.
Do not embed acceptance thresholds here; an admission floor remains a G.4 AcceptanceClause cited by R1.
- L0 — Anecdotal. Claims exist; lanes sparse; examples ad‑hoc.
- L1 — Worked‑Examples. Multiple worked examples with lane tags and Scope slices declared; no replication yet.
- L2 — Replicated. Independent replications identify their distinct bearers or operating conditions and declare the claim scope and selected slices, source and method editions, and qualification windows used; lane separation is observed and decay windows are explicit.
- L3 — Benchmark‑Severe. Repeated wins or parity on community baselines or severe tests. Declare any F.9 Bridge on which this comparison relies, with its loss notes.
Optional rung (for QD/OEE‑heavy families; ordinal, closed enum):
- L4 — QD‑Hardened. For an active QD use, establish archive stability under the declared InsertionPolicy/DistanceDef editions and reproducible IlluminationSummary improvements under controlled budgets. For an OEE generator, establish severe tests of its EnvironmentValidityRegion. An OEE use without QD does not acquire QD archive or illumination requirements; retain the exact branch and evidence behind the rung.
Norms.
M1. The ladder is lane‑aware (TA/VA/LA) and freshness‑aware; it is not a global numeric score. Declare Scale kind=ordinal and the closed enumeration of rungs; register the enum at UTS (twin labels; editioned).
M2. Transitions MUST be justified by EvidenceGraph paths (once G.6 is available) and published at UTS; missing anchors ⇒ no advance.
M3. A maturity floor used for admission belongs in a G.4 AcceptanceClause. Cite its policy edition, scope, qualification window and verdict. For a floor of “at least L2”, apply the order declared for those rungs under G.8:4.5. A larger numeral alone cannot establish precedence. If the policy requires independent replication, check the evidence for replication or an applicable rule that entails it. Repeated wins by one group on severe benchmarks can support a benchmark claim while leaving independent replication unestablished.
M4. Declare the MaturityCard reference plane. If an admission decision relies on a relation to another plane, cite that exact obtaining plane relation, its direction and loss, and the applicable policy edition; any loss assessment or assurance penalty follows R4.
Rationale note. Treating maturity as a poset aligns with B.3's requirement for lawful comparisons and avoids scalarisation across ordinal/ratio scales; assurance calculations follow R4; maturity and admission retain their separate rules.
Unknowns & Shift classes (tri‑state discipline)
U1. (LEX). Enumerations for Degrade(mode) and Maturity rungs MUST be declared as closed value sets and registered at UTS (twin labels). Lexical SD (E.10) applies.
U2. A live S2 characteristic or predicate admits unknown only when its C.22 value rule permits it; unknown MUST map to a branch (Degrade or Abstain) declared on the family (no coercions). Each branch publishes a branch‑id and (where used) a mode from a closed enum registered at UTS.
U3. ShiftClass semantics follow C.22. For a known ShiftClass ∈ {covariate‑shift, concept‑drift, adversarial}, the default is Degrade(scope‑narrow) unless a CAL.AcceptanceClause explicitly guards the regime. A live unknown follows U2 and its declared Degrade or Abstain branch. All these branches remain subject to R0, applicable R0.QD and R2/R3.
Publication & wiring
W1. Register the SoS-LOG rule ids. Publish or cite a MaturityCardDescription for an established maturity judgement under §4.3; use §4.2.1 when an early stop leaves that judgement unavailable. RSCR tests cover Admit, Degrade, Abstain, and unknown paths, including early reports with missing bases or unperformed judgements. Relation and loss-policy ids appear only where a branch relies on them.
W2. Admissibility Ledger. Publish an editioned AdmissibilityLedger. Each selector-facing row identifies the exact MethodFamilyId, G.5 registry edition, TaskSignature, RuleId and rule edition, intended admission use, claim scope, qualification window, decisive branch, and decision result.
Record its MaturityRung, EvidenceProfile, AcceptanceClause and policy references, verdicts, evidence paths, DominanceRegime, and PortfolioMode according to §4.2.1: retain established or applicable reused values and explain any required but unresolved reference or unperformed judgement. An explanation of a missing or unperformed result accompanies its unfilled value position; it is not a substitute verdict or rung.
Include obtaining relation and loss-policy ids only when used, and G.6 path ids under R5's condition. UTS registers the row vocabulary; the ledger records the admission result and its basis.
W3. Strategy composition. For a selection composition called a strategy, cite its governing G.5 rule. Cite an E/E-LOG policy when the composition uses that policy over a still-live pool.
W4. The dispatcher (G.5) consumes the rules and reports each decision with its conditions and grounds. To assign a job, identify an applicable member or an authorized way for the performer to choose one in time. Check the required result, means, authority and execution conditions under A.15. A probe-only decision supports the specified probe within its budget and guards.
Archetypal Grounding
The following are constructed branch examples. For each later branch, stipulate that the named family and policy editions resolve and R0 passes; the QD case also passes R0.QD. Every Admit below additionally requires the complete R1 Eligibility, evidence and Acceptance basis. If a stipulated prerequisite is missing, use the early-Abstain result in §4.2.1 instead.
Continuous dynamics (ODE task).
S2 excerpt. DataShape=ODE; stiff?=unknown; Hamiltonian=unknown; Size≈10^3; Objective={↓error@ratio, ↑throughput@ratio}; Jacobian_sparsity=high; Missingness=MAR. A declared safety predicate tests the required ordinal safety level.
Families. Implicit‑BDF vs Explicit‑RK vs Symplectic.
Rules.
Implicit‑BDF: withJacobian_sparsity=high, its declared unknown-handling branch permitsDegrade(probe-only)under exploration policyEE-ODE-1and Acceptance-bound guardsProbe-ODE-1. An applicable L3 maturity judgement remains cited. If later observations establishstiff?=trueandHamiltonian=false, and the complete R1 basis for the intended use holds, that new input can receiveAdmit; the original unknown is not silently converted to true.Explicit‑RK: requiresstiff?=falsefor ordinary use. Its declared unknown branch permitsDegrade(sandbox)after R0 and eligibility are checked for that bounded use.Symplectic: requiresHamiltonian=true; its declared branch for the liveunknownreturnsAbstain, without reporting a false Hamiltonian judgement.
Didactic anchor. This mirrors C.22’s typed‑signature discipline and CHR legality (no ordinal means; unit alignment for ratio).
Rackauckas and Nie (2017) describe DifferentialEquations.jl as a common problem-and-solve interface for different differential-equation methods. That interface illustrates how task description can remain stable while solver choice varies. The C.23 admission conditions above are FPF's conditions. (Journal of Open Research Software)
Planning/scheduling (MIP task).
S2 excerpt. DataShape=MIP; NoiseModel=deterministic; Objective={↓cost@ratio, ↑service_level@ordinal}; Size≈10^5 vars; convex_relaxation=available.
Families. MILP (branch‑and‑bound), Constraint‑Programming, Heuristic meta‑search.
Rules.
MILP: Eligibility requiresconvex_relaxation=available; the cited L3 MaturityCard edition names the registered family, evidence profile, benchmark basis, claim scope, qualification window, and intended selector use ⇒Admit.Constraint‑Programming: MaturityCard=L2; the ordinalservice_level≥BAcceptance predicate is met. Baseline parity is an additional CAL condition, not a missing R0 minimum in this example. Its declared failure behavior permitsDegrade(scope-narrow)while parity is unresolved, after R0 and eligibility pass for the narrowed use. If that parity basis were required by R0, the result would beAbstain.Heuristic meta‑search: MaturityCard=L1. The applicable AcceptanceClause and family failure policy decide between a permittedDegrade(sandbox)branch andAbstain; L1 alone supplies neither outcome.
Using these results. Hand the dispatcher the admission decision and permitted conditions for each family. A family admitted for a sandbox trial can enter that trial. Before selecting it for ordinary work, establish its eligibility for that use. A shortlist can retain alternatives eligible for the same intended use. To claim that a member is non-dominated, compare performance on cost and service under the stated order and common conditions (C.18:4.4). The admission results above provide no such comparison.
Dunning, Huchette and Lubin (2017) describe JuMP as a modeling language that separates model expression from optimization-solver choice. This supplies a concrete software example of preserving a problem description while considering different solvers. (SIAM Review)
QD archive (policy search).
S2 excerpt. PortfolioMode=Archive; CharacteristicSpaceRef(d=2); ArchiveConfig(CVT, res=1k cells, K=1, DistanceDefRef.edition=v2, InsertionPolicyRef=dyn‑elite); EmitterPolicyRef=v3; DominanceRegime=ParetoOnly.
Rules. After Admit, G.5 declares a Shortlist drawn from the QD archive, with exact members, ordering and basis under G.5:4.4b–4.4e. The archive remains its source. Illumination is reported separately; changes to DistanceDef or the emitter are editioned in SCR; dominance remains ParetoOnly.
Open‑ended GeneratorFamily (POET‑class).
S2 excerpt. GeneratorIntent{GeneratorFamilyRef=GF‑01, EnvironmentValidityRegion=EVR‑A, TransferRulesRef=TR‑A, CoverageMetric=…}; PortfolioMode=Archive.
Rules. After Admit, G.5 declares a Shortlist over eligible {environment, method} pairs with its source and selection basis; Degrade(scope-narrow) if TransferRules=unknown, as permitted by the family policy under R7. Telemetry publishes the declared coverage and regret with their edition and policy basis. This use requires no QD search or illumination report. If a later receiving use needs IlluminationSummary, obtain its C.17:5.1 measurement basis and retain the governing policy id on improvements.
Bias‑Annotation
Maturity is evidence-relative: a rung established for one family, scope and use does not establish suitability for another. A declared bounded exploration branch can remain useful below an exploitation floor, while keeping its own prerequisites and limits.
Conformance Checklist (normative)
Consequences
- Explainable admission. Every Admit/Degrade/Abstain is backed by anchored evidence and explicit unknown handling (selector reports are SCR‑linked).
- Run‑safe pluralism. Multiple families can co‑exist with policy‑governed exploration (E/E‑LOG) and maturity‑aware gating.
- Cross-Tradition reuse. The account identifies the obtaining relation, its limitations, and any justified assurance penalty under R4.
Rationale
For a named admission-assurance claim, B.3 requires each of F, G, and R that the use consumes to have a declared bearer, meaning, and scale. Claim scope and selected USM slices remain explicit; WLNK and Φ(CL) are used only under an applicable, calibrated aggregation or loss rule. Evidence-typed maturity rungs preserve the declared ordinal distinctions and admissible comparisons.
Separately, locally declared DesignRunTag values retain their temporal predicates and C.27 adequacy conditions; cite a GateCrossing only when the selected E.18 transformation-flow structure contains it.
Relations
Builds on: G.5 (selector consumes these rules), G.4 (Acceptance & EvidenceProfiles), C.22 (S2 typing), C.18 NQD‑CAL, C.19 E/E‑LOG, B.3 (named assurance claims and declared aggregation, including WLNK only when applicable), A.10 (evidence basis) and E.10 (naming). G.6 supplies EvidenceGraph paths when used; E.18 applies when a selected structural crossing is current.
Publishes to: UTS (MaturityCards, rule ids), SCR/RSCR (branch coverage; parity hooks).
Constrains: G.8 (LOG Bundling must cite MaturityCards), G.9 (parity harness draws baselines per rung), G.11 (refresh windows per rung & decay), G.5 (Open‑Ended Family mode for GeneratorFamily).
Outcome. Admissibility logic for MethodFamilies combines LOG shells, the maturity poset, degrade modes, and publication requirements with CG‑Spec legality rules, CHR guard‑macros, and CAL acceptance mechanics.
C.23:End
Plan Tool or Service Calls for a Fixed Action (C.Agent-Tools-CAL)
Type: Calculus (C) Status: Stable Normativity: Normative
Plain-name. Plan tool or service calls for a fixed action.
Intent. Help a practitioner turn an already fixed action or option into a budgeted call plan, then revise that plan or return a bounded checkpoint without confusing planning with selection or execution.
Depends on.
A.15and its planning and Work patterns for Methods, descriptions, plans, performed Work, and attribution;A.15.7for a situation-responsive next-action decision;C.11for a fixed choice among a currentOptionSet;C.2.1when the relied-on decision or checkpoint needs a persistent episteme;C.18for candidate-generation records;C.19for live-pool policy;C.19.1for a scale-based comparison or waiver;C.16for measured comparison inputs;OPS.11.1:4.2–4.3for constructing a route from required contributions and reconciling its resource conditions;G.6when a call trace must remain replayable;B.3only when a named assurance use needs one bounded assurance result.
Coordinates with. U.PromiseContent for service acceptance conditions, C.28 when a planned call is intended to support a causal use, and E.17/E.24.PUB when an already obtained result is being published or made available.
Use this when
Use A.15.7 first when ongoing Work still needs the next action to be chosen from current facts within a domain Method. Enter C.24 only after that action is fixed and tool or service calls must be planned. A call plan is neither the situation-responsive decision nor proof that the chosen action was performed.
Use C.24 when an applicable domain prescription or a completed decision has fixed the action or option and the practical question is now:
- which admitted Methods to call, in what order;
- which applicable time, compute, cost or risk budget to reserve;
- what stops or replans the route; and
- whether the useful output is a
CallPlanor aCheckpointReturn.
Do not use it to generate candidates, keep a live pool, choose among unresolved options, execute calls, or score completed Work.
What goes wrong if missed
- a route is scheduled by an opaque heuristic, so its budget use and stop conditions are hidden;
- unresolved choice or pool-policy work is hidden in a plan;
- a route description is mistaken for a Method, a plan for performed Work, or a successful probe for a decision to adopt and execute the route; and
- replanning loses the basis that fixed the action and its conditions.
What this buys
- one small, tool-neutral plan that cites the accepted action basis;
- visible budgets, stop conditions, and replan triggers before calls are made;
- recoverable call facts and provenance when later use needs an execution account; and
- one bounded checkpoint stating whether to probe again or seek a decision to adopt the proposed route.
Primary working object. One ATC.CallPlan : U.WorkPlan. Each step selects a U.Method. A route description may help locate or constrain that Method, but remains a separate U.MethodDescription. Actual calls are dated U.Work and remain outside this planning result.
First useful move. Recover the action and completion conditions fixed by the domain prescription, A.15.7 decision or C.11 choose-now result. Establish which required results are already available and what still has to be done with them. If a usable result can be delivered as ordered, plan its delivery. If the order requires a fresh performance, include it. Reuse an adequate known route for the remaining work.
Otherwise, work backwards from the required result to admitted Methods and their needed inputs, then follow the proposed route forwards from the available inputs as in section 4.1.
Return the supported ordered calls, applicable budget and stop or replan conditions, or the particular missing contribution or condition. Add route-description refs where they are needed to understand or qualify the route.
Not this pattern when. Use C.11 while fixed-option choice is unresolved, C.19 while treatment of a live pool is unresolved, G.5 when the current task is selector-facing result declaration, A.15.5 for work-entry readiness, and A.15.1 when the question is what Work actually occurred or which Method it enacted.
First-minute questions
- Which applicable domain prescription, A.15.7 decision or C.11 choose-now result fixes the action or option now being planned?
- What still has to reach the recipient? Work back to the available inputs and choose the admitted Methods whose results can supply the missing contributions.
- Which budget is current: a still-upstream probe budget or an enactment/call budget?
- What event stops or replans the route?
- Is the useful output a plan, a checkpoint, or a return to a neighbouring pattern?
First output
The first useful output is one of these:
nextPlannedAction and recommendedNextAction are local fields. Exactly one action-basis branch is present.
Add one of the branch-specific refs in C.24:4.4 only when that constraint still affects the plan. A plan with no current policy branch needs no policy placeholder.
If neither output can cite its accepted action basis and state what happens next, the C.24 work is unfinished.
If the domain prescription, its applicability or the action it requires changes, reopen that domain basis before revising the plan. If the A.15.7 decision changes, is withdrawn, or no longer fixes the action, reopen the plan and return to A.15.7. If the C.11 ChoiceResult changes or no longer says choose now, return to C.11. A changed live-pool branch returns separately to C.19. Do not revise the call plan as though its action basis were still settled.
Problem frame
Tool-using Systems may plan across web services, local programs, instruments, robots, or human-operated routes. The implementation may be an LLM agent, a search system, a conventional planner, or a fixed program. The planning problem is the same: turn a fixed action or option into an ordered and bounded route without hiding route grounding, budget, or stop logic.
A local system-role kind or assignment is recorded only when that separate fact matters.
When planning, revision, or a call is claimed as precise performed Work, recover each performer's identity through A.13 and let A.15.1 independently admit the dated Work from its performer, Method, interval, and containment facts.
Add the exact A.2.1 assignment reference and F.6 only when the plan or receiving use expressly consumes precise assignment-bound attribution through the same obtaining A.13 assignment; F.6 identifies neither assignment nor performer, and missing or failed F.6 attribution alone leaves the Work intact while its independent A.13/A.15.1 grounds still hold.
Problem
We need a tool-neutral way to produce or revise one call plan under explicit budgets and policy while keeping Method, route description, plan, performed Work, service promise, trace representation, the action basis that fixed the action, and any assurance result distinct.
Forces
Solution
Local objects and boundaries
ATC.CallRouteDescriptionis aU.MethodDescriptionwhose EntityOfConcern is the selected admitted Method and whose claims explain how that Method is carried out through the route, under A.3.2. When it carries vendor-local route data, it states the vendor or source scheme, exact scheme or API edition, intended use, and selected Method ref before any access details, inputs, outputs, or route limits.ATC.CallPlanis aU.WorkPlanfor intended calls. Its steps select Methods and may cite route descriptions.ATC.CheckpointReturnis a C.2.1 result episteme stating what was tested, what budget was burned, and what route action is recommended next. It is not the tested Work.ATC.CallGraphRefcites the applicableG.6trace representation over actual call Work. The representation records or points to facts; it creates none of them.
actionBasis contains exactly one of three branches:
domainPrescribedActionidentifies the applicable domain Method, the exact prescription and its required action, applicability and stop conditions. An optional intended WorkPlan may carry that intention; it establishes neither authority nor permission. No decision wrapper or ongoing Work is required for this branch.situationResponsiveDecisionEpistemeRefrefers to an episteme identified under C.2.1 because this plan relies on an A.15.7 decision; the episteme states the selected action, deciding System, intended performer, action-changing fact, relevant Method limit, and stop or feedback condition.fixedOptionChoiceResultRefrefers to a C.11ChoiceResultwhose result ischoose now.
Consuming either decision result here preserves its existing result kind.
There is no catch-all ATC.PolicyRef. When a constraint branch is current, cite its object: C.19 PoolPolicyResult or EmitterPolicy, a C.19.1 probe, comparison, local-policy, or waiver result, or a domain constraint whose kind and defining pattern are named. The plan retains applicable time, compute, cost and risk ceilings and its stop or replan conditions.
Owned planning operations
C.24 owns only planning and replanning:
A.3.1 supplies Method admission. The action basis fixes the action or option being planned; it does not admit the Methods chosen for plan steps.
A domain-prescribed basis carries its Method, prescription, required action, applicability and stop conditions. Permission and authority remain independently governed. An A.15.7 basis keeps the selected action, deciding System, intended performer, action-changing fact, relevant domain-Method limit, and stop or feedback condition. A C.11 basis is a ChoiceResult whose lawful result is choose now; probe again, reject current set, and reroute do not fix an action for C.24.
C.18 may supply generated candidate or front material, and C.19 may supply a live-pool treatment that informed the decision; neither record admits a Method. Comparison comes from the selected evaluation Method and, when scale preference is claimed, C.19.1.
A.15.1 governs the actual call or observation Work; G.6 provides a trace representation when those independently established facts, results and provenance relations must remain replayable. C.24 only constrains what the plan or checkpoint must retain for those later uses.
To construct the route, use the backward construction and forward check in OPS.11.1:4.2–4.3 for this fixed action:
- Recover usable contributions. Read the available Methods and any needed route descriptions for the results they supply, required inputs, preconditions, relevant effects and limits. Resolve missing interface meaning from the maintained description or provider. Begin with the action's receiving completion conditions.
- Work backwards from the required result. Choose an admitted Method that can supply it. For each input not already available, find a qualified preceding Method or an available source that supplies it. Continue until the route reaches available inputs. Keep required joins and enabling conditions. If a contribution cannot be supplied, return that particular gap rather than adding an ungrounded call.
- Check forwards from the available inputs. Follow the proposed outputs into the next inputs. Check consequential units, scope, access, timing and other use conditions. Reconcile shared resources and applicable budget. Elapsed time is not the sum of resource expenditure when calls overlap. Reuse established interface and performance bounds where adequate. Compare the final output with the result promised to the recipient. If a call only prepares the input for delivery or acceptance, include the work still needed and its time and entry conditions in the plan.
- Return the route or the unresolved question. State the calls, useful intermediate results, required ordering or safe parallelism, shared resources and stop or replan conditions. When a later call depends on an earlier result, specify what the person or system making that choice will know in time and what it can then do. For a guaranteed result, check the remaining budget and completion conditions on every reachable branch. If the deciding information arrives after an irreversible choice, establish a route that works across the cases still possible then, or report the missing condition. Use §4.2 for a bounded scout when its result could resolve that question at justified cost. A changed action returns to its selecting Method.
Bounded scout or probe cycle
When the action remains fixed but an unresolved route condition could change the continuation, first use adequate interface documentation, existing observations or provider information. Select a bounded scout only when its attainable answer can change the warranted continuation enough to justify its burden, within the applicable budget and permission. State what observation would retain, revise or reject the route before planning that scout. Its actual performance can support a CheckpointReturn; an unfamiliar route alone does not require probing.
- If another probe could still change which option survives the
OptionSet, the budget remains a C.11 probe budget and planning returns there. - If changed live facts or domain-Method limits could change an A.15.7 action, return there instead.
- If a prescribed action loses its applicable prescription, return to the domain Method before further route planning.
- If the action or option remains fixed and only route shape or rollout order is uncertain, the probe uses enactment budget and its checkpoint belongs here.
commitTrigger states the condition for considering a checkpoint decision to end route probing and adopt an identified CallPlan for the fixed action. Name the proposed plan and the work it schedules in recommendedNextAction or in the accompanying explanation. When the trigger holds, the responsible planner considers that transition using the probe evidence, residual budget, and any required safety or assurance result; the planner may adopt the plan, require a further bounded probe, or stop. Satisfying the trigger alone does not make that decision.
Adopting the plan settles which route to use. Permission to perform its calls remains subject to the applicable work-entry conditions, and actual performance is separate U.Work. Where an autonomy budget governs the route, use the checkpoint decision and guard or override path in E.16-S1.A; the trigger grants no wider budget or scope.
Planning laws
ATC-1 — Plan the call, not the app. A plan step selects a Method. A route description, endpoint, service promise, trace row, or response does not become that Method or an actual call.
ATC-2 — Use the C.19.1 branch that applies. A bounded call plan with no scale/generalization claim or declared generality policy can proceed under its task, budget and guards. When that claim or policy is current, start with C.19.1's scale-claim probe. Consume its first result: no scale claim yet, local analogy or policy, bounded scale comparison, or full Scale-Audit selected. Only the latter two open comparison or audit work.
A completed comparison of usable responses over the receiving budget range, with uncertainty and applicable floors, may then warrant a bounded preference or no scale-based preference. A larger slope alone is insufficient; a generality tie-break needs its separate policy basis.
Keep a BLP-waiver separate: it is used only when a declared generality preference would otherwise decide the use, and it records rationale, the admitted review System, the direct waiver-review responsibility or missing governor, and expiry or review.
If comparable evidence is absent, stop the empirical preference; do not invent a slope vector or treat a waiver as evidence.
ATC-3 — Make budgets and harm limits visible. A CallPlan states its applicable planned ceilings. A CheckpointReturn or Work-side record states actual burn. The admitted System stops or replans when a named ceiling or safety condition is breached.
ATC-4 — Keep live-pool exploration declared. Cite a C.19 PoolPolicyResult only while treatment of that still-live pool constrains this plan. Cite its exact EmitterPolicy only when the plan uses that profile; then record explore_share, including 0 when the current profile explicitly plans none. Do not fabricate either ref after the fixed action or option has made pool treatment irrelevant, and do not silently turn illumination or novelty telemetry into a decision criterion.
ATC-5 — Preserve the execution basis needed by later use. An actual call claim retains its dated Work, performer, enacted Method, interval and containing-System relation under A.15.1. When later use needs a replayable account, retain the relevant plan ref, actual budget delta, inputs and outputs subject to privacy, route-description edition used, and G.6 trace reference. The plan and trace do not establish the represented facts by themselves.
ATC-6 — Add assurance only for a named use. When a planning or rollout decision depends on assurance, name the target claim and use, then cite the B.3 result with its basis, disposition, limits, and reopen condition. No policy label or confidence level substitutes for that result.
ATC-7 — Bind vendor routes to the selected Method. Vendor-specific tokens belong in an edition-pinned ATC.CallRouteDescription that recovers the vendor or source scheme, exact scheme or API edition, intended use, and selected Method ref. Access details, inputs, outputs, and limits may follow. An arbitrary profile, executable adapter, or F.9 Bridge does not satisfy this binding. When executable adaptation is current, identify the Method, MethodDescription, System, and performed Work through their direct patterns. Cite an F.9 Bridge only when its relation independently obtains between two local meanings.
Policy and comparison branches
Add only the branch that still constrains this plan:
Each comparison tolerance names its characteristic, bearer, scale, evidence basis, and window. Graduation, rollout, or widening uses a condition defined by the cited result or direct domain pattern.
Plan-local ceilings and stop conditions need no policy object. If another domain constraint is current, give its field the result-kind name and cite its defining pattern; do not put it in a catch-all constraint ref.
When a condition relies on assurance, cite the exact B.3 result and its supported scope; no universal assurance level is inherited from C.19.
Causal action-use field
Add the causal field only for a named causal use in which the planned calls are intended to observe, intervene, collect counterfactual-rung evidence, simulate for a causal claim, condition a counterfactual policy, or evaluate a policy causally:
The field states the planned causal use and any support already consumed. It does not estimate an effect, prove identification, certify fairness, or turn simulation output into realized counterfactual evidence. Use C.28 for those support questions.
Public quick card
An ordinary written plan can supply this content; the schemas above are optional forms. Retain an existing applicable plan when it already states:
- exactly one action-basis branch: the grounded domain-prescribed action, an A.15.7 decision episteme, or a C.11
choose nowChoiceResult; add the objective and ordered Method refs; - route-description refs only when needed, with their source scheme, exact edition, intended use, and selected Method binding;
- dependencies, safe parallelism and any earlier result that governs which later call can be made, when these change the route;
- applicable time, compute, cost or risk budgets, with stop and replan conditions;
- next planned action; and
- an exact C.19 or C.19.1 result, B.3 assurance result, causal-use result, provenance ref, or named domain-constraint result only when that branch is current.
This is enough for an ordinary plan. Do not fill the heavier branches merely to make the record look complete.
Closure and worked cases
Close as a CallPlan when route order and budgeted enactment are the current question. Close as a CheckpointReturn after a bounded route probe, stating whether to probe again, seek the checkpoint decision to adopt the proposed plan, or stop. Return to the domain prescription, A.15.7 or C.11 when the corresponding action basis reopens; return to the applicable neighboring pattern when pool treatment, selector declaration, readiness, execution, or publication becomes the current question.
Domain-prescribed action before Work.
In a constructed catalogue-maintenance case, CatalogueMethod_3 requires rebuilding a digest when the catalogue’s source checksum changes. That condition is established; no choice of action remains and the rebuild has not begun. Plan the two admitted Methods directly:
The prescription fixes the action, while work-entry permission remains separate. If a permitted route probe is needed, its CheckpointReturn retains this same domainPrescribedAction branch and states the tested Methods, findings, spent/residual budget and next action. A changed prescription or input condition reopens the domain basis. Neither form invents an A.15.7 decision or completed Work.
A.15.7 decision into a known route.
During ongoing repository-repair Work, changed source facts make produce_patch_and_verify the next action under the current repair Method. Using the steering Method in A.15.7, the responsible maintainer makes that decision. The retained decision names the repair agent as intended performer, the changed-source fact, and test failure as the stop and feedback condition. Because the call plan relies on the decision later, the team retains it in one episteme identified under C.2.1. The episteme describes the situation-responsive decision; it is not a C.11 ChoiceResult.
An actual call needs its independent A.15.1 grounds; add assignment-bound attribution and a trace representation when the receiving use consumes them.
Constructing a route from a fixed result.
The fixed action is to deliver a weekly report and obtain the recipient's acceptance. The maintained interface specifies the timestamp unit, the recipient's week boundary and time zone, and the required report format. Four admitted Methods extract the rows, convert their timestamps, aggregate by week, and render a conforming report. For the specified input size and one sequential performer, they take at most 2, 3, 4 and 1 minutes respectively. The team has 12 minutes.
Work backwards from acceptance. The four calls prepare the report in at most 10 minutes. Their outputs establish that it can be ready to send. The recipient also provides an admitted submission Method. Under its stated conditions it delivers a conforming report and returns the recipient's acceptance within 2 minutes. The intended performer has access, the receiver is available for the full interval, and the prepared report meets its entry conditions. The five-call plan therefore fits the 12-minute budget. The time bound warrants adopting the plan under the stated conditions. To close the actual job by its deadline, establish when this recipient accepted this report. Acceptance after the deadline means the commissioned job finished late. Delivery itself may have been on time. An undated acceptance can establish receipt and acceptance while leaving timeliness open.
If the export already supplies compatible timestamps, the remaining calls fit in 9 minutes. If the submission service promises only to queue the file within 2 minutes, the team can promise timely submission on that basis. A deadline for recipient acceptance still depends on what happens after queuing.
If the timestamp unit or week boundary is missing, recover that interface or receiving condition from its authoritative source. The magnitude of sample numbers does not establish a unit. A changed report requirement returns to the action basis before replanning.
Unfamiliar route.
In this constructed repair case, ci_route_choice_09 fixes the action: produce a candidate patch and run the targeted tests. The proposed ci_patch_plan_09 orders EditCandidateMethod_2 and TargetedTestMethod_7, has an upper duration bound of 25 minutes, and stops for replanning if the reproduction changes or the tests fail twice. Adoption here requires a supported guarantee that the repair fits the enactment time left after route probing.
Before adopting that plan, the maintainer probes whether the failure can be reproduced reliably enough to guide the patch. For this case, MinimalReproductionMethod_5 requires the same failure in two consecutive runs on the fixed repository revision, with the trace and reproduction evidence collected on that revision. One such run has completed. The enactment budget covers both this route probing and the planned repair.
If the repeat establishes the trigger, the maintainer decides whether to end route probing and adopt ci_patch_plan_09: produce the patch, then run its targeted tests. The maintainer compares the time remaining after that repeat with the plan's 25-minute upper duration bound and checks the applicable safety or assurance conditions.
A 3-minute repeat leaves 27 minutes, which covers that bound. A 6-minute repeat leaves 24 minutes, so the current bound does not establish the required guarantee. A tighter justified bound for the same route, a qualified alternative route or an independently permitted budget revision may establish fit; otherwise stop.
If the plan is adopted, the repair agent may start its first call only with the required work-entry permission. The checkpoint itself records neither that permission nor a performed call.
Two vendor routes with one token.
Vendor A and Vendor B both publish a route called search. vendor_a_search_v2 states scheme VendorA API, edition 2026-07, intended use repository text search, and selected Method RepositoryTextSearchMethod_3. vendor_b_search_v5 states scheme VendorB agent tools, edition 2026-08, intended use web source retrieval, and selected Method WebSourceRetrievalMethod_8. The shared token identifies neither binding; the description fields do. An executable adapter, if used, remains distinct from the Method it implements, and its execution remains separate Work.
Scale comparison, when current.
The cheap C.19.1 probe for BatchSearchMethod_3 and IndexedSearchMethod_6 returns bounded scale comparison for the same repository-search task and 10k–100k files window. The comparison then uses elapsed time and missed-match rate from repo_search_benchmark_12, including uncertainty and cost limits, and warrants a preference for IndexedSearchMethod_6 only inside that window.
If the evidence for one Method covers only small text files and the evidence for the other covers only large mixed repositories, the task difference leaves their relative performance on the receiving workload open. Return no scale-based preference. A tie-break written for established equivalent performance does not apply to that finding. The project can use a local policy that covers the actual case or seek a waiver when overriding a preference that would otherwise decide the use.
Near misses. A route label with no recovered Method remains probe material. A plan with no Work is still intent. A trace row does not prove performer, assignment, Method, or service acceptance. A satisfied commitTrigger with no checkpoint decision leaves the proposed route unadopted.
Transfer examples. The same result shape works for research assistance, program repair, and lab automation. The Methods and safety conditions differ; the plan/checkpoint boundary does not.
Bias-Annotation
Keep notation and vendors out of the conceptual contract. Do not average unlike scales. Do not let a route description, plan, trace, response, or confidence label stand in for a Method, performed Work, evidence result, or assurance result.
Conformance Checklist
- Every result states exactly one grounded action basis: an applicable domain-prescribed action with its Method, prescription and conditions; an A.15.7 decision episteme; or a C.11 choose-now ChoiceResult. A WorkPlan alone supplies no permission or authority, and a prescribed action needs no decision wrapper or ongoing Work.
- Every planned step names an A.3.1-admitted Method that realizes or supports the fixed action. The action basis fixes the action or option; selecting it does not establish Method identity. Route-description refs remain separate and optional.
- The plan records its applicable time, compute, cost or risk ceilings and stop or replan conditions.
- C.18 candidates or front material and C.19 live-pool treatment may inform the action basis but do not admit Methods; C.24 owns only planning and replanning results.
- A scale branch first cites one C.19.1 probe result, then any selected comparison or Scale-Audit result; a
BLP-waiverremains separate from evidence. - Every vendor-bound
ATC.CallRouteDescriptionidentifies source scheme, exact edition, intended use, and selected Method; an arbitrary profile, adapter, or Bridge cannot substitute. - Each current policy or constraint ref resolves to its C.19, C.19.1, B.3, or domain-defined object; a plan with no such branch remains valid.
- A
CheckpointReturnstates tested Methods, evidence, burned and residual budget, next action, and commit trigger. - Actual call claims retain their A.15.1 Work grounds. A receiving use that needs replay also retains the relevant plan, budget delta, source editions and G.6 trace refs.
- A causal action-use branch uses the current C.28 question, support-component, and support-result contract and grants no downstream authority.
- The ordinary quick path remains readable without ontology or assurance apparatus.
Common Anti-Patterns and How to Avoid Them
- Planning the whole tool lifecycle. Keep candidate generation, selection, execution, scoring, and publication outside C.24.
- Route description as Method. Recover the Method or keep the route in probe state.
- Plan as execution. Put actual burn and call facts in Work-side results and the trace.
- BLP slogan as comparison. Use C.19.1's probe and any selected comparison; keep a waiver separate or return no scale claim or no scale-based preference.
- Catch-all policy or profile ref. Cite the PoolPolicyResult, EmitterPolicy, C.19.1 result, B.3 result, or domain-defined constraint, or omit the branch.
- Confidence threshold as assurance. Use a direct condition and cite B.3 only for a named assurance use.
- Executable adaptation by implication. Store the binding in a route description; identify any executable adaptation independently.
- Successful probe as commitment. Keep the trigger, decision to adopt the proposed plan, and permission to perform its calls distinct; satisfying the trigger supplies no missing decision or permission.
Consequences
Tool use becomes inspectable before execution: the result shows which applicable prescription or accepted decision fixed the action or option, which Methods are planned, what budget is reserved, and what changes the route. Identical vendor tokens remain distinguishable by source scheme, edition, intended use, and selected Method. The cost is explicit Method grounding and branch-specific constraint refs. Heavy assurance, causal, or scale-comparison records appear only when their use justifies them.
Rationale and current practice
Qualification window. This comparison uses the source set as of 2026-08-21. Reopen it when a later result changes the relative value of explicit planning, route grounding, active information gathering, checkpoint use and replanning, multidimensional evaluation, or long-horizon budget and dependency handling for the declared use.
These contributions distinguish route grounding, plan quality, execution and replanning. Keep only fields that change the route, stop, reliance or replay. Reconsider the plan when new knowledge changes one of those decisions; another benchmark alone need not change an adequate plan.
Relations
A.3.1supplies admitted Method identity; neither decision branch admits a Method merely by selecting an action.- The steering Method in
A.15.7is used to reach the situation-responsive decision cited throughsituationResponsiveDecisionEpistemeRef; when this plan relies on the decision, its episteme has the identity conditions defined in C.2.1. - A C.11
ChoiceResultwhose result ischoose nowis cited throughfixedOptionChoiceResultRef. C.18supplies generated candidate or front material, andC.19suppliesPoolPolicyResultorEmitterPolicyonly when live-pool treatment still constrains the plan. Neither admits a Method.C.19.1supplies the scale-claim probe, any selected comparison or Scale-Audit result, and any separate local policy orBLP-waiver.A.15,A.15.1,A.15.2,A.2.1, andF.6keep Method, description, plan, Work, performer, and attribution distinct.G.6supplies the trace representation cited byATC.CallGraphRef.B.3supplies one bounded assurance result only when a named assurance use is current.C.28supplies causal-use support when the plan is used for causal evidence, intervention, policy, fairness, or counterfactual work.C.27evaluates temporal claims about speed, narrowing, recovery, or stop/replan rate. More calls or faster narrowing is not success by itself.E.23may use C.24 plans and checkpoints inside improvement Work; C.24 does not restate the improvement loop.E.10.MOVE,E.11.PUR, andA.15.5recover project moves, pattern-use recommendations, and work-entry readiness when those questions are not plan-local.
C.24:End
Q-Bundle: Express a Quality Claim as One Characteristic or a Structured Bundle
Type: Definitional (D) Status: Stable Normativity: Normative unless marked informative
Plain-name. Quality-bundle normal form.
Builds on.
C.2.1for the enclosing quality-claim episteme;A.2.6for scope algebra;A.6.1for operation-declaration references when current;C.16/A.18for Characteristic and Scale legality.
Coordinates with.
C.17-C.19for quality-related measurement families;C.16.Pwhen characteristic/scale/score wording is not yet recoverable;A.15for method, work-plan, or work-occurrence gating;C.16.Qfor quality/evaluative-characterization wording before the endpoint is one explicit characteristic, Q-Bundle-shaped claim content, objective, or another governing pattern.
Use this pattern when an engineering-quality family such as availability, resilience, security or maintainability may be hiding several differently typed contributors and the reader needs one claim that keeps them distinct.
First useful move. Ask: what would make this quality claim false? If one measure on one declared Scale answers the question, state that one Characteristic and stop. Use Q-Bundle-shaped claim content only when several differently typed contributors—such as a measure and scope, or measures plus a load-bearing window or mechanism—jointly determine the answer.
First result. Write one readable quality claim about one exact bearer and include only the contributors on which its truth or the next receiving action depends. Ordinary prose can express that structure; no separate template is required. Omit an optional slot unless changing it could change the current claim or receiving action.
Nearest non-use. Stay with the direct Characteristic pattern when one measure and Scale carry the claim. Use the direct scope, measurement, evidence, assurance, gate, publication, viability-envelope, or temporal pattern when that neighboring question—not quality-family decomposition—is the current work.
Problem frame
Engineering quality language repeatedly drifts into one of two invalid simplifications: either every -ility is treated as one scalar characteristic, or every engineering-quality statement is left as loose evaluative prose. A conforming engineering corpus therefore needs a uniform discipline that keeps admissible measurements, scope declarations, mechanisms, statuses, and evidence visibly separated without inventing a new kernel ontology.
Problem
Without a normal form for engineering quality families:
- Composite families are scalarized illegally. Terms such as resilience, security, or maintainability are treated as if one number exhausted them.
- Scope is confused with measurement.
A claim's
ClaimScope/WorkScopeis spoken of as if it were a magnitude rather than a USM set-valued applicability object. - Mechanism and status are mistaken for evidence or metrics. Presence of redundancy, certification, or audit controls is described as if it were itself a measurement value.
- Guards become unstable. Admission checks silently mix scope coverage, numerical thresholds, mechanism presence, and evidence freshness in one phrase.
- The governing pattern for the quality claim remains unclear.
After
C.16.Qrepairs a bare quality term, orC.16.Prepairs characteristic, scale, score, metric, or proxy wording inside that term, the admissible endpoint is unclear unless FPF distinguishes single-CHR cases from bundle-shaped quality families.
Forces
Solution - Q-Bundle normal form
C.25 defines a lightweight normal form for the claim content of engineering quality families. A publisher facing a quality term first decides whether one claim episteme should state:
- one admissible CHR characteristic, or
- one structured quality bundle whose measurable slots, scope slots, mechanisms, statuses, and evidence remain explicit.
Endpoint split
Use the single-characteristic branch when one exact U.Characteristic, one declared Scale, and the ordinary CHR laws carry the quality claim. The claim-bearing result is still one C.2.1 episteme about its exact bearer; C.25 adds no bundle record.
Use the Q-Bundle branch when several differently typed contributors are part of one quality claim. The result is one C.2.1 episteme whose ClaimGraph contains the record-shaped Q-Bundle content below.
Q-Bundle shape and identity boundary
The full escalation form is:
Q-Bundle := <Name, QualityBearer, ClaimScope?, WorkScope?, Measures[CHR], QualificationWindow?, Mechanisms?, Status?, Evidence?>
Q-Bundle names a C.25-local record-shaped portion of one episteme's claim content. It is not a new Kernel kind, an independently identified world object, or a second identity beside the enclosing episteme. That episteme supplies the claim content, one independently identified QualityBearer as its EntityOfConcern, and the effective reference scheme under which the quality claim is read.
The ? marks an optional slot: omit that slot unless changing it could change the current claim or receiving action. A bundle may therefore contain only Name, QualityBearer, Measures, and one load-bearing scope or window. The full tuple is an escalation aid, not a form every author must fill.
Changing any bundle content that changes the quality claim changes the ClaimGraph and therefore identifies another episteme under C.2.1. A changed layout, form, publication occurrence, or carrier can leave that episteme unchanged.
A gate, publication, proxy, comparison, or roll-up cites the exact episteme or one exact C.2.1 ClaimAddress, meaning the exact edition plus an intrinsic claim identity declared by that edition's ClaimGraph. Later ClaimAddress uses in C.25 mean that same value; a field list or raw record reference is not enough.
Field meanings
- Name. The engineering quality family label inside the claim content, such as
Availability,Resilience, orSecurity; it is not an identity key. - QualityBearer. The one independently identified EntityOfConcern of the enclosing claim episteme. It may be a
U.System,U.PromiseContent,U.Episteme, or another entity under its direct identity pattern. When selected organization is the subject, use oneA.22U.Structurewith exact constituents, selected obtaining relations, applied constraints, and one selection-use frame. A list of local system-role kinds and assignment occurrences does not by itself identify a bearer. - ClaimScope / WorkScope. USM sets over
U.ContextSlicedescribing where the claim holds or where the capability's deliverability claim may be evaluated. These are set-valued scope objects, not characteristics. - Measures[CHR]. One or more admissible CHR characteristics, each bound to one declared scale.
- QualificationWindow. The temporal policy under which the quality claim is judged.
- Mechanisms / Status. References to realizations of operation families declared in
U.Mechanismepistemes, control presences, certification states, or other named gating prerequisites. They are not measurements. - Evidence. Anchors that justify the measures, mechanisms, or scope claims.
Guard reading
A quality guard is conjunctive only over the truth conditions that the current claim declares. For example:
declared scope covers TargetSlice AND declared measures satisfy their own laws AND each other declared prerequisite holds
An optional field can be left out when the claim does not depend on it. When another party uses that claim to make a decision, recover the conditions of that decision and compare them with what the claim covers. A supplier may document response time while the receiving service also requires current information. Meeting the response-time target settles that part. The service can accept both conditions when it also has grounds for freshness. An established breach of either bound rules out acceptance under this joint requirement.
Each measure keeps its own scale and comparison law. Use the declared decision rule for alternatives, trade-offs, weighted combinations or partial orders. If the typed decomposition cannot express what makes the claim true, use the pattern that governs the comparison.
Archetypal Grounding
A quality family is not automatically one metric. Use one Characteristic when one measure and Scale carry the claim; use a Q-Bundle only when several differently typed contributors are jointly load-bearing.
Minimal availability case.
Under ServiceQualityScheme-v4, the claim says: CheckoutAPI maintained at least 99.9% availability for customer-facing request handling over the rolling 30-day window. Its exact bearer is the independently identified CheckoutAPI System. In this example the receiving use separately checks coverage of customer-facing requests and the rolling-window qualification as well as the ratio.
Its Q-Bundle content has Name: Availability, ClaimScope: customer-facing request handling, Measures: AvailabilityRatio[%] >= 99.9, and QualificationWindow: rolling 30 days. WorkScope, Mechanisms, Status, and Evidence are omitted because this drafting use does not rely on them.
If evidence reliance, a failover prerequisite, or a gate later becomes current, add only the direct relation or slot that question needs. If the receiving question instead asks only for AvailabilityRatio under an already fixed scope and measurement window, the single-Characteristic branch is sufficient.
Escalation examples. A resilience or security claim often needs several measures, scenario or attack-class scope, mechanisms or control statuses, and a qualification window. Those contributors belong in the bundle only when they are part of that claim's truth conditions; treating the family as one scalar score would erase which contributor failed.
Bias-Annotation
The pattern biases authors toward explicit decomposition. It is better to publish a visibly structured quality bundle than to gain short-term convenience by collapsing scope, measures, and mechanisms into one overloaded quality label.
Conformance Checklist
-
CC-C.25-1If an engineering quality claim is intended as one measurement characteristic, the publisher SHALL bind it to one namedU.Characteristicwith one declared scale. -
CC-C.25-2If the claim requires multiple measures, scope slots, mechanism slots, status slots, or qualification windows, the publisher SHALL use Q-Bundle-shaped ClaimGraph content rather than an undeclared scalar surrogate. -
CC-C.25-3ClaimScopeandWorkScopeSHALL remain USM set-valued scope objects; they MUST NOT be treated as ordinal or numeric quality levels. -
CC-C.25-4Mechanism or status slots MUST NOT be conflated withMeasures[CHR]. -
CC-C.25-5Any scalar comparison or thresholding inside a Q-Bundle SHALL apply only to declared CHR measures, not to scope slots. -
CC-C.25-6When cross-context comparison is current, the publisher SHALL align the exact bundle heads or slots, resolve the two exactF.17local senses, and, when their<ReferenceScheme, LocalSenseClaim>projections differ and the comparison needs semantic correspondence, test the directF.9Bridge predicate and state a separate bounded-use claim only if that Bridge obtains.The comparison and its reliance MUST NOT change any Q-Bundle slot; ordinary reliance uses
A.10, whileB.3opens only when a named assurance claim is current. -
CC-C.25-7A materialized Q-Bundle SHALL be recoverable as content of one exactC.2.1episteme with one independently identified QualityBearer as EntityOfConcern and one effective ReferenceScheme. A gate, publication, comparison, proxy, or roll-up MUST NOT cite a field list as though it were an independently identified bundle object.
Common Anti-Patterns and How to Avoid Them
Consequences
Rationale
Engineering quality language is useful because it groups recurring concerns under memorable family labels. The same grouping becomes dangerous when those labels are mistaken for one universal metric. C.25 preserves the family labels but forces the underlying structure to stay typed and visible.
SoTA-Echoing
The comparison below selects lines by the quality-family problem they can solve.
FPF-local synthesis. C.25 combines the contributions needed to express the local quality claim: one exact bearer; a single-Characteristic exit; otherwise a typed separation of load-bearing measures, scopes, windows, mechanisms, statuses, and evidence references; and a guard over only the conditions the claim states. The tuple and this conditional guard are FPF synthesis, not a claim that contemporary practice already shares one universal Q-Bundle.
Defeating and reopen conditions. Prefer a direct domain pattern when it already supplies a clearer composite-quality identity, aggregation law, and practitioner route. Reopen C.25 when:
- a useful quality claim cannot be stated through the available typed contributors without inventing filler;
- a non-conjunctive trade-off or dependency cannot be named under its direct pattern;
- the record costs more than the receiving decision warrants; or
- a proxy repeatedly becomes the decision object despite the source claims remaining load-bearing.
Relations
E.21 specialises Q-Bundle-shaped claim content for FPF pattern-quality claims. C.25 remains the general endpoint pattern for engineering quality families; E.21 governs the claim when its exact EntityOfConcern is one FPF pattern version evaluated as action-guiding FPF text.
C.27 temporal-claim relation.
- C.27 may flag: a quality-family statement where agility, resilience, adaptability, recovery, or robustness depends on braking, redirection, stabilization, recovery rate, or rhythm under effort.
- This pattern keeps: quality-family bundle structure, scope, mechanism/status slots, evidence, qualification window, and failure mode.
- Non-admissible use: temporal adequacy alone does not establish quality adequacy; speed, recovery, or rhythm becomes quality content only when it contributes to a declared quality-family claim. Include scope, mechanism/status slots, evidence, and failure mode only when the claim or receiving action depends on them.
- Coordinate with C.27 only when a temporal claim is being used to change action; do not make every quality bundle carry dynamic slots.
Other relations.
- Builds on:
A.2.6for scope algebra,A.6.1for operation-declaration references when current, andC.16 / A.18for CHR legality. - Coordinates with:
C.2.2a,A.16.0,A.10for ordinary reliance on a bounded cross-context use;B.3only when a named assurance claim is current;A.15for gate use;C.16.Pfor unresolved characteristic, scale, score, metric, or proxy wording inside a quality-family statement;C.16.Qfor overloaded quality or evaluative-characterization wording;C.33,C.34, andC.35when captured structure, lost structure, preservation, or generated-result adequacy becomes part of a composite architecture quality family;C.17,C.18, andC.19for adjacent quality-family measures;F.9when bundle comparison needs semantic correspondence between local senses with different interpretation bases, orF.9.1when Bridge stance annotation is required.
- Constrains: engineering quality authoring whenever a quality term would otherwise drift between single-CHR and composite-bundle readings.
Endpoint function in evaluative classification
Use C.25 directly when the engineering quality claim and its contributors are already recoverable. Use C.16.Q for unresolved evaluative wording and C.16.P for an unresolved Characteristic, Scale, score, metric or proxy; then return to the endpoint their result supports. qualityTermAscription(...) may remain a transitional repair record, but it is not the universal resting place when the admissible result is a claim about one Characteristic, one episteme with Q-Bundle-shaped claim content, or an explicit objective-oriented quality claim under its own endpoint pattern.
Decision Test: Single Characteristic or Bundle?
The most common authoring failure is not in the bundle syntax itself; it is in choosing the wrong endpoint shape. The quickest useful test is to ask what would make the quality claim false.
Use one U.Characteristic when
A quality claim should terminate in one admissible CHR characteristic only when all of the following hold together:
- one measurable aspect is doing the evaluative work,
- one declared scale is enough to compare relevant cases,
- the bearer and scope are already clear without introducing extra quality slots,
- mechanism or status presence is not itself part of the core quality head,
- and downstream gates can read the claim without needing a bundle decomposition.
Examples include a narrowly declared AvailabilityRatio[%], a specific latency percentile, or one response-time threshold under one fixed window.
Use a Q-Bundle when
A quality claim belongs in C.25 when one family label is standing over several distinct typed concerns, for example:
- several measures are needed together,
- scenario or claim scope is load-bearing,
- mechanism presence or certification state constrains admissibility,
- qualification windows alter the reading materially,
- or one scalar head would hide which part of the family is failing.
The bundle is the explicit authoring form for claims whose truth conditions are already composite.
Borderline cases
Some quality families contain both a bundle-shaped form and a narrow single-characteristic form. For example, a service team may use:
- one CHR characteristic for a narrow uptime commitment, and
- one Q-Bundle for the broader service-availability family that includes scope, windows, failover mechanisms, and evidence.
This is legitimate as long as the text states clearly which head is currently in play. The single-characteristic form does not replace the broader family; it selects one evaluative slice of it.
Slot Interaction Law
C.25 names the slots and defines how they interact.
Scope and measure remain orthogonal
ClaimScope answers where or under what contextual slice the quality claim holds; WorkScope answers where the capability's deliverability claim may be evaluated. Measures[CHR] answer how a measurable aspect behaves. A broader scope is not a larger measurement value; a narrower scope is not a penalty value. Scope is governed by set inclusion and coverage, not by scalar order.
Mechanism and status are gating slots
Mechanisms and statuses may be load-bearing for admissibility, but they do not become measurements merely because they matter. A redundancy mechanism may be required for claiming a resilience bundle, and a certification status may be required for external publication, yet neither slot is itself the Measures[CHR] head.
This matters because many quality arguments fail by turning mechanism presence into an implicit score.
Declare windows that affect the claim
A quality claim that depends on rolling windows, observation periods, maintenance intervals, or disruption horizons must publish that temporal qualifier explicitly. If the truth of the quality claim changes when the window changes, then the window is part of the declared bundle record rather than optional commentary.
Report-only summary proxies
A publisher may compute a report-only summary proxy, such as a compact quality summary value. Identify the source quality-claim episteme or exact ClaimAddress, and state which Q-Bundle slots the proxy summarizes and omits. The proxy is a separate Characteristic or claim under its direct pattern; it does not replace that source in a norm, gate, comparison, endpoint classification or cross-context use.
Worked Quality Families
Availability family
A narrow service commitment may use AvailabilityRatio[%] as one characteristic. A broader availability family usually still needs a bundle because the claim depends on:
- declared service and time scope,
- observation and qualification window,
- one or more mechanism slots such as failover or redundancy,
- and evidence tying the measurement to declared observation conditions.
The bundle form makes it possible to distinguish "the measurement fell short" from "the measurement is fine but the declared mechanism prerequisite was absent".
Resilience family
Resilience is almost never one scalar. It commonly binds together:
- disruption scenario scope,
- observed restoration measures, such as mean time to restore under a declared observation protocol, and the recovered data point,
- recovery mechanisms and preparedness states,
- and scenario-specific evidence about drills, restorations, or incident traces.
Keep the observations distinct from their objectives: RTO bounds permissible recovery duration, while RPO specifies the required data-recovery point. Neither is an observed recovery result. Compare the observations with the applicable objectives under the declared rule.
Trying to compress this into a single resilience value can hide the difference between fast recovery in one scenario and structural fragility in another.
Security family
Security claims routinely combine:
- trust-zone or attack-class scope,
- measurable characteristics such as patch latency, control coverage, or response interval,
- control-set and certification slots,
- and evidence from audit, observation, or incident review.
C.25 therefore treats broad security-family claims as bundle-shaped unless the claim has already been narrowed to one admissible CHR characteristic.
Maintainability and evolvability
For maintainability or evolvability, recover the particular claim and the contributors on which it depends. These may include:
- the system or change-class scope;
- measures such as change lead time, defect reintroduction rate, restoration interval or review load;
- enabling mechanisms such as modularity rules or a test harness; and
- the observation window or other support used.
Keep measures distinct from mechanism and scope claims. A narrow claim carried by one Characteristic can stop there; a broader claim retains every load-bearing contributor, not every example in this list.
Authoring and Review Guidance
For authors
Begin with what would make this claim false? Then:
- identify the exact bearer and the quality-family label used in the claim;
- if one measure on one declared Scale carries the claim, state that Characteristic and stop;
- otherwise add only the differently typed contributors that jointly carry the claim;
- omit scope, window, mechanisms, status, or evidence when changing that slot would change neither the claim nor the receiving action;
- identify the enclosing C.2.1 episteme through its claim content, bearer, and effective ReferenceScheme; and
- add a proxy, gate, publication, evidence relation, or assurance result only when its own receiving question is current.
The schema remains available for a demanding case; it is not the authoring order for every bundle.
For assessors
A checking reader should ask:
- whether the chosen endpoint shape is admissible,
- whether any scope slot is described as a scalar,
- whether mechanism presence has been mistaken for a metric,
- whether the window is optional or load-bearing,
- and whether any summary proxy is being used to replace the underlying bundle.
For gate designers and assurance leads
Resist a guard such as resilience must be high. Cite the exact quality-claim episteme or addressed claim and name only the slots the decision uses—for example one scope, one measure threshold, a load-bearing window, or a required mechanism. Do not require an absent slot merely because the source claim uses Q-Bundle-shaped content.
Repair and Boundary Notes
Repair from bare quality requirement prose
Keep a familiar label such as security requirement when its content already makes the intended claim clear. If the label hides what is required, recover the endpoint:
- choose the endpoint shape first,
- then bind the requirement or commitment to that explicit head.
C.16.Q may still be the entry repair for overloaded quality wording, and C.16.P may repair characteristic, scale, score, metric, or proxy wording inside the same statement; C.25 is the resting place only after the engineering quality family has been made explicit.
Boundary to cross-context use and reliance
For cross-context comparison, use the bundle alignment and conditional correspondence operation in section 18.3. That comparison changes neither the endpoint's Characteristic-or-bundle kind nor the bundle's claim content.
Boundary to publication convenience
A report, summary publication, or executive summary may express only one slice of the selected quality-claim episteme. Keep the selected episteme and any exact ClaimAddress distinct from its publication occurrence, form, and carrier under E.24.PUB. A coarser form does not collapse the source claim content, while changed Q-Bundle claim content identifies another episteme even when the file or layout stays the same.
Serviceability and supportability
Serviceability, supportability, and adjacent family labels often look simple in prose but become composite as soon as operational use is declared. An admissible bundle for this family may need:
- support-scope slices,
- measured restoration or service intervals,
- mechanism slots for support mechanisms, access discipline, or replacement procedures,
- and evidence from service traces or support records.
The lesson is the same as elsewhere in C.25: once the truth of the family claim depends on several typed contributors, the bundle should stay explicit.
Boundary to description-side and selector-side evaluation
C.25 is for engineering quality families whose bearer is a system-side, promise-side, or explicit quality-bearing artifact. It does not automatically cover:
- viewpoint-fit or grounded architecture adequacy claims, which may belong in viewpoint or evaluative-ascription patterns,
- or selector/objective heads where quality means use-value under a search or portfolio frame.
This boundary matters because the same word quality appears across those zones. C.16.Q repairs overloaded quality wording, C.16.P repairs characteristic, scale, score, metric, or proxy wording when that is the hidden object, and the resting endpoint depends on what is being evaluated.
Bundle Decomposition and Comparison Law
Local decomposition rule
A family label may remain stable while its internal slots differ materially across contexts. Conforming comparison therefore starts by aligning the bundle decomposition: scope slots with scope slots, measure slots with measure slots, mechanism/status slots with their own kinds, and evidence/window slots with their own kinds. Comparing one bundle's measure directly to another bundle's mechanism claim is a category error even if both sit under the same family label.
Narrow slice versus whole family
A publication may expose one narrow Characteristic claim from a broader Q-Bundle-shaped claim episteme, but it must identify that addressed claim as only one contributor to the broader family. It must not cite the slice as though it exhausted or reidentified the source episteme.
Cross-context family comparison
Cross-context comparison of quality families starts with explicit bundle alignment: compare scope with scope, measures with corresponding measures, mechanisms or statuses with their own kinds, and windows or evidence only when the receiving comparison uses them.
For each semantic correspondence the comparison needs, resolve the two exact F.17 senses and test the direct F.9 predicate only when their <ReferenceScheme, LocalSenseClaim> projections differ. Cite a Bridge only when it obtains, then state the proposed use separately with its direction, correspondence rule, tolerated loss, and polarity.
Keep observed loss in evidence, use A.10 for ordinary reliance, and open B.3 only on its own assurance trigger. None of this changes the Q-Bundle or supplies an automatic penalty.
Gate, Proxy, and Reporting Discipline
Report-only summary proxy
Apply the report-only proxy rule in section 14.4 to the exact source episteme or addressed claims.
Gate binding rule
When a gate uses a quality family, take its acceptance conditions from the receiving decision. Cite the quality-claim episteme or its addressed claims and identify which conditions each supports, such as a measure in a specified window or a required mechanism. Reuse other evidence for any condition the supplied claim leaves open. The gate's own rule determines what to do with a known failure or an unresolved condition.
Roll-up caution
A roll-up is a separate claim-bearing episteme. Cite the exact source epistemes or ClaimAddresses and explain how their claims support the property the recipient needs. A direct observation of the service may already establish the property. When drawing a conclusion from the parts, identify how their states or contributions determine the property in question. Suppose a service can accept a request precisely when both stages are available. Each stage was available on 99 of the same 100 request opportunities. If their outages coincided the service was available on 99 opportunities. If they occurred separately it was available on 98. These records settle a target of at least 98. To decide a target of 99, establish whether the outages coincided or use records of service availability itself. When a summary is used for local engineering action, return to the source claims and the Q-Bundle slots that action depends on.
Review Matrix and Repair Tests
A checking reader can test a Q-Bundle with five questions:
- Is the endpoint shape admissible? One characteristic where one characteristic is live, one bundle where several typed contributors are load-bearing.
- Are scope and mechanism slots kept distinct from measures?
- Is any summary number being used to replace the bundle?
- Would a gate still be auditable if the family label were removed?
- If comparison needs cross-context semantic correspondence, is the
F.9Bridge predicate tested separately from the family bundle?
Repair from bare family prose should therefore recover the endpoint shape first, then choose whether any narrow slice deserves a separate CHR publication.
Viability-envelope, quantum-like, and temporal-claim relation note
Use C.25 for the quality-family decomposition needed by a quality claim, proxy metric, trade-off, gate, or report. A viability claim should not become quantum-like merely because it involves uncertainty, feedback, several qualities, or changing operating conditions; a temporal claim should not become a Q-Bundle merely because the working phrase mentions speed, cadence, rhythm, or recovery.
Practical reading:
- Decide whether one Characteristic answers the quality question; if it does, stop there.
- If several differently typed contributors are load-bearing, identify the bearer and include only those measures, scopes, windows, mechanisms, statuses, or evidence anchors.
- If one proxy or this bundle answers the receiving question, stay in
C.25. - Open
C.26.3when several conditions must hold together under a disturbance and the question is whether the viable region is retained, violated or unresolved, or how a proposed change could restore it. - Open
C.27only when a temporal claim is being used to change action—for example, when rate-change, effort, window, resistance, recovery, or cadence changes that action.
For an envelope question, use C.26.3:4.5 to recover the exact bearer and protected work or result, derive the relevant conditions from ways that work can fail, and obtain bounds from the receiving requirements and actual provision. Keep each bound's subject explicit. Compare the available observations with those bounds under the disturbance. A supported diagnosis can finish the use: a known violation establishes failure of the conjunction even while another condition is unknown; a positive viability claim needs all required conditions supported.
For proposed regulation, identify the intervention and which constraint each predicted effect concerns. Retain the grounds, unknown effects, displaced work, trade-offs and adaptation costs needed to judge that proposal. Recover needed authority or execution conditions for the proposed action. Unknown acceptability limits remain a requirement question or an explicit scenario assumption; observed averages do not supply normative bounds. These conditions are part of the claim and its use, not a compulsory envelope form.
Useful outputs:
- one
C.2.1quality-claim episteme with Q-Bundle-shaped content when the issue is quality decomposition; - a
C.26.3supported envelope diagnosis or conditional regulation proposal, with the grounds required by that result; - a
C.27temporal-claim adequacy card when a claim used to change action says that an intervention changes a rate, recovery, or cadence; state its effort, window, and resistance where load-bearing;
Architecture-decision Q-Bundle boundary
C.32.P2S, C.32.PAD, and C.32.ADA may cite exact C.25 quality-claim epistemes or ClaimAddresses as architecture-characteristic inputs, accepted-loss structure, guardrail rows, feedback concerns, or adequacy concerns. C.25 keeps their Q-Bundle claim content, bearer, scope, measures, mechanisms, qualification window, and evidence distinct from the problem-to-structure architecturing flow, project architecture decision relation, and ADR-like publication projection.
C.25:End
Quantum-Like Modeling Lens
Type: Architectural pattern Status: Stable Normativity: Normative unless marked informative
Problem frame
FPF already has local patterns for decisions, boundaries, bridges, work, measurement, search, and quality bundles. Some architecture cases break when those patterns are applied as if every read, question, dashboard, workshop, bridge, or simplified representation were a passive view of a stable state.
Use this pattern to examine a possible contextual-model obstruction that the ordinary subject account does not yet settle and that could change the inference or action.
Retain a QL use only when a named obstruction is established under stated assumptions—for example, failure of a proposed joint probability law, incompatible probe algebra or an order-sensitive instrument result. A label, diagram, ordinary locality or model plurality does not establish that result. The claim is relative to the specified construction and use; it is not proof that every non-QL treatment must fail.
What goes wrong if missed. A dashboard, workshop, metric, bridge, export, or coarsened model is treated as a passive faithful readout even when the probe, frame, publication, or representation shortcut changes what can be inferred.
What this buys. The user keeps the ordinary FPF pattern in charge and adds the minimum quantum-like lens needed to prevent that representational mistake.
Identify the claim before the lens. Name its subject and the interpretation, scope, probe/model frame and comparison conditions on which it depends. Recover an exact C.2.1 episteme or U.ClaimScope when the receiving use consumes that identity. For an empirical-grounding claim, test the exact covered claims and direct mappings under C.2.1:4.3; known absence and an unresolved grounding assertion remain different.
If a viewpoint matters, record one U.ViewpointRef that resolves to the U.Viewpoint episteme P.
When a particular dated evaluation Work is claimed, recover the actual performer System through A.13 and let A.15.1 independently admit the dated Work with its enacted Method.
Cite the same obtaining A.13 assignment occurrence, its declared species, and F.6 only when the receiving use consumes precise assignment-bound attribution. Name a non-performing participant by its evaluation relation and position. Keep these neighboring values separate from the QL-use fields.
In FPF, quantum-like names a detached mathematical and representational lens, not a physical claim about the target. It is comparable in use to probability, calculus, optimization, or state-space modeling. A QL-lite note can supply a small recognition or conditional comparison. Choose additional mathematical or empirical support from what the receiving use needs the claim to establish, following :12b.
Ordinary modeling questions that can expose a QL need are listed below. C.26.1–C.26.3 also supply independently useful ordinary accounts of probe effects, distributed-state evidence and viability envelopes. Using one of those accounts does not activate QL; the lens requires the residual obstruction stated above.
Plain glosses:
quantum-like: a detached mathematical or representational lens, not a claim about what the target is made of.probe: an operation that both produces an output and may change the represented state or admissible use of the output.frame: the exact probe frame, measurement frame, comparison frame, or model frame selected by its subject pattern; the effective reference scheme remains a separate part of the claim-use account.state: the represented condition relevant to the current decision.state update: a typed update claim. When load-bearing, name the kind of update: system change, work change, epistemic reading update, carrier update, emitted-output update, formal model update, or update-law change.context: when locality may matter, recover the exact claim scope, reference scheme, local-sense endpoint, selected model-use structure, qualification window, viewpoint relation, or direct subject relation that the sentence needs.export: a carried representation whose use may lose timing, coordination, system-role or participation relations, use conditions, confidence, or relation structure.coarsening: an intentionally cheaper state representation with declared loss and reopen conditions.
Phrase hygiene:
Example style:
Read these terms by the distinctions the model needs.
Problem
Teams make five recurring representational mistakes that this lens addresses:
- They treat a probe as a neutral read when the probe changes later answers or behavior.
- They combine two posterior-looking outputs as if both came from one shared sample space.
- They export a team state, dashboard value, or context-map result as if it were a faithful-enough export for the intended use.
- They compress a large state representation for speed and then reuse the shortcut outside its admissible-use scope.
- They let words such as
quantum,entanglement,collapse, orfieldimport ontology that the model never earned.
For example, the team may:
- approve a release from a dashboard whose publication and operational use changed the work it was supposed to report;
- average results produced in incompatible local algebras;
- reuse a local decision under a different effective reference scheme after the admitted bridge lost load-bearing meaning;
- claim a speed gain because the representation was low-bit, linear, symbolic, or compressed without naming the loss.
Forces
Solution
Start with the ordinary subject question and the account already available. Identify the claim, its subject and the conditions that affect its meaning. Examine only the remaining contextual-model question that can change the receiving answer.
Keep local probe results in their own algebras unless a justified comparison or joint construction is available. A supported ordinary answer finishes without QL; a named obstruction can justify a bounded QL contribution under the conditions below.
Application sequence:
- Name the ordinary FPF pattern that already carries the baseline question.
- Recover the subject, reference scheme, scope and relevant probe/model and comparison frames. Use an exact episteme, scope object, grounding relation or viewpoint only where the claim depends on that distinction.
- State the proposed inference: for example, reading pre-probe state from an output, combining local probabilities, exporting a state claim or using a coarsened representation. Identify the premise that is unsupported or contested.
- Use the available observations, procedures or formal objects to test that premise. The ordinary subject method may settle it. For a joint-probability question, compare local outcomes, their overlap marginals and the conditions of the proposed common law; :5.1 gives a finite construction and impossibility argument. For a probe effect, use C.26.1's procedure-and-timing reconstruction; for another formal obstruction, use its stated definition and argument.
- Return the supported result: an adequate ordinary account, an unresolved premise, or a named obstruction and its consequence. A QL-lite note can carry the last result in ordinary prose; the vocabulary alone supplies no argument.
- Identify what the result will be used to establish. Add the applicable account when prediction, model adoption, comparative performance, assurance or another relied-on conclusion needs it. A reusable conditional explanation within the same assumptions can remain small; apply :12b and reuse adequate existing support.
C.26 ordinary output: produce one of these, then stop or select the neighboring applicable pattern body:
- an adequate ordinary account without QL;
- the exact unresolved premise that prevents the requested inference;
- a QL-lite explanation of a supported obstruction and its bounded consequence;
- a C.26 child result, which may stand without QL;
- further mathematical, empirical or assurance work only when its contribution is needed by the receiving claim.
Keep the entry cost proportional to the use. A QL situation does not begin with a full record.
The everyday pattern move is a small recognition note plus a bounded action. Add a decision-bearing or assurance account when its receiving use requires one.
Choose detail by the receiving question:
Apply the same selection when writing, checking or reusing a pattern. The role of the reader does not determine the form or amount of support. Use C.11.DUA when deciding whether additional work can improve the receiving decision enough to justify its cost.
Checking discipline:
Cluster maxim: retain the support adequate for the receiving question. Add assurance work when a consequential unresolved premise requires it; reuse, publication or a formal notation alone does not change what has to be established.
Use a cited pattern for the contribution needed by the claim you are making. A citation to C.26 opens the lens only for the residual contextual-model obstruction. C.26.1, C.26.2 and C.26.3 can instead supply their ordinary probe, evidence or envelope result without QL. The citing pattern's action guidance remains primary; neither citation requires every field or every child pattern.
Model-use structure and crossing boundary. Select a BoundedModelUseStructure only when the organization of one exact model episteme, admitted model-use holons, obtaining applicability/use/coherence relations, applied constraints, invariants, and one named receiving use changes the decision. Compare two such structures only after each is independently selected on that basis. Assert a subject crossing only when an exact direct governor makes one direction-sensitive crossing occurrence obtain among those exact structures. When the direct governor is absent, return the missing-governor blocker.
QL boundary selection:
The following questions can organize a retained QL use in ordinary prose or a record. Include the identity, grounding and viewpoint distinctions only where the claim or receiving use consumes them.
Examples of the decision difference to recover, whether the ordinary account is sufficient or a residual QL obstruction remains:
Minimum ordinary-pattern return from the QL activation test:
The changed priorities can be assessed under the work and measurement patterns. Reopen C.26 when a named contextual-model obstruction survives that account and changes a conclusion or action. Existing evidence can settle the decision.
Finish with the useful result the receiving question needs. A mathematical consequence or corrected reading may be sufficient. Use C.11 only when a choice remains.
Retire QL when the residual cue disappears. If A.6, F.9, C.16, A.10, B.3, A.15, C.25, A.6.3.CSC, A.6.3.RT, or another ordinary FPF pattern now carries the claim without a false passive read, false shared frame, false faithful export, unsupported distributed-state reading, or QL-specific coarsening residue, remove QL wording from the active working note or pattern prose.
Retain the lens only when both sides of the activation test support its use. C.26 remains active only when one named contextual-model obstruction survives the ordinary subject patterns and changes an admissible engineering inference or action: for example, a no-global-section result, an incompatible-probe algebra, or an order-sensitive instrument effect.
Bridge loss, feedback, coupling, openness, compression, coarsening, vocabulary, graph shape, and DDD locality are not QL cues by themselves.
Preserve each local result in its own algebra unless an independently admitted joint-comparison route exists; do not assume a global frame or infer a structure crossing from comparison.
Cue families for examining a proposed obstruction. Every row remains subject to the activation test above; the listed effect alone is insufficient.
Apply both sides of the activation test:
Keep incompatible-probe outputs in their own algebras. A common label or diagram does not supply a joint probability space, comparison relation, grounding relation, or cross-structure occurrence.
Read the observation and its intended use before choosing a more precise formulation. Ordinary FPF methods can settle the following questions:
- For “the metric reported readiness,” identify which check passed. To claim that publication changed behavior, trace the change in the work through which the metric was produced or used.
- "two risk scores disagree" becomes "the two scores may come from non-shared comparison frames with no declared admissible joint comparison route";
- For “the workshop discovered the split,” identify what participants recorded or agreed. Attribute a change to the order of questions only when the procedure and evidence support that effect.
- "the team knows" becomes a bounded claim about the coordinated work evidenced, with its relevant interpretation limits;
- "this smaller model is enough" becomes a claim that its retained distinctions suffice for the named use, with any actual loss and limit stated.
Inherited QL boundary
Invariant QL-NQ: within FPF, quantum-like is a detached mathematical and representational modeling lens. It may use quantum-theory-derived structures such as contextual probability, Hilbert-like state spaces, non-Boolean logic, instruments, operator-like update, order effects, open-system descriptions, or incompatible probes.
Quantum-like does not assert physical quantum substrate, microscopic quantum process, qubits, quantum computation, physical entanglement, nonlocal causality, literal collapse, mystical observer effects, social substance, or collective mind. A physical-quantum claim is a different claim and needs separate physical or empirical support outside this pattern cluster.
Child patterns inherit QL-NQ. They should not restate the global boundary as local guidance unless they are repairing a specific confused phrase.
Pattern selector
Causal-use exit before QL retention
Before retaining QL-lite, QL-NQ, or a quantum-like framing for a claim being made, check whether the question is intervention, counterfactual comparison, causal effect, causal fairness, causal policy, off-policy causal evaluation, or realizability of counterfactual-rung data. If so, redirect the claim or question to C.28 before any quantum-like retention.
What changes in practice: apply the relevant causality-ladder rung declaration, causal identification, causal evidence support basis, and counterfactual sampling realizability requirements to the causal-use claim.
C.26 keeps quantum-like modeling discipline; C.28 governs causal-use support, including counterfactual material.
Use this as a diagnostic sequence before retaining QL wording. DDD, microservice domain analysis, and direct boundary, model-use, local-sense, and Bridge subject patterns stay first for service cuts, integration points, and exported meaning. Retain QL only when one named contextual-model obstruction survives those subject patterns and changes what can admissibly be inferred.
-
Measurement, metric, scale, method, evidence, or assurance load goes first to measurement and evidence patterns:
C.16,A.10, orB.3. -
Bridge, translation, publication availability, rendering, or exported-loss question goes first to its applicable subject pattern:
F.9for an exact SenseCell Bridge;E.24.PUBfor publication occurrence, form, and carrier;E.17only for a current multi-view publication form or face; andE.17.EFPonly for a current explanation-faithfulness claim. -
A causal intervention, command, or routine question goes first to its pattern.
For precise Work enactment, recover each actual performer System through A.13 and let A.15.1 independently admit the dated Work with its enacted Method. Cite the same obtaining A.13 assignment occurrence, its declared species, and F.6 only when the account or its receiving use expressly consumes precise assignment-bound attribution; F.6 identifies neither assignment nor performer, and missing or failed F.6 attribution alone leaves the Work intact while its independent A.13/A.15.1 grounds still hold. A non-performing relation participant stays with its relation and position.
-
Boundary or interface wording, service-interface typing, bridge endpoint, relation precision, or lexeme-collision question goes first to the subject pattern:
A.1for holon delimitation or boundary crossing;A.6.Pfor relation precision or service/access recovery;A.6.0orA.6.5for signature or slot claims;A.6.Mfor module-interface claims;A.6.Ffor functional ports or elements;A.6.Conly when recovered contract, SLA, protocol, or agreement-like wording bundles promise, utterance or publication, governance, Work or consequence, or evidence claims;A.6.Bonly for L, A, D, or E statement classification inside a boundary package;A.7,E.10, orF.18for wording-use repair.
-
Quality, viability, feedback, or control-tuning question goes first to quality, dynamics, and measurement patterns:
C.25,U.Dynamics, andC.16. -
Suspect option menu, unknown alternative, local plateau, basin movement, or candidate-generation question goes first to search and regime patterns:
B.5.2,C.18,C.19, orA.19. -
Retain QL only when the remaining cue identifies a contextual-model obstruction under :4 and changes the admissible inference or action.
C.26 can mark that the available readings sit in non-shared comparison frames or lack a declared admissible joint comparison relation. Choice, generation of missing alternatives, and C.11 decision quality remain with C.11, B.5.2, C.18, C.19, or A.19.
The last column locates a possible remaining modeling question. Apply the :4 activation test to it; an ordinary probe effect, missing comparison or export loss alone is insufficient.
Escalation by evidence or authority demand
When a conclusion relies on QL model adequacy or advantage, compare the relevant rival explanations. Failure of a simple Bayesian or passive-read model does not establish QL necessity. The following families supply alternatives when they address the same question; a conditional mathematical consequence does not require empirical comparison with every family.
Math reveal sequence:
Use the least formal account that establishes the needed result; a mathematical consequence may require M2 even when it needs no empirical assurance.
Mathematical formality and evidence use are separate choices.
Select the evidence-use class from the question the result must answer. QLP-0 and QLP-1 cover recognition and working support. QLP-2 and QLP-3 supply the additional comparison or assurance needed by a receiving decision. The classes identify useful contributions; reuse an adequate available account.
Evidence-use class scales by use:
Recognition case matrix
State-representation coarsening card
Use this section when a fuller state representation is too costly or detailed for the work and you need to establish what a coarser representation can still explain or support. Begin with the ordinary account of the reduction. Use the QL lens if the remaining obstruction meets the conditions in section 4.
For a shorter rendering of the same claims, use A.6.3.CSC. For another way to represent the same selected entity, use A.6.3.RT. When comparing a more abstract model with a more concrete one, use B.5.MR:4.2–4.3 to establish what the concrete model preserves and which alternatives it excludes. A.19 and U.Dynamics govern state and change. Use C.26 when the decision also encounters the contextual-model obstruction tested in section 4.
Question-to-pattern map:
To decide from a coarser reading, identify what detailed states or histories could produce it under the model. If none can, check the reading and the model before using them together. If every compatible history permits the proposed action, the missing detail cannot rule out that action. A shared reading may still leave open which action is best. If one compatible history permits an action and another rules it out, recover what distinguishes them before relying on the reading for that choice. For a decision based on probability or expected value, use the applicable probability model and its evidence.
When the commission is to explain a result, follow the proposed explanation through the histories that the coarser representation treats alike. Different mechanisms can warrant the same immediate action. A common ground for the result may explain it across all those histories. If the recipient needs to know which mechanism operated on this occasion, recover the distinguishing relation or report the alternatives the evidence still supports.
For a bounded coarsening use, start with these questions. An adequate ordinary explanation can answer them without a separate card.
When the receiving use needs a more detailed account of the shortcut, add the relevant distinctions below to that same explanation.
For a linearity claim, identify the operator or equation, its domain and the argument establishing linearity. A proof of the required property can finish that question; any performance advantage needs a separate comparison.
For a comparative speed, cost, compression, stability or tractability claim, identify the baseline and proposed representations or routes, the operation and quantity being compared, and the conditions under which the comparison holds.
Preserve the meaning of the compared quantity and any resource or cost basis it uses, such as execution time, memory use, measurement or coordination effort, energy use, risk, cognitive load or privacy loss.
Supply the mathematical or empirical grounds for the claimed difference. Explain and support a mechanism when it forms part of the claim.
Retain any actual loss or error that matters to the comparison or intended use, with the applicable budget and use limits. Compare other serious alternatives and state return conditions when they can change the conclusion or decision. Reuse an adequate existing account.
If the shortcut carries a transition-speed, stabilization or control claim, supply the dynamics distinctions on which that claim relies:
Archetypal Grounding
Tell: A reliability dashboard says "Ready" after a new readiness metric is published. Before publication, teams treated incidents as local triage. After publication, they change priorities to satisfy the metric, while unmeasured recovery work gets delayed.
Show, System side: the delivery system, teams, dashboard, incident-handling cycle, and release decision form one operational situation. The dashboard is not only a window; it is part of the work ecology because it changes attention, escalation, and behavior.
Show, Episteme side: the ordinary account uses C.16 and A.15 for the metric and changed work, A.10 for evidence use, and B.3 only if the release-assurance question is current. Its useful result is to treat the dashboard as evidence affected by publication and behavior, not as a passive readiness read. No surviving contextual-model obstruction is supplied, so this case returns without QL wording.
Second ordinary-pattern return: a large state-space model is too expensive for triage, so the team uses four typed operational states. Use the direct modeling and state-abstraction method to establish the retained distinctions, loss, bounded use, and reopen trigger. Reduction to four states does not activate C.26; a named residual contextual-model obstruction would be a separate premise.
Matching local distributions can still block a joint model
In this constructed mathematical example, a modeler wants to export three specified pair-measurement distributions as one probability distribution over binary variables A, B and C. Only the pair measurements AB, BC and AC are available in the stipulated model. The proposed export requires each repeated variable to have one shared meaning and value across its two pair accounts; this is an explicit assumption of the comparison, not an inference from a repeated label.
The following probabilities are stipulated model inputs, not observed frequencies:
Each row sums to one. Each variable has probability 1/2 of either value in both accounts where it appears. Every local outcome is possible, so checking only allowed assignments leaves all eight binary triples available. These local checks do not establish a joint probability law.
Use C.29 for the declared mathematical correspondence and C.29.1 for the proposed transfer: projecting the joint distribution onto each pair must recover that row. The missing comparison is whether any such joint distribution exists. This is the probabilistic global-section question in Abramsky and Brandenburger, §§2.4–3 and 4.3. A measurement or cross-context meaning claim would separately need its C.16 or F.9 basis.
Derive the obstruction. For one binary triple, let D count unequal pairs. If all three values agree, D = 0; otherwise one differs from the other two and D = 2. Every probability distribution on triples therefore has E[D] ≤ 2. The supplied pair accounts instead require
Thus no nonnegative joint probability distribution reproduces all three rows under the stated identification. This is a probabilistic no-global-section result; it does not say that no individual triple is locally possible.
Use the result. Keep the pair accounts separate and reject the proposed joint export. A statistic needing that joint law remains unsupported. Changing the pair laws, the shared-variable assumption or the requested statistic reopens the comparison; adding a joint table without that change cannot repair it. The construction establishes no observed probe effect, physical quantum realization or empirical adequacy.
Compare an unobstructed case. Give each of the six nonconstant triples probability 1/6 and give 000 and 111 probability zero. Every pair then has probabilities (1/6, 1/3, 1/3, 1/6). These revised local laws have the displayed joint model. Pairwise availability alone therefore does not activate C.26.
The first useful result here is the exact obstruction to the live joint-export proposal. Ordinary probability theory or C.29.1 can supply the same argument; use an already adequate argument directly without an extra QL note. The C.26 contribution is this named contextual-model comparison, not a claim that QL is the only or faster way to compute it.
Bias-Annotation
This pattern biases authors toward ordinary FPF patterns before QL vocabulary. That bias prevents prestige use of the word quantum-like and keeps the mathematical lens focused on its declared representational payoff.
It also biases authors toward minimal admissible outputs. In ordinary use, the right result is often "apply the neighboring FPF pattern", "do not merge these comparison frames", "mark this dashboard as an instrument", or "return to the source representation if the shortcut fails".
The pattern may under-admit some mathematically valid QL models when the author cannot explain the practical payoff. That is acceptable for FPF pattern prose: a model that cannot say what it buys the working reader does not yet support this pattern's practical use.
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Near misses remain ordinary uses unless the separate :4 activation test establishes a contextual-model obstruction.
Cluster conformance scenarios
Use these as quick applicability tests. The ordinary result, including a result supplied by C.26.1–C.26.3, may be sufficient. Only a further named contextual-model obstruction activates C.26; the scenarios do not establish one merely by mentioning a child pattern.
The ordinary accounts and any supported QL finding can inform design options to test under the subject patterns:
AI and LLM work-cycle route examples
LLM-mediated work cycles can involve false passive reads, overused summaries, unsupported shared comparison frames and shortcuts without loss limits. The routes below recover the ordinary claim first. A C.26 child can supply its ordinary result; QL still requires the :4 activation test.
Consequences
This pattern gives FPF a single place to define QL-lite and the inherited non-quantum boundary. That reduces repeated disclaimers in child patterns and makes ordinary use lighter.
The gain is a better-bounded reading: probe-dependent evidence is not treated as a passive copy, comparison uses a shared frame, and export names what the target use may rely on. Where ordinary patterns already carry the claim, their result finishes the question.
The additional work is to establish any surviving contextual-model obstruction and state the inference or action it changes. Coarsening may make a state description cheaper to use; its declared losses and reopen conditions limit that gain to the intended use.
Rationale
The common lens relates a contextual-model obstruction to the inference it constrains. Probe effects, distributed-state evidence and viability envelopes have their own objects and operations in the child patterns. Their ordinary results can stand alone; a surviving obstruction determines when they need the common lens.
The lens earns its use when a named contextual-model obstruction changes the admissible inference or action. The inherited non-quantum boundary keeps that mathematical contribution separate from a physical claim.
Literature supports the modeling move; local evidence supports the local state, export, or probe claim. A source anchor can justify why order effects, contextual probability, instrument-like readings, or open-system modeling are legitimate modeling patterns.
Claims that this dashboard changed this team's state, this workshop changed a boundary, or this export lost the live coordination need their own proof or evidence under A.10, C.16, A.15, B.3, and the ordinary pattern for the local claim.
SoTA-Echoing
Selected operational source anchors
The following anchors support the pattern's operational modeling moves.
Relations
C.28 causal-use relation.
- C.28 governs causal-use question, causality-ladder rung, causal estimand, identification, counterfactual sampling realizability, causal evidence support basis, causal-use verdict, causal fairness, causal policy, and causal method parity.
- This pattern keeps residual quantum-like probe, frame, order, export, or coarsening discipline after ordinary causal-use explanation has been tried.
- Non-admissible use: intervention, causal effect, causal fairness, causal policy, counterfactual comparison, causal method parity, or counterfactual-rung-data realizability do not activate quantum-like modeling by themselves.
- Exit: when the question under repair is causal, cite
C.28before retaining QL-lite or QL-NQ.
C.27 temporal-claim relation.
- C.27 may flag: ordinary state/rate/rate-change, effort-window, rhythm, braking, coasting, or intervention-timing claims before any quantum-like cue is considered.
- This pattern keeps: residual quantum-like probe, frame, order, export, or coarsening discipline.
- Non-admissible use: discreteness, finite differences, typed states, state-space reduction, tokenization, dashboards, probes, measurement plans, speed words, rhythm words, or Dyn2 words do not activate quantum-like modeling by themselves.
- Boundary: use C.27 and ordinary FPF patterns first; use C.26 only where residual probe, frame, order, export, or coarsening cue remains after those relations are named.
Other relations.
- Builds on:
E.8,E.9,C.11,C.16,C.25,A.6,A.6.P,F.9,E.24.PUB,E.17,E.17.EFP,A.15,A.10,B.3,A.3,C.18,C.19,A.19. - Constrains: QL wording in
C.26.1,C.26.2, andC.26.3. - Carries: state-representation coarsening as a card inside
C.26:4.5, not as a separate pattern. - Does not cover: physical quantum claims, a generic probe ontology, a generic state ontology, a service/cell pattern, or a field-like synchronization pattern.
- Name boundary:
Quantum-Like Modeling Lensis a pattern label for a modeling lens and modeling discipline.
C.29 mathematical-lens use relation
C.26is a C.29-compatible specialization for quantum-like modeling. It carries a pre-filled adequacy profile: preserved structure includes order, probe and contextual-probability effects when supported; loss names distinctions omitted by the selected mapping; physical quantum ontology is not implied by that mapping; the canonical stop condition remainsQL-NQ.A QL-lite note retains its lightweight use. Apply C.29:4.4's reliance rule when more is needed: a conditional comparison or reusable explanation within the same assumptions may stay small; reliance as an adequate phenomenon model for prediction, a consequential decision, model adoption, benchmark/assurance input, Bridge-dependent use or transfer to further cases needs the full applicable account.
Reuse the supplied C.26 profile and adequate existing validation rather than filling a blank FullCard.
C.26:End
Probe-Coupled Boundary Interaction
Type: Architectural pattern Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when a boundary read, meeting, metric, API read, dashboard, workshop, survey or message is used as evidence of state, and a possible effect of that interaction on the reported state could change the decision. Recover which state the output supports before deciding whether the read is passive, probe-coupled, or unresolved.
The useful move is to recover what the interaction read, what it changed and which state its output can support. That ordinary result can finish the use. Add QL only under the separate C.26 activation test.
Plain glosses:
passive read: treating a workshop, metric, dashboard, API read, survey, or message as if it only reports state.probe-coupled: the read/export/intervention participates in the represented state enough to change the lawful decision.coupling channel: the concrete workshop, metric, message, API, dashboard, meeting, bridge, or event stream through which the effect travels.export loss: what the carried output cannot faithfully carry into another context or decision.
Problem
Architecture work often treats boundary interactions as passive. A dashboard "shows readiness". A workshop "discovers the context map". An API read "returns state". A meeting "collects alignment". A message "transfers a decision".
Sometimes that is good enough. Sometimes it is false for the current decision. Publishing the dashboard changes readiness behavior. Running the workshop changes alignment and local meaning. Reading the API changes timing assumptions. Sending the message changes trust, escalation, or error handling. Splitting the service changes the viability envelope that the split was meant to optimize.
When the false passive reading survives, the team may keep the wrong boundary, trust the wrong export, combine incomparable outputs, or redesign the system around an artifact that partly created what it reports.
Forces
Solution
Examine how the interaction produced the output and whether its effects matter to the state claim being used. A supported passive reading can finish the inquiry. A supported probe effect calls for a qualified reading; insufficient information leaves that interpretation unresolved.
A probe-coupled interaction may be useful and intentionally state-shaping. Its output may still report a pre-probe quantity accurately. The relevant question is which quantity was read, at what point in the interaction, and how the result is used.
Start with these recognition questions; ordinary prose is sufficient when it preserves their answers.
Use the fuller account below when the receiving decision or a substantive dispute needs distinctions missing from the recognition note. Reuse an already adequate account; publication or reuse alone does not require a new record.
Distinctions for a fuller decision-bearing account:
Activation boundary
Use this pattern while a possible state-changing interaction leaves an unresolved question about the interpretation of its output. A positive probe-coupled claim needs grounds for the relevant effect; suspicion alone does not establish it.
A meeting that changes attention may still accurately report a fact observed before discussion. A mutating API operation can return the old or the new value according to its interface semantics. Identify that timing and meaning before treating either output as a false passive read.
Performative-prediction analysis can explain behavior changed by a deployed prediction, score or metric. Use that account directly, with C.16 for measurement, A.10 for source recovery and bounded reliance, and B.3 for a named assurance claim. C.26.1 can connect the supported probe effect to the corrected use of its output without QL; use it only while that question remains unresolved.
Finish conditions
The pattern emits one of these results:
Probe-coupled context-cut worked use slice
The following constructed case shows a DDD workshop that changes the meaning and use of its own result.
Ask whether the bounded-context cut was only discovered or whether the interaction changed alignment, ownership, vocabulary, export validity, or viability. Examine the workshop, dashboard use, API extraction, bridge use, split, merge, or orchestration involved.
Boundary interaction under concern and operational sequence
The boundary interaction under concern is used as evidence, export, comparison input or architecture decision input. Identify the concrete operation, any instrument or channel it uses, and the output being interpreted.
The useful result is a supported reading of that output, a corrected use, or the unresolved premise. A redesign is a further result only when it is needed and justified.
Operational sequence:
- State the quantity or condition the receiving decision needs, its bearer and relevant time. Locate the output in the procedure, interface description, question sequence or available trace.
- Reconstruct the order of operations that can affect that quantity: preparation, sampling or questioning, state changes, output construction and later use. Read interface contracts, query definitions, timestamps, question wording or observed actions for those steps. Keep an unavailable step explicit rather than filling it with an assumed passive read.
- Trace a plausible effect from the operation to the reported quantity. For example, a query may populate a cache, publishing a score may change what people report, or discussion may change the agreement later summarized. Use existing mechanism or causal grounds;
C.28.CMcan develop an unresolved causal account. Temporal succession alone does not establish the effect. - Determine whether the output concerns the state before, during or after that effect. Compare the intended interpretation with the procedure and observations. An operation that changes later state does not automatically invalidate an earlier reading. Distinguish an actual change in the subject from a changed label, carrier or interpretation.
- Retain only effects, export losses, order, memory or frame conditions that change the current use. If material alternatives remain, seek the existing trace or procedural fact that distinguishes them. A new comparison or probe is warranted only when its attainable result could change the receiving action enough to justify its cost. If the needed distinction cannot be made, state the unresolved interpretation and its consequence.
- Return the supported reading and choose any needed continuation: qualify the output, redesign the probe or export, change the boundary arrangement, or use the subject method for the remaining question. A sufficient reading finishes here.
For example, suppose an API request measures a cold-cache response and then fills the cache. The first response time can still support a cold-cache claim; a later response time cannot be used as that same cold-cache observation without further grounds.
Recover the cache initialization, write point and measured interval from the procedure or trace. When those are unavailable, the unresolved point is which cache regime the timing represents.
Well-formed probe-coupled boundary state
A positive probe-coupled claim identifies the interaction, the relevant state or output meaning it changes, the grounds for that effect, and the consequence for the intended reading. Include the boundary, instrument, carrier, state history and confidence qualification when they carry that claim. A before/after reconstruction may be partial; do not turn missing information into an asserted change.
A useful result remains incomplete when the receiving use still depends on an unexplained output-producing step, an unsupported effect or a lost condition. State that specific gap. Evidence, command, measurement and semantic correspondence retain their distinct relations when more than one is used.
A minimal supported result may be:
- "this dashboard value evidences readiness behavior while the displayed score is used to allocate staffing";
- "the workshop changed agreement, so its closing note is evidence of the resulting alignment";
- "this API response contains the value sampled before the write."
Add a limit or further route only for a live unsupported use.
Probe, observable, output, and carrier split
Distinguish what is being read from the method used to read it, the resulting output, and the output carrier.
This split prevents a common mistake: "the dashboard says ready" hides at least four objects. The dashboard definition, the displayed result, the behavior it changes, and the readiness decision are distinct.
Reroute and disambiguation guide
Use this guide when a draft says that a boundary, system, team, or service is "coupled", "aligned", "interacting", "measured", "exported", "synchronized", or "read".
Use the direct relation pattern when the relation and participants are known, A.6.P when their meaning remains unclear, and F.18 when the settled relation needs a reusable name.
Positive examples and near misses
The positive examples establish probe-coupled interaction readings. Use those results without QL wording when the ordinary subject patterns carry the case. Retain QL only when the additional C.26 activation test establishes one named contextual-model obstruction that survives those patterns and changes the admissible inference or action.
Evidence posture for probe-coupled claims
Use the least-committing evidence posture that still supports the intended use.
Recognition or local working use may remain at QLP-0 or QLP-1 when its account is sufficient. Select additional support from the receiving conclusion under C.26:12b; publication, reuse, or formal wording alone does not require escalation or recollection of adequate evidence.
Archetypal Grounding
A team runs a service-boundary workshop after a series of incidents. The first map says “Payments is separate from Checkout.” During the meeting, participants agree that failure to update a customer's payment status will enter checkout incident triage. The signed decision records the new triage rule. The next week's incident log cites it when two such incidents are redirected.
The meeting record establishes the new agreement. The later log shows two applications. To decide whether earlier incidents of the same kind already went to checkout triage, examine earlier routing records for failures to update payment status. Payments can be a separate service while Checkout handles those incidents. Establish the service boundary from the relevant interfaces and responsibilities if the decision requires it. The records leave open whether a different question order would have changed the agreement.
Use the recorded rule for the handoff it governs. If another team will apply it, compare which incidents each team calls a failure to update payment status and whether a difference changes routing. Use F.9 for a cross-context correspondence when needed.
Bias-Annotation
This pattern biases authors toward examining how an interaction affects the interpretation of its output. That bias is useful when dashboards, workshops, surveys, and API reads are treated as if they carry state without changing it.
The pattern also biases authors against overusing QL. When no probe-coupled reading remains, the ordinary questions of intervention, message passing, bridge loss, API semantics, and work enactment stay with their existing FPF patterns.
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Consequences
This pattern corrects the interpretation used by a boundary decision. It turns "the workshop showed the split" into "the workshop both showed and changed the split-relevant state." It turns "the dashboard says ready" into "the dashboard is probe-coupled evidence with a limited decision use."
Dashboards, workshops, surveys, API reads and messages remain useful. A supported probe effect changes only the readings it affects; an unresolved effect limits the dependent conclusion.
Rationale
An output depends on the operation that produced it. When that operation also changes the represented state, the user must recover which state the output supports before using it in a decision. A supported ordinary account can establish that relation without QL.
Use this account while the interaction leaves a material uncertainty or error in the interpretation of its output. An already adequate subject account is sufficient.
SoTA-Echoing
Apply these source contributions:
- start from the ordinary FPF pattern before QL wording;
- show the concrete operation that produced the output;
- show the supported state update or export loss that changes the decision, or the unresolved premise;
- keep relation wording local; use
A.6.Pfor unresolved relation meaning andF.18for a reusable name only after the relation is settled under its direct pattern; - keep source-formalism language as modeling support, not as pattern-body ontology.
Relations
- Builds on:
C.26,A.6,A.6.B,A.6.P,F.9,A.15,C.16,A.10,B.3,C.25,A.1.1. - Coordinates with:
C.26.2when the interpreted output contributes evidence about how parts of a collective acted together;C.26.3when the boundary interaction changes a viability envelope. - The worked context-cut case in
C.26.1:4.3connects a workshop's effect to the corrected interpretation of its output.
C.26.1:End
Enacted Distributed State Evidence
Type: Architectural pattern Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when you need to establish how a collective system acted from what its parts did and how their actions were connected. A single incident record can show one team sending a particular item and another receiving it. Several reports of activity can leave that same handoff open. Recover the actions and relations the claim requires, wherever their evidence is recorded.
The pattern supports a bounded claim about the collective's current organization or behavior from the evidence available across its parts. Distinguish what those observations establish about the state from what explains its occurrence, and preserve uncertainty in either question.
Plain glosses:
collective bearer: the declared team, organization, service mesh, market slice, or otherU.Systemwhose coordinated work, behavior, or trace pattern is being read.coordinated work or behavior: role-work, service behavior, market participant traces, routine, commitment, artifact use, or coordinated action.distributed-state reading: a minimal evidence-bound interpretation of coordinated work or behavior.carrier: a log, trace record, dashboard, report, API response or other artifact that makes relevant work inspectable. A commitment, routine or change in work is distinguished from the record or observation used as evidence of it.faithful-enough export: a representation good enough for the current decision; when it is not faithful enough, state what was lost and why.
Problem
Teams often need to talk about latent alignment, readiness, market posture, service-mesh behavior, or an enacted "we already decided" before one explicit representation exists. Without a pattern, they oscillate between two bad options.
One option is unsupported collective-state language: "the organization knows", "the culture decided", "the market wants", or "the service mesh understands", used as if the wording established the state. The other option is false reduction: a survey answer, policy sentence, dashboard, or report is treated as the whole state.
Both fail. Coordinated behavior may evidence a working state, but it does not automatically identify a durable mind, internal representation, causal mechanism, or faithful-enough export for the intended use.
Forces
Solution
Model the claim as evidence-bound U.Episteme over a declared collective U.System. Distinguish the collective's distributed state from the surveys, dashboards, reports, and API responses used to examine it.
If the proposed bearer is a market slice or service mesh, declare whether it is a collective U.System, delivery system, trace population, or evidence set. Continue with EDSE only when a collective U.System bearer is established; otherwise retain any supported trace-population or evidence-set claim. Do not infer systemhood from coordinated-looking traces alone.
Ground the reading in observed coordinated work or behavior and its evidence. Distributed cognition, routine dynamics and team cognition supply the primary interpretive accounts. These can support an EDSE result without QL; add that lens only when the C.26 activation test establishes a residual contextual-model obstruction.
Recover the enacted pattern before asking how it arose:
- Locate the participants, actions and relevant times in the available observations or records. Distinguish an observed action from a report, plan or copied statement about it.
- Connect the actions through the relations that support the proposed collective reading: for example, work handed between teams, jointly held releases, coordinated routing or use of a shared artifact. Preserve gaps in that connection; simultaneous traces alone may support only parallel activity.
- State the narrow state claim those observations support for the declared bearer and window. A routine, command or shared stimulus may explain that same state rather than refute it. A claim about unspoken agreement, a new mechanism or durable culture needs the additional grounds that distinguish it from those explanations.
- Examine serious alternatives only where they change the current reading or its use. Return the supported claim, or name the missing observation or relation that prevents it. If an existing routine or subject account already answers the receiving question, use it directly.
This separates two questions: what the collective was doing, and why it did so. The account need not posit an unexplained coordination residue.
Start with these recognition questions. Ordinary prose is sufficient when it retains their answers.
Use the fuller EDSE record below when the receiving claim or action needs distinctions that the recognition note does not retain. Reuse adequate existing content; publication, reuse, or leaving the immediate discussion alone does not require a fuller record.
Full EDSE record:
Minimal claim principle
Preferred wording stays close to observed work:
- "During window W, the incident-response organization acted consistently with a rollback-readiness posture, evidenced by release timing, escalation criteria, and shared trace use."
- "The service mesh exhibited a coordinated throttling regime under policy P, evidenced by route changes and saturation traces."
- "Market traces support an expectation-shift claim under probe Q, with media amplification still a rival explanation."
Avoid wording that asserts a heavier claim, such as "the organization decided", "the market knows", or "the team has a distributed mind", unless another FPF pattern and evidence requirement independently support it.
Finish conditions
Finish with one of these results:
Rival explanation rule
Consider these alternatives when they change the proposed claim. Explaining observed coordination does not erase the coordinated state.
Export-loss discipline
The phrase "not faithfully exportable under current probe and bridge conditions" is supported only when the text says:
- what export was attempted;
- what would have counted as faithful enough for the current decision;
- what state, coordination, evidence path, timing, role alignment, option structure, survivor relation, or use limit was lost;
- whether the loss comes from bridge, measurement, representation, audience, timing, state change, or another cause;
- whether a more discriminating probe, bridge, representation, or time window could produce a higher-fidelity export.
Compound-state decomposition card
When the reading depends on relations among parts, recover that organization explicitly. The table helps distinguish local observations, their supported connection and any explanation or uncertainty that matters.
Evidence-bound state reading and claim floor
The evidence-bound state reading is a U.Episteme reading over coordinated work, behavior, or trace patterns by a declared collective U.System. The pattern does not govern an inner group entity, a culture substance, a market mind, a hidden service intelligence, or a reusable kernel state kind.
The useful state claim retains the collective bearer and boundary, the observed pattern, the evidence supporting it and its time window. Include persistence, probe or export conditions, serious alternatives and further subject work only when the receiving claim depends on them.
Read the timing condition from the decision. Some decisions combine separately completed checks. For an ordered handoff, follow the same item from completion by one part to receipt by the next. For joint readiness, identify the required instant or period. A single observation can establish the joint state at an instant. Observations made at different times can also suffice when other evidence establishes the necessary overlap. One verified interruption during a period of required continuous readiness defeats that claim. When every history compatible with the observations satisfies the specified timing condition, accept it. When every such history violates it, report failure. When some satisfy it and others do not, retain the local readings and leave joint readiness open. If the observations conflict so that no history fits them, resolve that conflict before using them together.
This pattern is useful because many work states are enacted before they are articulable. For example:
- A team may behave as if a release freeze exists before any single person states it cleanly.
- A service mesh may exhibit one routing posture before any one dashboard carries the whole situation.
- A market may shift expectations before any one survey faithfully exports the shift.
These enacted patterns can support bounded state claims. The carrier remains distinct from the state.
Operational evidence sequence
- Name the collective bearer as a declared
U.Systemand state its boundary; a bare social label does not establish that System. - Name the coordinated work, behavior, or trace pattern being read: actions, routines, commitments, role-work, service behavior, market participant traces, artifacts, or timing. For a particular dated Work claim, use the
A.15alignment route throughA.13and independentA.15.1admission; addF.6only when the receiving use needs precise assignment-bound attribution through the same obtaining assignment. - Identify the observations or records and what each supports. Use
A.10when a replayable source-to-use account is needed; the empirical claim always retains its required support. - State the time window and any serious alternative that changes the reading. Add persistence support, decay/refresh conditions and reprobe cost only when they affect the claim or receiving action, following :4.11. When considering another observation, weigh what it could add against its effects on participants and on the evidence. For example, a quick follow-up in a small team could reveal who gave a response or change what participants report next.
- Connect the observed actions under the declared bearer and formulate the smallest supported collective-state claim. Keep its causal explanation distinct; a supported routine or command may explain that state.
- When an attempted export carries the claim, state its fidelity criterion and loss. Otherwise retain the supported local reading.
- State the supported use and any material limitation. Name another subject question only when it remains needed.
- Add
B.3assurance only when a named assurance claim is current;QLP-3requires that claim and aB.3assurance result for the target claim and receiving use. Consequence level, audit, release, or accountability use may also impose a direct domain requirement for that claim.
Return the supported distributed-state reading and its necessary qualifications, or the gap that prevents it. A useful result may preserve only observations or a narrower trace-population claim when collective-system identity or coordinated organization remains unestablished.
For an adequately supported positive case:
During incident window W, the incident-response organization acted consistently with rollback-readiness posture P, evidenced by deployment queue changes, escalation messages, rollback artifacts, and support-routing changes; this reading is not faithfully exported by survey S because S loses timing, role-work, and trace linkage.
That sentence is narrower than "the organization decided". It supports adjusting incident communication and release-triage posture during window W; it does not support a durable culture claim or release-readiness assurance claim.
Well-formed EDSE record
When a record is useful, the following shape can carry the account. Optional fields are included only for a claim or action that consumes them.
The syntax is illustrative. A decision-use record retains the distinctions on which its supported claim or action depends. For an explicitly momentary, local use, follow :4.11 and omit persistence, decay, or reprobe-cost detail when those values cannot change the claim or action.
Well-formedness constraints:
collectiveBeareris a declared collective system, not a metaphorical subject.evidenceCarriersmake the relevant work inspectable through observations, traces, logs, commitment records or other work-result records; the claimed occurrence is not identified with its record.timeWindowbounds the claim; persistence beyond that window needs its own support.ordinaryRivals, when needed, contain serious alternatives that change the current reading.minimalSupportedClaimstates the collective pattern established by the observations and relations. A causal explanation is an additional claim with its own grounds.unsupportedUse, when needed, identifies a live use that exceeds those grounds.
Carrier, probe, report, and state split
The pattern keeps four objects separate:
A survey operation probes; its response records carry outputs. A policy claim can explain an observed routine, while a publication of that policy carries its text. A dashboard can display evidence and act as an instrument in a behavior-changing probe. In each case, recover the operation, content and carrier used; none is automatically a faithful representation of the collective state.
Case bank and near misses
Evidence posture and confidence
EDSE claims become useful when the text says how much consequence the evidence can carry.
Publication or reuse under unchanged assumptions does not by itself raise the evidence posture. Select additional support from the receiving conclusion under C.26:12b.
For a momentary claim, name the bearer, observations or carriers, window and minimal supported reading. Add the temporal, alternative or export detail that changes the receiving action; local use, reuse or a fuller form alone does not create that need.
The ordinary EDSE claim is minimal but actionable. It says: "this coordinated work or behavior supports this bounded reading for this use."
Archetypal Grounding
Tell: In a constructed incident case, the independently identified incident-response System includes Release, Support and Operations teams. Between 14:00 and 14:30, their deployment logs show non-critical releases held, support-routing records show incident traffic redirected, and rollback artifacts are prepared. No written freeze decision is found for that interval. Later survey answers about whether "we had decided to freeze release" conflict.
Show, System side: the bearer is that incident-response System during the stated half-hour. Preparing rollback and changing support routing are observed work; logs, messages and artifact records make those actions inspectable. The observations are not replaced by the later survey.
The incident coordinator can report the three actions as consistent with freeze preparation during this half-hour. To report a coordinated response, use work records that connect them. For example, one incident entry could link the release hold to the request to redirect support traffic and to the rollback prepared for the held deployment. That link would support coordination even if the actions occurred in sequence. A claim that all three teams were ready together has a different test: a trustworthy joint observation can establish readiness at its recorded instant. To report readiness throughout the half-hour, establish how long it held.
An unwritten directive, playbook, shared alarm or dashboard effect remains a rival explanation; the case does not establish a separate causal mechanism. If a playbook fully explains the actions needed by the receiving use, retain that ordinary account.
If a next-day plan relies on the posture persisting, obtain the additional support for that later window. The half-hour reading supplies neither persistence nor release-readiness assurance.
Bias-Annotation
This pattern biases authors toward minimal evidence-bound claims and explicit evidence. This makes distributed-state language usable.
It also biases authors to keep primary grounding in distributed cognition, team cognition, organizational routines, socio-technical work, and evidence practice. The QL lens is secondary and additionally requires C.26's contextual-model obstruction test; ordinary export loss or changed probing alone is insufficient.
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Consequences
This pattern lets FPF discuss enacted collective states. It gives authors a way to use traces, routines, coordinated work, and export loss in one minimal claim.
The cost is that many claims become narrower. Minimal evidence-bound claims are often more useful than confidently asserted claims that lack support.
Rationale
Existing FPF patterns can carry parts of the support requirement, but no single ordinary pattern makes the combined minimal distributed-state claim easy to write. A.15 carries Work alignment, A.10 carries source-to-use evidence accounts, B.3 carries named assurance claims, F.9 carries cross-context semantic Bridge loss, and C.16 carries formal measurement. C.26.2 uses those contributions to establish what the actions of the parts and their connections show about the collective system. A report that already establishes the needed connection can supply that evidence directly.
SoTA-Echoing
Apply these source contributions:
- ground the claim in coordinated work before QL vocabulary appears;
- state the evidence carriers and time window before stating the state reading;
- examine alternatives where they change the state claim or receiving use;
- treat survey/report/dashboard outputs as carriers or as instruments of a stated probe, not as the state;
- escalate to measurement, evidence, assurance, or authority patterns when the use requires measurement, evidence, assurance, or authority support.
Relations
- Builds on:
C.26,A.15,A.10,B.3,F.9,C.16,E.17.EFP,C.11. - Coordinates with:
C.26.1when the probe changes the state being evidenced and its output is being used as a passive read;C.26.3when the coordinated state is part of viability-envelope regulation. - The distributed-state reading concerns the declared collective bearer and remains bounded by its supporting observations and time window.
C.26.2:End
Viability-Envelope Boundary Regulation
Type: Architectural pattern Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when protecting a function across boundaries depends on several conditions holding together under disturbance. Start with the work the recipient needs kept possible. Its requirements determine what state to assess, what a proposed method must preserve, and what the people running it will need to observe and do.
What goes wrong if missed. The team treats one dashboard value, stability slogan, or local metric as viability, while another envelope variable, intervention cost, boundary condition, or failure mode is already breaking the protected promise or function.
What this buys. The practitioner can derive the envelope from the work or promise being protected, locate the basis of its limits, and judge a disturbance or proposed change against all relevant constraints. A supported diagnosis can finish without a remedy; missing limits or effect grounds leave a bounded open question.
Use C.26.3 for the general envelope-regulation claim when several characteristics must remain inside a viable region under a disturbance and a candidate intervention, boundary condition, adaptation cost, or failure mode matters. Continue to use the direct control, quality, SRE, causal, measurement, boundary, and work patterns for the objects and claims they define; C.26.3 governs the resulting envelope claim.
QL is an optional coordination branch. Use C.26 and its QL vocabulary only when a probe, frame, export, coarsening, order, incompatible representation, or measurement-changing-state issue remains after the ordinary patterns have carried their part and leaves a specific contextual-model obstruction that changes what may be inferred or done. FEP, allostasis, and active inference remain source analogies rather than a second entry condition.
Plain glosses:
-
viability bearer: a local lens position, not a kind or relation.- If the bearer is a System, cite that System's A.1 identity.
- If it is a selected organization of systems, system-role kinds, and assignment occurrences, identify one A.22
U.Structurefrom independently identified constituents, selected obtaining relation occurrences, constraints as applied, and one named selection-use frame. Kind declarations or assignment occurrences listed together do not identify a Structure. - A population or market slice instead needs a declared domain and effective reference scheme, membership or scope, and identity basis.
If no branch supplies one object, stop.
-
protected promise / function: the separately governedU.PromiseContent, stakeholder-value claim, function claim, operating-regime claim, commitment payload, or delivery promise whose continued satisfaction or realization the regulation decision is meant to protect. It is not a slot or part of the object in the local viability-bearer position. -
serviceor market wording: the wording does not itself identify the viability bearer. Apply the A.1 System branch, the four-discriminator A.22 Structure branch above, or the population/market-slice branch, as applicable. Keep claims about promise content, access points, assignments, commitments, Work occurrences, evidence, and direct relations separate. If no branch identifies one object, stop; do not turn the phrase or a list of role kinds and assignments into a bearer kind, situation kind, or bundle. -
viability envelope: the region of declared characteristic values within which the object remains inside the viability bounds stated for the current protected claim or use. -
envelope variable: one characteristic that must stay within bounds, such as latency, reliability, support load, compliance exposure, safety margin, energy, or operator attention. -
actuator/candidate intervention: actuator is a control-theory or source label, not an FPF kind and not a synonym for Work. Use candidate intervention only as a local prompt until its proposal-side object is recovered:- a Method;
- a
U.MethodDescriptionor policy episteme; - a proposed setting change;
- a
U.WorkPlan; - an access or permission claim;
- a Bridge proposal or description.
Separately identify any dated
U.Work, independently groundedU.Transformation, obtaining relation occurrence, or resulting state claimed to exist. Keep these objects distinct. -
allostasis: preserving function through separately governed changes to settings, environment relations, boundary conditions, or operating regime when circumstances change.
Problem
Teams often collapse viability into one dashboard value or fixed target. They optimize latency and increase operator load. They improve availability and increase compliance exposure. They preserve one metric while exhausting the team, hiding risk, or making recovery slower.
A second failure is passive sensing. A metric, probe, dashboard, alert, or health check is treated as a neutral window into viability, even when its use or publication changes behavior through separately grounded Work, interaction, or governance, or hides unmeasured dimensions.
A third failure is static stability. Teams say "keep the system stable" as if stability always means holding one internal variable fixed. In architecture work, preserving viability may require a candidate intervention.
Recover whether the proposal concerns a Method or description, setting proposal, WorkPlan, access or permission claim, or Bridge proposal or description. Separately identify any dated Work, actual change, obtaining relation occurrence, or resulting state claimed to exist; words such as caching, staffing, routing, protocol, and measurement design do not choose among them.
Forces
Solution
Use C.26.3 when the work must regulate a multi-characteristic viability envelope under disturbance. Use C.25, U.Dynamics, measurement, boundary, causal, and work patterns to state the qualities, changes, observations, relations, and Work on which the envelope claim relies. Add C.26 / QL only under the residual-obstruction condition in §1. If no such obstruction remains, omit the QL fields and checks and retain an otherwise useful envelope-regulation result.
Start with these recognition questions; ordinary prose can supply the answers.
Use the fuller account below when the receiving viability decision needs distinctions missing from the recognition note. Reuse adequate existing content. A supported diagnosis can finish without an intervention proposal.
Full envelope-regulation record:
Homeostasis and allostasis reading
Homeostasis means keeping a parameter or bundle inside viable bounds. Allostasis means preserving functioning through separately governed changes to internal settings, external relations, boundary conditions, or operating regime when circumstances change.
Use these terms when their distinction changes the interpretation of the viability decision.
Finish conditions
Finish with one of these results:
Metric-induced distortion
Name sensors, probes, dashboards, alerts, metrics, disturbances, and candidate interventions in the envelope claim when they matter. They participate only in world-side relations defined by their direct patterns; C.26.3 does not infer a generic viability relation from their appearance in the same card. A probe or dashboard may still affect behavior, but that effect needs its own grounded claim.
Conditional dynamics detail
When rate, acceleration, second-order change, inertia, damping, resistance, effort, or the strength and latency of a recovered intervention or resulting actual change is load-bearing, state:
- what rate or acceleration matters;
- what slows or speeds the change;
- whether the rate of change itself is changing, rebounding, overshooting, or damping out;
- which inertia is useful and which is harmful;
- which recovered intervention object is proposed, what Work, relation, or setting change can lawfully realize it, and which independently grounded actual change affects the envelope fast enough;
- which evidence shows the dynamic state.
If those variables are not load-bearing, the short recognition note or the full envelope-regulation record is enough.
Claim identity and operational sequence
Identify an envelope account under C.2.1 when its claim-bearing identity is used: its EntityOfConcern is the viability bearer, its claim content states the envelope, and its effective ReferenceScheme determines how that content is read. A record is one possible presentation.
Keep the object of a proposal clear. A setting or policy proposal may be part of that account's claims; mentioning it does not require another episteme. A separately identified proposal, U.WorkPlan, Method, description or relation retains the identity supplied by its direct pattern. In particular, a modal F.9 proposal designates possible endpoints and a profile; it does not establish an obtaining Bridge.
The first useful move is to replace a one-scalar stability account with an inspectable envelope-regulation claim for the decision.
Envelope-regulation sequence:
-
Starting from the work or result to be protected, follow a possible failure to the participants and connections that can affect it. Identify which object the viability claim concerns.
- For a System, cite its A.1 identity.
- For selected organization, cite one A.22
U.Structurethrough independently identified constituents, selected obtaining relation occurrences, constraints as applied, and one named selection-use frame; a role-kind/assignment list is insufficient. - For a population or market slice, state its declared domain and effective reference scheme, membership or scope, and identity basis.
Then name the separately defined promise or function being preserved. If no branch identifies the bearer, stop.
-
Recover the protected work or result from the receiving requirement, promise, operating procedure or current failure. Follow what must remain possible: successful outputs, necessary inputs, supporting work, resources and timing. For each relevant way it can fail, identify the quantity or condition that distinguishes continued operation from that failure. Keep the subject of each constraint explicit; support-team load is not automatically a characteristic of the service System.
-
Obtain each bound from its governing basis. Reuse an applicable customer requirement, commitment, safety or operating limit, or an established resource constraint. Where the constraint is capacity, relate the demand in the stated window to the people, equipment, time or other provision actually available under that same workload and service condition.
A historical average describes performance; it does not set acceptable performance. If a needed limit is absent, recover it with the party entitled to set the requirement or make the trade-off. Until then, a proposed threshold remains a scenario assumption: return the effects or feasible range and identify the unresolved acceptability question.
-
Identify the disturbance or regime change and the observations needed to locate the current state against those bounds. Use existing observations where adequate. Recover what each sensor or probe measures, including timing, coverage and a supported behavior-changing effect; use
C.26.1for a material unresolved interpretation. Match the observation to the time covered by each bound. An hour's total can settle a limit on that total. One verified breach in a five-minute interval defeats a bound required in every interval. To confirm compliance throughout, cover all the intervals by observations or by a justified bound on the unobserved ones. -
If regulation is needed, trace the failure to an available change in the mechanism, resource, operating rule or boundary condition. Reuse a suitable domain Method. When the dependency is unresolved,
C.28.CMdevelops the causal account;C.28.MRderives consequences of a specified mechanism change, andC.29contributes when formulation or transfer needs mathematical work. Recover the proposal's object and distinguish it from any performed Work, actual transformation or obtaining relation. An F.9 endpoint/profile change requires testing the resulting Bridge candidate; Work does not revise a fixed Bridge occurrence. -
Follow each proposed change to every affected constraint, including displaced work, transition time and adaptation cost. Use a calculation, simulation, case evidence or other justified account appropriate to the decision. Keep its assumptions and uncertainty visible. Improving one variable is sufficient only if the remaining required conditions also hold.
-
Return the supported diagnosis or conditional proposal. A known violation can rule out restoration of the envelope even while another effect remains unknown. A positive viability claim needs all its required conditions supported; an unresolved limit or effect may instead warrant a restricted action, further inquiry or no change. New inquiry is conditional on its attainable contribution to the receiving decision, not a required extra step.
-
Add rate, inertia, damping, latency, resistance, acceleration or a reprobe condition only when it changes that result.
Ordinary output: explain the viability bearer, variables and bounds, disturbance, observations and failure condition. When regulation is proposed, add the identified proposal, trade-off and adaptation cost. A record is useful when its layout helps recover those distinctions; ordinary prose can preserve them.
For a diagnosis, return the supported state and any limit on that conclusion. To design a way to regulate the service, work back from a threatened condition to a change within the responsible performer's reach. Establish which observations can arrive before action is needed. For a guarantee, determine whether an available response protects the result across the situations still indistinguishable at that time. If they require different responses, find a timely distinguishing observation or another way to protect the result. For a risk or frequency target, assess the response under the conditions and tolerance that target specifies. A qualified existing operating method may already supply the answer. For an operating assignment, use the resulting method in the work, observe the protected result and respond to departures while the obligation lasts. The proposal, its actual performance and the service obtained each have their own grounds.
Viability envelope record
When a record is useful, it can present the envelope claim and its grounds. Retain only the fields consumed by that claim or receiving action, while keeping its subject and any needed episteme identity recoverable.
Changing the ClaimGraph content changes the claim and therefore identifies another episteme under C.2.1. A new layout, publication occurrence, form, or carrier can leave the episteme unchanged. New evidence can change the support for the claim without changing that claim; if the asserted ClaimGraph changes, the result is another episteme.
Well-formedness constraints:
- the local viability-bearer position points to one object and introduces no kind or relation: a System has its A.1 identity; an A.22
U.Structurehas independently identified constituents, selected obtaining relation occurrences, constraints as applied, and one named selection-use frame; a population or market slice has its declared domain and effective reference scheme, membership or scope, and identity basis; a list of system-role kinds and assignments satisfies none of these branches; - service or access wording names each current object and relation separately—the viability bearer, promise content, system, system-role assignment, commitment, Work occurrence, evidence, or another direct relation—through the pattern that defines that object or relation; the wording creates neither a root bearer nor a bundle;
- this pattern's multi-characteristic branch keeps every dimension needed by the protected result; if one characteristic already answers the question, use the adequate single-characteristic account;
- at least one candidate intervention is named when the text proposes regulation rather than only diagnosis, and its proposal-side Method, description, setting proposal, WorkPlan, access or permission claim, or Bridge proposal or description is recovered under the subject pattern; any dated Work, actual transformation, obtaining relation occurrence, or resulting state is identified separately;
- authority and latency are stated only for an object to which they apply; keep descriptions, Methods, plans, setting labels, Bridge descriptions, and resulting states distinct from actors and Work;
- when regulation or adaptation is proposed, its adaptation cost is named, because allostasis hides cost when phrased as "stability through change";
- the failure mode is named to show what would count as loss of viability.
Sensor, probe, candidate-intervention, and metric split
Do not let one dashboard value stand for the whole envelope.
A metric value or dashboard carrier is neither Work nor an actual change. Its use, publication, or a surrounding governance routine may participate in a separately grounded behavior-changing claim. When Work is asserted, recover each actual performer's A.13 core and independently admit the dated occurrence under A.15.1. If the claim must also identify the assignment under which the Work was performed, name that assignment and check the relation separately through F.6. Name any changed setting, actual transformation, access or permission relation, or boundary relation separately. Repairing one envelope variable may still damage another.
Homeostasis, allostasis, and architecture work
Homeostatic wording is useful when the claim concerns keeping a variable or bundle inside a stable range. Allostatic wording is useful when preserving the named function requires one or more separately governed setting, boundary, environment, access, staffing, routing, protocol, cache-policy, or operating-regime changes. The wording does not decide whether each item is a Method, description, plan, Work, transformation, relation, or resulting state.
Use the minimal reading that carries the case:
Use allostasis only when stability through change is the useful architecture reading.
Case bank and near misses
Source-to-pattern translation
Allostasis, active inference, FEP, Markov blankets, and computational-boundary sources are useful here only after translation into FPF architecture terms:
Use the translated source contribution to explain an envelope constraint or choose the next needed operation—for example, compare a probe redesign or a proposed setting change. The operational sequence and direct subject patterns govern the resulting proposal, observation or actual change; a source analogy supplies none of their missing grounds.
Archetypal Grounding
In this constructed case, a platform team must preserve checkout service during a traffic spike. The receiving service requirement specifies both timely responses and sufficiently current payment-failure status. The team traces a proposed increase in cache duration: serving more requests from cache may reduce delay, while retaining old status longer can mislead customers and increase support contacts.
CheckoutSystem-1 is independently recognized under A.1 as the deployed System. CheckoutPromiseContent-1 separately states the checkout outcome and reliability on which customers may rely. For this example, the receiving requirement sets the latency limit at 250 ms for the highest five-minute p95 in the incident hour and the stale-status limit at 1% of statuses in that hour. The support limit is derived from the provision reserved for this incident: four available specialists can each handle up to five comparable contacts within the hour under the stipulated workload, giving reserved capacity for 20. These are constructed requirement and capacity premises, not limits inferred from the observed averages.
Compare two matched one-hour windows under the stipulated spike condition, before and after the cache-setting change:
In this constructed case the highest five-minute p95 falls from 350 to 180 ms. In the second hour the status and support bounds are both exceeded. That hour therefore fails the joint requirement.
The decision combines properties of CheckoutSystem-1 with demand on the incident-support team. The support figure counts contacts requiring handling; a count of completed handling would concern actual support Work. Operator attention, if added, would concern the people handling those contacts. Fulfillment of CheckoutPromiseContent-1 still needs its own evaluation.
Suppose the acceptable stale-status bound is missing. Demand of 45 contacts already exceeds capacity for 20, so the joint requirement failed in this hour. Keep the observed stale-status rate of 3%. Recover the freshness requirement if deciding whether that rate also fails or which remedy to pursue.
A dashboard query is an observation operation unless a behavior-changing effect is separately established.
For this hour the team can report that service fell outside the required bounds. To decide whether changing the cache setting caused the difference or would restore service in another spike, examine how the setting affected reads and updates and what else differed between the two hours. That account can guide comparison of a shorter cache duration and a bypass for payment-failure updates, including their effects on upstream work and latency.
Recover the proposal's object, adaptation cost, applicable authority and realization latency before choosing it; a planned enactment belongs in a WorkPlan. Missing intervention evidence leaves the diagnosis useful.
If support capacity is independently revised to 50 comparable contacts per hour under the same conditions, the same demand of 45 contacts satisfies that revised constraint, but the 3% stale-status result still exceeds the retained 1% requirement.
Suppose the team is also asked to operate checkout through the next spike. An alert calculated at the end of the incident hour comes too late to prevent stale replies in that hour. Work back from those replies to the reads that can return an old payment status. The team can consider bypassing the cache for those reads while keeping it for other requests. To use this rule, establish that these reads can be recognized before the response and that the source reached by the bypass is current often enough to meet the 1% stale-status limit for the service. Check that latency and support demand also meet their bounds over the load conditions the service promises to cover. Once those conditions are established, the authorized operator can enable the rule in time, check the three bounds in use and respond when conditions change. A claim about the service actually delivered during the spike relies on observations or bounds covering that period.
Bias-Annotation
This pattern emphasizes constraints that a favorable scalar can hide, including damage to another envelope dimension.
It also biases authors toward actionable architecture work. The pattern asks which direct object a boundary, access, protocol, staffing, cache, throttle, bridge, or measurement proposal denotes and how quickly its separately governed effects can matter. For a particular dated Work claim, establish each actual performer's A.13 core and independently admit the dated occurrence under A.15.1. Add F.6 attribution only if the account must also identify the assignment under which that Work was performed. Any actual transformation is grounded separately.
Use C.25 alone when one quality bundle or metric can be handled without envelope, disturbance, boundary condition, recovered candidate intervention, adaptation cost, or viability failure mode.
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Markov-blanket routes:
Consequences
The envelope account makes trade-offs visible: preserving one characteristic can violate a constraint supplied by another claim. It supports a diagnosis, or comparison of independently identified proposals, while the direct subject patterns continue to govern the proposed objects and any actual changes.
An envelope account requires more detail than an account based on one metric. That detail is warranted when the metric hides another envelope dimension, an intervention cost, a boundary condition, or a failure mode.
Rationale
Ordinary quality-bundle work does not always show boundary conditions, candidate interventions and their recovered direct objects, disturbances, adaptation cost, and failure modes together. C.26.3 coordinates those elements while preserving ordinary FPF patterns.
The QL lens is secondary. It matters when the way viability is probed, exported, or coarsened leaves a contextual-model obstruction that changes the state reading or admissible use of the representation.
SoTA-Echoing
Apply these source contributions:
- state the envelope before importing source terminology;
- use a source analogy only through the subject claim and relations it supports;
- keep sensors, probes, metrics, candidate interventions, and every recovered direct object distinct;
- state adaptation cost and failure mode;
- apply ordinary quality and measurement patterns to one-scalar quality concerns.
Relations
C.27 temporal-claim relation.
- C.27 may flag: braking, throttling, cadence, recovery, or stabilization moves in claims such as slow rollout protecting support capacity, request throttling preventing collapse, or cadence change preserving attention/team health.
- This pattern keeps: viability bearer, protected promise/function, viable region, disturbance, sensor/probe/candidate-intervention/direct-object split, adaptation cost, and failure mode.
- Non-admissible use: stabilization wording is not a viability envelope, and C.27 is not the pattern for all stability-through-change claims.
- Exit: if the claim being made is only better quality, healthier team, or more resilient service without a declared viability envelope, use C.25, E.13, or the relevant quality/proxy/value pattern rather than C.26.3 or a C.27 profile.
General relations.
- Builds on:
C.26,C.25,U.Dynamics,A.6,A.15,C.16,A.10,B.3,A.3,A.19,C.18,C.19. - Coordinates with:
C.26.1when a state-changing probe's output is used as a passive read for the decision;C.26.2when the observations of parts and their connections must support a claim about the collective. - Does not replace: ordinary quality-bundle patterns, generic control theory, full FEP doctrine, or biological homeostasis claims outside FPF bridge and loss discipline.
- Name:
Viability-Envelope Boundary Regulationnames architecture work over a viability envelope and boundary/action conditions.
C.26.3:End
Temporal Claim Adequacy: State Readings, Temporal Trends, and Intervention-Sensitive Change
Type: Claim-adequacy pattern Status: Stable Normativity: Normative unless marked informative
Use This When
Use this pattern when you need to establish what a claim about speed, rhythm, throughput, recovery, convergence, rollout, adoption, braking, coasting, redirection or stabilization supports. The assessment can answer a request to explain a temporal pattern or inform what someone will do.
The practical question is:
Are we only reading a state, only reading a rate, or claiming that an intervention changes a rate, rhythm, recovery, or regime?
If the claim is only a state or rate reading and that answers the receiving question, stop. For an intervention-sensitive claim, recover the proposed input, relevant time, resistance or cost, and grounds for the predicted change. Return the supported use and any material limitation; add a reopening condition when later reliance needs it.
What this buys. The practitioner distinguishes a trend from an intervention model. Braking, pausing, stabilizing, redirecting, coasting, widening, narrowing, or slowing rollout remain available when acceleration would be the wrong move.
Primary working object. The temporal claim whose adequacy matters to the conclusion, explanation or action the recipient needs. Recover its content and subject from the available plan, report, model, observation or other account. An exact C.2.1 episteme or ClaimAddress is needed when the receiving use relies on that independently referable identity. Here ClaimAddress means exact edition plus intrinsic claim identity declared by that edition's ClaimGraph.
When an adequacy assertion is independently identified under C.2.1, its EntityOfConcern is the source episteme or addressed claim, not the reference used to designate it. An ordinary assessment can state its conclusion directly. Publication under E.24.PUB is separate from that claim and identity.
First useful move. Recover what is claimed to change over time, for which subject and conditions. Use C.27.TA when its temporal meaning needs clarification, then classify the claim:
Dyn0, Dyn1, and Dyn2 classify authored claims. They do not classify Systems, teams, services, Methods, or practices as kinds of things.
Not this pattern when.
- The temporal wording is ordinary explanation and changes no practical use.
- The result needed is only a positive temporal-aspect claim. Use C.27.TA.
- The question is measurement construction or comparability. Use C.16.
- The question is a transition law, simulation, prediction, or control model. Use A.3.3.
- The question is a bounded transformation. Use A.3.4.
- The question is planned or performed Work. Use A.15.2, A.15.1, and F.6 as applicable.
- The question is causal use, benchmark parity, a promise, assurance, value, harm, viability, scaling, adaptation, search health, publication, or residual QL. Use the direct pattern for that question; keep C.27 only if a separate temporal-claim adequacy question remains.
- The reader has only a cue such as braking difficulty, rhythm mismatch, demand accumulation, divergent event traces, or more activity without better results. Keep the cue through A.16, A.16.1, B.4.1, or B.5.2.0 until a temporal claim can be stated.
Problem
The recurring failure is:
An author measures or names a rate and then assumes that this is enough to know how to change that rate.
Typical consequences include:
- a past slope is used as a future control law;
- visible throughput rises while hidden queues, rework, quality loss, service demand, or burnout worsen;
- a local rate change is projected across scale without an aggregation relation or bearer continuity;
- rhythm is used without an identified bearer, temporal reference, window, or supported use;
- a planning assumption substitutes for the grounds needed for causal proof, benchmark superiority, a release gate, a public promise, or an assurance claim;
- more reviewers, data, tokens, tool calls, model capacity, or parallelism is assumed to produce linear improvement;
- slowing, braking, pausing, or recovery is treated as failure merely because speed is the implicit value;
- probe, frame, token, dashboard, or active-sensing language activates QL before ordinary patterns have carried their questions.
C.27 repairs one authored-claim failure. It is not a dynamics model, a measurement calculus, or a physics analogy promoted into FPF ontology.
Faster is not automatically better. A temporally adequate claim is not automatically valuable, safe, legal, ethical, feasible, promised, or assured. Direct value, quality, harm, promise, legal, safety, and assurance patterns carry those claims.
Forces
- Cheap recognition versus costly overformalization. Keep the state, rate and intervention-sensitive distinction easy to recover. Use C.27 when a temporal distinction can change what the recipient may conclude, explain or do.
- Effort profiles versus derivative words. Interventions can be impulses, scheduled pushes, feedback policies, adaptive regimes, brakes, pauses, coasting periods, or redirections. The useful question is what input occurs over which window.
- Domain-specific resistance. Lag, queue pressure, habit, coordination cost, technical debt, service demand, physical inertia, or another domain-local resistance can matter. None becomes a new mass or force kind.
- Temporal meaning and coupling. Preserve the temporal claim's bearer, predicate and timing under C.27.TA; add an exact separate reference when needed. Cross-bearer phase, synchronization, coupling or entrainment appears only when the receiving use relies on that relation.
- Readable practice versus hidden claim inflation. Ordinary actor and practice wording should stay readable. Exact System, assignment, authority, Work, causal, benchmark, promise, and assurance details appear only when the current claim relies on them.
- Preservation versus copied neighboring schemas. Rare uses need trigger questions, but C.27 should cite the result of C.16, C.28, G.9, C.26.3, or another direct pattern instead of reproducing its schema.
- Value neutrality versus acceleration bias. C.27 must allow acceleration, deceleration, braking, redirecting, coasting, pausing, stabilizing, recovering, sustaining, widening, and narrowing without deciding which is valuable.
Solution
Establish the temporal conclusion the recipient needs. An explanation of the observed course can be complete even when the same course of action remains appropriate. A recommendation to change that course also needs grounds for the expected effect.
- Recover the temporal claim from the available account, keeping its subject, quantity or condition and relevant time explicit. Cite an exact episteme or ClaimAddress when that identity is consumed by the use.
- Resolve what the timing means: a dated observation, elapsed duration, order, recurring cadence, validity condition or another temporal property. Use
C.27.TAif this is unclear; a separate temporal-aspect artifact is unnecessary when the claim already states it adequately. - Classify the claim as Dyn0, Dyn1 or Dyn2. Compare the decision with what the observations cover. Two dated counts of the same stock under the same counting rule can establish its net change. If duplicate entries were removed between counts, establish how much of the recorded change came from the correction before crediting it to completed work. The same endpoints can follow a steady decline or a rise followed by a sharper fall. To judge the path between them, use records or a supported model of the intervening period. To attribute a change to an action, compare what the relevant quantity would do under that action and under its alternative.
- If it is Dyn0 or Dyn1 and no intervention-sensitive question remains, stop or use C.16 for the unresolved measurement question.
- For Dyn2, connect the proposed input to the claimed temporal effect. Recover the baseline, relevant window and resistance or cost; identify whether the basis is observation, a planning assumption, a model consequence or a causal result. Reuse adequate grounds. When the effect is the unresolved causal question, use C.28 and its construction or mechanism-replacement method as needed.
- Return the supported conclusion and only the neighboring result on which it depends. Unknown resistance or another missing premise can justify a narrower or unresolved result.
- Add more detail only when it can resolve a remaining receiving question; an incomplete form is not itself a reason to open a larger profile.
A Sufficient Dyn2 Account
A compact account states the input, proposed temporal change, relevant window, resistance or cost and grounds. Its length follows the explanation needed by the receiving use.
For example: "Adding two reviewers is expected to increase completed reviews during the next two weeks. That forecast assumes sufficient independent review capacity; the shared approval bottleneck remains unmeasured, so doubling throughput is not yet supported."
The optional record below is useful when separately identified claims must be referenced or reviewed:
When an independently identified C.27 adequacy episteme uses this record, its EntityOfConcern is the source episteme or addressed claim denoted by sourceTemporalClaimRef.
positiveTemporalAspectClaimRefidentifies the temporal claim used by that reference-bearing account. Ordinary prose may instead state the temporal aspect directly, keeping its meaning under C.27.TA.- move names the claimed temporal change, for example accelerate, brake, coast, recover, stabilize, widen, narrow, or a domain-local move.
- claimedInterventionOrInput says what is claimed to affect the temporal behavior. It does not assert performed Work, authority, capability, or causal effect.
- interventionWindow is omitted when the C.27.TA window is enough. Add it only when the input occurs over a different interval.
- resistanceOrCost names the relevant lag, constraint, stored work, queue pressure, coordination cost, residue, or another domain-local obstacle. Unknown is an acceptable local answer.
- reasonForReading names one exact evidence relation, measurement result, model assumption, planning assumption, diagnostic judgement, or direct neighboring result. These alternatives do not become one generic evidence kind.
- supportedUse states the result's practical reach; unsupportedUse names a live overreach only when it needs to be excluded explicitly.
- reopenTrigger is included when continued reliance needs a condition for revision, new evidence or another direct result.
One local window may stand for claim, sampling, intervention, rhythm, and validity when the difference changes neither the claim nor the receiving action. Split them when, for example, evidence is sampled over a different interval, input precedes the observed outcome, comparison needs a baseline, follow-up occurs later, or validity expires sooner than the trace.
Unknown resistance can support a local diagnosis or planning discussion. It cannot support durable acceleration, causal, benchmark, promise-like, gate, or assurance use without the direct evidence or assumption boundary required for that use.
Readable Actor and Intervention Boundary
Ordinary practitioner prose may say, for example, “the engineer slowed the rollout” when it recognizably names the System acting in the situation.
If the receiving claim relies on performed Work, identify the actual System actor, recover its A.13 core, and independently admit the dated Work under A.15.1. Add F.6 afterward only when the temporal claim also needs precise assignment-bound attribution.
If the claim relies on a local system-role kind, System classification, or assignment, add each distinction separately. An assignment does not act and does not supply authority; cite its directly declared relation species and exact obtaining occurrence while still naming the holder System.
For a non-system input, name its actual direct relation to the temporal behavior, or keep it as an unresolved or source-side intervention claim. Keep authority, WorkPlan, capability, performed Work, and claimed effect separate.
Rhythm, Coasting, and Reversibility
Observed cadence can remain Dyn1. A rhythm claim becomes Dyn2 only when effort pattern, coordination, recovery, stabilization, or another intervention-sensitive use changes the supported action.
For coasting, ask what continues after effort changes or stops, why it may continue, over which window, what use that reading supports, and what change reopens it. Possible bases include habit, automation, stored work, queue pressure, learned capability, commitment momentum, social norm, physical inertia, or unknown.
Keep coasting and debt distinct. Coasting describes continued movement or stability. Debt or hysteresis describes what remains and how costly reversal or recovery is. A claim may need both. Rework, service demand, quality loss, burnout, hidden queues, risk, or coordination cost can appear after acceleration. Reversibility may be unknown; that bounds use.
Boundary-Crossing Header
An adequate compact account can finish the work. When the receiving use depends on another pattern's result, include its contribution and an exact reference where that identity is required. This optional header can organize separately referable results:
When the profile is used to identify an adequacy episteme, its EntityOfConcern is the source episteme or addressed claim denoted by sourceTemporalClaimRef. The profile's references preserve the direct results it consumes; they are not substituted for those results.
Each activeNeighborResults entry names the result kind supplied by its direct pattern. Do not replace a C.16 measurement result, C.28 causal-use result, G.9 parity result, C.26.3 envelope-regulation claim, Work occurrence, evidence relation, or assurance claim with a generic C.27 field.
Archetypal Grounding
Backlog Reduction Planning
Source claim: “Adding two reviewers for two sprints will double the backlog-reduction rate.”
For this case, CheckoutReviewSystem-1 is already independently identified under A.1. The source claim is one exact claim about that System and its backlog measure. The planned staffing change is a WorkPlan input; no performed Work or causal effect is asserted.
Suppose the team needs at most 40 items left at the end of the second one-week sprint. It starts with 120 comparable items, has a baseline net reduction rate of 20 per week, and expects no new arrivals or reopened items. At 40 per week, exactly 40 would remain. This calculation settles whether the proposed rate would be enough. To select extra reviewers on the ground that they can achieve that rate, establish what they can complete and what the shared approval step can absorb. If every one of the 120 items still needs a new approval before it can leave, and the approver can pass only 20 per week, at most 40 can leave in the two weeks. Extra reviewers alone cannot bring the backlog down to 40. If 40 items have already been approved and await final review, that same approval limit permits 80 departures. Check which stages hold the starting backlog before deciding where to add capacity.
If the measured first-sprint reduction is 20, there is progress, but the uniform 40-per-week estimate is not met in that sprint. Reassess ramp-up and the remaining estimate before relying on the second-sprint forecast; the first sprint alone does not determine the two-sprint outcome. No boundary-crossing profile is needed for this local use.
Braking to Protect Viability
Source claim: “Slowing rollout for two weeks will keep CheckoutSystem-1 inside its usable envelope.”
The temporal claim states the reduced rollout cadence and two-week window. The C.26.3 account states the envelope and identifies CheckoutSystem-1 through its A.1 System identity. This example uses separate exact references because the temporal assessment consumes both claims.
Knowing the bounds identifies what to check. To select the slower cadence on the ground that it will preserve viability, establish how it affects the constrained quantities over the two weeks, including effects that could defeat the proposal. A C.26.3 account that already supports that intervention consequence can supply the basis. Where the effect remains uncertain, the team can compare scenarios or consider a bounded trial under its own risk and authority conditions.
Practice Rhythm
Source claim: “Daily twenty-minute drills stabilize the learner's task rhythm.”
The plan for LearnerSystem-1 prescribes one twenty-minute drill daily for four weeks. The instructor wants to know whether the learner starts comparable work sessions at more regular intervals. The drill schedule and the observed starts of those work sessions concern different activities.
To judge regularity, specify which sessions count, how start-to-start intervals will be compared, and what timing matters to the work. The same number of starts per week can come at evenly spaced times or in a cluster. Comparable records before and during the scheduled period can establish a change in the observed spacing even if every drill was cancelled. To assess the effect of doing the drills, establish which drills took place and for how long. Compare the observed rhythm with what the learner’s work schedule and other conditions support in the absence of those drills. A new duty roster could itself fix session starts. If all that is needed today is the drill plan, its prescribed cadence suffices. For replication, state what timing and effort the learner has to reproduce.
Compact Case Bank
Optional Boundary-Trigger Reference
Skip this section for ordinary local diagnosis and planning. It is a trigger-and-destination map. The examples in each row are representative rather than exhaustive.
Pattern-Use Notes
- A local resistance value of unknown is allowed. It blocks stronger use without the direct evidence or assumption boundary required for that use.
- A historical trend does not supply a control horizon, update rule, constraints, or stability.
- Evidence from policy A does not carry policy B merely because both policies concern the same rate.
- Equal final scores do not erase unequal adaptation windows, effort, rework, validity, or recovery.
- Scalar throughput does not describe a whole multi-object work cycle when interactions and aggregation change the result.
- Metric improvement does not establish System improvement.
- A temporal metric does not become value merely by publication or target use.
- Measurement as action does not make QL relevant by itself.
- A causal, benchmark, promise, gate or assurance conclusion retains the direct grounds required for that use. Cite the neighboring result only when the temporal assessment relies on it.
Bias-Annotation
Lenses: Onto, Prag, Epist, Arch, Gov.
- Onto: source claim, described subject, temporal-aspect claim, adequacy claim, direct relation, publication occurrence, form, and carrier remain distinct.
- Prag: state and rate readings remain sufficient when they answer the question; Dyn2 adds only the explanation needed for its receiving use.
- Epist: evidence, measurement, model assumption, planning assumption, diagnostic judgement, and causal result are not interchangeable.
- Arch: C.27 keeps temporal-claim adequacy and cites the direct pattern for every other question.
- Gov: a local card does not widen into a budget, gate, benchmark, promise, assurance, or public claim without the direct result required for that use.
Plain speed, acceleration, effort, inertia, rhythm, agility, process, practice, service, and tool-call wording may help recognition. In a technical claim, recover the rate, temporal window, input or Work, resistance, exact bearer, and direct relation.
Conformance Checklist
Common Anti-Patterns
Consequences and Reopen Conditions
C.27 makes intervention-sensitive temporal claims inspectable. An adequate state or rate reading can remain sufficient; a stronger forecast needs its temporal and effect grounds.
Refresh follows the receiving use:
- include a reopening condition when later reliance depends on it;
- a transferred claim may need a validity interval, current evidence or another stated condition;
- G.11 carries refresh orchestration, and B.3.4 carries evidence decay and epistemic debt.
Reopen when, for example:
- a claim, sampling, intervention, baseline, follow-up, or validity window changes;
- a rhythm bearer, temporal reference, rhythm window, proxy, or coupling changes;
- effort, resource, actual System actor, assignment, authority, or availability changes;
- resistance changes because tooling, work mix, queue topology, constraints, or service demand changed;
- a measure becomes a target, gate, dashboard, incentive, or public comparison;
- cross-scale transfer is attempted;
- results reverse, overshoot, oscillate, stall, or become unstable;
- hidden queues, rework, burnout, quality loss, service demand, safety demand, or coordination debt appear;
- the use changes from local diagnosis or planning to benchmark, causal, assurance, promise-like, publication, gate, or formal-model use;
- a coasting, braking, recovery, or stabilization claim outlives its evidence or assumption.
Rationale
The source contribution is the shift from “what is the speed?” to “what effort or input, over which windows, changes rate, rhythm, direction, or stability against what resistance and cost?”
C.27 adopts:
- the state, rate, and intervention-sensitive rate-change distinction;
- effort profiles over time rather than one acceleration word;
- braking, redirection, coasting, recovery, stabilization, and rhythm as legitimate working questions;
- interval-sensitive practice transfer, including the error introduced by copying static poses or rate words without the timing and effort pattern.
C.27 uses physical vocabulary only for the temporal property or dependency claimed. Choose calculus, control modeling or another subject method when its operation is needed to answer that question.
SoTA-Echoing
The source set below is the June 2026 basis for C.27's claim-adequacy rules. Reopen the source use when newer work changes one of the named obligations.
Source references:
- D2-SRC-1: Statics and the dynamics of the first and second derivatives (in Russian).
- D2-SRC-2: Learning-Based Model Predictive Control: Toward Safe Learning in Control, Review on model predictive control: an engineering perspective, and Goal-oriented safe active learning for predictive control using Bayesian recurrent neural networks.
- D2-SRC-3: A Survey of Constraint Formulations in Safe Reinforcement Learning, A Review of Off-Policy Evaluation in Reinforcement Learning, Conservative Q-Learning for Offline Reinforcement Learning, Methods in dynamic treatment regimens using observational healthcare data, and Safe Continual Reinforcement Learning Methods for Nonstationary Environments: Toward a Survey.
- D2-SRC-4: Causal Inference: What If and Causal Inference About the Effects of Interventions From Observational Studies in Medical Journals.
- D2-SRC-5: Performative Prediction, Performative Prediction: Past and Future, and Categorizing Variants of Goodhart's Law.
- D2-SRC-6: OCEL 2.0 and Object-Centric Event Logs: Specifications, Comparative Analysis and Refinement.
- D2-SRC-7: Active Inference: A Process Theory and Embodied decisions as active inference.
- D2-SRC-8: Neural entrainment underpins sensorimotor synchronization to dynamic rhythmic stimuli, A review of psychological and neuroscientific research on musical groove, and Finding the rhythm.
- CT-TIME-SRC: David Deutsch and Chiara Marletto, Constructor theory of time.
Relations
Recover the temporal claim and its grounds, return the smallest adequate assessment, and use a direct neighboring result only for a remaining question. Stop when that answers the receiving use.
C.27:End
Temporal Aspect: Time Windows, Rhythm, Cadence, and Currentness
Type: Definitional pattern Status: Stable Normativity: Normative except where a section is explicitly informative
Use This When
Use this pattern when a project needs to state a positive temporal-aspect claim about one object or claim—for example, its time window, cadence, freshness, recovery timing, or currentness.
Use it when the working question is:
- which time window, interval, duration, latency, cadence, rhythm, synchronization, currentness, freshness, validity window, recovery timing, stabilization timing, trajectory, effort over time, inertia, or refresh condition matters;
- which bearer has that temporal property: system, episteme, work plan, work occurrence, claim, source, benchmark, architecture-selected structure, method description, publication, or project-world object;
- which temporal reference makes the statement reviewable: calendar time, clock time, event order, cycle, sprint, epoch, release train, sampling interval, follow-up interval, or domain-local timing reference; and
- whether the property is merely stated, measured, used in a temporal claim, used in a transformation claim, or used in a work, evidence, or decision relation.
Primary EntityOfConcern. The EntityOfConcern is the independently identified bearer or claim being qualified. A temporal label such as cadence, freshness, or recovery timing is a predicate or qualifier in the ClaimGraph; it is not a second entity.
C.2.1 and publication boundary. A materialized temporal-aspect statement is ClaimGraph content in one C.2.1 episteme whose effective ReferenceScheme makes the temporal terms interpretable. Changing that claim content identifies another episteme. A changed layout, publication occurrence, form, or carrier can leave the episteme unchanged; those publication objects remain separate under E.24.PUB when availability matters.
When the claim depends on another direct relation, cite that relation's declaration or independently established occurrence. A PatternID remains an ordinary rule citation and is never a relation reference.
First useful move. Identify the bearer or claim, state its temporal property, and supply the reference and extent needed to interpret it. An observed cadence needs the period observed; a general ordering rule can instead name the event types and their relative order.
First result. A practitioner can state the timing claim: “CheckoutSystem-1 released at least once in each reporting week of R14.” This gives a colleague a criterion to apply. To report that it was met, check the release events against those weeks as in section 5.1.
Open the fuller statement only when the receiving use also depends on measurement, a scheme or scope, a selected Structure, a source or use boundary, currentness, reopen conditions, coupling, another direct relation, or an explicit rule citation or blocked overread.
What goes wrong if missed. A cadence has no bearer, a currentness claim has no reference date, or a recovery estimate omits its starting event. Readers then cannot tell which temporal claim could support the receiving action.
What this buys. A practitioner can state one positive temporal-aspect claim before selecting C.27, A.3.4, A.3.3, A.15.2, A.15.1, C.16, C.28, G.9, or the relevant evidence, source, gate, or assurance pattern for the receiving use.
Not this pattern when.
- If the question is adequacy or supported use of an authored temporal claim, use
C.27. - If the question is bounded transformation under conditions, use
A.3.4. - If the question is a state-space and transition-law episteme, use
A.3.3. - If the question is work planning or dated work, use
A.15.2orA.15.1. - If the question is measurement construction, rate construction, scale, score, or metric comparability, use
C.16and related characterization patterns. - If the question is causal use of an intervention or policy, use
C.28. - If the temporal phrase is ordinary prose and no practical use changes, do not introduce a C.27.TA statement.
Problem Frame
Distinguish two concerns. One concern is temporal-claim adequacy: whether an authored claim about speed, rhythm, rate-change, recovery, or stabilization can carry a named use. The other concern is positive temporal subject matter: windows, duration, cadence, synchronization, freshness, currentness, inertia, effort over time, recovery, stabilization, and trajectory as aspects of objects or claims.
C.27.TA addresses the second concern. It lets a practitioner state what temporal aspect is in play without immediately opening an adequacy card, a dynamics model, a work plan, a causal-use record, or a transformation statement.
Problem
Common failures:
- Cadence and rhythm lack a bearer or timing. A text says "release cadence" or "team rhythm" without naming the bearer, timing reference, or the extent to which the claim applies.
- Freshness lacks a reference date or use condition. A source, benchmark, dashboard, or claim is called current without a reference date or the conditions that make it current for the intended use.
- Recovery and stabilization hide their interval. A claim says "recover faster" or "stabilize" without saying after which disturbance, over what duration, and for which bearer.
- Effort and inertia lack a subject or course of change. A text speaks about momentum, residue, stored work, adaptation cost, or resistance without linking it to the relevant object and course of change.
- Transformation statements leave timing implicit. A transformation statement names a change but leaves timing and ordering implicit, so method, mechanism, work, evidence, and temporal claims get tangled.
Forces
Solution
Definition
A temporal-aspect claim says that an object or claim has a time-bearing or order-bearing property. Its temporal reference and extent must make that property interpretable: these may be a dated interval, a duration from an event, or a relation among recurring event types. For example, "each approval follows its corresponding review" specifies relative order without selecting a calendar window. A separate adequacy judgement under C.27 is needed only when its answer could change the receiving use.
Typical temporal predicates and qualifiers include:
timeWindow;duration;latency;freshness;currentness;validityWindow;cadence;rhythm;synchronization;trajectory;recoveryTiming;stabilizationTiming;effortOverTime;inertiaOrResidue;refreshOrReopenCondition.
Temporal Aspect Statement
Use this fuller statement when the receiving use needs the corresponding references or distinctions:
When the statement is materialized, its claim content belongs to a C.2.1 episteme. entityOfConcernRef resolves the bearer or claim being qualified; aspectPredicate states the property; temporalReference supplies the clock, event order or other timing basis. Include windowOrInterval when the claim is restricted to that extent or its interpretation requires one. Preserve any interval already asserted by the claim.
Add the remaining fields only when their values change the claim or receiving action. PatternID citations identify applicable rules. A claim about another direct relation needs that relation's declaration; an assertion of an obtaining occurrence also needs its actual participants and conditions. Scheme, scope, Structure, source and use references retain the identity and conditions defined by their respective patterns.
Direct Use and Rule Citation
Rhythm, Cadence, And Synchronization
A rhythm or cadence claim needs its bearer, recurrence or ordering property and timing reference. An empirical assertion such as a weekly cadence observed during R14 also needs that observation window. A planned rule such as "run a review every Friday while this plan applies" can state its recurrence and applicability directly. Add phase, synchronization, entrainment or other cross-bearer relations when the claim depends on them.
Choose the timing rule by what the recipient needs. An average number of occurrences per week, at least one occurrence in each calendar week, and a maximum seven-day gap answer different questions. For weekly coverage, place dated events in their reporting weeks. For a gap limit, compare elapsed times between successive events over the period the agreement covers. When a missing record is meant to establish a missed occurrence, check which events the recording procedure covers. A plan supplies the intended timing and can be used to schedule work.
Optional reference-bearing form:
The optional fields appear only when the receiving use relies on them. A PatternID citation identifies the rule used to judge a claim; it is not the coupling relation.
A plain "release cadence" or "workshop rhythm" may remain ordinary prose. It needs C.27.TA when cadence or rhythm changes transformation, work planning, benchmark, source, assurance, coordination, or claim-use decisions.
Currentness, Freshness, And Validity Window
A currentness or freshness claim identifies the bearer, the property asserted, and the reference time or edition for the intended use. State a validity interval when the claim promises or establishes validity over that interval. "The registry entry was current when checked on 3 June" establishes no later validity period. A source, benchmark, model, dashboard, or claim may be fresh enough for one use and stale for another.
Add a source-use boundary, currentness condition, refresh condition, or reopen condition only when the receiving use changes when that value changes. Use the direct source, evidence, benchmark, assurance, or refresh pattern for the separate provenance, parity, assurance, or refresh-work claim.
Recovery, Stabilization, Inertia, And Effort Over Time
For recovery, stabilization, inertia or effort over time, identify the bearer and property, then recover the relevant starting event, timing reference and duration or observation interval. State the extent supported. Use C.27 when an unresolved question about the claim's adequacy changes the receiving action; a sufficient temporal-aspect claim can finish here.
Add the disturbance or starting condition, measured reading, effort, resistance, residue, inertia relation, rule citation, or direct-relation reference when the receiving use needs that distinction.
Archetypal Grounding
Release Cadence
Suppose R14 comprises four full reporting weeks, Monday through Sunday. Operations needs at least one release in each week. Four dated release events, one in each week, establish that requirement for R14. A total of four alone establishes an average of one per week and leaves their distribution open. A week with no releases defeats the weekly requirement. A blank in a log establishes that absence when the log covers every release in the week.
A support agreement might instead limit the elapsed time between releases to seven days. Consecutive releases on Monday of one week and Sunday of the next meet the one-per-week requirement for those two weeks while leaving thirteen days between releases at the same hour. Two log entries thirteen days apart leave room for intervening releases. To report a breach, establish an interval longer than seven days with no release. A series of confirmed releases spaced at most seven days apart can show coverage even if additional releases went unrecorded. Include releases on either side of R14 when the agreement covers gaps crossing those boundaries.
Source Freshness
A benchmark comparison uses a model report from April and a competitor report from June.
Identify what each report needs to be current about for the comparison, the relevant date or edition, and any period over which validity is claimed. April versus June publication dates alone establish neither staleness nor comparability. If no currentness basis is available, retain only the supported dated-report claim. G.9 and the applicable source, evidence and benchmark patterns govern the separate parity, provenance and reliance questions.
Architecture Recovery Timing
An architecture move is expected to reduce an interlevel conflict after two release cycles.
C.27.TA states the expected recovery timing for the selected structure discussed by the C.30 architecture claim, under the release-cycle reference and window. Use A.22 and the applicable architecture pattern to identify that structure independently; the claim episteme is not its proxy. Use A.3.4 for the proposed structure transformation and evidence/result patterns for any observed effect.
Work Rhythm
A review service promises a final answer within two elapsed days of submission. Two reviewers work in sequence. To state the service's timing claim, follow the same submission from receipt through the handoff to the final answer. Even if each reviewer answers within two days of receiving it, the final answer can arrive four days after submission. Suppose the first stage starts at submission, each stage takes at most one elapsed day, and the transfers to the second reviewer and to the final recipient are immediate. The two one-day bounds then cover the entire promised interval. Waiting before the first stage or after the second has to fit that same interval. For a completed review, the agreed submission and delivery times can give the duration directly.
A turnaround commitment allows irregular arrivals and deliveries. If the question concerns the rhythm of the work, identify the events meant to recur and the relation their timing must preserve. For example, a second review scheduled before the next publication depends on the phase of the first handoff within each publication cycle. Check that relation over the cycles covered by the claim.
If the agreement instead gives each reviewer two days from receipt, assess the two obligations on their respective intervals. If the task is to establish synchronization, recover the relation the work requires between the events. An account of how the participants bring about that timing also needs grounds for the coordinating mechanism.
Bias-Annotation
Lenses: Onto, Prag, Epist, Arch, Gov.
Distortions to watch for:
- unqualified rhythm: rhythm or cadence appears without its bearer, timing reference or the extent needed to interpret the claim;
- freshness-as-permission: currentness is treated as permission, evidence, or gate passage;
- time-as-transformation: timing language is treated as the transformation relation;
- temporal aspect mistaken for dynamics: a temporal aspect is treated as a state-space or transition-law episteme;
- temporal aspect mistaken for measurement: a temporal aspect is treated as a completed measurement construction.
Conformance Checklist
Common Anti-Patterns
SoTA-Echoing
Consequences
- C.27 addresses adequacy and supported use of authored temporal claims.
- A.3.4 identifies the transformation; C.27.TA states its temporal aspect.
- A.3.3 stays the dynamics episteme pattern.
- Use the direct patterns for work planning, actual work, source currentness, benchmark parity, and evidence use.
- Users gain one positive temporal-aspect claim before adequacy, dynamics, causal, benchmark, or assurance patterns are needed.
Relations
- Builds on:
E.24,A.6.5,A.7,C.2.1. - Coordinates with:
C.27,A.3.4,A.3.3,A.15.2,A.15.1,C.16,C.28,G.9, evidence, source, assurance, refresh, and publication patterns. - C.27 consumer boundary:
C.27assesses the temporal claim for its receiving use while preserving its bearer, predicate, timing, coupling and currentness conditions. Use an episteme orC.2.1 ClaimAddresswhen the receiving relation depends on that identity; an ordinary adequate explanation does not require a separate temporal-aspect record. - Used by: patterns that need a positive temporal-aspect claim without making a temporal-claim adequacy judgement.
C.27.TA:End
CausalUse-CAL: Causal-Use Questions, Identification, and Realizability
Type: Calculus (C) Status: Stable Normativity: Normative unless marked informative
Plain-name. Causal-use calculus.
Intent. Develop a causal answer to the question the practitioner faces, or assess what a proposed causal claim is supported to say. Establish the conditions under which the explanation or intervention consequence follows.
Primary EntityOfConcern. The causal-use question arising from an outcome to explain, a proposed intervention or an offered claim. What would change under the intervention, what explains the observed difference, or what can be said about the counterfactual?
Scope. Use the domain's causal models, intervention and outcome definitions, and scientific evidence to determine which statement is supported and under which limits.
Use This When
Use C.28 when a recipient needs to explain an outcome, derive what an intervention would change, or assess an offered result about a causal effect, counterfactual, fairness claim, policy or benchmark. Common cues include:
- “method A improves the outcome”;
- “users who received X did better, so X works”;
- “this policy would have prevented the failure”;
- “the model shows what would have happened”;
- “this fairness metric proves the intervention is fair”; and
- “this benchmark shows that one causal method is better”.
The cue opens a question, not a verdict. Ask what claim is being supported and what use of the evidence depends on that support.
Not this pattern when. If the task only reports a measurement, temporal change or model output without causal reliance, continue with that direct task. Section :4.11 locates the relevant neighboring pattern when a return is needed.
Simulation at entry. “The simulator produced these traces” can finish as a model-output report. “These traces support what would happen under policy P” opens C.28: identify the model, assumptions, validation and the causal use they support.
What Goes Wrong If Missed
- association becomes an intervention-effect claim;
- a changed metric becomes causal fairness;
- a simulated trace becomes realized counterfactual evidence;
- an estimated number is treated as proof that its estimand was identified;
- support for one population or environment is transported to another without an endpoint or assumption; or
- a support verdict is mistaken for permission to publish, deploy, or certify.
What This Buys
The first result is a supported statement with its limits and the next useful step. Section :4.0 locates additional support components by the question each answers; open a specialist profile only when its result is needed.
First-Minute Questions
- What outcome needs explanation, which change is under consideration, or what proposed claim needs assessment? What causal question must the answer settle?
- Is the requested statement about an observed association, an intervention, or a counterfactual?
- What observations, experiments, model assumptions or derived results are available?
- Which live threat could overturn the conclusion: for example, confounding, time order, missing comparison cases, interference, measurement error or transfer to another population?
- What statement is supported under those conditions, and what further evidence or calculation would change it?
When the causal question requires relations that have not yet been constructed, use C.28.CM to turn the subject account and material alternatives into explicit causal models. Return with their premises, conditional consequences or a precise missing relation. A verdict on the supplied evidence can complete a request to assess that evidence. If the recipient needs an explanation, give the mechanism and its grounds using an adequate existing model or develop the missing account. If the explanation remains open, state what the available evidence has settled and which relation still needs to be established.
First Output
Ordinary first result. Suppose the available comparison says that self-selected teams using method A completed more tasks than teams not using it, while task difficulty and prior team capability were not controlled. Report the observed association; the claim that A caused the improvement remains unsupported by that comparison. The next useful question is whether a design or existing evidence can distinguish the method's effect from those rival explanations.
This sentence-level result can finish the task and remain reusable. When a receiving use needs its distinctions separately addressable, the following local form can retain them:
supportedUse states the causal statement or evidential reliance supported under the named limits. unsupportedUse states the nearby stronger statement or reliance left unsupported by that evidence.
Triage can complete a request to assess an offered claim by stating what the evidence warrants. For a commissioned causal account, use that result to locate the remaining explanatory work. A short statement can carry a complete answer.
Problem Frame
The practical question is “what does this evidence support us to say about this causal question, and what would overturn that conclusion?”
Problem
Three kinds of collapse can produce causal overclaim:
- Rung collapse: observation, intervention, and counterfactual comparison are treated as the same question.
- Support collapse: data regime, identification, estimation, direct sampling, and simulation are treated as one alternative-valued “basis”.
- Authority collapse: an evidential conclusion is treated as publication, choice, deployment, fairness, or assurance authority.
C.28 keeps those distinctions visible while allowing a cheap stop.
Forces
Solution
Use the smallest result that answers the current question:
- recover the question and its subject, and assess any claim already offered;
- retain the question in ordinary prose or a small card when it must be reused;
- run the common threat screen;
- add only the specialist result needed now; and
- issue a small
CausalUseSupportResultwhen another pattern must consume the conclusion.
Public contract and support components
When a question or result must be referenced, recover its content and use the corresponding contract:
CausalUseQuestionRefidentifies the exact question content, normally a C.2.1 episteme.CausalEstimandRefidentifies the mathematical target or the episteme that describes it under its direct pattern.PotentialOutcomeContrastRefidentifies the exact contrast or its description.CausalUseSupportResultRefidentifies one C.2.1 result episteme defined below.
Support is composable. A result may use several components:
These fields answer different questions. Do not compress them into one exclusive value. Each optional specialist ref identifies the exact result or record used. Keep a component's own assumptions, uncertainty or sensitivity, supported and unsupported uses, and reopen information with that component when its defining contract requires them; do not copy fields merely to complete a standard list.
The common CausalUseSupportResult still states its own supportedUse, unsupportedUse, limits, optional evidence window, and reopenCondition. Naming a specialist subject in prose does not make its result available to a consumer.
CausalEmpiricalDataRegime is a local classification used only when it helps distinguish evidence:
realizedCounterfactualSamplingData is used only when an A.10 evidence path cites dated sampling Work and the resulting sample or data. A realizability result or WorkPlan alone establishes no empirical regime. Model output is recorded separately through simulationResultRef, not as an empirical regime.
Causality-Ladder Rung
- observational: a distribution or association under observation of the naturally occurring process;
- interventional: an outcome distribution or effect under a specified intervention, experiment or policy;
- counterfactual: a joint comparison across alternative interventions, or an intervention outcome conditioned on the factual history of the same case.
Potential-outcome notation alone does not select the third rung. The population average effect E[Y(1)] - E[Y(0)] uses two interventional marginals and can be a second-rung question. A joint claim such as P(Y(1) > Y(0)), or what would have happened to this case given its observed history, needs the corresponding third-rung information. See Yang and Bareinboim, §2 for the hierarchy's distinction among those distributions.
Lower-rung data may contribute to a higher-rung result only through a replayable identification, bound, or other specialist result. The rung label itself supplies no support.
The rung names the question put to a causal model. The levels in B.5.MR:4.3 name what a model specifies as it is abstracted or refined. A discipline can model a process and ask what was observed, what a specified intervention would produce, or what would have happened in a particular episode. Naming an intervention in a question does not establish that it was performed or reveal its effect. Even a documented action establishes an observed outcome only when the outcome itself is available.
Causal-Use Claim Kind
Choose the kind by the claim being supported, not by the tool or source. Simulation for a causal claim uses the appropriate claim kind plus simulationResultRef; it does not need a simulation-only claim kind.
Question cards and support result
A question that must remain separately recoverable can use this local card; reuse an existing account when it already preserves the needed content:
For a reusable or consequential claim whose receiving use needs the fuller distinctions, use this durable card:
When another pattern needs a stable conclusion, issue this C.2.1 result episteme:
Its identity and reference follow C.2.1. A receiving decision consumes it under :4.9.
Common causal-validity screen
Run only the questions relevant to the current claim. A live threat either points to an existing specialist field/result or lowers the support result; it does not trigger a mandatory dossier.
Ordinary effect case. A randomized treatment study records interventionWellDefinedOrConsistency=clear, temporalOrdering=clear, positivityOrOverlap=clear, interferenceOrSpillover=notApplicable, selectionCensoringOrMissingness=clear, and measurementErrorOrConstructShift=clear for its declared target and window. The screen points to the trial and estimate results; it does not repeat them.
Countercase. An observational cohort has the right rung label and a plausible estimand, but records exchangeabilityOrConfounding=liveThreat and positivityOrOverlap=liveThreat because severity is unmeasured and one treatment region has no comparator. The claimed effect remains unsupported until a suitable design or new evidence closes those threats; a suitable bound can instead support a correspondingly bounded claim. “Observational data” was classified correctly; that label does not establish validity.
Identification result
Identification answers whether the estimand can be expressed or bounded from the available data and assumptions. The bounds for nonidentified counterfactual queries in Raghavan and Bareinboim, 2026, §5 belong to this identification problem. The conclusion must be replayable:
An identified label without an identifying expression or derivation is incomplete. A bounded result cites the bound. A nonidentified result exposes the obstruction or failure witness. Identification is neither a numerical estimate nor direct physical sampling.
Replayable identified case. For treatment_effect_in_population_P, AdjustmentSet_Z is justified as blocking the relevant back-door paths. backdoor_adjustment_derivation_7 states the identifying expression in ordinary terms: compare treated and untreated outcomes within each Z group, then average those differences using the target population's Z distribution. The result cites the data regime, assumptions, expression, and the confounding or overlap change that would reopen it.
Replayable nonidentified case. In a constructed binary treatment/outcome example, the observed law is P(X=0,Y=0)=0.5 and P(X=1,Y=1)=0.5. Two models allowed by the stated assumptions reproduce that same law:
- Model M0 has an unobserved fair binary input U,
X=UandY=U. Replacing X by either constant leavesP(Y=1)=0.5, so the average intervention contrast is 0. - Model M1 has the same U and
X=U, butY=X. Replacing X by 1 gives Y=1; replacing X by 0 gives Y=0, so that contrast is 1.
The two compatible models give different values of E[Y|do(X=1)]-E[Y|do(X=0)]. An exact value therefore cannot be identified in a class admitting both. More observations of that same joint law cannot settle which effect is correct.
The same information does settle a weaker decision. Assume binary outcomes for the same target population under two well-defined assignments, with consistency: the outcome under the assignment actually received equals the observed outcome. In this stipulated population law, everyone receiving 0 had outcome 0, so switching such a person to 1 could leave the outcome unchanged or raise it to 1. Everyone receiving 1 had outcome 1, so switching such a person to 0 could leave the outcome unchanged or lower it. In both groups the contrast “outcome under 1 minus outcome under 0” lies between 0 and 1. The population mean contrast has those same bounds. M0 attains 0 and M1 attains 1. These premises rule out a negative mean effect. They leave open whether the mean improves by at least 0.25. For an empirical sample, establish what its distribution supports about the target population before using this bound there. This calculation uses the bounded outcomes for each assignment to put limits on the unobserved contrast. Manski, 1990, pp. 319–321.
For a supplied causal model, C.28.MR derives the consequence by replacing the selected mechanism, retaining the other mechanisms and input law, and solving the relations needed by the query. The following case illustrates that operation and its difference from observation.
Replayable sensor case: observation and intervention. Let H denote binary high load and S a binary alarm. Stipulate P(H=1)=0.5, P(S=1|H=1)=0.9 and P(S=1|H=0)=0.1. Bayes' rule gives P(H=1|S=1)=0.9 and P(H=1|S=0)=0.1. These are observational questions about the stated joint distribution.
For the query P(H=1|do(S=0)), additionally specify a structural model: H is determined by an exogenous random input; S is a noisy measurement of H with separate independent noise; S has no influence on H. Replacing the S mechanism by the constant zero preserves the H mechanism and its input distribution, giving P(H=1|do(S=0))=0.5. This derivation is an identified model-based result under those assumptions. The corrected Pearl, Glymour and Jewell primer, p.55 explains the mechanism-replacement operation.
If the alarm controls cooling, specify the intervention time and the subsequent load mechanism before answering a later-load question. F.0.2:5.5 uses this distinction when comparing source theories for an explanation.
Counterfactual sampling realizability
Use this result to answer whether a declared target distribution can be sampled under current constraints. Raghavan and Bareinboim, 2025, Definition 3.4 and Theorem 3.5 give a construction-or-failure decision for their graph and available-action setting. The result is prospective: it does not say that sampling was planned, performed, or yielded data.
A realizable result cites the sampling construction that the decision Method accepts for the exact target and constraints. A nonrealizable result exposes the obstruction or failure witness. unclear names what remains unresolved. Bounds on a counterfactual probability belong to the identification result at :4.4; they constrain what can be inferred about that probability, while this result asks how draws from the target distribution can be obtained. “Realized counterfactual sampling” never means observing incompatible outcomes for one unit in one realized world.
If the team plans to draw samples, use a separate A.15.2 WorkPlan. If sampling occurs, recover every precise performer's A.13 core and independently admit the dated Work under A.15.1. Add F.6 only when the sampling claim also needs precise assignment-bound attribution. If the samples are used as evidence, cite the resulting data through an A.10 evidence path. Actual sampling support requires both the dated Work and resulting data or evidence ref; neither realizable nor a WorkPlan can stand in for them. Identification from those data, when claimed, is another CausalIdentificationResult.
Applied profiles
Target trial. The TARGET Statement, 2025 supplies reporting guidance for an observational study explicitly emulating a target trial. Use it for the protocol-to-data account below, with causal validity checked separately.
An observational emulation keeps the protocol and its mapping to available data as separate results:
Every mapping field identifies the mapping. protocolToDataGapAccountRef points to one account that lists the observed gaps or states that none was found within the declared source and window. The residual-confounding and sensitivity fields remain present even when their bounded result is favourable. Reporting completeness is not a risk-of-bias, identification, or estimate verdict.
Constructed target-trial mapping. HypertensionEmulationMap-2025 maps HypertensionTargetTrial-1 to ClinicRecords-2022-2024: age and diagnosis fields implement eligibility; prescription records distinguish the two treatment strategies; the prescription date supplies assignment and time zero; encounter records map the twelve-month follow-up; and the recorded systolic-pressure field maps the outcome. GapRecord-17 states that adherence after prescription is not observed, ResidualConfoundingAssessment-17 retains unmeasured severity as a live threat, and SensitivityMap-17 points to the negative-control and quantitative-bias analyses. The result supports construction and review of this emulation. It does not by itself establish identification, low bias, or a transportable effect; new severity or adherence data reopens it.
Estimation.
At least one identification or explicit design-based basis is required before the estimate supports a causal use. Orthogonal scores, nuisance models, and cross-fitting belong in methodSpecificDetailRefs only when a DML Method is selected. estimationConsistencyResultRef points to the consistency result defined by the selected estimation Method or its direct evaluation pattern.
Counterfactual fairness. Before D.5 uses a counterfactual-fairness support result, its C.28 components cite the identification result and the extra assumptions needed to connect the available data to that counterfactual question. When the fairness conclusion depends on an estimate, they also cite the estimate and its estimationConsistencyResultRef. Without those conditions, return bounded or unsupported; more data, even an unlimited amount of the same data, does not repair missing counterfactual identification or an inconsistent estimator. Associative or interventional fairness claims use their own rung and do not inherit this stronger branch by label.
Constructed non-DML estimate. A randomized trial cites random_assignment_identification_4, trial_data_8, DifferenceInMeansMethod_2, standard_error_result_5, and its attrition sensitivity check. It needs no orthogonal-score, nuisance-model, or cross-fitting fields. The estimate supports only the declared population, outcome, assignment, and follow-up window.
Transport. For the first-moment population-measure problem under covariate shift, use Boughdiri, Berenfeld, Josse and Scornet, A Unified Framework for the Transportability of Population-Level Causal Measures, 2025, §§2–4 with its internal trial-validity, overlap and selected exchangeability conditions. Other endpoint changes need their own identifying result.
Identify every endpoint dimension that differs in the current claim. Population, domain, environment, data-generating regime, and semantic scheme answer different questions. A shared label does not establish transportability; a semantic Bridge is added only when its F.9 relation independently obtains.
Off-policy evaluation.
Causal representation. Use this record only when variables are learned, selected, or abstracted rather than supplied by the domain:
When the outcome is an aggregate, establish how its constituents contribute to the outcome the question asks about. When the intervention itself is specified on an aggregate, establish which changes to its constituents count as implementing it. Follow the permitted implementations to the requested outcome under the retained mechanisms. If their answers differ, identify the intended implementation, justify the distribution over implementations, or give the supported range. If every permitted implementation meets the required threshold, that bound can answer the question. State the interventions and queries for which the representation preserves the required conclusion.
Constructed representation case. WardStateRepresentation-4 derives three state variables from monitor traces through WardStateAbstractionMethod-2. Its intervention-validity result covers dosage interventions, its invariance result covers the two hospitals in the training and hold-out comparison, and its query-preservation result passes the declared one-step counterfactual query but fails the long-horizon query. The record therefore supports the one-step policy comparison only; a new hospital, sensor scheme, intervention family, or long-horizon claim reopens it.
Graph and calculus names
Use C.28.CM when the causal relations, material alternatives or observing process still need to be modeled. It supplies a model with explicit premises and a useful consequence or unresolved distinction. Use C.28.MR for a required intervention derivation in a supplied model. These contributions return to the support question here; constructing a graph does not establish its empirical adequacy.
Use specialist names only when the result depends on them. For a counterfactual graphical-model derivation, use the conditions and calculus in Correa and Bareinboim, 2025 and cite the derivation used:
These values classify the formal support form. References point to the model, diagram, derivation, assumptions, or proof. A graph-class label is not a proof and does not replace the plain statement of what was identified or bounded.
Causal evidence design and Work
Use CausalUseEvidenceDesignRecord when additional evidence could change the support boundary:
The three optional specialist refs are included only when an existing target-trial mapping, off-policy evaluation, or causal-variable representation result shows what additional evidence could change the support boundary. Before execution, cite a MethodDescription or WorkPlan only when used. After execution, cite every precise performer's A.13 core and the independent A.15.1 Work admission; cite F.6 only when precise assignment-bound attribution is also current. If performed counterfactual sampling is used as evidence, also cite the resulting sample or data through evidencePathRefs; Work without output data and data without its Work and provenance path each remain incomplete for that claim. Do not copy performer-kind, assignment, or occurrence mechanics into this record unless one of those facts changes causal validity, safety, authorization, or supported use.
Additional evidence is worth planning only when it can change a material causal statement or downstream decision enough to justify cost, risk, and delay, or when safety or release rules independently require it.
Support is not authority
CausalUseSupportResult.verdict has four values:
supported: the named causal statement or evidential reliance is supported under the stated limits;bounded: only the narrower statement or reliance is supported;unsupported: the claimed causal statement or reliance is not supported;undecided: the available work does not establish a causal conclusion.
When another pattern uses the result for a decision, it checks whether the named support and limits answer its question and applies its own remaining decision conditions. For undecided, it chooses whether to seek evidence, abstain or use an available non-causal result. “Report association only” limits the evidential claim; any decision to publish, choose or deploy remains with the receiving pattern.
Causal action policy class
Use CausalActionPolicyClass only when the action-selection regime changes the causal question. Identify the decision variables, horizon and natural mechanism; say whether the question compares one specified rule or an admissible family. Keep that rule or family with the question. When both behavior and evaluation policies occur, identify which one the field summarizes.
The mechanism distinctions follow Maiti and Bareinboim, Sequential Causal Games, Definitions 2.3, 2.5, 2.7–2.8. Let X° be the natural action and Z the allowed pre-action information. Order actions causally; Z excludes descendants of the current or later action variables.
For this interface, when a concrete rule is represented in the wider counterfactual family, record its simpler natural or interventional case when that reduction holds throughout the stated scope. Keep mixedPolicy when the specified combination matters. If selection itself requires the current natural proposal, expose that counterfactual dependence. For a family comparison, retain the declared available family and any restrictions. Thus the scalar is an informative summary of the supplied rule or family, not a disjoint ontology of policies.
Constructed three-rule example. For binary X°, compare:
The last evaluation needs the natural proposal and its relation to outcomes; a fixed-action evaluation does not answer it. A mixed example chooses with equal probability, using an independent coin, between following the natural mechanism and replacing its action by zero. Retain both component choices and their probabilities.
unknown records unresolved classification, not another member. Omit the field when it changes no support, comparison or receiving decision.
Neighbor selection
Examples of bounded results
Payoff check
Keep a causal-use record only when it changes the supported causal statement, blocks a concrete overclaim, changes evidence work, or supplies a basis to a downstream decision. Remove fields that do none of those things. Prefer triage when it preserves the same boundary.
Publication-unit boundary
When only wording inside one publication unit is unclear, use the publication and wording patterns. Open C.28 only when the wording is relied on causally.
Examples of unsupported causal claims
Archetypal Grounding
System. A team records higher completion among recipients of X than among other eligible users in the same period. This supports an observed group difference. If only recipients were observed, report their outcomes and any comparable within-group change. To decide whether X would improve the outcome under a proposed deployment, examine the assignment process, the comparison and the conditions of intended use.
Fairness. A report claims counterfactual fairness after a policy change. C.28 identifies the rung and estimand, exposes the additional counterfactual-identifiability assumptions, and cites an estimate with its consistency result when the audit relies on that estimate. Missing identification or consistency lowers the support result even with more of the same data. D.5 carries the BiasAuditReport@Context and makes the audit conclusion.
Policy. Logged behaviour data are used to evaluate a new policy. The result names both policies, horizon, confounding and overlap checks, transport endpoints when changed, estimate and uncertainty, supported regime, and unsupported unqualified optimality. C.11 or another policy pattern makes the choice.
Causal RL. An online learner combines logged behaviour, interventions, and a counterfactual-data source. The sampling-realizability result explains whether that source can be produced; dated Work and the resulting data path show whether it was produced; a separate identification or estimate result says what follows from it. Replay reward does not become an optimal-action claim.
Evidence Work. A lab's CounterfactualSamplingRealizabilityResult cites its decision Method and positive construction. That result supports planning but claims no sample. The later WorkPlan remains prospective. After sampling, the lab cites independently admitted dated Work and the resulting data in an A.10 evidence path before using realizedCounterfactualSamplingData; it adds precise assignment-bound attribution only when the receiving support claim uses it. Identification from those data remains a separate result.
Simulation. A simulator supports rehearsal and sensitivity analysis under named assumptions and validation. The support result blocks realized-sample and intervention-effect wording. A pure simulator-output report exits C.28 earlier.
Transport. The population is unchanged but the care environment and measurement mechanism differ. The transport result names source and target environments and data-generating regimes, then states the assumptions and formula. A population ref alone would miss the shift.
Benchmark. G.9 compares an observational predictor, intervention optimizer, and counterfactual policy only after it checks rung, estimand, support components, window, and endpoints. The admissible result may be a selected set or abstention rather than a scalar winner.
Bias-Annotation
Watch for these errors:
- using causal terminology as a substitute for support;
- presenting simulation as realized or interventional evidence;
- using a metric proxy in place of assessing the causal claim;
- treating a graph as sufficient support;
- treating feasibility as performed work;
- citing data without the Work that produced it;
- substituting a support label for its basis;
- reducing a causal benchmark to one score despite material differences in its comparisons.
Recover the question, support components, live threats, supported statement, unsupported statement, and reopen condition in the shortest form that remains replayable.
Conformance Checklist
- The outcome, intervention or offered claim and the causal question it raises remain identifiable through to the supported answer and its limits.
- Every reference required by the current use resolves to the exact question, target or result defined at :4.0; a sentence-level triage can finish without such references.
- Data regime, identification, estimate, sampling realizability, performed sampling evidence, simulation, and transport remain distinct and may be combined.
- A downstream decision uses the causal-support result within its stated limits and checks its remaining conditions under the receiving pattern, as required by :4.9.
- An identified result cites an expression or derivation; a bounded result cites a bound; a nonidentified result cites an obstruction or witness.
- A causal estimate cites an identification or explicit design-based result. Method-family details appear only when that Method is selected.
- The common threat screen routes every live ordinary threat or lowers the result; it is not a mandatory dossier.
- Non-causal simulator reporting and simulation-supported causal use take different routes at first entry.
- A sampling-realizability result cites its decision Method, any derivation used, and the sampling construction or obstruction required by its status;
unclearnames the unresolved question. Counterfactual-quantity bounds remain in the separate identification result. A prospective result claims no Work or data. - Performed counterfactual-sampling support cites independently admitted dated Work and resulting data or evidence; it cites exact assignment-bound attribution only when the receiving support claim uses it. A WorkPlan or
realizablelabel cannot satisfy this branch. - Before execution, evidence design cites a MethodDescription or WorkPlan only when used. When it cites performed Work, it identifies each precise performer under A.13 and the dated occurrence independently under A.15.1. Performed counterfactual sampling used as evidence also cites the resulting data through A.10. F.6 is required only when that account uses exact assignment-bound attribution.
- Transport identifies every changed population/domain/environment/data-generating-regime endpoint separately from semantic schemes.
- A counterfactual-fairness escalation exposes its additional identification assumptions and, when an estimate is used, estimation consistency before D.5 consumes it.
CausalActionPolicyClassidentifies the specified rule or available family through :4.10's mechanism and information conditions; reductions, combinations and unresolved classifications stay explicit. Consumers use the same meaning and omit an unused field.- Every specialist field changes support, a downstream decision basis, evidence work, or a reopen condition.
- A target-trial mapping result identifies the observational source and every protocol-to-data mapping, gap, residual-confounding assessment, and sensitivity mapping needed for its bounded use.
- Every retained specialist result that can independently change support can enter
CausalSupportComponentRefs; when it shapes further evidence, the evidence-design record can cite the same result without copying it. - The whole pattern remains understandable to a practitioner without requiring the formal graph vocabulary on the ordinary path.
Common Anti-Patterns and How to Avoid Them
Consequences
The pattern makes unsupported causal claims easier to lower while keeping ordinary triage cheap. Consequential claims become replayable across question, support components, threats, limits, and source window. The cost is additional specialist work only where a stronger causal statement or downstream decision depends on it.
Rationale
Temporal change, a higher metric, a convincing graph, or a plausible simulator can all be useful without supporting a causal effect. Conversely, observational data can support a causal estimate when an explicit identification result closes the inferential gap. C.28 therefore separates the question from the components that support it and separates that evidential conclusion from downstream authority.
SoTA-Echoing
Which question does the available model answer? In :4.4, the same observed alarm distribution gives P(H=1 | S=0)=0.1, while replacing the alarm mechanism gives P(H=1 | do(S=0))=0.5 under the stipulated no-feedback model. Reusing the observational answer is cheaper but answers the wrong question when the user proposes to set the output. Mechanism replacement, as explained in Pearl, Glymour and Jewell's corrected primer, p.55, supplies the operation used in the second derivation.
This comparison selects :0.3's question distinction and :4.4's requirement for an identifying expression or derivation. The added cost is stating the causal mechanism and assumptions needed for the requested intervention, instead of relying on the joint distribution alone. If the requested statement is observational, its existing answer suffices. If the alarm initiates cooling, the later-load question requires the changed mechanism and time order; reopen the earlier intervention result.
What supports a counterfactual-fairness conclusion? The analysis by Ma, Melnychuk, Frauen and Feuerriegel, 2026 identifies two failures in counterfactual-fairness baselines: missing counterfactual-identifiability assumptions and inconsistent counterfactual estimation. Their analysis uses identification up to a measure-preserving indeterminacy and a compatible consistency condition. More of the same data does not generally supply either missing condition.
For a receiving fairness audit, an improved metric or a large dataset therefore leaves a different question from the counterfactual guarantee. Section :4.6 selects a separate identification result and, when the conclusion uses an estimate, its Method's consistency result before D.5 consumes the support. The additional work is justified by that stronger question; an associative disparity report can finish at its own rung. The source supplies this failure analysis and a method-specific remedy, so the consumed guarantee retains its assumptions and scope. Reopen the support when those assumptions, the estimator or the fairness question changes.
How much interface is useful? A domain analyst's ordinary causal report can already state a question, assumptions, result and limitations. For the self-selected-team comparison in :0.4, that short report is sufficient: C.28's thin path returns the association and the live confounding question. Requiring the complete specialist profiles would add target, model and result declarations unused by that conclusion.
When several receiving uses need the same conclusion, the alternative is repeatedly extracting its question and limits from a larger report. Sections :4.0–:4.2 instead provide references to the independently used results and one common support conclusion. This costs explicit identification of those results and their conditions. It can avoid copying an identification derivation, estimate or sampling construction into each receiver. Use that structure when the receiving work needs it; reuse an existing result directly when it already supplies the required information.
These are bounded selections for the illustrated questions. They preserve cheap prose, expose a mathematical difference when it changes the answer, and require stronger support for a stronger fairness claim. Reopen the selected form if it hides a live causal distinction or requires information that the receiving use does not consume.
Relations
-
C.28.CM constructs and challenges causal models when the mechanism relations or material alternatives are still missing; the returned model retains its assumptions and evidential limits.
-
C.28.MR derives an intervention consequence within a supplied causal model, with the replacement, retained conditions and solution needed by that query.
-
C.16keeps measurements and scales;C.27keeps temporal-claim adequacy. -
A.10keeps evidence paths and provenance and may cite C.28 support components and result. -
A.2.4classifies how an episteme is used; it cannot promote simulation output or association into stronger causal evidence. -
A.15keeps Method, plan, Work, and attribution for interventions, target trials, and sampling. -
B.3may cite a C.28 result as one basis for a separate bounded assurance result. -
C.11,C.19, andC.24keep choice, pool treatment, and call planning and consume only the needed causal refs. -
D.5keeps bias/fairness audit and usesBiasAuditReport@Contextwhen a causal fairness question is consequential or reusable. -
G.5keeps method dispatch;G.9keeps parity and benchmark conclusions;G.11keeps refresh planning. -
C.26is used only for a residual quantum-like modelling issue after ordinary causal explanations are tried.
C.29 mathematical-lens relation
C.29 may describe a mapping as abstraction-like, quotient-like, coarse-graining-like, simulation-like, or macro-model-like. It does not decide causal support. When intervention, policy, counterfactual, causal explanation, or causal decision use is current, apply C.28; otherwise record no causal-use claim or the blocker.
C.28:End
Construct and Challenge a Causal Model
Type: Method pattern Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when you need to explain an outcome and the mechanism is open, or when you have a proposed cause and need to work out its consequences. Begin with the outcome and the work or process that could have produced it. Trace the conditions, actions and connections that could account for the observed difference. Use subject knowledge and observations to develop a candidate mechanism or challenge one already proposed.
Begin with one outcome, the contrast that matters, and the relevant observations and subject knowledge. Express two materially different accounts when both remain plausible. Work out a consequence on which they differ, or identify the premise that prevents that comparison. An annotated sketch and a conditional explanation can be the first useful result.
The object is the causal model, or family of alternative causal models, being constructed for that question. This is the causal branch of broader modeling and explanation work. The model describes proposed mechanisms and the conditions of a consequence; it remains distinct from the situation, its observations and the evidence for those mechanisms.
Use a sufficient existing model directly. A calculation under known physical relations can finish in B.5.FM or C.28.MR. A disagreement about whose objective should govern needs that practical or normative decision; constructing more causal arrows will not supply it. If a subject relation cannot be recovered, ask for that relation or retain the missing premise. Neither a literature review nor a numerical model is a universal prerequisite.
Problem
“Missing information caused the delay” can conceal several accounts. The information may be needed to perform the work. Both the omission and the delay may arise from overload. The recorded omission may describe a logging failure rather than the information actually available. Studying only reported incidents can introduce a further difference.
These accounts can support different responses while fitting the same initial story. Filling a field, changing the generating process, improving registration and temporarily supplying another route are different actions. A model makes their assumptions and consequences inspectable before one narrative becomes the explanation.
A second difficulty arises when a diagram is mistaken for evidence. Drawing a mediator does not establish that the mechanism operated. Leaving out an arrow makes a substantive assumption. A model can be useful for exposing such a premise before it supports an empirical effect claim.
Forces
Solution
Construct the account around the question, then challenge the relations on which its useful consequence depends. The following moves can return to one another. A discovered measurement error may change the outcome definition; a missing mechanism may change the question.
Fix the contrast and the time that matter
State what needs explaining or changing. Name the outcome and the cases it concerns, the relevant time, and any proposed action or comparison. “Why did this queue start?”, “Why does it remain?” and “What will shorten it now?” can require different mechanisms.
Distinguish an observed comparison from an intervention query. Observing X=x supplies information about the mechanisms already operating. Imposing X=x replaces a mechanism. For a question about a past case under another action, use factual observations to constrain the underlying conditions of that case. In the alternative, recompute the quantities that depend on the changed action. An outcome observed in the factual case is evidence about that case and may take a different value under the alternative. C.28.MR develops the calculation and C.28 determines what supports its use.
Choose a contrast that can answer the recipient's question. Two accounts can justify the same immediate protection while explaining the onset or persistence of the trouble differently. The common action can settle today's protection. For a request to explain the trouble, derive what the accounts establish in common and which disputed mechanism still matters. Use C.11.DUA when deciding whether obtaining further evidence warrants its cost.
Turn the story into variables with subject meaning
Choose quantities or distinctions that can vary across the relevant cases or times. Give each a meaning, possible values and observation time. A particular late delivery is a case; its lateness is a value of a selected variable. Whether a variable is binary, ordinal or continuous depends on the question and measurement.
Separate the subject quantity from its registration when that difference matters. For example, distinguish an identifier present in the submitted report, a log entry reporting its absence, and inclusion of that report in an incident sample. Recover how the observation is produced through C.16 when needed. A copied ticket is another report of the same event, not another occurrence of it.
When the question concerns a whole, work back from its outcome to the activities, handoffs or physical interactions through which it is produced. Select the constituents whose differences could change the explanation or the consequence being asked about. For example, a publisher adds one reviewer qualified for initial review and final approval. Assigning that reviewer to the first review or to final approval counts as the same staffing increase. If final approval is the only queue limiting completion, the two assignments can yield different publication dates. Keep the assignment and the handoff in the model when the difference matters. If all allowed assignments meet the requested deadline under supported bounds, the coarser description can settle that threshold. A request to explain why a particular issue waited can still require the stage and the history of its handoff.
Propose mechanisms and serious alternatives
For each consequential relation, explain how changing a proposed cause could change its effect under the relevant conditions. Use the process arrangement, physical law, subject theory, observations or qualified source that supports that proposal. Mark a conjecture as such. A chronological sequence can suggest a mechanism while leaving common causes unresolved.
Construct another account when it remains materially plausible: for example, another cause, a common influence, a recording difference, or a model in which the proposed direct influence is absent. Causes may coexist. An account of overload need not remove an effect of missing information. Keep the outcome and horizon comparable, and specify the same intervention in each model when that is the question.
A directed arrow X → Y represents a proposed direct causal influence in the chosen model. “Direct” is relative to the variables retained. Explain a consequential omitted influence as carefully as an included one: excluding it may be what makes the answer follow.
For an equation-based model, write the relevant mechanisms in a form such as Y=f(X,U), where U represents inputs left outside that mechanism. Specify shared inputs or their dependence when needed. A graph supplies qualitative restrictions; it does not supply an omitted functional form, effect size or disturbance law. Ask for these only when the receiving calculation needs them.
Inspect the whole relevant structure
Follow possible paths between the proposed cause and outcome. Look for common causes, intermediate mechanisms, measurement processes and selection of cases. A common cause can coexist with a causal path. The role of a node is relative to a path and question, not a permanent label attached to the variable.
For a directed acyclic graph (DAG), the following path test makes that inspection precise. A path joins distinct nodes along edges, ignoring arrow direction when finding the path. At an internal node, two arrowheads meeting there make it a collider on that path; other internal nodes are noncolliders.
Given a conditioning set W disjoint from the endpoints, a path is open when every noncollider on it is outside W and every collider is itself in W or has a descendant in W. Otherwise it is blocked. If every path between two variable sets is blocked, the sets are d-separated by W. For a distribution satisfying the graph's Markov property, d-separation implies the corresponding conditional independence. An open path permits dependence; it does not guarantee it for every parameter choice. Inferring graph structure from observed independences needs further assumptions, often faithfulness (no extra independences beyond those entailed by the graph), as well as adequate data.
Apply this rule to all relevant paths, including those opened by the way the sample was selected. Conditioning on an incident being reported can matter even when “reported” is absent from the regression.
Consider the complete graph:
There are three simple X-to-Y paths. X → M → Y carries the proposed causal mechanism; X ← Z → Y carries a common cause; X → S ← Y is blocked at S before conditioning. Conditioning on Z blocks the common-cause path and retains the causal path. Conditioning on M blocks that causal path. Conditioning on S, or its descendant D, opens the collider path. Inspecting only the fork would miss that selection effect.
For a total effect in an appropriate causal DAG, the back-door criterion provides one sufficient adjustment rule: choose measured covariates that are not descendants of X and block every path entering X through an arrowhead. This leaves the directed causal paths available. The resulting identification relies on the model's causal interpretation, suitable data and support for the required comparisons. Failure of this sufficient criterion is not proof that the effect is unidentified. Use C.28 for the identification question rather than inventing an adjustment from one recognizable three-node shape.
The ordinary d-separation rule above applies to DAGs. If feedback matters, distinguish times: an outcome at t may affect workload at t+1. A finite time-unfolded model needs its initial conditions and omitted influences justified. An equilibrium with simultaneous causal feedback needs an appropriate cyclic model, its solution conditions and its separation rule. Do not erase feedback merely to obtain an acyclic picture.
Derive a consequence that can distinguish the accounts
Work the same question through each retained model. Start with a direction, a possible or impossible outcome, a conditional independence, or an observation that one account can explain only by adding another premise. Show which relation makes the consequence follow.
Keep three operations distinct. Conditioning restricts attention using an observed value. Intervention replaces the specified mechanism. Omitting a variable from a report or marginalizing it out leaves its possible influence in the subject model. Treating all three as “fix the variable” can reverse the conclusion.
When the mechanisms are specified, C.28.MR derives an intervention consequence while retaining the relevant other mechanisms and underlying inputs. Use C.29 when expressing or computing the model requires further mathematical construction. If the needed function or input law is missing, return that gap along with any qualitative result that survives.
Two graphs may entail the same observed independences. Two mechanisms may produce the same measurements in the available range. Retain both when the present material does not distinguish them. A causal-discovery procedure can contribute within its declared model class and assumptions; its returned graph does not remove those conditions.
Challenge the consequence with available evidence
Compare the consequence with a relevant observation, contrasting case, or argument about the modeled process. An intervention that could distinguish two explanations gives the team a test it can run. To use that test as empirical evidence, obtain its result or use a suitable result already available. First check that the observation concerns the modeled variable, cases and time. A contradiction can arise from the mechanism, measurement, implementation of a proposed action, or a changed operating condition.
Distinguish a contradicted implication, an untested implication and compatibility with the inspected material. A graph with no relevant testable implication cannot acquire support merely from an absence of contradiction. Conversely, failure to reject an implied independence does not establish the graph. Deterministic constraints, measurement error, missing cases and sampling uncertainty can all change the interpretation.
Revise only as far as the finding warrants. Preserve a valid observation when its causal interpretation fails. Restore a rival when its rejected premise changes. If an additional study could change the answer, C.28:4.8 helps specify the evidence question and C.11.DUA helps choose whether to obtain it. A useful model may end with an unresolved distinction.
Return the model with the use it can support
Give the recipient the question, the relevant mechanisms and alternatives, their source basis and assumptions, the derived consequence, and the uncertainty that changes its use. A short explanation can carry this result; no universal record form is required.
Recognition can stop at “these two accounts imply different responses”. Consequential reliance requires the corresponding subject and evidential grounds. C.28 qualifies causal support. A causal estimate needs its identification or design basis; a bound needs its derivation and an estimation result only where it relies on estimated quantities. A controlled direct effect is a different query from a total effect. A claim about the cause of one historical outcome or responsibility also needs its chosen concept and subject standard; a population effect does not settle it.
Return the next useful contribution by what it must establish: for example, whether a log reflects an actual omission, whether a controller reads a display, or whether an apparent AI effect survives comparable task selection. Temporary mitigation can remain available under its own evidence, costs and authority while the causal question is open.
Archetypal Grounding
The following constructed cases demonstrate reasoning under stated premises. They do not report field effects.
Separate an omission, its record and the incident sample
Four incident tickets mention missing order identifiers after a reporting-template change. A planner asks whether supplying the identifier before import would remove manual matching. All four tickets came from problematic reports; the frequency among all reports is unknown.
Choose M for an actually missing identifier and Y for manual matching after import. X denotes template use and L high process load. One account proposes X → M → Y, with load affecting template deployment and identifier omission. A second account proposes that load produces both missing identifiers and an additional wrong join key K; K causes matching work even after the identifier is supplied. The second account has L → M and L → K → Y without an M → Y mechanism.
These accounts make different conditional predictions. If missing identifiers are the operative joining failure and supplying one changes no other input, supplying it removes that modeled obstacle. If K remains wrong, the same correction leaves the K-related work. A report with its identifier restored but the same failed join can challenge the first account's claim of sufficiency. It does not establish the second account merely by eliminating one rival.
Now recover the observing process. Let R be a log saying the identifier is missing. A logging defect can make R=1 while M=0. Inspecting the submitted report and the importer’s actual inputs can therefore resolve a recording question before further causal research is useful.
Let S=1 mean a report entered the incident sample, either because R=1 or because matching work was severe. The structure R → S ← Y means that analyzing only S=1 conditions on a collider. The four tickets cannot by themselves establish the population association or the effect of correcting identifiers. Obtain the needed comparison cases if they could change the decision; retaining a qualified unresolved answer is also possible.
The first return is specific: determine which joining inputs were actually missing or wrong, then distinguish the two modeled failure mechanisms. If all retained accounts support a cheap temporary manual check under the decision's own conditions, using it does not establish which account caused the incident.
Onset, persistence and amplification. A queue first appeared while a specialist was absent. This dates its onset. To decide whether the absence produced the queue, compare the service those arrivals actually received by the time the queue formed with the service they would have received by then under the dispatch rule with the specialist present. If the rule holds requests for a batch, they may have waited even with the specialist on duty. Enough capacity over the whole day could still leave work waiting when several requests arrive together.
After the specialist returns, suppose the backlog is 12 comparable items. Ten items arrive each day. While enough eligible work is waiting, the service uses its capacity of ten completions per day. Under B(t+1)=max(0,B(t)+A(t)+R(t)−C(t)), with arrivals A=10, rework R=0 and usable capacity C=10, the backlog stays at 12. Two extra items of rework each day make it grow by two per day under the same conditions. Preventing that rework would stop the growth in this model. To reduce the total backlog to zero, cumulative completions must also cover the initial 12 on top of arrivals and rework. These conclusions about the period after the return can guide action even if the cause of onset is still open.
Construct the measurement and device mechanisms separately
An idealized regulated supply has command C=12 volts and a fixed 6-ohm resistive load. Its display reads D=14 volts. The practical question is whether correcting the display can reduce the load current.
Construct separate variables for delivered voltage V, display D and current I. Use the load relation I=V/6. Two accounts fit the displayed value:
Setting the displayed number to 12 replaces the display mechanism in either model. It leaves V and I unchanged. Changing C to 6 instead gives V=6, I=1 in the first model and V=8, I=4/3 in the second, if the stated offsets and load relation remain valid. The observation D=14 alone does not choose between the accounts.
A separately qualified voltage or current observation can discriminate these idealized models. The need is a measurement of the physical quantity with a suitable independent basis, not another copy of the same display.
Change the situation: the displayed value is fed into an automatic controller for the next command. The earlier absence of a display-to-device path no longer applies over that horizon. Add D(t) → C(t+1) → V(t+1) and obtain the controller law before predicting the later current. The current instant and later controlled behavior are different questions. C.28.MR then performs the chosen replacement in the model constructed.
Separate AI assistance from assignment and selection
A team reports that AI-assisted tasks had a higher success rate. The question is the effect of using assistance A on success Y for the same eligible task population. Let D denote pre-existing task difficulty and H the worker's prior skill.
One model contains A → Y, D → A, D → Y, H → A and H → Y. The rival keeps the assignment and difficulty/skill relations but omits A → Y: the observed difference could arise through who used assistance and on which tasks. Success rates alone do not settle that difference.
The models permit different intervention consequences even when they fit the same aggregate report. Under the explicit assumption that D and H suffice to block common-cause paths, with comparable treatment versions and adequate overlap, adjustment might identify the chosen effect. Those conditions are additional premises, not results of drawing D and H. If workers choose assistance using an unmeasured expectation of difficulty, preserve that possible influence and return the identification gap.
Suppose inclusion in the showcase S depends on both use of assistance and success. A → S ← Y adds a selection path; conditioning on showcased tasks opens it. Reconstruct the eligible set and inclusion process before using the selected comparison for the population question.
Now randomize an offer Z while leaving use A voluntary. The offer's effect is a different estimand from the effect of A. If the offer also teaches a technique used without the AI, Z has a route to Y outside A. Using the offer as an instrument to learn the effect of use would require it to affect success only through use, among other assumptions. The additional route violates that requirement. The model returns the identification question and keeps observed performance descriptive until the required result is available.
The useful return can be a revised comparison population, a retained unmeasured cause, or a direct-effect route that invalidates a proposed design. A larger benchmark or more detailed simulation does not resolve those missing premises by itself.
Bias-Annotation
The scope is construction for a bounded causal question. Epistemic discipline separates a mechanism proposal from its evidence; pragmatic discipline stops at a useful conditional result; ontological discipline separates variables, their values, observations and the subject being changed.
Narrative closure favors the first fluent explanation. Diagram authority favors a tidy graph. Observed-variable bias omits causes that were not measured. Remedy these tendencies by constructing any materially plausible rival, explaining consequential omissions, and checking how observations and case selection were produced. Retaining every imaginable model can also prevent useful work; narrow the family by subject grounds and the receiving question.
Conformance Checklist
- The outcome, cases, time and required contrast can be recovered.
- Each consequential variable and relation has a subject meaning; registration and selection are explicit when they change the answer.
- Important included and excluded influences have stated grounds or remain assumptions.
- Material alternatives receive comparable questions, including coexisting causes where relevant.
- The graph or equation class supports the rule used; DAG path inspection covers all relevant paths and conditioning, including selection.
- A consequence follows from named premises, or the missing premise is returned.
- Compatibility, contradiction and absent discrimination retain different meanings.
- The returned use distinguishes recognition, model consequence, evidential support and action choice.
Common Anti-Patterns and How to Avoid Them
Consequences
The practitioner obtains explicit alternatives, a usable conditional consequence and, when a premise is missing, a statement of that premise. This can prevent spending effort on an intervention that changes only a record or on a study that cannot distinguish the live accounts.
The cost is recovering mechanisms, meaningful variables and competing assumptions. Limit that cost to distinctions that can change the question's answer or a consequential use. Some questions remain unidentified or underdetermined; a clearer statement of that limit is a useful result.
Rationale
Constructing a causal account and judging its support answer different questions. A coherent hypothetical model can expose a decisive experiment or invalidate an inference without establishing that the model describes the subject. Preserving both the constructive and evidential questions makes that intermediate result usable.
A model family preserves a live disagreement when the available grounds do not select one account. Explicit measurement and selection processes prevent uncertainty about observing from disappearing into a confident story about the subject. Time distinctions make the same discipline usable when an initiating event and an ongoing mechanism differ.
SoTA-Echoing
The comparison concerns usable construction and criticism for a bounded causal question. It does not rank causal-discovery algorithms or claim a universally best empirical workflow.
Relations
- B.5.FM constructs a general first model; this method develops the missing causal relations and returns their conditional consequences.
- B.5.2 generates and compares hypotheses. Causal-model construction can change their plausibility grounds or expose an unresolved rival.
- C.16 qualifies measurement and registration; C.27 qualifies temporal claims when those questions are needed.
- C.28 governs causal-use support, identification and evidence questions; C.28.MR derives a consequence of changing specified mechanisms.
- C.29 supplies needed mathematical representation and computation.
- C.11.DUA compares worthwhile inquiry with available action; A.15.9 obtains a missing contribution from another practice.
C.28.CM:End
Derive an Intervention Consequence by Mechanism Replacement
Type: Method pattern Status: Stable Normativity: Normative
Problem frame
Use this pattern when a causal model is available and the working question asks what follows if a mechanism is replaced: a controller uses a different rule, a variable is held at a chosen value, or an action occurs at a specified event. The model may be small enough to calculate by hand.
The subject is the intervention query in that model. The Method constructs the modified model and derives the requested value, distribution, contrast or bound under its assumptions. The practitioner needs the subject meaning of the variables and enough mathematical or computational help to solve the relevant equations.
First useful move. Point to the equation that represents the proposed change. Replace that equation and re-evaluate the variables needed for the question. If a display merely reports voltage, changing its displayed number leaves the supply equation in place. Changing the supply command replaces an input to regulation. Recalculate the delivered voltage and the current through the load. Two commands can yield the same output when the supply is at its current limit.
Use an adequate existing derivation directly. Observational reporting can finish without this transformation. If only an observed joint distribution is available, recover a causal-model assumption or an identification result through C.28 before claiming a particular intervention consequence. Identifying an effect from data is a separate task from deriving it in a supplied model.
Problem
An observed value tells us something about how the existing mechanisms operated. An intervention changes a mechanism. Conditioning on the observed value can therefore answer a different question from imposing that value.
A second difficulty appears when a mathematical or software operation has an unclear physical interpretation. Assigning a value to an indicator, replacing a controller and forcing a physical quantity can change different mechanisms. Reusing the same variable name hides that difference.
Even a clearly specified replacement can change the solvability of a model or its subsequent dynamics. The useful result requires both the model transformation and a derivation for the question asked.
Forces
Solution
Choose the intervention and question → recover the mechanisms and input law → replace the selected mechanisms → solve the affected relations → obtain the requested consequence → return its interpretation and useful limit.
These moves form a derivation. A few equations can carry the complete result.
Specify what is changed and what is asked
Name the modeled variable or mechanism, the proposed replacement and the result needed for use. Include the time or horizon when it changes the question. “Set the alarm to zero now” and “keep its output at zero for the next hour” describe different interventions in a model with later feedback.
For a constant intervention, use do(X=x) to mean replacing the equation for X by the constant x. For a changed rule, state that rule and the information it may use: for example, a controller's command as a function of the readings available before the command. C.28:4.10 distinguishes the relevant action-policy families when that choice matters.
Specify a comparison only when one is wanted. A value under one intervention, a difference between two interventions and a difference from natural behavior have different required inputs. A question about the same historical case under another action also needs the factual observations and the resulting information about that case's underlying inputs. Infer those inputs using the factual mechanisms and observing procedure before replacing the mechanism. Carry the inferred distribution or compatible possibilities into the alternative calculation, preserving the dependence that makes it the same modeled case. C.28 governs what supports that counterfactual use; a population intervention average can answer a different question.
Recover the model that gives the replacement meaning
Write or recover each relevant mechanism in a form such as X_i = f_i(parents, U). The function tells how the modeled quantity is determined by other modeled quantities and by U, the inputs left outside these mechanisms. Recover their domains and the joint distribution of U, or their given values for a deterministic calculation.
Keep the subject interpretation with the equations. An equation's left-hand side is selected by the causal account. Algebraically rearranging I=V/R to V=IR preserves the equality but does not decide which physical component an intervention replaces. In :5.2 the supply controls V and the resistor relates V to I.
Preserve common inputs and their dependence. If two disturbances arise from one shared event, retain that event or their joint law when calculating both effects. A simulation must generate one compatible joint input for a case; independent resampling of its components would change the model.
Recover only the mechanisms needed to determine the queried consequence and their relevant conditions. A missing actuator relation may prevent a physical consequence while allowing a useful comparison inside an assumed model. State that assumption where the result uses it.
Construct the modified model
Replace each targeted mechanism by its stated constant or function. Retain the other mechanism functions and the supplied input law. Then let their output values follow from the new inputs to those functions.
For a randomized replacement, give its randomization law and its dependence on existing inputs. An independent randomizer is one possible specification. If the proposed change also alters the environment or population, include that change in the compared model instead of treating it as part of a constant intervention.
Keep a replacement's scope visible in time. In a discrete-time model, identify the affected update steps. In an event-driven model, identify the triggering condition and the state change it initiates. A.3.3.TR helps construct the relevant continuation rule from interactions, events and subject laws when that rule is missing.
For physical interpretation, ask whether the proposed action also changes another modeled mechanism. Changing a supply command and attaching a device that forces its output can have different consequences for current limits and other connections. Add the relevant effects before using the modified model to answer that physical question.
Solve what the intervention query requires
For an acyclic model, evaluate the retained and replaced functions in dependency order. For a time-stepped model, propagate from its initial conditions through the selected horizon. For simultaneous equations or continuous dynamics, use the solution method and conditions appropriate to those relations.
Check the transformed relations when a replacement can change existence or select among several solutions. A solver's first returned branch may leave other admitted answers. If all admitted solutions agree on the queried quantity, that agreement can answer the question. If their answers differ, retain the possible answers, derive a useful bound, or expose the condition needed to choose among them. If the model has no relevant solution for inputs the query includes, return that obstruction and repair the model or question.
For probabilistic results, the selected solutions must define the requested random quantity under the input law. Existence of one numerical trace supplies less than a uniquely determined distribution. A general proof is unnecessary when a finite calculation settles the present query.
For a dynamically triggered jump, recover the event-order or flow/jump choice if it changes the result. A model's steady-state solution alone does not determine what happens during a transient. The hybrid-system approach discussed in :11 supplies one treatment with explicit conditions; the subject model determines which treatment is appropriate.
Calculate the consequence under the retained conditions
Let Y_g(u) be the queried output obtained with replacement g and input u. For a finite input population with probabilities p(u), its mean is
E[Y_g] = sum_u p(u) Y_g(u).
Use the corresponding integral for a continuous input law when it exists. For a contrast between g and h in the same population, evaluate both under that population's law. A paired simulation can reuse each joint input u for both computations. Preserve the dependence of inputs within each computation.
A symbolic expression, an inequality or a bound can be sufficient. Numerical precision should serve the receiving choice or explanation. If an estimate is required, distinguish its computational or sampling error from uncertainty about the causal model and its parameters.
Return the result and reopen only what changes it
Return the consequence together with the replacement, model assumptions and comparison or horizon needed to interpret it. The equations and answer already produced may suffice. Use C.28's reusable support form only when another use needs a referenced conclusion.
Keep the mathematical result conditional on the supplied causal account. When the receiving question concerns an actual device or organization, connect the proposed action to the modeled replacement and ask which unmodeled interaction could change this result. Obtain further evidence only when that unresolved question can change the use. B.5.MPC connects the physical account, mathematical construction, computation and observation when this passage needs work.
When a parameter, intervention scope or mechanism changes, revisit the dependent derivation. Preserve unaffected relations and calculations. A changed event time can require another continuation; a changed source of disturbance can require another joint input law. Stop when the requested result or a useful supported limit is available.
Archetypal Grounding
Observing an alarm and forcing its output
Let H mean binary high load and S mean a binary alarm. Take independent uniform inputs U_H,U_S on [0,1] and the supplied mechanisms
H = 1[U_H < 0.5]
S = 1[U_S < 0.1 + 0.8 H].
Here 1[condition] is one when the condition holds and zero otherwise. In high-load cases, the alarm sounds with probability 0.9. In low-load cases, it sounds with probability 0.1.
The observational probability of no alarm is 0.5*0.1 + 0.5*0.9 = 0.5. High load with no alarm has probability 0.5*0.1 = 0.05. Thus P(H=1 | S=0)=0.05/0.5=0.1: the observation changes what we infer about the load.
For do(S=0), replace the second mechanism by S=0. The equation for H and the input law are retained. Therefore P(H=1 | do(S=0))=0.5. Forcing this indicator removes its information about H without changing H in the supplied model.
For a particular past episode the answer uses what was observed in that episode. Suppose the alarm sounded and load was unobserved. The factual mechanisms give P(H=1 | S=1)=0.45/(0.45+0.05)=0.9. In the alternative where the alarm was silenced at that instant, its replacement leaves H unchanged for each compatible underlying input. The probability of high load in that alternative given the original alarm is therefore 0.9. Drawing a fresh episode from the population would give 0.5. Treating the imposed silence as an observation of a naturally silent alarm would give 0.1. Each number answers a different question.
All three calculations concern current load under the stated absence of an S-to-H influence at that instant. If alarm output controls subsequent cooling, a later-load query needs that later mechanism and the duration of the override.
Change a display or a supply command
Consider an ideal regulated supply and a resistive load. C is the command in volts, V the delivered voltage, D the displayed voltage, R a fixed resistance, I the load current and W its power. In the operating region being modeled,
C=12 V; V=C; D=V; R=6 ohm; I=V/R; W=V I.
The natural result is I=2 A and W=24 W.
Replacing the display equation by D=6 V leaves V=12 V; current and power remain 2 A and 24 W. Replacing the command equation by C=6 V instead gives V=6 V, I=1 A and W=6 W. The two changes share a numeral but replace different mechanisms.
Suppose the supply's current limit is now relevant. Use the revised positive-command model V=min(C, I_max R) with I_max=1.5 A. The 12 V command now gives 9 V at the load and 13.5 W. The 6 V command still gives 6 W. Relative to natural behavior, the calculated power reduction is now 7.5 W, instead of 18 W in the earlier operating-region model.
To use this comparison for a physical supply, establish its regulation, load and limiting behavior. Editing a displayed number can be a real change to the display. The calculated effect on load current applies to a change of supply command. Establish which mechanism the interface changes and whether a controller later acts on its reading.
A definite intervention answer with an indefinite baseline
Take real-valued variables with mechanisms X=Y and Y=X. Every equal pair satisfies the baseline equations. Without another condition, the model does not determine the baseline value of Y.
Replace the first mechanism by X=1. The retained equation now gives Y=1. That intervention consequence is determined. A difference from the baseline remains undetermined because the baseline's value or distribution has not been supplied.
If the requested result is only Y under the intervention, return one. If the requested result is its change from natural behavior, return the missing baseline condition. Resolving the first question does not require inventing that condition.
Bias-Annotation
Small equation examples favor deterministic reasoning and can hide shared disturbances, partial observation and delayed effects. Carry the joint input law and intervention horizon into stochastic or dynamic applications. The directness of a software assignment can also bias the modeler toward treating any variable as independently controllable. Recover the physical or organizational mechanism that would implement the proposed change.
The examples illustrate derivation under supplied assumptions. They do not estimate how often those assumptions hold in another practice. A user can retain a useful conditional result while investigating a particular assumption that matters to its application.
Conformance Checklist
- The intervention and queried consequence refer to the intended model and population, with their times or horizon explicit where they affect the answer.
- The replaced equation or rule is recoverable, including any inputs used by a policy or randomizer.
- Other mechanisms remain in force; their values are recomputed where the replacement affects them.
- The joint input law and common causes needed by the query survive the computation.
- The transformed model determines the reported answer, or the result retains the relevant alternatives, bound or obstruction.
- A comparison uses the required baseline and conditions; an intervention outcome alone is not reported as an estimated change.
- The result's physical use connects the action to the modeled replacement and any consequential additional effect.
- The amount of calculation, recording and new evidence serves the receiving question.
Common Anti-Patterns and How to Avoid Them
Consequences
The practitioner obtains a reproducible model-based consequence and can locate which proposed change produces it. This supports comparisons of control rules, experimental proposals and organizational mechanisms under their stated assumptions.
The method also exposes useful partial results. A fixed intervention outcome may be available when a baseline contrast is not. A bound may already separate two choices. These results preserve progress without requiring a complete empirical programme.
The main effort lies in recovering the mechanisms and conditions needed to answer the query. More elaborate equations can increase solving cost while leaving a decisive omitted interaction untouched. A small model is preferable when it already resolves the receiving question under adequate assumptions.
Architectural Rationale
The construction works on mechanisms because the question concerns what would change if a different way of determining a variable operated. Statistical conditioning instead uses information about values generated by the current ways. Both operations are useful; :5.1 shows when they answer different questions.
Retaining functions while recomputing values carries the effects of a local replacement through the model. Keeping the input law retains the population and common causes against which those effects are compared. This makes the difference attributable to the specified modeled change rather than a change of background.
Acyclic evaluation is the simplest executable case. Coupled constraints and dynamic events require attention to the transformed system's solutions. Query-specific determinacy keeps that requirement proportionate: the relevant question can have one answer even when another part of the model is unresolved.
Physical enactment, empirical identification and model-based derivation provide different needed results. A supplied model is enough for the conditional derivation. Applying that consequence to the world additionally needs a suitable connection between action, model and observations. This arrangement lets a theoretical construction contribute immediately while exposing the particular additional result its physical use needs.
SoTA-Echoing
The corrected Pearl, Glymour and Jewell primer, p.55 provides the classical mechanism-replacement account and explains why an intervention with additional direct effects needs a richer model. It supports the ordinary constant-intervention construction here.
Bongers, Forré, Peters and Mooij (2021), Foundations of Structural Causal Models with Cycles and Latent Variables, particularly :2.4 and :3.4, shows that solvability can change under intervention. The model transformation and its induced distribution therefore need separate consideration. Its perfect-intervention results do not automatically establish the properties of every replacement policy.
Zane et al. (2025), A Counterfactual Semantics for Hybrid Dynamical Systems, :3-4, treats instantaneous, state-triggered interventions as transformations of flow and jump constraints. Under stated assumptions it preserves existence, uniqueness and measurability over the modeled horizon. This informs the event and solvability boundary; it does not impose one hybrid-system formalism on every causal query.
The synthesis retains ordinary substitution as the small executable case, adds the solution question where it matters, and connects the result to the subject interpretation. Domain-specific causal discovery, identification and estimator choices remain governed by their methods and evidence through C.28.
Relations
- C.28 recovers the causal-use question, distinguishes intervention and counterfactual reliance, and combines the support results needed by a receiving use. This member supplies the derivation inside a given model.
- C.28:4.10 distinguishes action-policy regimes when the replacement's available information changes the causal question.
- A.3.3.TR constructs continuation rules from interactions, operations and subject laws when time or events matter to the replacement.
- B.5.MPC connects the physical account, mathematical construction, computation and observation used to interpret the consequence.
- B.5.RR revises the affected reasoning when a premise or mechanism changes.
- C.29.1 checks transfer of a mathematical result into another representation or use; this member constructs the intervention result that can later be transferred.
C.28.MR:End
Mathematical Lens Use
Type: Architectural pattern Status: Stable Normativity: Normative unless marked informative
Plain-name. Mathematical lens use.
Primary EntityOfConcern. The use of a mathematical representation to answer a stated working question, with an explicit correspondence to the phenomenon and limits on the resulting inference.
Use this when. Use C.29 when choosing or transferring a mathematical representation could expose a needed relation, invariant, obstruction, approximation or resource limit, or when an existing representation is being relied on beyond what its correspondence supports. The related Methods in C.29.1, C.29.2 and C.29.3 respectively construct a result transfer, a computation and its realization. Enter the Method whose contribution is missing; B.5.MPC connects these contributions to a physical question.
What goes wrong if missed. The reader either misses a useful mathematical construction or carries a result into a situation where a needed assumption or distinction has been lost.
What this buys. A mathematical construction that makes the needed inference or operation possible, with grounds for using its result in the work that called for it.
Not this pattern when. An adequate local equation, algorithm or domain model already answers the question and no separate representation or transfer issue remains. Complete that work without a C.29 note. A one-off metaphor used only for orientation also needs no card.
First mathematical move
Start with what the recipient needs to understand or do. Work back to a relation from which they could obtain that result. Which objects take part, what may vary, and which differences must their representation retain? Choose an adequate mathematical construction, derive its consequence and use the correspondence to answer the original question. A familiar law can supply the whole argument.
For waiting work, a queue candidate turns “slow flow” into questions about arrivals, service and waiting. For a proposed dense state array, counting its entries can already rule out an implementation. For a familiar local calculation such as V = IR under its accepted conditions, use the equation directly: the calculation does not require a lens-use record.
If no concrete object can yet be chosen, retain the working cue and the next observation or construction as a candidate. B.5.FM helps when the participants or interactions needed to formulate that mathematical question are still unclear. A family label remains a discovery cue until a concrete object and correspondence are supplied. Recording options in :4.4 serve later inspection or reuse of the result; the mathematical operation comes first.
Problem frame
A mathematical representation can make a working question answerable by exposing the relations needed for an inference. It may combine source cases into a summary, change coordinates without losing distinctions, or embed the source in a larger mathematical domain. Each construction supports different uses. A queue may expose flow restrictions while omitting rework; an extension of the rational numbers to the reals makes limits available while retaining rational arithmetic.
C.29 addresses this representation choice and its use across a stated correspondence. It begins either with a working problem that needs a useful construction or with a proposed representation whose consequence needs to be derived, limited or rejected. Domain theory supplies the mathematical laws and application conditions.
A useful mathematical lens makes a needed inference possible through an explained correspondence.
Ask what the correspondence preserves, omits or introduces, which operations it supports, and how the resulting conclusion answers the working question.
First-minute working situation
A production manager sees waiting work but cannot tell whether an extra station would help. The first task is to distinguish an arrival restriction, a service bottleneck and a batching effect. The candidate queue has to make at least one of those distinctions calculable or observable. The discovery cues in :4.2b help choose an object; they are not a required tour of mathematical families.
Minimum scenario and anti-case set
Positive scenario. In :4.4's two-station queue, the departure recurrence distinguishes a five-minute latency reduction from a change in sustained output. The next observation focuses on the slower station. Changed service time or finite-buffer blocking withdraws the affected calculation; actual capacity reliance requires checking the model against the line.
Anti-case. “The organization is a quantum system” supplies no mathematical object, correspondence or derived consequence. Keep a local metaphor when it helps orientation, or construct the needed model and its use boundary through C.29; apply C.26 when a residual contextual-model obstruction warrants that branch.
Under-lensed anti-case. “The work stream has dynamics” or “this portfolio is a network” is offered as a diagnosis, but no model supplies its claimed consequence. Construct a candidate that could answer the diagnostic question, or keep the sentence as ordinary orientation.
False-positive scenario. A Markov kernel appears inside accepted local reliability modeling. When no separate representation or transfer question is open, complete the A.3.3 result without a C.29 output. If such a question arises, use the reliance rule in :4.4.
Intended FPF use-value
The useful result is an answer, bound, obstruction, diagnostic split or next observation that the working reader could not recover from the prior account. Keep the correspondence and limitations inspectable when someone will reuse that result. A more elaborate record is useful only for the information its receiving use needs.
Problem
A broad name such as “graph”, “field”, “dynamics” or “quantum-like” does not yet identify the mathematical object or yield a consequence. Conversely, a valid derivation inside a model may be transferred to a case whose necessary structure the representation does not preserve.
The practitioner therefore needs both construction and criticism: make the relevant structure explicit, perform the mathematical operation, and test whether its result answers the original question under the stated conditions.
Forces
Solution - selected answer
Construct and test the correspondence
- Recover the work the mathematics must support. What was requested, who is to do it, and which choices were left to them? Establish what counts as the required result, under which conditions, over what period and with what precision. If plausible interpretations give different answers, use their conditions to distinguish the answers or establish which interpretation governs the work. An explanation needs a derivation connecting the premises to the outcome. To construct a way of working, supply operations people can perform and show how their results fit together. For a particular execution, use or obtain the result of that execution. To keep a service working as conditions change, determine which property must continue to hold and what those responsible can observe and do in time. Identify the mathematical contribution that the work still needs.
- Construct a representation for that use. Start with the relation or law that could supply the result. When the answer depends on composition, hold a part fixed and vary how the whole uses it. Then hold the reported whole fixed and vary how its parts produced it. Which variation could change the required result or the grounds for explaining it? Follow each relevant kind through its own constituents. The stages specified by a method, the operations performed on one job and the records of those operations can have different compositions. Connect them by the relations the work requires. Ask whether the question concerns what holds for each constituent, how one affects another or what they produce together. At each relevant scale, keep distinctions on which that use depends. Try treating the remaining cases alike and work the proposed inference. Restore a distinction if cases so grouped require incompatible results or different grounds the recipient needs. Across different models for the same inquiry, identify the alternatives each admits and which conclusion transfers in which direction. Stop when the correspondence supports the work. C.39.SA develops this choice when the range is open. Several representations can suffice. If a mathematical object is still missing, retain the candidate and the next useful construction or observation.
- Establish the correspondence. Say what each relevant element or operation represents and in what direction the inference is used. Distinguish an analogy, a fitted representation, a simulation and an exact structure-preserving map. State the assumptions, scale and context that the correspondence requires.
- Determine what the correspondence preserves, omits or introduces. Identify the structure on which the intended result relies and any omitted distinction that could change it. Establish the preservation claimed for the needed operations. If the receiving domain adds objects or operations, determine which of their results answer the source question. C.29.1 constructs these comparisons, including cases with no loss of source distinctions and cases where a receiving solution has no source counterpart.
- Do the mathematical work. Calculate, derive, construct or demonstrate the obstruction. For a missing computational formulation or procedure, use C.29.2; for an unsettled connection to an executing system, use C.29.3. Return the result through the correspondence to the working question. A statement that a queue “exposes bottlenecks” is not yet the bottleneck calculation.
- Return the result to the work. For an explanation, show which relation and premises make the claimed outcome follow. For a construction, return the object or operations and establish the properties required of them. To perform the work, apply those operations and return the requested outcome. For continuing use, identify which changed condition affects preparation, execution or interpretation and how to preserve or restore the required result. A conclusion about an actual system needs grounds for the correspondence and premises there. If a choice is required with several states still possible, an action that works in all of them can suffice. A later observation can support different actions when it reaches the chooser in time. A bound or an existing result can complete the commission. Use :4.4 when someone needs to recover the reasoning.
Transfer the result through the operations
A source operation or result is available, and you need to use it through another representation. Use C.29.1 - Mathematical Result Transfer to recover the operation and its permissions, compare performing then mapping with mapping then performing, and test whether choosing another represented source case changes the answer.
Its Solution constructs a transferable consequence, a justified bound or a specific repair of the correspondence. It also separates a receiving operation that gives a useful relaxation from one whose result can be realized as a source action. The reservation and route-cost entries in :7.1–:7.2 lead to its worked constructions.
Construct a computation for the question
The needed answer is identified, but the available representation and operations do not yet explain how to obtain it. Use C.29.2 - Computational Formulation to select the distinctions retained in computational state, construct a procedure or solver formulation, establish the claimed result and estimate the cost that matters to its use.
Its Solution starts from either a question or available computational means. Worked cases develop an interpreter, a bounded root calculation, a memory-limited quantum-state representation, two questions about the same circuit, and a probability estimate. The result is a usable computation under stated conditions or the missing construction that prevents it.
Realize the computation and read its result
A computation is available, but how a concrete system performs it is unsettled. Use C.29.3 - Computational Realization to connect input preparation, system actions and result interpretation, then compare the required result with the interpreted execution. For a design calculation, the comparison uses the supplied system model; a claim about an actual run uses its observations.
The result is a conditional realization, an interpreted result or a located failure in preparation, operation or readout. Digital, analog, stochastic and manual arrangements can enter through the same question. Its robot, analog addition and admission-card cases show repairs of range, scale and shared-state conditions. B.5.MPC coordinates these results with the physical account, mathematical construction and receiving action.
Mathematical Lens Use Principle
A mathematical lens is useful for a stated question when its correspondence preserves the structure needed for a consequence and makes the material losses explicit. A proof under mathematical assumptions establishes that consequence inside the model. Reliance on the phenomenon additionally requires the correspondence and application conditions to hold; prediction and other model-bearing uses require their validation under :4.5a.
Plain phrases such as “what survives transfer” can guide recognition. For a used result, make the correspondence, preserved structure, loss and stopping condition recoverable. Include a blocked overread only when it passes F.19's plausible-reader test.
When mathematicalization pays
Introduce a representation when it changes what can be derived, compared, observed, constructed or ruled out for the working question. An adequate ordinary calculation may already supply the answer. If no useful consequence or next inquiry follows, keep ordinary prose or the domain result; no C.29 output is required.
Discover a candidate from the working cue
Choose the row that fits the problem, or use a closer domain construction. The menu is informative. Recover the elements and conditions of the particular construction selected; alternatives in a row do not all require the same structure. The shared test is :4.1.
For a constrained extremum or stationary construction, construct or reuse a family of allowed candidates and calculate the resulting change in the target quantity. Distinguish an improving candidate, a necessary stationary condition and a justified optimum; carry the resulting conclusion and its assumptions into the working question.
For a symmetry argument, follow the transformation through the problem's data, conditions and required answer. Derive the transferred solution, restriction or obstruction to selection, and say which problem that consequence answers.
When a balance must be constructed or its boundary changes, C.29.BB identifies the additive quantity, included stores and crossing transfers. It combines compatible accounts and returns the total, a bound or the missing contribution.
For a first candidate, compare with ordinary prose, direct observation or the accepted domain model before a broader survey. When the question is a tradition-scale source synthesis, use G.2; C.29 needs only the candidate or rival relevant to this working question.
When expected information gain determines which observation to obtain, choose a feasible way to estimate it. If density approximations are used, allocate samples between fitting them and computing the expected gain. In high dimensions, consider reducing the parameter or observation space and account for the information lost by that reduction. Compare estimation cost and error with the distinction needed for the choice. Li, Baptista and Marzouk (2026) develop these sample-allocation and dimension-reduction choices.
Bounded-observer structural-information lens
Use this subcase when a mathematical lens estimates, compresses, codes, compares, or otherwise exposes how much selected structure a bounded observer can recover from a description, relation trace, generated graph, model, or reusable-structure accounting result. Typical examples include MDL-like two-part codes, epiplexity-style extracted-structure estimates, compression-complexity comparisons, and information-functionals over relation graphs.
C.2.8 defines extractable structural information for the episteme, the particular expression conveying it and the observer under stated conditions. When this lens output is used to estimate that characteristic, state how its mathematical objects, admissible models, selected structure and resource bound correspond to that expressed account, observer and comparison conditions. A comparison that uses an adequate domain method directly needs no mathematical lens.
For an epiplexity estimate, distinguish the selected model's description length from the residual data description, the model-execution bound from estimation effort, and conditional model information from all familiar structure a reader can recover. Use the existing source, mapping, preserved/lost-structure and stop fields to state the correspondence. The formal model and its application conditions are explained in C.2.8:4.6; a numerical measurement claim also uses C.16.
Content to recover for this use, in the sufficient form selected under :4.4:
When the target is a physical, organizational, or project-world situation, the account must say whether the structural-information claim is observational, postulated, simulated, or only a description-local compression. When the lens is used in architecturing, C.30 governs architecture as EntityOfConcern, C.30.ASV governs structural-view adequacy, C.30.AD governs architecture descriptions, and C.31 or C.31.RSA governs modularity or reusable-structure accounting. C.29 records only the declared lens use: what recoverable structure the mathematical lens makes visible, what it loses, and where that use stops.
Architecture-local lens descriptions
Architecture work may use C.29-local descriptions for graph, flow, control, structural-information, RG or coarse-graining, and multilevel-learning or frustration lenses.
MLU.Description@RGArchitecture applies only when the use names a declared aggregation scope, scale variable or scale window, coarse-graining rule, preserved structure, lost structure, source-return condition, declared use, and stop condition. Include a blocked overread only when it passes F.19's plausible-reader test. If the claim becomes a scale-preference claim, C.31.ASAP governs the architecture preference side; C.29 keeps the declared mathematical-lens use.
Content of an RG architecture description, using an adequate existing account where available:
For architecture work, a common RG-shaped candidate object is:
InterfaceSpecificationRefs_l contains only U.EpistemeRef values that resolve independently identified InterfaceSpecification epistemes under the identifying rule located at A.6.M. The references, their resolution, and the specification content remain separate; changing a lens token or retargeting a reference does not edit the specification.
The index _l selects the aggregation scope declared for this representation. State which parts form the whole at that scope and which distinctions the coarse-graining rule leaves open for the requested inference. The index can group system components while the referenced work or specification concerns a different whole. Use the relation to that work or specification when its result supplies a premise. Return to the finer account when a distinction it retains changes the conclusion.
MLU.Description@MultilevelLearningFrustration applies only when the use names declared holon levels or declared scopes, a mapping between them, conflicting constraints or residuals, preserved structure, lost structure, and the nearest neighboring FPF pattern for any measurement, causal, evidence, assurance, work, selected-set, or decision claim.
Content of a multilevel-learning and frustration description, using an adequate existing account where available:
Declared lens use: triage, explanation, candidate generation, rival-lens comparison, scale-window reasoning, source-return triggers, and architecture-decision rationale only when the neighboring pattern defines or constrains any non-C.29 claim.
Stop or return when the mapping, scale window, preserved structure, or source basis no longer supports the use. A causal-proof, assurance-score, or necessary-complexity-growth claim needs its own argument and source basis. Apply C.11, C.28, B.3, C.16, G.5, or the applicable stakeholder or ethics pattern when its respective non-C.29 claim is current. Include a project-wide-optimizer or other blocked overread only when it passes F.19's plausible-reader test.
Use boundary
Use C.29 for a consequential choice or transfer of mathematical representation. Ordinary equations, data structures and accepted domain models remain usable directly when no such issue is open. A separate causal, measurement, dynamics, semantic-bridge or other claim uses its subject pattern under :4.4.6.
In discovery prose, name the correspondence established. Use structure-preserving representation when the required preservation is justified; name equivalence or identity only when that stronger relation has been established.
Record the result for its receiving use
First complete or delimit the mathematical move in :4.1. Record its question, concrete object, correspondence, retained and lost structure, consequence and next action when another reader or later reliance needs them. Choose the smallest sufficient form below; a form does not replace a missing construction, derivation or observation.
One reliance rule. A conditional derivation, diagnostic comparison or reusable explanation may remain a MiniCard when it retains the same object, assumptions, correspondence, loss and narrow use, and its consequence is used to understand the model or choose the next inquiry. A FullCard is required when the result is relied on as an adequate model of the phenomenon for prediction, an operational or consequential decision, model adoption, benchmark or assurance input, Bridge-dependent reliance, or transfer as a reusable model to further cases. It then needs the applicability, rival and validation information relevant to that reliance. Publication or repetition of the same conditional explanation alone does not change its class.
Keep adequate existing domain/model documentation and refer to it; FullCard means a recoverable full account, not recopying that information into a blank form. Reassess the same rule whenever the intended use or reliance changes.
Application output classes:
Micro-template examples:
Architecture and P2W first-use slice:
The heat balance is the proposed first mathematical operation. A formal vocabulary/law declaration uses A.6.0 only when that separate declaration is needed; carrying accepted problem-side distinctions into later work uses E.18.1. The receiving conditions are stated in :4.4.6.
Worked conditional queue comparison. A production manager asks whether speeding station A would raise the line's output. The manager may need the sustained production rate or the number of finished jobs by a specified deadline. Keep both questions open until the intended use selects one. To choose the model, hold the two station times fixed and ask what could change the answer. If each station independently completes a different product, there is no handover between them. If every item must pass through both in order, release from A makes that item available to B, and B may still be occupied by the previous item. For the latter arrangement, keep each item's arrival and each station's availability. They determine when the next completion can occur. For a count of yesterday's departures, a complete and reliable dispatch record could already supply the answer. Here the manager asks about a proposed change, so the model must also represent what the change would leave in force. Use a stipulated case: identical jobs arrive at a_n = 10n minutes, starting with n = 0, and pass through A then B. Both stations have one server, first-come-first-served non-preemptive service, no failures or rework, and an unbounded intermediate queue in the model. For a finite run the same calculation applies while its actual buffer does not fill. Transfer time is zero. A takes 10 minutes per job; B takes 15. The line starts empty.
The mathematical object is a deterministic two-station tandem queue. A job represents one part; an arrival is its release to A; service is uninterrupted processing at the named station; a departure from A is the arrival at B. This preserves order, routing, occupied service time and interstation waiting. Variable product mix, stoppages, rework and finite-buffer blocking are omitted. The network correspondence and the importance of blocking are introduced in Wu's queueing-network lecture, slides 9 and 23–24; the following deterministic numbers and derivation are constructed here.
A job can start only after it has arrived and the previous job has left that station. With previous departures initialized to zero, its departure times therefore obey:
The recurrence gives d_A(n) = 10n + 10 and d_B(n) = 15n + 25: after its first arrival B remains occupied. The output spacing is 15 minutes, hence the long-run rate is 4 jobs/hour. The station-capacity bound is min(6, 4) = 4 jobs/hour, attained in this stipulated run. Latency is d_B(n) − a_n = 25 + 5n minutes. It keeps growing; the calculation supplies no finite steady-state mean delay.
Now halve A's service time to 5 minutes while leaving the arrivals and B unchanged. The same recurrence gives d_A(n) = 10n + 5 and d_B(n) = 15n + 20. The rate is still 4 jobs/hour; each job's latency falls by 5 minutes but still grows with n. The useful distinction is between a shorter initial traversal and a higher sustained output rate. For a requirement of two finished jobs by minute 35 after an empty start, count the departures up to and including that minute. The original line has one, at minute 25. With faster A it has two, at minutes 20 and 35. For sustained output, investigate B’s service, interruptions and queue growth. For the deadline, also establish that the actual line starts and operates as modeled. An observed-output dashboard is the ordinary fallback and a complementary check; output counts alone do not derive the two station-change scenarios.
Use and return. This is a MiniCard-level conditional derivation and comparison, reusable with the same premises. It directs observation, not a capacity commitment about an unvalidated line. Actual prediction, equipment selection or operational reliance uses :4.4's full applicable account and :4.5a's validation. Keeping A's original 10-minute service, if an omitted inspection adds 5 minutes to every B service, B's service becomes 20 minutes: d_B(n) = 20n + 30, the rate is 3 jobs/hour and latency is 30 + 10n. Withdraw the 4-job/hour and 25 + 5n results. If finite storage instead blocks A, reconstruct the blocked-service recurrence before reusing the schedule.
For MathLensUse.OneLine:
VisiblePayoffsays what the lens makes visible, such as a bottleneck, invariant, obstruction, incompatibility, loss boundary, or diagnostic split.NextLensUseActionstates the reader's next action within the result's limits, such as computing a local quantity, comparing only inside a declared structure, running a validation slice, applying a neighboring pattern, keeping the phrase as local metaphor, or removing the phrase from claim-affecting use.ObservationOrReadoutNeeded?names the missing observable, readout, assignment, outcome, validation slice, or scale point needed before the repaired line makes the stated action usable.OrdinaryRivalOrFallback?records a relevant available alternative when its comparison changes selection or use: ordinary prose, accepted local domain theory, direct measurement, a causal model, a queueing model instead of a quantum-like metaphor, anA.19space declaration instead ofC.29, or anF.9bridge instead of category-like wording.
If two mathematical lenses already change the next action at this cheap-output class, add one ordinary-language note about the disagreement and use MathLensUse.MiniCard or MathLensUse.FullCard before claiming a reusable rival-lens relation.
No extra rival is required for an already justified conditional result; if a consequential choice lacks a known adequate fallback, retain that limitation rather than invent one.
MathLensUse.LensCandidateNote records a first candidate. The next result may be ordinary work with no further C.29 output, MathLensUse.OneLine, or the result of the neighboring subject pattern. Retain a no-lens or neighboring-pattern note only when its explanation is useful.
Do not use MathLensUse.OneLine with an empty CandidateMathObject. If the candidate object has not yet been named, use MathLensUse.LensCandidateNote first, keep ordinary prose, or write a NeighborGoverningPatternNote when a non-lens claim is being made.
Cheap stop: if the mathematical phrase does not affect any claim beyond orientation, do not use the full card. If the first output is NoMathLensUseNeededNote, that is a successful C.29 result, not an underfilled card.
Output set and declared-use boundary
Use the single output/reliance rule in :4.4. The form states the mathematical account; a used empirical, causal, semantic-bridge, assurance or decision result additionally needs its subject-pattern basis in :4.4.6.
LensMappingMode, LensUseBoundaryValue, and declared lens use are separate fields.
LensMappingMode names construction, not permission. Typical local values include representation, abstraction, quotient, coarse-graining, embedding, homomorphism, isomorphism, functor-like transfer, simulation, and learned or fitted representation. A broad family name such as graph, field, category, geometry, quantum-like, variational, or Bayesian is only a prompt until the concrete construction and preserved structure and lost structure are named.
LensUseBoundaryValue declares only a limited lens-use boundary:
State the declared lens use in declaredLensUse and its stopping or return boundary in StopCondition. Elegance, familiarity, source prestige, and mapping type supply no substitute for that declaration. Include blockedLensOverread? only when it passes F.19's plausible-reader test.
From lens to local action
Use the mathematical consequence at the point where the recipient needs it. For example, to explain a line’s throughput limit, follow the jobs to a station each must visit and derive the maximum rate at which that station can serve them. Attaining that rate also requires a compatible flow through the other stations. To construct a schedule for successive jobs in a given order, carry forward each station’s next available time and start a job when it has arrived and the station is free. For a question about actual deliveries, a complete dispatch record may directly establish the count. These uses concern the same line and draw on different grounds. Where a separate subject practice supplies a missing contribution, use the corresponding entry in :4.4.6 and bring its result back to the work.
Worked local bottleneck: dense-state storage. A proposed dense pure-state representation for 50 qubits contains 2^50 complex amplitudes. At a stipulated 16 bytes per amplitude, its array alone requires 16 × 2^50 = 2^54 bytes, or 16,777,216 GiB. A 64 GiB implementation cannot hold it. Counting the representation's elements and multiplying by storage per element identifies this bottleneck before implementation.
Recover the requested observable or property, then reconsider which structure must be represented. A restricted case, another representation or a justified approximation may change the resource demand; estimating each candidate's storage and work can change the choice. When a particular representation change has been chosen, A.6.3.RT:4.1 supplies the conversion and preserved/lost-content check. Until a suitable domain algorithm is supplied, the present result is rejection of the dense implementation and a specific representation question. Ordinary storage arithmetic alone needs no lens card.
No-lens entry: choosing a first candidate lens
Use this when a mathematical construction could help but no adequate object has been named. Recover the working cue and the consequence needed, choose a plausible construction from :4.2b or a domain Method, and obtain its first object or relation through :4.1. A LensCandidateNote can retain that proposal and the next useful action. Compare an available ordinary fallback when it can settle the choice; broaden the search only for a remaining consequential question.
If the difficulty is not yet articulated, use C.2.LS, A.16, A.16.1, B.4.1 or B.5.2.0 for the missing contribution. A concrete observation, variation or model sketch can help articulate the question; do not require a completed mathematical object to begin that inquiry. Construct a measurement or experiment only when the receiving question needs that result.
First C.29 entry cases
For E.11-style first-entry recognition, distinguish the working entry case before choosing an output:
False-positive bank and entry stops
An ordinary ODE in physics, a Markov kernel in local stochastic dynamics, a graph data structure, an A.19 distance/topology/order/embedding, a category-theoretic proof internal to its domain, or a one-off teaching metaphor needs no C.29 output merely because mathematics appears. The same applies to Markov-blanket wording used only to recognize a physical interface or boundary already recovered through its subject pattern.
Enter C.29 when a separate representation or transfer issue affects the result: unexplained waiting may need a queue construction; an important comparison may need a distance with explicit losses; transferring the same graph between contexts may need both mathematical correspondence and F.9 semantics. A learned representation used for scientific explanation needs :4.5a's observation and validation conditions. A scale claim needs the applicable scale-law or preference argument in :4.4.6.
If no stable structural question can yet be stated, continue only the useful articulation or inquiry identified in :4.4.3. An inconclusive candidate, a rejected lens, an ordinary local answer and a receiving subject-pattern result are all useful stopping points.
Subject-pattern boundary table
Use this table when the mathematical result contributes to a separately governed question. Name that question and the first receiving action; cite the existing result when it already supplies the needed basis.
CandidateMathObject names the mathematical object used in the representation. State its correspondence and preserved and lost structure in the accompanying account.
For a relation claim with clear participants and meaning, apply the current direct relation pattern's rule and use its result. If the rule needs an unavailable case fact, identify that fact and leave the claim unresolved pending it. Use A.6.P when the relation or participant meaning remains unclear; use A.6.RCD only after recovery when no current direct predicate can state the needed claim. Explicitly identify an obtaining relation occurrence under its direct identity rule only when a receiving claim or operation must distinguish that occurrence.
Use A.6.0's FormalSubstrate profile when a separate declaration of vocabulary, laws, imports and applicability is needed. Apply A.6.1 for mechanism import or realization of that declaration, and E.18.1 when accepted problem-side material needs the declaration carried into later work. The same mathematical object may be designated in several epistemes or uses; select the subject pattern for the object and claim.
MathLensUse.Card@Context shape
The full card collects the account needed for a declared reliance under :4.4. MathLensUseOutputRef may reference any applicable C.29 output; its use does not require a FullCard. Local naming conditions are in :6.1a.
Read MathLensUse.Card@Context through three aspects:
Conditional fields apply only when the corresponding neighboring claim, claim-bearing use, or publication use is being made:
Plain card gloss. Read the account as an argument for why this representation supports this consequence for the intended use, and which changed condition would defeat that use.
Conditional overlays
Apply overlays for the reliance selected in :4.4. A conditional calculation states and checks its mathematical assumptions. A claim that a model is adequate for the phenomenon additionally requires the relevant validation. For a fitted or learned representation, recover its data, mapping and approximation conditions; generalization and validation claims require the corresponding grounds when they are made.
Use the validation overlay for the FullCard reliance in :4.4: prediction about the phenomenon, an operational or consequential decision, adoption of a model, benchmark/assurance input, Bridge-dependent model reliance, or transfer as a reusable phenomenon model. This includes a scientific claim of model adequacy. A published explanation of a conditional derivation needs its derivation and assumptions; it does not require a claim of empirical adequacy. LensUseBoundaryValue alone is insufficient for the stronger reliance.
Use the evidence appropriate to the claim. Verification checks whether the equations, algorithm or implementation deliver the specified mathematical behavior. A proof can cover an input class. A comparison with a known solution can expose an implementation error or establish an error estimate for the conditions tested. Validation examines whether the model represents the phenomenon well enough for the proposed use. Calibration estimates parameters or aligns a readout using observations. Agreement on the observations used for calibration can establish that fit. For a prediction under changed conditions, examine evidence for predictive accuracy there. An explanation shows how the representation makes a distinction intelligible. NASA-STD-7009B, §§4.2.3–4.2.6.
Recover known failures and counterexamples relevant to the proposed use, and perform the failure or robustness examination that use requires. Where no counterexample is known, do not manufacture one or treat its absence as validation.
To evaluate a probabilistic prediction, choose a scoring rule appropriate to its forecast form. Use a proper rule when the score should favor reporting the assessed distribution without distortion in expectation. The Brier loss for binary events and logarithmic scores for predictive densities are examples. State which direction is better and compare forecasts against the same observations. Gneiting and Raftery (2007) explain these scoring choices.
Use the learned-lens overlay when the mathematical object is fitted, learned, latent, simulation-trained or otherwise derived from data. Identify what the representation is being used to obtain. A description of the fitted data can remain at that scope with grounds for its claimed fit; an assertion about new cases needs its stated generalization claim and supporting validation. Making these fields conditional does not reduce the validation required for the phenomenon-model reliance in :4.4.
For DataOrTrainingRegime, identify the data's origin, what the collection includes and omits, how observations were collected and transformed, and the relevant use limitations. Use those facts to qualify the result and any proposed generalization. Datasheets for Datasets supplies questions for recovering these conditions; select those relevant to the present use.
Use the following learned-lens stop variants when the declared use reaches the corresponding boundary. Include a separate guard only when it passes F.19's plausible-reader test:
This is not a first-class causal abstraction card. It is a lightweight check: when LensMappingMode is abstraction, quotient, coarse-graining, macro-model, or simulation, and declaredLensUse would include intervention, policy, counterfactual, or causal explanation, apply C.28 for causal-use question and verdict.
For causal explanation through a learned representation, state which variables and interventions correspond between the models, then compare their results under those interventions. For approximate agreement, specify the similarity measure, the distribution of evaluated interventions and the way similarities are aggregated. Decoding a variable from an activation shows that the decoder can recover it; a claim that the variable affects the model's behavior needs the intervention comparison. Geiger et al. (2025), §§2.4, 3.2 and 3.6.3 develop these distinctions.
Repair decision table
Reopen a used result
C.29 output-change conditions:
Smallest source-return and output-change conditions:
AI-assisted thin-echo result rule:
C.29 edge-case boundary results:
When a source changes the used result
C.29 separates source-use relations from source-use disposition. Adopt, Adapt, Reject, and candidate-stress-test disposition say what FPF does with the source; SourceUseRelation says what work the source may perform inside a C.29 application.
Local SourceUseRelation slot discipline:
- source material reference or locator;
- declared C.29 output, lens-use boundary, or
LensUseBoundaryValueaffected by that source; - source-use disposition for the substantive use: adopt, adapt, reject, candidate stress test or recognition cue;
- currentness, supersession, contradiction, narrowing, or demotion condition;
- output-change condition for the C.29 result;
- stop or return condition, and any blocked overread that passes F.19's plausible-reader test.
Field meanings
Naming, ontology, and epistemic precision-restoration account
Name
Name: C.29 — Mathematical Lens Use.
Local namespace: MathLensUse = Mathematical Lens Use. The pattern-local card and reference namespace uses MathLensUse; checklist IDs use CC-C29-*.
The stable name is Mathematical Lens Use because C.29 governs a declared use and its use boundary. A mathematical lens may summarize, re-express or extend an account. For a recorded use, CandidateMathObject, LensMappingMode, PreservedStructure, LostStructure, LensUseBoundaryValue, and StopCondition describe the correspondence and its limits.
C.29-local naming guard
MathLensUse.* instruments are C.29-local unless separately admitted.
When one C.29 application needs a mathematical-lens name to become reusable outside that application, use F.18 local-first naming; when it quantifies over a class of described entities, use C.3 Kind-CAL; when it creates or reuses a durable concept or record family, use F.8 minting or reuse and E.9 design-rationale discipline.
Subject and representation
Distinguish the subject from its description and carrier. A mathematical object may itself be the subject, or it may represent another subject through a correspondence. State what that correspondence establishes: representation alone proves neither identity nor source-domain ontology; an established identity, equivalence or embedding remains usable at its actual strength.
Archetypal Grounding
Worked micro-cases by failure mode:
A speculative learning-dynamics model can be tried as a candidate with a concrete mathematical object and a testable correspondence. The bounded source use is in :13.2.
Preserve a reservation operation
A display that retains on-hand quantity n but omits reserved quantity r merges (n,r)=(1,0) and (1,1), although only the first permits another reservation. C.29.1:5.1 constructs the available quantity F(n,r)=n-r, compares reservation before and after mapping, and establishes the shared permission condition. It returns a quantity that supports the reservation question while retaining separate totals when another question needs them.
A route summary changes the cost question
Routes costing 1 and 4 can have the same endpoints without having the same cost. C.29.1:5.2 shows why the cost of a chosen route cannot be assigned to their common endpoint summary, then constructs a different answer: the minimum over admissible routes. A shared continuation preserves that minimum; a continuation available only after the more expensive prefix defeats minimizing the first stage alone. The repair retains the continuation condition or compares compatible complete routes.
Make calculation rules available as data
The same executor can perform different integer calculations when their instructions are supplied as data. C.29.2:5.1 constructs its state and instruction rules, obtains 8 and 7 from two orders applied to input 3, and gives separate arguments for the result, termination and cost. Adding jumps, interaction or larger stored integers returns to the corresponding behavior or resource question.
A correct distance calculation meets a finite command range
A calculated count of 63,662 motor increments cannot be sent as a positive signed 16-bit command. C.29.3:5.1 derives the count from a stated motion model, shows the wrapped command's contrary motion, and tests splitting the count under relative and absolute command meanings. It returns a command procedure with its supported range and motion conditions; B.5.MPC:5.1 shows the joint physical, mathematical and computational reasoning.
Realize a capacity bound with a shared stock of cards
Three unique admission cards can support a three-visitor bound when their possession and transfer rules control entry. C.29.3:5.3 follows free, reserved, inside and awaiting-return states. The free-card count gives a sufficient occupancy bound during handover; an occupancy equality additionally needs the intermediate states. A second entrance must share or partition the same stock. The worked construction shows how copying the stock defeats the bound and how controlled transfers restore it.
Bias-Annotation
Conformance Checklist
Use this checklist after constructing or delimiting the result in :4.1. Its conditions concern the correspondence, consequence and receiving use; a completed form alone does not satisfy them.
Common Anti-Patterns and How to Avoid Them
Consequences
Rationale
The useful representation is question-relative. The two-station recurrence distinguishes output rate from latency while omitting details of the work. An embedding can instead retain the source distinctions and introduce mathematical objects useful for an inference. C.29.1:5.4 uses a real-number construction to obtain a rational approximation, then examines a request that requires a rational solution. These uses depend on different return conditions. A reader with an adequate construction, correspondence and limitations can use them directly; C.29 helps recover whichever contribution is missing.
The cost is explicit construction and a check of material losses. A queue calculation, geometric transformation or learned model still needs its own mathematical Method. C.29's discovery cues help locate a candidate; its reliance rule determines which missing application information must be supplied for the result's use.
SoTA-Echoing
Question and selected line. How should a practitioner choose and bound a mathematical representation before relying on its result? Adopt problem-first domain modeling as the first-use Method: recover the question and required precision, choose an adequate construction from known structure and available data, compute its consequence, and check the application conditions. Adapt C.29 as assistance with missing correspondence, loss and transfer conditions, not a preliminary hierarchy of forms.
Same-case comparison. Use :4.4's queue question: does halving A's service time increase sustained output? Give both approaches the same arrivals, station times, routing and service assumptions, and a reader who can use maxima, a short recurrence and minutes-per-hour arithmetic. Give both the same bounded work allowance: a four-job prefix, the two specified service-time variations and their use limits. Neither approach receives a trained surrogate, extra observations or a larger literature review.
The first action-changing difference from an output-first procedure is the departure construction, not a more complete form. The retained transfer check can change a reused conclusion or leave reliance unresolved, but it is also part of competent domain modeling. C.29 supplies a reusable entry for that missing work; it is not selected over an already adequate domain Method.
Source comparison and limits. Meng et al., §2.1.2 compares conventional PDE solvers with physics-informed learning through confidence in equations, precision, cost and incomplete knowledge. Adapt that choice discipline, not a claim that the PDE Methods solve this queue. Known equations and a small exact computation favor the direct construction here; missing dynamics or a costly many-query problem can instead make a learned or hybrid candidate worth testing.
Dietrich and Schilders, §§2 and 5 explain problem-specific combinations of structure and data and the limits of generalization, robustness and physical consistency. This supports :4.1's application conditions, :4.2b's plural discovery and :4.5a's learned-model checks; it does not establish a universal best representation.
Mitchell et al., §4 supplies tailored reporting for trained models, their context and stakeholders. Adapt intended use, evaluation conditions and limitations for actual model reliance. Reject a universal FullCard requirement inferred from model reporting: :4.4 allows the conditional explanation to remain small and strengthens the account when reliance changes.
Trade-off and reopening. The selected line spends effort on the mathematical operation and only the application information needed by the receiving use. It gives up the uniform appearance of a full form for every example. Reopen the choice when a competing construction gives the needed result with lower effort or loss, when observations defeat a relied-on premise, or when the intended reliance changes.
Structural-sameness recognition examples
Sandberg's examples, available in the Math section of Links for 2026-05-12, suggest several discovery cues used in :4.2b: Stokes and boundary/exterior-derivative relations; de Rham and cohomological obstruction; a CLT fixed-point view; Lawvere-style diagonal constructions; Noether's symmetry/conservation relation; and Legendre, potential-duality and tropical-limit viewpoints. Use the example to find a candidate, then return to the chosen mathematical result and its assumptions before relying on it.
In the CLT-as-RG or fixed-point viewpoint, the Gaussian is an attractive fixed point for finite-variance distributions under the usual normalization; other stable laws are other fixed points under suitable normalization.
Metric/noise coupling as a candidate lens
Vanchurin, Geometric Learning Dynamics, v3, §§2–6 studies learning dynamics with a trainable-space metric and noise covariance. Its g ∝ κ^α regimes and proposed interpolation provide a candidate for a question about how that coupling changes an update process. The stationary-entropy-production construction has its stated loss constraint; the Schrödinger-like case additionally depends on a discrete shift symmetry.
For that question, specify the trainable variables, update/loss model, covariance and metric, then test the chosen relation and time-scale assumptions. Compare with the ordinary update model under the same data and intended use. Retain only a conditional candidate until the correspondence and validation support further reliance. The paper's proposed physical and biological interpretations do not by themselves establish that correspondence for a particular system. Its §6 explicitly leaves rigorous phase-transition analysis open.
Plural mathematical structures
Rodin's plural-foundations discussion supports considering several interpretable structural families rather than selecting by a foundational label. Its role here is a discovery prompt. The construction still has to preserve the law needed by the working question under :4.1.
When a mathematical equivalence, interpretation or homomorphism supports later formal work, name the exact objects, correspondence and preserved law. Use A.6.0 when a separate formal vocabulary/law declaration is needed, and E.18.1 when accepted problem-side material must be carried into later work. The object and receiving conditions are stated in :4.4.6.
Applied category theory
Adopt the operation-preservation discipline explained by Fong and Spivak, §3.3.2: a functor maps objects and arrows while preserving identities and composition. C.29.1 makes the corresponding comparison available before a result is reused, including operation permissions, representative independence and weaker bounds. The authors' §2.5.3 also shows how composition and choice compute route costs; C.29.1:5.2 uses a constructed case to expose when an omitted continuation condition defeats that reduction.
Applied category theory remains one organizer for composition, interfaces and transfer. The book's databases, electric circuits and dynamical systems provide source applications; adjoint functors, enriched categories and toposes provide further constructions to study when the question needs them. The trade-off is the work of defining the objects and operations and establishing their laws. When an ordinary domain calculation already provides the correspondence and consequence, use it directly. A failed comparison can instead identify the next required distinction or construction.
Obstructions to compositionality
Adapt the obstructions and failures-of-compositionality perspective into LostStructure and StopCondition: a lens can be useful precisely because it exposes where transfer fails, not only where it succeeds. In plain language, a good lens does not only say "this transfer holds"; it also names the boundary where transfer stops.
Computation and its physical realization
Turing 1936, §6 constructs a machine that interprets encoded machine descriptions. C.29.2 adopts the rules-as-data construction. Its finite interpreter obtains results from the listed instructions; Turing's broader universality result uses his machine-simulation construction.
Horsman, Stepney, Wagner and Kendon 2014, §§VI–VIII connect abstract computation with physical preparation, evolution and interpretation. C.29.3 adopts this comparison for digital, analog and stochastic realizations. Its source discussion distinguishes the resulting computational claim, the system model and performed execution.
These Methods are a methodological synthesis. Their worked cases are constructed under the stated models and conditions. Algorithm design, numerical analysis, learning, coding and distributed computation supply further construction techniques, guarantees and cost analysis when the working question needs them.
References
The comparison above selects the first-use Method. The references below provide further source returns for the particular discovery cues and model conditions in :4.2b/:4.5a; they do not rank those families for a working problem.
Relations
-
Construction and argument recovery: B.5 recovers the inputs, operations and dependencies needed to obtain or understand a result. C.29 tests which consequence can be carried through the proposed mathematical correspondence.
-
Related Methods: C.29.1 constructs mathematical result transfer; C.29.2 constructs a computation; C.29.3 connects computation to concrete execution. Each has its own working entry. They can be composed when one result supplies another's input; a computational question can also begin with an already adequate mathematical representation.
-
Boundary balance: C.29.BB constructs and revises an additive change account, preserving shared transfers and the information needed for the requested total.
-
Joint reasoning: B.5.MPC connects the physical account, mathematical question, computation and realization, starting from whichever contribution is available and returning to the contribution whose conditions fail. A.3.3 supplies state and continuation semantics; A.6.1 supplies the realization relation; C.39 helps find or develop a missing operation.
-
Extractable structural information:
C.2.8defines the characteristic and observer conditions consumed by a structural-information estimate. C.29 supplies the particular mathematical-lens correspondence and its limits. -
Architecture lens boundary:
C.32.P2S,C.32.PAD, andC.32.ADAmay cite C.29 lens outputs for preserved structure, lost structure, structural information, epiplexity, scale mapping, residual mapping, or source-return. -
Structural-information adequacy boundary:
C.33,C.34, andC.35may cite C.29 outputs when mathematical-lens results expose captured structure, preserved structure, lost structure, or discovery adequacy. -
Builds on:
A.1.1,A.6.P,A.6.RCD,A.3.3,A.19,A.10,A.15,B.3,C.16.P,C.16,E.17.EFP,E.17.ID.CR,A.6.3.RT,A.6.3.CSC,F.9. -
Constrained by:
E.8,E.10,F.19,C.2.P,E.19. -
Coordinates with:
A.6.0,A.6.1,E.18.1,C.11,A.15.1,A.15.2,A.15.4,C.18.1,C.19.1,C.26,C.27.TA,C.27,C.28,C.31.ASAP,G.5,G.9,G.2,G.10. -
Specialization relation:
C.26is selected as a C.29-compatible specialization for quantum-like modeling, with affordability qualifications. The conditional receiving questions and first contributions are collected in :4.4.6.
C.29:End
Mathematical Result Transfer
Type: Method pattern Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this when you have two mathematical accounts of a working situation, or a proposed change of representation, and want to carry a result from one to the other. The difficulty is whether the correspondence preserves the operation, distinction or condition on which that result depends. A stock total is to support reservations; route summaries are to support cost calculations; new coordinates are to make a coupled calculation easier.
Mathematical Lens Use, C.29, covers choosing a mathematical account and returning its consequence to a working question. Within that broader use, this pattern develops transfer of a mathematical result: construct the correspondence, compare the relevant operations, and obtain a consequence that survives the change. The work is to establish when a result obtained in one account can serve as an answer or premise in the other.
What changes in practice. Instead of accepting a familiar formula because the symbols look similar, you identify what its inputs represent and show why the intended conclusion follows after the change. The first useful result can be an equality, a sufficient bound, or two allowed cases that the representation merges although they require different answers. That last result tells you what to retain or which question to weaken.
For example, one item on hand can be unreserved or already reserved. Both situations have stock total one, but only one permits another reservation. The pair exposes why total stock cannot by itself decide availability. You can then construct available stock as on-hand stock minus reserved stock and compare the reservation updates in the two accounts.
Ordinary non-use boundary. Use a familiar result directly when the correspondence and its conditions are already established for the present use. A symbol rename or conversion under an established notation rule normally needs only A.6.3.RT. If the mathematical object or result is still missing, construct it in the relevant mathematical practice; C.29 helps choose the account and B.5 helps organize the missing reasoning. Transfer cannot provide an absent theorem by itself. If the mathematical relation is established but its physical interpretation or execution remains unresolved, use B.5.MPC to connect those contributions.
Be able to identify what the source quantities mean and follow the operation being transferred, or obtain its explanation from a suitable mathematical contributor. The worked cases explain their elementary constructions.
Enter here with an available proof, calculation, model, summary or proposed mapping. There is no preliminary requirement to repeat lens selection or fill a full lens card.
Problem
A useful change of representation alters which distinctions are visible and which operations are easy. That is also why it can alter an answer.
Agreement on object names or on a few numerical examples does not establish agreement of operations. An update can change the source while leaving a proposed summary unchanged. A summary can combine histories that permit different next steps. A transformed equation can be correct on represented inputs but be applied to other inputs that have no source counterpart. Even an exact change of coordinates can be followed by a lossy projection that no longer answers the original question.
To reuse the result, the practitioner constructs the correspondence, recovers the result's dependencies, and compares the operations in the direction of use. When exact transfer fails, the same work should reveal whether an added distinction, a smaller domain or a weaker conclusion is sufficient.
The central question is: what can be concluded in the receiving account, and returned to the working question, from the correspondence that has been established?
Forces
Solution
Recover the intended conclusion and its source → construct the correspondence → compare the relevant operations and their conditions → determine what merged cases can still answer → derive the exact or bounded consequence → return it, or repair the lost connection.
The arrows show dependencies. An existing map or proof can supply several steps. A counterexample can send the work back to the choice of representation. A sufficient bound can end the inquiry before an exact answer is constructed.
Recover the result and the direction in which it will be used
Start with the receiving question. Say what is to be determined and what difference the answer makes: whether another reservation is allowed, the cost of a chosen route, the least possible route cost, or a temperature bound at a specified time. These questions can concern the same objects and require different retained information.
Recover the source result together with the ingredients that produced it:
- the allowed inputs and the meaning of their quantities;
- the operations or relations used in the inference;
- the premises and operation conditions on which the conclusion depends.
Use an existing derivation when these ingredients are available. If an explanation gives only a final formula, recover the missing construction or argument through B.5:4.2 and B.5:4.3. The transfer comparison begins once the required source step is clear enough to reproduce.
Then name the direction of use. There are several common possibilities.
The last two uses are stronger than mapping source cases forward. Keeping the direction explicit prevents an answer about a relaxed problem from becoming an unsupported action recommendation.
Construct the correspondence from the required distinctions
Choose a source domain X and a receiving domain Y. Construct a map F from the allowed source cases to their receiving representations. Explain its action in the working terms before relying on its notation: F(n,r) = n − r retains available stock and forgets the separate totals; a route endpoint summary forgets which route was taken.
A correspondence can use several maps. An operation may take one kind of input and return another, so its input map and output map can differ. For an operation with several inputs, establish which tuples are allowed in the source. If inputs describe constituents of one whole, ask what makes them parts of the same admissible arrangement. Map those combinations into the receiving account. Taking all combinations of the separate summaries can introduce arrangements with no source counterpart. A guarantee proved on this larger set holds on the represented subset when the property carries back. To return a feasible arrangement, recover one compatible source combination with the required properties. Where the intended account relates one case to several possible representations, use that relation explicitly. Selecting one representation for each source case defines another construction; establish that the selection supports the intended conclusion.
Construct the correspondence by asking what the receiving question needs. Identify a candidate variable, relation or operation that carries that information. Compute its value on source cases. Determine which cases it combines. Try to express the receiving operation using only what remains. If the expression still depends on an omitted quantity, either retain that quantity or establish why it cancels for this use.
For a change of coordinates, construct the inverse when the receiver needs the complete source state. For a summary, identify the source cases compatible with each receiving value. It can give the exact answer to one question while leaving other properties open. Moving from a part to its whole changes what the account describes. To compare levels of abstraction of models, fix the question and interpretation and compare the possibilities each model admits for it (B.5.MR). For the transfer of a particular conclusion, establish the direction and the property it uses.
Keep the domain visible. A division requires a nonzero denominator; a square-root substitution may impose a sign choice; a route may exist only for certain endpoints or histories. A map justified on one part of X supports conclusions on that part. A receiving value outside F(X), the set of represented source cases, needs a further argument before it is used as a source possibility.
The first result of this step is a usable correspondence with an explained meaning. It may be a formula, a small table, a diagram, or an already defined mathematical map. Its form follows the work.
Compare performing and representing in both orders
For a source operation U and a proposed receiving operation V, carry out two constructions on the same allowed input:
- Perform U in the source account, then represent its output.
- Represent the source input, then perform V in the receiving account.
With input map F and output map G, the exact comparison is:
Read this as an equality of the outcomes relevant to the question. The letters can denote numbers, states, paths or other mathematical objects. For a state update with the same representation before and after, G is F. Use different maps when input and output representations differ.
Derive the equality from the definitions or use an applicable preservation theorem. Numerical examples can discover an error or make the relation understandable. For a general claim, establish the comparison throughout its declared domain. Exhaustive enumeration can establish a claim about a specified finite domain when every permitted case has been included.
Also compare where each operation is available. For a forward representation of a source operation, every source step used by the claim has a defined receiving counterpart. If the receiving account is to decide whether a source step is allowed, its answer agrees with the source condition for the represented case. If a receiving action is to be returned to the source, construct an allowed source action with the required outcome.
An equality on inputs where both sides happen to be defined leaves those availability questions open. This matters in reservations and in routes with restricted continuations.
For a sequence of operations, follow the intermediate representations. If each step has the required correspondence and passes an allowed intermediate result to the next, composing the equalities transfers the sequence. If a later operation depends on a distinction discarded earlier, the stepwise construction exposes where the summary has become insufficient.
An invertible coordinate change offers a constructive route. Given F and its inverse, define the receiving update by V = F ∘ U ∘ F⁻¹ on the represented domain. This definition yields the commuting comparison there. If V was proposed independently, compare it with this expression. A bijection between states alone does not determine whether that proposed update agrees.
The same comparison can relate transformed inputs and outputs within one model. Set V = U, choose F for the input transformation and G for the output transformation: G(U(x)) = U(F(x)) means that transforming a result agrees with applying the operation to transformed inputs. This is equivariance. Invariance of a quantity q under F means q(F(x)) = q(x). To obtain another solution of the original fixed problem by such a transformation, establish that its defining data and conditions are preserved. When the data change, carry that change into the receiving problem.
Always name the transformation whose effect is being compared. Relabelling two components can preserve a quantity which the time update changes. To establish preservation during evolution, compare q(U(x)) with q(x) for that update.
Determine whether a merged representation defines the answer
Suppose the source question has answer q(x), and you want a receiving function g with:
Such a g is well-defined on F(X) exactly when all allowed source cases having the same representation give the same answer:
If g exists, both source answers equal g of the same receiving value. Conversely, if the source answer is the same for every case represented by y, define g(y) to be that common answer. Choosing another representative then changes nothing.
Use this as a construction, not only as a test. First try to rewrite q in terms of the retained variables. If the unwanted variables disappear, derive the resulting g. If they remain, look for two permitted source cases with the same representation and different q values. That pair proves that the requested answer cannot be recovered from that representation alone. One such pair is enough to refute the proposed function. Failure to find a pair is not a proof of independence.
For an update, ask whether cases sharing the current summary have the same required next summary. For a summary that is also to decide availability, compare the operation conditions across those cases. For a prediction after several steps, compare what the permitted continuations can do. The relevant comparison follows the requested result; it does not require preserving every property of the original object.
The requested result may instead be one acceptable action. For a reachable summary y, take the source cases compatible with y and consider the actions allowed and sufficient in each. To return one action with a guarantee on y, choose one the recipient can carry out that meets the requirement in every such case. The complete sets of acceptable actions may differ. Even if every pair of cases shares some action, the whole group may have none in common. If the work permits a decision after further information arrives, construct the rule at that later decision point using what its performer can then know and do. MMP.8:4.2–4.3 develops that formulation. When only the mathematical existence of a choice was asked for, the corresponding existence argument can suffice.
Distinguish defining a value on a merged class from changing the question. “The cost of this route” requires the chosen route's cost. “The least cost among these routes” combines several costs by taking a minimum. The second is a new, potentially useful function; it does not make the first independent of its representative.
Derive the consequence in the strength that survives
Derive the receiving consequence from the comparison.
For an exact calculation, substitute the represented inputs and derive the receiving output. When returning an answer, express it in the quantity originally asked for. For a transported derivation, identify which premises and inference steps are covered. For example, an identity built from compositions of the compared operations gives the corresponding receiving identity on represented inputs. If a proof also uses order, division or an existence premise, establish how that contribution applies in the receiving account. A quantified claim depends on the cases over which it ranges.
A map that combines source cases can erase differences; equality of their images does not establish equality of the original cases. A receiving domain can contain cases outside the map's image; a result proved only for represented cases does not cover those additional cases. Use an inverse, a separate argument, or a narrower conclusion when the receiving claim requires it.
When the exact queried value is not determined, construct the values compatible with the retained information. For a represented value y, these arise from source cases satisfying F(x) = y and the stated premises. Prove a lower bound L(y), an upper bound U(y), or another relation that holds for all those cases. The resulting interval can answer a threshold question even when it cannot identify one value.
For the threshold q ≤ b, an upper bound U(y) ≤ b establishes the threshold for every compatible case. A lower bound L(y) > b excludes it for every compatible case. When L(y) ≤ b < U(y), the enclosure alone leaves the answer open for one unspecified case. Its bounds can include unattainable values. To establish that the retained information cannot decide the threshold for that case, exhibit a compatible case meeting the limit and another exceeding it, or prove both outcomes are possible. C.16.IR applies this test to quantities inferred from readings. A single permitted case above b already refutes a guarantee that every case meets it.
Preserve the domain, units and relevant time of the comparison. A bound for one sampled instant can settle the decision for that instant. To guarantee a condition throughout a period, also establish what can happen between samples.
A relaxation offers another useful transfer. Let S be the allowed source solutions, let T be receiving solutions, and suppose every solution in S has an image in T with the same cost. If the receiving account permits additional solutions, minimizing there can give a lower bound on the source minimum. When the minima exist:
The reason is that the receiving search includes a cost-preserving image of every source option. Maximization gives an upper bound under the analogous assumptions. An optimal receiving solution that has no source counterpart still supports the bound; it does not supply a feasible source plan. A returned plan requires an allowed source witness. A budget decision may need less than an optimum. For minimization, any feasible source plan at or below budget establishes feasibility. A proven lower bound strictly above budget excludes every plan. If a lower bound equals the budget, establish that the source permits a plan at that cost before promising delivery. Even an infimum can be approached by feasible costs without being attained.
Where the comparison is approximate, derive the error relation for the needed operation and propagate it through later steps. For example, if a later scalar operation h satisfies |h(a) − h(b)| ≤ K|a − b|, with K ≥ 0, on the relevant interval, an input error at most ε contributes at most Kε at its output. Any additional error introduced by computing h is added to that contribution. Adding the errors of successive steps without accounting for amplification can understate the final error.
Return a sufficient result or repair the failed correspondence
Return the conclusion together with the assumptions that change its use. A short calculation can contain the complete transfer. When another person or later use needs to recover the reasoning, preserve the source result, correspondence, comparison and conclusion in the smallest adequate form. C.29:4.4 supplies recording options when a receiving use calls for them.
If the comparison fails, use the failure to choose the next construction.
After a repair, repeat the affected comparison and the later steps that depend on it. Keep conclusions whose premises and correspondences remain applicable.
Stop when the exact consequence or bound answers the working question. A located failure can also settle the immediate question: “This summary cannot determine availability; retain reserved stock or use available stock.” If the missing contribution is a physical model, a measurement relation, a numerical method or a new proof, name that contribution and return to its practice. B.5.MPC handles the wider coordination when several of those contributions must agree.
Archetypal Grounding
The following constructions use stated mathematical models. Their conclusions show what follows from those models and correspondences. Applying them to a store, transport arrangement or thermal system also depends on the relevant accounting, compatibility or physical premises.
Transfer a reservation update to available stock
Situation and question. A reservation account stores on-hand quantity n and reserved quantity r. Both are integers, with 0 ≤ r ≤ n. One reservation is allowed when n − r ≥ 1. While this operation is performed, there are no deliveries, shipments or cancellations. The question is whether a smaller display can both decide that operation's availability and calculate its effect.
The source reservation update is:
Start with the proposed display H(n,r) = n, labelled “available.” The allowed states (1,0) and (1,1) have the same display, one. In (1,0), another reservation is allowed; in (1,1), it is not. Thus availability is not a function of H alone. Moreover, a permitted reservation leaves H unchanged, whereas the amount still reservable decreases.
The first result is that the total on hand omits a distinction needed by the reservation question. H remains useful as the on-hand total.
Construct the repaired display F(n,r) = a = n − r. Define its receiving reservation operation by V(a) = a − 1 for integer a ≥ 1. Now compare both orders:
The operation conditions also agree: U is defined exactly when F(n,r) ≥ 1, which is the condition for V. After an allowed update, 0 ≤ r + 1 ≤ n and a − 1 ≥ 0, so the result remains inside each account's domain.
For n = 5 and r = 2, the source becomes (5,3), and the display moves from 3 to 2. After k reservations, the same reasoning gives available stock a − k, for integer 0 ≤ k ≤ a. Each intermediate reservation is allowed; the next reservation at k = a is not.
Returned result. Available stock is sufficient to decide and update this reservation operation. It does not determine n and r separately: (5,2) and (7,4) both give a = 3. If the next question asks how many items are reserved, retain that distinction. If the next operation is shipment or cancellation, recover its definition and compare it separately. Correct transfer of the reservation equation does not establish that a physical stock count is accurate.
Separate route cost, minimum cost and a relaxation bound
Situation and assumptions. Two routes p and q go from A to B, with costs 1 and 4. A route r goes from B to C, with cost 2. These are the only elementary routes considered, costs add when routes are composed, and initially either p or q may be followed by r. The question is the least cost from A to C.
First examine a coarser, different question: whether A can reach B. Both p and q answer yes. A summary that records only their endpoints can answer this reachability question. It cannot define “the cost of the route” by choosing a representative: choosing p gives 1 and choosing q gives 4.
To answer the least-cost question, construct a different operation. Take the minimum over alternatives, and add the cost of a permitted continuation. In the stated case:
The equality holds because the same continuation is available after both alternatives and adds the same amount. For arbitrary finite costs a and b and common continuation cost c, adding c preserves their order, so min(a+c,b+c) = min(a,b)+c. This explains why the lower-cost prefix can be retained for this continuation. The source witness p followed by r has cost 3.
Change the compatibility premise. Now r is allowed only after q; taking p consumes a permission needed for r. The two arrivals at B have the same location but different permitted continuations. The allowed complete route is q followed by r, with cost 6.
The arithmetic min(1,4)+2 = 3 still holds. Its use as the attainable minimum fails because its selected prefix p cannot be followed by r. Equality of locations did not preserve route composition.
Repair the state at B by retaining whether the continuation permission remains. Let p arrive at (B,0) and q at (B,1). Only (B,1) has the r transition to C. Minimizing over the allowed complete routes in this expanded account yields 4 + 2 = 6. Returning q followed by r supplies an allowed source witness.
There is also a useful result from the coarser account. With the permission restriction removed, every allowed source route remains available at the same cost. The relaxed minimum of 3 rules out a source route within a budget of 2. For a budget of 4, the lower bound alone is inconclusive. The permission rule and the stated complete list of routes show that the only permitted complete route costs 6. If the budget is 6, returning q followed by r gives the planner a route it can use.
Returned result. Endpoint reachability, the cost of a selected route and the minimum over compatible routes are different questions. Retain history only insofar as it changes future compatibility or cost. When a coarser account is cheaper to use, its lower bound may still answer the receiving question without constructing the optimal source route.
Change temperature coordinates, then choose what can be forgotten
Situation and physical meaning. Consider two bodies with equal, constant heat capacity C > 0, each represented by one uniform temperature, T₁ or T₂. The bodies exchange heat only with each other. For one chosen sampling interval, stipulate a heat transfer from body 1 to body 2 of Q = αC(T₁ − T₂), with fixed 0 ≤ α ≤ 1/2. Negative Q means transfer in the opposite direction. Temperature differences use kelvins; temperature values below are expressed in degrees Celsius.
This is a supplied discrete model. Equal heat capacities and the opposite heat changes give:
The range of α makes each new temperature a convex combination of the old temperatures, without reversing which body is warmer. The model does not prescribe the continuous temperature history inside an interval.
The working question is whether the warmer body's modeled temperature is at most 60 °C after two intervals. Direct iteration is possible. A coordinate change can also expose which information the answer needs.
Construct the coordinates:
The inverse recovers both temperatures. Thus (m,d) loses no distinction between admitted temperature pairs. Substitute the source updates and compare:
The receiving operation preserves the mean and scales the contrast. It is a simpler expression of the same discrete update. Equal heat capacities explain the physical significance of the preserved mean: the two heat changes cancel. The warmer temperature is M = max(T₁,T₂) = m + |d|.
Relabelling and evolution. Let S(T₁,T₂) = (T₂,T₁) swap the bodies, and let U denote the update above. Equal heat capacities and the shared α give S(U(T₁,T₂)) = U(S(T₁,T₂)): both sides equal (αT₁+(1−α)T₂, (1−α)T₁+αT₂). The same update therefore applies after the swap. The warmer temperature M is invariant under S. It changes under U: for α = 1/4, (20,80) becomes (35,65), so M falls from 80 °C to 65 °C. Preservation of the mean during that update follows from the separately derived identity m′ = m.
Now consider discarding d and retaining only m. The identity update m' = m is still exact, but the warmer temperature is no longer determined. The states (0,100) and (50,50) both have m = 50. For α = 1/4, the first becomes (25,75), then (37.5,62.5); the second stays at (50,50). They give opposite answers to the 60 °C threshold question after two intervals. Mean preservation alone is insufficient.
For α = 1/4, the contrast halves each interval. Starting from (20,80) gives this trajectory:
A bound on the contrast suffices after two intervals. Suppose the individual temperatures are unknown, but m = 50 °C and |d₀| ≤ 30 K are known. Repeatedly applying the receiving update yields dₖ = 2⁻ᵏd₀. Therefore:
Temperature differences are added to a temperature on the same scale. At k = 1, an initial pair of (20,80) gives a maximum of 65 °C and refutes a 60 °C guarantee covering all compatible pairs. An initial pair of (50,50) stays at 50 °C, so for one unknown pair the retained information leaves the answer open. At k = 2 the upper bound is 57.5 °C. Every compatible pair now meets the 60 °C limit at that sampling instant.
If an exact maximum is wanted, retaining (m,|d|) is sufficient under this symmetric update. The sign of d is needed only for questions distinguishing which body is warmer. Retain the distinction needed by the question, rather than restoring both coordinates automatically.
Returned result and return condition. The bound answers the stated question at the second sampling instant under the supplied model. It does not assert that the bodies were always below 60 °C: the initial pair (20,80) was not. Unequal heat capacities, external heat exchange or a changed transfer rule require a new update and a new comparison. Establishing that this discrete model predicts a pair of bodies requires physical and measurement work; B.5.MPC connects that work to the mathematical result.
Return a rational result from a real-number construction
A calculation accepts positive rational settings x and needs |x² − 2| ≤ 0.01. If a rational candidate is already supplied, substituting it can settle this question. The following construction uses an available positive real root of x² = 2 to obtain a rational candidate.
The inclusion of the rationals in the reals retains their equality, order, addition and multiplication. The larger domain also contains limits of rational sequences that have no rational limit. This supplies new mathematical objects without merging distinct rational inputs. The needed return is still a rational setting satisfying the original inequality.
Compute rational bounds:
Squaring is increasing on the positive reals, so the root lies between those endpoints. Choose the rational setting x = 1.414 and check the requested result: |x² − 2| = 0.000604 ≤ 0.01. The returned setting and its error calculation answer the original question. C.29.2 develops a procedure when obtaining such bounds requires one.
Now change the requirement to a rational setting with x² = 2. Suppose x = p/q is in lowest terms, with integers p and nonzero q. Then p² = 2q², so p is even. Substituting p = 2r shows that q is also even, contradicting lowest terms. The real root therefore has no rational counterpart. Return that obstruction; the requester can retain a tolerance or change the allowed number domain.
The extension supports a useful approximation and an existence argument in the larger domain. Which result can be returned depends on the requested property and the allowed source settings.
Bias-Annotation
Bias toward convenient summaries. Aggregates are easy to display and compute. Test them against the receiving question by constructing cases with the same aggregate and different answers. In the thermal case, the mean is sufficient for mean evolution and insufficient for the maximum.
Bias toward familiar structure. A recognized matrix, graph or algebra can make a proposed correspondence feel established. Recover the represented participants and operation conditions before applying its theorem. The route graph needs permission state when permission changes continuation.
Bias toward exact answers. Insisting on reconstruction of every source detail can hide a cheaper sufficient answer. Derive a bound over the compatible cases and compare it with the decision threshold. If the bound crosses the threshold, a tighter argument may still settle it. Establish which threshold outcomes the compatible cases can actually attain before claiming that both are possible.
Bias from machine-generated fluency. An AI can produce a plausible mapping and a correct algebraic simplification while omitting the premise that connects them. Ask a contributor for the domain, operation comparison and returned conclusion. Use subject expertise to assess an unfamiliar proof; a persuasive explanation does not settle its premises.
Conformance Checklist
These checks concern a claimed use of mathematical transfer. Recognition of a promising representation starts the work; the comparison and argument establish what conclusion it supports. Select the checks that apply to the claimed use.
A complete argument about a mathematical model establishes its stated conditional consequence. Confidence that an observed system satisfies the premises, or that an implementation performs the operation, comes from the corresponding subject work.
Common Anti-Patterns and How to Avoid Them
Consequences
Useful consequences. A transferred result comes with the correspondence that makes it usable. A failed transfer becomes a mathematical finding: two merged cases disagree, an operation lacks a counterpart, a domain is uncovered, or a bound is insufficient. Each finding suggests a repair that is narrower than discarding the whole account.
The method also makes productive simplification possible. Available stock omits separate totals while preserving reservation behavior. A route relaxation omits compatibility while providing a lower bound. Mean and contrast simplify a thermal update, and a bounded contrast suffices for a later threshold decision.
Costs. Recovering operation conditions and proving a comparison can take more effort than repeating a calculation in the source account. For a small one-off calculation, direct derivation may be cheaper. A useful summary can also require richer state than expected when later operations depend on history.
Remaining limits. The pattern does not supply the domain theorem, physical law, numerical algorithm or implementation correctness that a transfer uses. It can expose a missing premise and make that next contribution precise.
Architectural Rationale
Why operations, conditions and questions are considered together
A result comes from a construction and the premises it uses. Mapping the named objects alone leaves open what happens to that construction. Comparing the two orders identifies the requirements for transfer.
Operation conditions belong to this comparison because partial operations are common in practice. A reservation consumes availability. A route continuation consumes or requires permission. Removing the condition can enlarge the receiving problem even while preserving the arithmetic of every allowed source step.
The receiving question determines which preservation is useful. Reconstructing a complete state, preserving a selected query and obtaining a bound have different mathematical requirements. Making that choice early avoids both retaining unnecessary detail and losing a decisive distinction.
Why representative independence is constructive
When a summary merges cases, the key question is whether the desired answer is constant among those cases. This yields both a proof method and a method of discovery. Rewriting the query can construct its receiving form; a conflicting pair can identify the information the summary lacks.
An invertible coordinate change supports complete reconstruction of a source state within its domain. Many useful transfers need less. Requiring an inverse for available stock would defeat the purpose of the summary. Conversely, a map's usefulness for one query does not establish that nothing has been lost. Query-specific preservation states the useful result without that stronger claim.
This also explains why changing the question can be a mathematical repair. The cost of a represented route may be undefined while the minimum over the represented routes is well-defined. That new minimum still needs its own composition and feasibility argument.
Why the method allows inequalities and returns
Exact preservation is one strong route to result reuse. Sound inclusion and bounds are another. If an account covers more possibilities than the source, a claim holding for all those possibilities also holds for the covered source possibilities. An apparent bad case in the larger account may require refinement before it becomes a source counterexample.
The method uses that asymmetry to retain useful consequences. A relaxation bound can settle a cost threshold even when the relaxed optimizer cannot be returned. A temperature interval can settle a sampled threshold without identifying both temperatures.
Returns therefore follow the failed dependency. The practitioner can refine a state, restrict a domain, recover a premise or ask a weaker question. The unaffected mathematical work remains available.
Why expression construction and joint reasoning remain distinct
A.6.3.RT helps construct an expression in a representation language suited to the operation. Defining m and d is such an expression move. Deriving their inverse and update equations establishes the mathematical transfer addressed here. The notation makes that derivation easier to perform and inspect; it does not determine its truth.
C.29 supplies the broader choice and use of a mathematical account. B.5.MPC connects a mathematical result with physical interpretation, computation and execution when the working question requires all of them. This pattern contributes the mathematical comparison inside that work and can also be used independently.
A contributor can construct the correspondence, derive an identity or search for a conflicting pair. The receiving practitioner still needs enough understanding of the domain, operation and conclusion to use the result or request the missing argument. B.5:4.5 and B.5.MPC:4.8 explain how to divide such contributions without losing the shared question.
SoTA-Echoing
The working problem is reuse of mathematical consequences across accounts at a cost justified by the receiving question. Established mathematical lines supply complementary methods: preservation of operations, coverage of possible results by a sound abstraction, and extension to a domain in which a needed construction is available.
The pattern's synthesis is the practitioner sequence connecting these obligations to the intended answer: construct the map, compare operations and availability, establish independence or coverage, and return the consequence. The reservation, route, thermal and rational-setting constructions derive their claims from their stated premises. The cited sources supply reusable mathematical lines, not evidence that a physical or organizational application satisfies those premises.
A direct proof in the original account remains an alternative. Prefer it when it is simpler than establishing and maintaining a transfer. Reconsider a chosen summary when a new query needs a distinction it omits, when composition introduces a new condition, or when a tighter justified bound changes the decision. Reopen the mathematical work affected by that change.
Relations
C.29.1:End
Computational Formulation
Type: Method pattern Status: Stable Normativity: Normative when this Method is selected; worked cases retain their stated assumptions.
Problem frame
Use this when. You know what a mathematical answer would mean, but the available expression, data and operations do not yet explain how to obtain it. An equation may characterize a solution without giving a procedure; a proposed procedure may need information its stored state has discarded; or a short calculation may expand into an unaffordable representation.
The practical question is: What should be represented, and how will the available computation obtain an interpretable answer? Begin with the distinction the answer must make. For example, if an executor must accept either “increment, then double” or “double, then increment”, storing only the two operation names loses a distinction: on input 3 the required answers are 8 and 7. Ordered instructions and a current position make that distinction operable.
The object being developed is a computational construction for a stated question: represented quantities or state, computational rules, and a way of obtaining and interpreting the requested result. Computational formulation names this work. The construction may leave evaluation order to a solver or describe a stochastic or continuing process. Its account can be a short derivation, pseudocode or a program with an explanation; these are ways to communicate the construction, not additional required records.
Intended reader. A practitioner who can state the working question and read its subject mathematics, and who needs to recover, construct or commission the computation. Elementary arithmetic and following a sequence of instructions suffice for the first case. A numerical or quantum-modeling case additionally supplies its mathematical assumptions locally; developing a production solver still needs that discipline's preparation.
Computational thinking is the wider practice. This pattern covers formulation: represented inputs and state, computational rules, a way to obtain and interpret the output, and the argument and resource account needed for its use. Algorithm design, numerical analysis, programming-language design, learning and distributed computation supply deeper construction methods.
Family relation. C.29 groups methods for constructing, transferring and realizing interpretable mathematical accounts. C.29.2 contributes a computational formulation by composition and result use; it does not inherit every step or recording option of C.29. C.29.1 supplies a needed result-transfer argument. C.29.3 connects a computation with a system that prepares, performs and exposes its result. Enter this pattern directly when the working question is already available.
First useful result. Return a formulation with an applicable way to obtain and interpret the requested result under stated conditions, or a demonstrated obstruction and the construction still needed. A justified restriction, error bound or resource rejection may finish the current question.
Ordinary non-use. Use an already adequate local calculation or implementation directly when its inputs, operations and result are understood. Open C.29.1 for an unsettled transfer between accounts, C.29.3 for an unsettled execution correspondence, or the subject method when that is the only missing contribution.
Problem
Knowing a condition that the answer satisfies does not always tell a practitioner how to get that answer. Writing y = argmin f(x) specifies a selection problem; it supplies a computation only together with a way to represent and search or otherwise solve the admitted problem. Conversely, a running procedure may return a value without establishing that it satisfies the requested condition.
Three failures make this gap expensive:
- A representation identifies cases that require different answers. No later procedure using only that representation can recover the missing distinction without another input.
- An operation such as “solve”, “update” or “sample” hides the construction that is unavailable. Giving it a name does not make it elementary or obtainable.
- A procedure is assessed under the wrong semantics or cost model. Exact integers, fixed-width words and rounded values have different operations; a single arithmetic instruction may manipulate a growing number of bits.
The repair is to connect the requested answer to a construction, then follow the construction's meaning and costs. It need not start by writing software. A hand calculation, an invariant or a storage count can locate the decisive obstacle first.
Forces
Solution
Establish how the computation obtains the requested result and why that result has the required meaning. Read the following steps as connected work with returns: an unaffordable state, an unavailable operation or a failed argument can change an earlier choice. Begin with any contribution already available.
Specify the answer before selecting its representation
State the admitted inputs and the answer the receiver needs. Say whether the result is a value, a witness satisfying a condition, a bound, an approximation or a continuing response to inputs. These are illustrative result forms; choose the one used by the task.
Distinguish the answer condition from the proposed calculation. For a value, ask how it will be obtained. For a witness, ask how candidates will be constructed and tested. For a negative answer, ask what establishes that no admitted witness was missed. A search that eventually finds a witness need not decide the cases in which none exists.
Make an approximation requirement operational. |y_hat - y| <= epsilon means an absolute error bound on the requested quantity; relative error uses a different comparison and needs care near zero. A probability guarantee must state what is random and the event whose probability is bounded. Neither a small residual in a different equation nor a favorable average automatically supplies the required result.
For stochastic computation, distinguish a requested distribution, samples from it and an estimated statistic. Those outputs require different constructions and resources. For an approximate distribution, name the events, statistics or distance over which accuracy is required. A distribution's definition does not by itself supply a sampler; a sample average needs an argument connecting it with the requested population quantity.
Derive the input distinctions from the operations the recipient needs. A representation for an exact value must distinguish inputs with different required values. If the order allows any satisfactory witness, a witness common to all inputs with the same representation can suffice. For an ongoing response, determine what its performer will know at each step and what subsequent inputs may require. Establish that the resulting computation can obtain an acceptable answer for the admitted inputs. Restore missing information, use an adequate alternative, or agree a change in scope when the present representation precludes the requested work. C.29.1 develops the preservation argument.
Choose represented state and operations together
Choose data in which the next operation can be expressed. Explain what each stored value means, how an input initializes it, and how a returned value will be interpreted. The same mathematical quantity may be represented by an integer, an interval, a symbolic expression or another subject-appropriate object; the operations must match that choice.
Construct the state from what the continuation needs. In a sequential procedure this often includes a control position and intermediate values. For each proposed next step, ask what it reads, changes and preserves. If two histories reach the same stored state but require different continuations, recover the omitted condition or summary. Use A.3.3 for the underlying state-and-continuation construction; keeping every past observation is only one possible repair.
Specify the meaning of an elementary operation at the level used by the argument. An exact integer addition is different from modular word addition. A comparison of exact rational numbers is different from comparing rounded observations. If an operation is available only through another procedure, expose that dependency when its conditions or cost can change this computation.
A formulation can specify equations or constraints without choosing an evaluation order. Identify the supplied quantities or boundary conditions, admissible solutions and the output the receiver needs. An applicable solver or modeled computational process must connect those relations to obtaining a result; its execution order may remain an implementation choice. When restructuring the equations, preserve the required solutions under the stated conditions and retain expressions that recover requested quantities removed from the computational state. If no such construction is available, the equations still characterize answers and the missing way remains a task under :4.3.
The available operations can also be the starting contribution. A collaborator working under C.29.3 may supply preparable inputs, controllable changes, readable outputs and their limits for a candidate system. Use those capabilities to propose computational states and operations that can answer a useful question; do not force them into an unsuitable instruction set. Keep the physical model and the experimental or conditional basis of that contribution visible. A new computational model or language needs meanings for its expressions as well as formation rules. A.6.3.RT helps construct expressions under available notation rules; designing missing rules is separate notation or language-design work.
Obtain the computational construction
If a known construction answers the question, recover its inputs, operative rules or steps and assumptions from its explanation and use it. B.5 supports that recovery. Check that the available operations can perform those steps and that their result has the required interpretation. A library or solver can supply the construction, but its accepted input class, result guarantee and failure behavior must fit the present use.
When the connection is missing, work back from the required output to an operation that could produce it. Which inputs would that operation need, and how could they be obtained from the data or state already available? A direct substitution can complete the construction. To build a procedure from smaller operations, determine what remains after an available step, retain the information needed by the next step and show how the joined results satisfy the original condition. A calculation on a small instance can reveal an intermediate value or dependency to carry into that argument. Recurrence construction, search, optimization, numerical discretization and other techniques have their own subject methods. Use C.39 to find or develop a missing way; use C.40 when a workable change-and-test operation is already available and branching search is the live difficulty.
For a still-missing contribution, return a substantive task: the available inputs and operations, the result or intermediate relation needed, the conditions it must preserve, and what would count as a useful solution. “Find a better algorithm” is usually too weak. After a dense-state rejection, for example: “Given this circuit family and this requested observable, provide a representation and update/readout procedure with a justified error bound and a peak-memory estimate below the stated budget; do not materialize the full amplitude array in a hidden conversion.”
Explain the result and progress
Follow the construction from its inputs to the result the recipient will interpret. A derivation from the operation rules can establish the result for their whole input class. Work through a particular input when the order calls for its result or a trace can resolve uncertainty about state, instruction order or output meaning.
Supply the argument appropriate to the claimed range. For a loop, find a statement relating the current state to the work already completed and the answer still sought. Show that initialization establishes it, each iteration preserves it, and the stopping condition makes the desired conclusion follow. This statement is the loop invariant. Separately explain why the loop reaches that condition, for example through a nonnegative integer that decreases at every iteration. For a recursive construction, explain its initial cases, the smaller calls and how their results give the caller's answer.
The proof form depends on the computation. A direct finite composition may need only substitution through its operations. A randomized procedure needs its probability argument. For an estimated statistic, state the sampling assumptions and connect the claimed error or uncertainty to the sample count. For a continuing interaction, state the preservation or response property required and the assumptions under which progress is claimed; global termination may be the wrong requirement.
Limit the claim to what the argument or observation supports. A trace establishes that traced case. A proof using exact arithmetic establishes the stated abstract procedure under exact arithmetic. A finite precision implementation or an executing device requires the additional comparison. Tests can expose failures and support selected empirical claims; a few passing tests do not prove an unrestricted input claim.
Resolve accuracy and computational limits that affect the use
A well-defined mathematical object need not have the finite representation or uniform procedure being assumed. Ask what information the input representation provides and which operations are effective on it. Replacing “all real numbers” by finite strings, an evaluation oracle or a family of increasingly accurate approximations changes the computational problem. If a computability or impossibility claim matters, obtain the applicable subject argument rather than inferring it from a failed attempt.
For a finite search domain with an effective test, explicit enumeration can provide a terminating baseline. It may be too expensive, but it separates an obtainable procedure from an open construction. For an unbounded search, failure to find an answer in the allotted time leaves a different result; it does not establish nonexistence.
For numerical work, connect the stopping test to the result the recipient will use. An error bound sufficient for reporting a value may still allow both answers to a threshold question. Compare the full error interval with that threshold. If it crosses the threshold, try a direct comparison, a tighter enclosure or a reformulation of the test. Locate the errors from input representation, approximation, arithmetic and output conversion and justify how they combine. To use the result for a physical decision, also account for consequential model and measurement uncertainty.
If finite precision can reverse a decisive comparison, increase precision, use a justified enclosure, reformulate the test or return that comparison as unresolved. If an iteration no longer changes its stored state, a limit of the ideal iteration does not show that the implementation will reach the requested tolerance. Return to the representation or stopping rule. Narrow the claim only when the narrower answer remains useful and the change is explicit.
Calculate costs from the chosen representation
Begin with the resource that can decide the choice. For a stored array, derive how many elements the representation requires and how much storage each element occupies:
payload storage = element count × bytes per element.
Identify which storage must coexist and which arrays share it. Two views of the same 40 GiB allocation can each expose the entire array while using that payload once. Two independent 40 GiB copies require 80 GiB while both are live. Include the original allocation if it is live at the same time. For a design allowed to choose the execution order, one enforceable order can establish a fit. For a guarantee over all permitted orders, bound the peak in each or establish one bound covering them all. Count each live allocation once, including workspace, indices and temporary copies that can affect the peak. A payload lower bound can reject a design. To establish that it fits, give a sufficient upper bound for the required execution.
In a model of discrete operations, count how often each operation is performed and what each performance costs. State the size parameters. An instruction count with one unit per arithmetic operation answers a different question from bit operations on growing integers, memory transfers or elapsed time on a machine. Explain the dominant term before using asymptotic notation. A bound for one representation is not a lower bound for every algorithm solving the mathematical problem.
A computation described by continuous dynamics may need a cost relation for duration and accuracy rather than an instruction count; obtain the estimate with C.29.3.
Include input conversion, preparation computation and output production when they can dominate the result. Precomputation can be worthwhile across many uses, but say how many uses amortize it. A compact internal state does not make an explicitly requested exponential-size output cheap to enumerate.
When resources fail, change one of the causes: represented structure, stored precision, retained data, procedure, admitted problem class, requested answer or proposed execution resources. Recompute for the chosen alternative. An exact structural reduction, a lossy approximation and a different cost model are different changes. C.29.1 supplies a needed consequence-transfer argument; C.29.3 assesses whether the resulting execution arrangement supports the selected computation.
Return the sufficient result and its next use
Return what the recipient asked you to obtain. To deliver the procedure, supply its operations, the meaning of their inputs and outputs, and the conditions under which its result and cost meet the requirement. For a result on supplied inputs, carry out the computation or use an existing result whose grounds apply. If a fresh execution is part of the order, that execution remains to be performed. For a continuing response, provide the state, operations and progress conditions needed at successive inputs. A short derivation can supply these connections.
If a contribution is absent, identify it at the failing connection. A missing state distinction returns to formulation; an unavailable algorithm returns to its construction; an unsupported numerical bound returns to the numerical method; an execution mismatch returns to C.29.3, which may in turn supply a reason to revise this formulation. The iteration can change both model and executing arrangement.
A conditional procedure, a restricted result or an obstruction can be sufficient. Keep “no procedure obtained”, “this representation exceeds the budget”, “no algorithm exists in the stated model”, and “the implementation failed on this case” as different conclusions, each with its own basis.
Archetypal Grounding
The cases construct an interpreter, a bounded numerical approximation, a resource-sensitive representation, two computations from one system of relations, and a probability estimate with an error guarantee. Their conclusions follow under the stated mathematical and physical premises.
Make calculation rules available as data
Question. A team must change an integer calculation without changing the executor. The input is an integer x0 and a finite sequence P consisting of n instructions chosen from increment and double, followed by one stop. Other sequences are outside this procedure's admitted input class.
Construction. Store the ordered sequence and use mutable state (p, x): the position of the next instruction and the current integer. Positions start at zero. The meaning of the output is the left-to-right composition of the arithmetic instructions applied to x0.
This explains the elementary operations and their order. A parser or caller admitting other strings must validate the sequence or define the additional cases; it must not silently interpret an unknown instruction as doubling.
Argument. After p arithmetic steps, x is the result of applying exactly the first p instructions to x0, and 0 <= p <= n. Initialization gives the empty composition. Each branch applies the next specified operation and advances the position, preserving that statement. While an arithmetic instruction remains, n-p decreases by one and cannot be negative. At p=n the next instruction is stop, so the returned integer is the requested composition. With P = [stop], the same procedure returns x0.
Cost changes with representation. Counting one step for each arithmetic instruction and the final stop gives n+1 instruction steps, apart from validation and input/output costs. That does not make arbitrary-size arithmetic constant-time. Represent the magnitude in binary and retain the sign separately. Let b0 be the binary length of |x0|, counting zero as one bit. Each operation increases the magnitude's length by at most one, so the current magnitude needs at most b0+n bits, plus one sign bit. The position and stored program need additional space.
With simple materialized binary arithmetic that scans or copies the current digits, an arithmetic step on b bits costs at most proportional to b. Summing the growing lengths gives an arithmetic-work upper bound proportional to n*b0 + n^2 for that implementation. Add program-access costs when they matter to the estimate. A representation that treats doubling differently requires another estimate. On a fixed-width integer implementation, enough increments or doublings can instead overflow; the exact-integer argument then requires a range restriction or a different realization.
What became possible. The executor can perform any calculation in this finite instruction class by receiving another sequence. Turing's 1936 construction, §§5–7, makes encoded computation rules available to an interpreter; this small case uses that constructive idea. Turing's universal-machine result concerns a richer simulation construction. Adding jumps or continuing input to this case changes its progress and cost questions and requires their own argument.
Turn a root condition into a bounded approximation
Question and representation. Return a rational approximation y to the positive root of z^2 = 2 with |y - sqrt(2)| <= 0.001. The equation characterizes the root. The requested rational output still needs a way to compute it. Use exact rational arithmetic in this construction.
The initial interval is [l,u] = [1,2] because 1^2 <= 2 <= 2^2. For nonnegative arguments squaring is increasing, so a comparison of the midpoint's square with 2 determines which half still contains the root.
Meaning and argument. The preserved statement is 1 <= l <= sqrt(2) <= u <= 2. The square comparison preserves it, and every iteration halves the interval width. After k iterations that width is 2^(-k). The returned midpoint therefore differs from the root by at most 2^(-k-1). Choosing the first k for which this is at most epsilon supplies both the stopping rule and a finite bound on the iteration count for every positive requested tolerance.
After nine halvings, return 1449/1024 = 1.4150390625. Its error is at most 1/1024 = 0.0009765625, which satisfies the requirement. The interval argument establishes the bound without requiring a previously calculated decimal expansion of the root.
Now suppose the receiver asks whether the root is at most 1.415. The accepted midpoint is slightly above 1.415, and the ninth bracket contains values on both sides. In this case a single exact comparison settles the new question. On the positive half-line squaring preserves order, and 1.415² = 2.002225 > 2, so the root is below 1.415. The midpoint remains a valid answer to the original approximation request.
Resource and accuracy consequences. There are nine midpoint-square comparisons in this case. For finer tolerances, the dyadic numerators and denominators grow; a count of comparisons alone does not include the growing cost of exact squaring. A finite precision version must preserve the bracket decisions and avoid a midpoint that rounds to an endpoint while the tolerance remains unmet. Output rounding also consumes accuracy. These are returns to arithmetic and representation choices, not evidence that the exact rational construction failed.
For a costly general continuous function, an established bracketed method using interpolation may save function evaluations. Bisection remains useful when a simple interval argument and predictable reduction are worth the extra evaluations. For a function not known to be continuous, a sign change alone does not justify this root argument. A demand for an exact finite decimal expansion of this irrational root changes the answer format to an impossible one; an exact symbolic expression or a rational approximation is a different, obtainable request.
Count a dense state before allocating it
Question. Can a proposed dense numerical pure-state array for 50 qubits fit in a 64 GiB memory budget? The representation stores one complex amplitude for each binary string of length 50. There are two choices at each position, hence 2^50 entries. Stipulate 16 bytes per stored complex value: two 8-byte components.
The payload alone is:
16 × 2^50 = 2^54 = 18,014,398,509,481,984 bytes,
or 16 PiB = 16,777,216 GiB, where 1 GiB = 2^30 bytes and 1 PiB = 2^50 bytes. Since 64 GiB = 2^36 bytes, the payload exceeds the budget by a factor of 2^18 = 262,144. Workspace, copies and indexing cannot reduce this payload requirement. This rejects the proposed dense allocation without building the simulator.
The count is tied to a representation. It says nothing by itself about the cost of every way of answering a quantum-modeling question. Using 8 bytes per amplitude halves the payload to 8 PiB and still fails this budget; it also changes numerical precision. Discarding small amplitudes requires an error argument for the requested output. Calling the representation sparse supplies no bound on the number of retained entries or on growth during its operations.
Construct a restricted alternative. Suppose the admitted states are products of 50 normalized single-qubit pure states, every operation is a single-qubit unitary gate, and the requested output is the probability that a named qubit is read as 1. Store the 50 pairs (alpha_i, beta_i) instead of the full array. The corresponding joint amplitude for bit string s is the product, over positions i, of alpha_i when s_i=0 and beta_i when s_i=1.
Initialize each pair from its supplied single-qubit state. For a gate with unitary 2-by-2 matrix U on qubit i, replace just that pair by U*(alpha_i,beta_i), retaining the old two values while computing both new ones. The tensor-product rule preserves the product form, and unitarity preserves the pair's normalization. For normalized pairs, return |beta_i|^2. The payload is now 50 × 2 × 16 = 1,600 bytes, with additional algorithm and representation overhead to be counted separately.
Starting with all pairs (1,0), apply the Hadamard operation (a,b) -> ((a+b)/sqrt(2),(a-b)/sqrt(2)) to the first pair. It becomes (1/sqrt(2),1/sqrt(2)), and the requested probability for the first qubit is 1/2. The construction uses a fixed number of complex operations per gate and stores only the pairs. Its structural factorization is exact under the admitted model; stored numerical coefficients still require their precision account.
An entangling gate can invalidate that representation. The two-qubit state with nonzero amplitudes 1/sqrt(2) at 00 and 11 and zeros at 01 and 10 cannot be one product: nonzero alpha_0*alpha_1 and beta_0*beta_1 would make all four factors nonzero, contradicting a zero cross term. The product procedure must therefore reject that extension or receive a richer representation and update algorithm. Even in the product class, requesting all 2^50 amplitudes explicitly restores an exponential output count.
First result and next contribution. The original dense proposal is ruled out. The factorized procedure answers the separately stated restricted question; it is not a replacement for an unspecified general circuit. To continue, recover the input-state and gate family, requested observable or samples, and tolerated error, then obtain and cost an applicable domain algorithm. The general method supplied the count and the construction question; quantum simulation supplies the representations and update/readout algorithms.
Obtain different computations from the same circuit relations
Model and questions. An ideal resistor and capacitor are connected in series to a voltage source. Let i flow toward the capacitor's positive plate, v_R be the resistor's voltage drop in that direction, and v_C the capacitor voltage. Use R = 10 ohm, C = 0.1 F and these relations:
The equalities do not assign a computational direction. One question supplies the source voltage and asks for current; another supplies a desired current and asks for the source voltage. Keep the component relations while changing which quantities are given and which must be obtained.
Voltage given: construct the trajectory and its readouts. Take constant v_s = 12 V and the initial condition v_C(0) = 2 V. Substitute the first two relations into the third:
dv_C/dt = (v_s - v_C)/(R*C).
The reduced differential state is v_C. Retain v_R = v_s - v_C and i = (v_s - v_C)/R as readout expressions, so eliminating those variables from the state does not remove the requested outputs.
Here tau = R*C = 1 s. Put z = v_s - v_C; then dz/dt = -z/tau and z(0) = 10 V. Thus z(t) = (10 V)*exp(-t/tau): differentiation gives dz/dt = -z/tau, and substitution at zero gives the prescribed initial value. Recover the original quantities as v_C = 12 V - z, v_R = z and i = z/R. At t = tau*ln(2), the exponential is 1/2, so the returned quantities are v_C = 7 V, v_R = 5 V and i = 0.5 A.
Current given: choose another computational dependency. Now require i(t) = 0.5 A and keep v_C(0) = 2 V; the source voltage is to be found. The same relations give dv_C/dt = i/C = 5 V/s, hence v_C(t) = 2 V + (5 V/s)*t. Then v_R = R*i = 5 V and v_s(t) = 7 V + (5 V/s)*t. At t = 0.1 s, return v_C = 2.5 V, v_R = 5 V and the required v_s = 7.5 V.
Release the earlier condition v_s = 12 V when making v_s an unknown. Keeping it would contradict the new question already at t=0, where the relations require v_s = 7 V. This change chooses another computation from the model; whether a source can deliver the resulting waveform is a C.29.3 question.
Structural reduction and numerical choice are separate. The substitutions remove algebraic unknowns and retain their reconstruction formulas under R,C > 0. They do not select a time-stepping algorithm. The closed form is adequate for the constant-voltage question above. A numerical variant could instead take a forward Euler step:
v_C_next = v_C + h*(v_s - v_C)/(R*C).
With h = 0.1 s, its first step gives v_C_next = 3 V, from which the readouts are v_R = 9 V and i = 0.9 A. The closed form gives v_C(0.1 s) = 12 V - (10 V)*exp(-0.1) ≈ 2.95162582 V; this step's voltage error is about 0.04837418 V. Selecting a step rule and size therefore needs the requested accuracy. The exact elimination of v_R and i did not cause that time-discretization error.
Consistent initialization is another problem. In the voltage-given question, the prescribed v_C(0)=2 V and v_s(0)=12 V force v_R(0)=10 V, i(0)=1 A and dv_C/dt(0)=10 V/s. An initializer's guess i(0)=0 A may be replaced while finding these values. Making i(0)=0 A an additional required condition instead contradicts the relations. The capacitor's prescribed initial voltage represents an initial physical condition in this model. A starting guess guides numerical search; the resulting numerical approximation is assessed against the initial constraints and required accuracy.
For a larger differential-algebraic model, obtain the needed consistent-initialization, tearing or index-reduction Method from that discipline. Return the equations, givens, initial constraints and requested readouts with the unresolved question. A numerical initialization failure alone does not establish that the constraints are inconsistent; the contradiction in this small case follows from the displayed algebra.
Construct a probability estimate with a specified error guarantee
Question and available operation. Estimate a fixed unknown probability p from independent binary observations X_1,...,X_n with the same probability p of 1. Require the procedure's probability of an error of at least 0.05 to be at most 0.05, for every p in [0,1]. The sampling operation and its independence are premises supplied to this construction; C.29.3 examines their realization.
State and procedure. Retain two integer counters: observations read j and ones observed s, initially zero. For each observation x, set s = s+x and j = j+1. After the selected n observations, return p_hat = s/n. The invariant is that s counts the ones in the first j observations. Thus the counters obtain the sample mean without retaining every observation.
Choose n from the guarantee. For a binary observation, E[X]=p and Var(X)=p*(1-p) ≤ 1/4. Independence gives E[p_hat]=p and Var(p_hat) ≤ 1/(4*n). On the event |p_hat-p| ≥ epsilon, the squared error is at least epsilon². Therefore
Taking n = 2,000 makes the bound 0.05 at epsilon = 0.05. The guarantee concerns repeated executions under the sampling model; it is not a posterior probability assigned to p after seeing one estimate. For instance, s=1,100 returns 0.55, while the guarantee still belongs to the stated procedure.
Suppose the recipient will act on one fixed sample of 2,000 observations when its mean is at least 0.55. If the true probability is at most 0.50, that action requires an estimation error of at least 0.05. The derived bound therefore limits the chance of taking that action under any such p to 0.05. Repeatedly inspecting interim means and acting at the first crossing changes the rule and requires its own error argument.
This construction consumes 2,000 observations and counter updates. Each counter needs 11 bits to represent values through 2,000; the output can be retained as the rational s/2,000. Include the cost of obtaining an observation when that cost matters. This conservative bound already supplies a finite construction; a sharper concentration argument can reduce the required observations when that saving is worth obtaining.
Change the sampling premise. If every read repeats one sampled bit B, then p_hat=B for every n. At p=0.5, its error is always 0.5, so the requested guarantee fails. More reads of that retained bit do not repair the construction. Obtain a sampling operation with the required independence, or recompute the error bound from the dependence supplied.
Bias-Annotation
Conformance Checklist
Apply these checks to the computational claim being made.
Common Anti-Patterns and How to Avoid Them
Consequences
The practitioner can now tell whether the computation answers the question, answers a useful restricted question, or still lacks a construction. Procedure recovery can expose missing information before implementation, and a representation-level estimate can reject an infeasible design before substantial resource use.
The cost is making the answer-producing connection explicit. For an ordinary small calculation this may take only a few lines. A large or delicate problem can require substantial algorithmic and numerical work; this pattern helps identify that work and connect its results. Retaining a simple reference procedure may cost extra implementation effort while providing an intelligible comparison for an optimized candidate.
Architectural Rationale
Why formulate around the answer and operations together? Starting with a data structure is convenient when it already supports the required query. Otherwise it can discard the needed distinction or force an unnecessarily expensive computation. Starting only with an answer predicate has the opposite defect: it can hide unavailable operations. Connecting the two makes each choice answerable to the same use.
Why recover a known construction before inventing another? Its procedure and argument may already resolve the question at low cost. A new representation or specialized algorithm becomes worthwhile when an input property, repeated use, precision requirement or resource limit changes the result. A direct formula, a library routine, exhaustive enumeration and a new algorithm are alternatives; none wins merely by looking more formal or sophisticated.
Why keep a simple procedure when a faster one exists? A small exact baseline exposes meaning and can provide expected results for selected implementation tests. Bisection's interval reduction is easy to inspect; a more elaborate solver can reduce expensive evaluations. The choice depends on the present cost of evaluation and the needed guarantee. A reference calculation is not required if an already adequate construction supplies those answers.
Why separate computability, correctness and feasibility? A finite procedure can be correct yet exceed available memory. A fast implementation can return the wrong interpretation. A failed search can leave computability open. Distinguishing these questions makes the return useful: obtain a construction, repair its argument, choose another representation, or change the execution arrangement.
Why allow formulation to return from realization? Available physical operations may suggest a better computational model, and a mismatch may reveal that an assumed primitive or output is unavailable. The connected use therefore permits revising the formulation and the proposed executing arrangement together. The formulation still needs an interpretable result and a justified computational claim; the physical comparison remains a separate contribution.
SoTA-Echoing
Practice question. What is the strongest usable way to turn a stated mathematical result into a computation at the effort warranted by the task? The selected line is to reuse an applicable construction when possible, make its input/output meaning and argument explicit, and compare alternatives using the accuracy and resources that can change the answer. There is no single best algorithm across the input classes in this pattern.
Meaning and correctness before a larger claim
The Dafny tutorial, “Loop Invariants” and “Termination” demonstrates constructing a preserved relation to the answer and proving progress separately. Adapt that practice in :4.4 and :5.1: a short manual argument can establish the finite interpreter's stated semantics; a few traces cannot establish its whole input class. Mechanized checking is an alternative when program complexity or assurance needs justify its specification and proof effort. Reopen the choice when the procedure or required assurance becomes too large for the retained argument.
Choose numerical methods by their guarantees and costs
The SciPy bisection documentation states the bracketing premises and its absolute-plus-relative termination criterion. Adopt explicit tolerance selection in :4.5; library defaults need not match the receiver's requirement.
The documented Brent routine is an alternative combining bracketing, bisection and interpolation. Adapt the choice in :5.2: retain bisection for its simple bound and small exact case, while considering Brent's method when function evaluations are costly. The trade-off is a simpler argument for potentially more evaluations. Reopen for a changed function class, tolerance, evaluation cost or finite precision failure.
Cost the represented objects, not the problem label
Qiskit Aer's simulator documentation supplies a comparator: dense state-vector storage, alternative simulation representations, and controls that discard matrix-product-state coefficients. Adopt representation-sensitive counting in :4.6 and reject treating a truncation setting as an accuracy guarantee for an arbitrary observable. The :5.3 payload arithmetic and restricted product-state construction expose the gain. The suitable general simulator remains unselected until the state family, operations, output and error are known. Reopen the calculation when those conditions or the algorithm change.
Let executing capabilities change the formulation
Kalita, Butler, Stepney and Kendon, Novel models of computation from novel physical substrates: a bosonic example, v1 (2026), §§1, 3–4, develop computational concepts, a language and reference implementation from a physical-model contribution. Their bosonic example includes probability distributions in the computational meaning. Adapt the reverse entry in :4.2 and the return in :4.7 instead of requiring every device to implement a preselected model. This is a research Method candidate with a bosonic illustration. Section 4.7.1 explicitly leaves physical implementation for later work; its simulation does not establish device execution.
Stepney, Co-designing the computational model and the computing substrate (2019), §§4–5, proposes jointly exploring model and substrate rather than fixing one permanently. Adapt that reciprocal return while retaining a settled model when it already answers the use. Joint design costs a larger search; undertake it when a capability or mismatch can change the useful computation. The paper's demonstrated reservoir-characterization work does not establish the proposed general co-design process. Reopen for a useful operation excluded by the chosen model or a formulation whose required operation cannot be realized.
Keep structural reformulation and numerical solution distinct
ModelingToolkit's model-building reference, “System simplification” and “Exploring the results of simplification”, describes reformulating equations and recovering eliminated variables through stored expressions. Adopt that pairing in :4.2 and :5.4. Hand substitution suffices for the small circuit; a symbolic compiler becomes useful when model size or changing equations make that work substantial.
Its initialization tutorial distinguishes required conditions from guesses and shows how changing the givens can require releasing a retained constraint. Adapt that distinction when parameters or initial quantities become unknowns. Consistency of an initial system and numerical success in finding its solution are different questions.
Dyad's transient-analysis documentation separates the initial-value problem from algorithm and tolerance choices. Adopt that separation: structural reduction supplies a computational problem and recoverable outputs; numerical analysis supplies an appropriate solution method and accuracy argument. Reopen the affected choice when equations, initial constraints, requested outputs or accuracy change.
Relations
- Member of C.29: supplies the computational formulation used when the mathematical account does not yet provide a way to obtain its consequence. Direct entry is available.
- Uses C.29.1 when needed: establishes the result-preservation or bounded-transfer claim for a representation reduction or changed operation.
- Exchanges results with C.29.3: supplies represented inputs, operations, output interpretation and computational conditions; receives execution limits or available capabilities that can revise them.
- Uses B.5: recovers an available construction and its argument from an explanation, and revises dependent reasoning when premises change.
- Uses A.3.3: constructs sufficient state and permitted continuations, including control and retained history when required.
- Uses A.6.3.RT: constructs expressions under available notation rules and checks the content of a representation change. Notation- or language-design methods supply missing rules.
- Uses C.39 and C.40 by their entry conditions: obtains a missing way, or develops branching search when a workable change-and-test operation is available.
- Uses subject Methods: algorithm design, numerical analysis, symbolic computation and other computational disciplines supply their constructions, proofs and cost models. C.16 and physical-modeling methods supply measurement and empirical conditions when the requested result depends on them.
- Coordinates with A.3.1, A.6.1, B.1.5 and B.1.6: Method identity, an explicitly needed operation declaration, composition of identified Methods and resource accounting for performed Work remain under those patterns. A description of a procedure is not a claim that its execution occurred.
- Uses C.2.1, A.10 and B.3 for the corresponding claims: C.2.1 governs episteme identity, A.10 governs evidence use, and B.3 governs assurance of the result and reliance being asserted.
C.29.2:End
Computational Realization
Type: Method pattern Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this when a computation is available, but the proposed executing arrangement has not yet been connected to its inputs, operations and interpretable results. A motor interface may accept a different number from the one calculated. A circuit may produce a voltage whose scale differs from the input scale. A material admission procedure may preserve a count only while several people share the same stock of cards.
Start with the result the arrangement must deliver. For an input or history covered by the order, explain how the system is prepared, which actions perform the computation and how its result is read. Compare that result with what the recipient needs. An operating law can establish the comparison over its whole range. A particular input can expose a failure, such as a command outside the interface's range.
The first result is a realization design under stated conditions, an interpreted result from an execution, or a failed connection with a proposed correction. The method helps construct and use that connection. The engineering of a circuit, controller, physical apparatus or administrative arrangement supplies the mechanisms that perform the computation.
The reader should understand the intended computation and be able to obtain explanations of the system's operations. The examples supply their arithmetic, elementary electrical account and card rules. More demanding uses need the corresponding knowledge of hardware, numerical methods, physical models or concurrent operation.
Use an adequate existing implementation when its inputs, behavior and result interpretation already meet the receiving need. Use C.29.2 when the computational procedure itself must be constructed. Use B.5.MPC when the open question also concerns which physical account and mathematical question can answer the original problem.
Problem
An abstract operation can be well defined while its proposed execution changes the question. The integer 63,662 can be the right calculated command and be delivered as -1,874 by an overflowing signed field. A circuit can correctly average two voltages while its user reads that average as their sum. Two attendants can each follow the same capacity rule while admitting twice the intended number of visitors through duplicated stocks.
The failure can occur before execution, during intermediate actions or when the result is interpreted. More arithmetic at the middle of the calculation leaves those failures in place. A successful trace of one input can also leave another required input outside the representable range, or omit an interleaving that breaks a shared-state assumption.
A practitioner needs to construct the correspondences, establish the property needed by the receiving use and locate which part to change when the interpreted execution fails to supply that property. The useful comparison can concern a value, a bound, a probability distribution or a property maintained throughout an interaction.
Forces
Solution
Build and compare two routes from an input to the result required by the use:
The second route must supply the result relation needed by the receiving work. This can be equality, a bound, agreement of output distributions or a property of continuing behavior. Input preparation and output interpretation can differ.
Recover the needed result → prepare a representable input → construct the executing actions → interpret the output or behavior → compare the routes → repair the failed connection → use the sufficient result.
Enter at an available contribution. A known device behavior can guide a revised formulation; a settled implementation can supply an operation within a larger computation.
Determine what the receiving work needs
State the inputs and the consequence to be supplied. For a terminating computation, identify the output and its meaning. For continued interaction, identify the property that must hold across the relevant histories: for example, whether another admission is permitted without exceeding capacity.
Choose the comparison accordingly. An integer command can require equality of counts. A physical readout can support an interval containing the computed value. A concurrent procedure can need a bound that holds during every permitted intermediate state.
For a sampling computation, identify the requested distribution or statistic and the discrepancy that the receiving use can tolerate. Individual draws from two adequate samplers can differ. If the computation supplies an estimate to a larger learning or control method, determine what accuracy that method needs from the estimate.
Keep the computational result and its later physical use distinguishable. A controller can calculate a command, and an actuator can then change a physical system. The calculation and the resulting movement consume different premises. In the robot example below, the interface must realize the intended count; interpreting that count as travel additionally uses the motion model.
Use the available premises for a conditional design. An assertion about a performed execution uses the observations and operating circumstances needed to establish that assertion. Additional testing is selected for what it could change in the receiving decision, through C.11.DUA.
Construct input preparation
Explain how a required input is set in the system. Identify the state of the executing arrangement that represents it: a stored value, an applied signal, an arranged material stock or another prepared condition.
Determine the admitted range and the relevant distinctions. A signed field has a finite range. A voltage input has units and a reference. A stock of three exclusive cards differs from two independent stocks of three. Recover initial state when the input is interpreted relative to it.
Apply the preparation to the required input. If it cannot be represented, choose among changing the encoding, changing the arrangement, restricting the input set or revising the computation. A large integer can sometimes be represented by several smaller values, but their execution must preserve the intended composition.
State rounding or other preparation loss in the same quantity that the receiving comparison uses. Retain a sufficient bound when it answers the question. Locate a lost distinction before deciding that a more precise number will restore it.
Construct the actions that execute the operations
For each operation whose realization is unresolved, identify what the system does and how its state changes. Connect sequences through the state they leave for the next operation.
When one logical operation spans several physical actions, examine the intervening states. Reserving a card, crossing a room boundary and returning the card are separate events. A visitor can be outside while the card is unavailable. Either show that the intervening behavior preserves the needed result, or change the model or the arrangement.
For shared operation, determine how the same resource or state is accessed. Exclusive possession means that each card can be held by only one participant at a time. Independent copies of an available-card count require a different coordination mechanism. The engineering or administrative method supplies that mechanism.
For repeated or timed operation, include completion, reset and ordering where they affect the result. Two relative increments add only when both are executed under the intended rule. Repeating an absolute target ordinarily asks for the same target again. An ongoing physical process can require a temporal model rather than a sequence of instantaneous assignments.
A physical evolution can itself perform the computation. Determine which initial preparation, controls and interval make its readout useful. A sampler may use a stationary law; a finite-time estimator may use a transient. Choose the evolution and reading rule for the requested result.
A.3.3 supplies the construction of state and permitted continuations. A.6.1:4.6 supplies the realization relation when a reusable operation declaration needs it. Their contributions leave the circuit, physical interaction or operating procedure to its subject method.
Construct result interpretation
Identify the state or indication that the receiver can obtain, and explain how it yields the computational result. Give its scale, reference, timing and aggregation rule wherever those change the answer. When several observations yield an estimate, include their dependence in the sampling argument; the number of readings alone does not determine the estimate's uncertainty.
Derive this interpretation from the executing relation. An input scale need not be the output scale. In the analog example below, each input uses one volt for ten units, while the output uses one volt for twenty units.
Check whether different possible outcomes have become indistinguishable to the readout. Counting unavailable cards can bound occupancy without determining it. A command acknowledgment can establish receipt without establishing completed motion. In each case, return the consequence the indication supports.
Use C.16 when reading the result requires a measurement model. Instrument loading, calibration, finite resolution or disturbances enter that model when they change the inference. A bound or another suitable indication can answer the question without reconstructing an inaccessible exact value.
Compare the routes over the claimed scope
Follow the inputs or histories covered by the claim through preparation, execution and interpretation. Establish the required relation from the operating rules, relevant evidence or both. A worked input can resolve an uncertain correspondence. A claim about one particular execution needs grounds about that execution, such as its recorded output under the agreed interpretation.
For a finite case, a complete case analysis can suffice. For an invariant, show it in the initial state and show how every permitted operation preserves it. For an approximate result, propagate the losses to the quantity used by the receiver. For a sampler or statistical estimator, compare the distributions or estimation properties. Use the mathematical, numerical or empirical method appropriate to that claim.
For example, a required binary sampler gives output 1 with probability 0.5. Suppose the device model gives probability p between 0.49 and 0.51 under the intended preparation. The absolute discrepancy in that probability is at most 0.01, so it meets a tolerance of 0.02. Comparing two individual draws cannot establish or refute that distributional result. Estimating p from a finite run requires the sampling assumptions and uncertainty calculation appropriate to those observations.
When the requested result contains several random components, compare their joint behavior. Suppose the target is a pair of independent fair bits. Preparing one fair bit B and returning (B,B) gives the right distribution at each output separately. But the target assigns probability 1/2 to unequal outputs; this system assigns probability 0. Using two independent draws gives each of the four pairs probability 1/4 and restores the requested result. The realization must provide that joint sampling behavior. When its observations are dependent, use the joint law or change how the second component is obtained.
When a case can distinguish competing accounts of the realization, choose it from features that matter to the requested result, such as an endpoint of the command range, a different initial state, overlapping actions or indication error. A permitted input with a confirmed output outside tolerance refutes a claim that every execution meets it. Establish the input, output reading and operating conditions needed for that comparison. A pair of different draws can be entirely consistent with a probabilistic guarantee. Use the sampled outcomes and their probability law for the distributional question.
Separate three possible outcomes:
- The interpreted execution supplies the required result under the stated conditions.
- It supplies a weaker result, such as an interval or occupancy bound, that may still answer the receiving question.
- A connection fails or remains unknown, and its location determines the next construction, observation or restriction.
Repair the connection and its dependents
Return to the part that changes the answer. Repair input preparation when a value cannot be represented. Repair the execution when permitted actions violate the intended operation. Repair readout when the available indication has been given the wrong meaning. Revise the physical or computational model when its retained state cannot express the relevant behavior.
Compare repairs by the result the work needs. Restricting a command range may be sufficient for one device. Splitting commands may preserve a larger range at the cost of extra execution time. Include the preparation, conversion, repeated execution and readout consumed by the proposed repair when comparing its cost with the available alternatives. A stronger sensor may be unnecessary when a conservative bound already settles the action.
Revisit contributions that consume the changed value, meaning or condition. Retain independent results whose assumptions still hold. B.5.MPC connects this local repair to the wider physical, mathematical and computational question.
When adapting the formulation and the apparatus together could improve the result, compare both directions. Ask which physical operations are available, how they can be interpreted computationally, and which change to the arrangement would make the useful operation easier or more reliable. C.29.2 develops the resulting formulation. Designing a new computational model or notation can require a substantial further method; name that task when the present construction reaches it.
Return the result at the strength obtained
Give the usable realization and the conditions that make its result interpretable. For an execution order, prepare the specified input, perform the operations and return the interpreted result. When designing continuing service, establish how preparation, operation and readout remain available across the admitted histories and how service can resume within the promised time. When entrusted with operating it, use the available observations to act or secure the required response from its responsible operator in time. Establish whether the promised behavior continued or was restored. A claim about service already provided needs evidence from the period concerned. A question confined to the design can be settled from its operating conditions.
A conditional design can finish before a device is built. An observed execution can establish a result under its conditions. An identified incompatibility can stop an unsuitable implementation and open another. Further information or assurance is obtained when the receiving use requires its contribution.
Archetypal Grounding
These constructed cases expose different realization failures. They establish consequences of the stated models; using an apparatus also requires physical and operating knowledge.
Carry a calculated count through a finite command interface
A robot's supplied motion model uses rolling without slip, effective wheel radius 0.05 m, ten motor revolutions per wheel revolution and one thousand commanded increments per motor revolution. Positive motor increments produce forward travel. Successful execution completes the requested relative increment.
For a requested distance d, calculate and round:
One metre gives 31,831 increments and a modeled displacement of approximately 1.0000003576 m. Rounding contributes at most half an increment, approximately 0.0000157080 m. This bound concerns command discretization under the supplied motion relation.
The interface accepts signed 16-bit values from -32,768 to 32,767. The one-metre input is representable. Two metres require 63,662 increments, which cannot be prepared as one positive value in that field.
Suppose the interface wraps modulo 65,536 and interprets the result as signed. Then:
The calculation of the desired count remains correct; input preparation changes the delivered count. An interface that rejects the value would produce a different failure, so the interface behavior matters.
One repair uses two completed commands of 31,831. Under the relative-addition rule, their total is 63,662 and their modeled displacement is approximately 2.0000007151 m. The preparation now supplies two representable values, and the execution argument uses addition across their completed displacements.
If a command means "reach this absolute count," repeating 31,831 leaves the same target. For an absolute interface, reconstruct the target from the initial count and check that the resulting target is representable. B.5.MPC supplies the wider comparison when the motion model or the count's physical meaning also changes.
A range restriction is another useful repair. The largest positive relative command corresponds to approximately 1.0294056648 m under this model. Whether that restriction is adequate depends on the requested motion. A deadline can make two commands unsuitable even when their displacements add.
The result is a command procedure with a stated input range and composition rule. To assess travel, determine whether the commands were completed and use a measurement or physical relation that accounts for consequential slip and effective radius.
Read an analog sum with the output scale it needs
An existing analog channel is intended to supply the sum of two numbers x and y in the range 0 to 10. Its input preparation sets voltages:
Each source is connected through an equal resistor R to one common node. The supplied electrical model has ideal voltage sources, equal resistances, settled behavior and a readout drawing negligible current. The resistor-current relation and current balance at the node give:
The arrangement produces an average voltage. To obtain the intended sum, derive its readout:
For x = 8 and y = 6, preparation gives 0.8 V and 0.6 V. The node gives 0.7 V. Decoding returns 14. Reusing the input scale, ten units per volt, would return 7.
The abstract and physical routes now agree. Input preparation and output interpretation use different factors because the intervening operation halves the voltage sum.
Now allow each prepared input voltage an error of at most 0.001 V. Suppose the readout indication has an additional error of at most 0.002 V under the same circuit model. The worst-case error in the indicated output is:
An absolute tolerance of 0.1 is therefore met under these bounds. A tolerance of 0.01 is not established by them. Better input setting or readout, a different realization, or a weaker receiving requirement would need comparison. The bound is deterministic; no cancellation or probability distribution has been assumed.
If the instrument loads the node or the resistor values differ, the stated averaging relation needs revision. Those changes belong in the circuit and measurement account, through C.16 where appropriate. The common realization method supplies the preparation/execution/readout comparison and the receiving error calculation. Electrical design supplies the physical law and component behavior.
Preserve a bound during material admission operations
A demonstration room has a capacity of three visitors. It starts empty with three distinct cards in a free stock. One card is issued to a visitor before entry; the visitor retains it inside and returns it after exit. Every entry follows this rule, each visitor inside holds one card, each card is exclusively assigned, and cards are not lost or duplicated.
The computation is the continuing admission decision and preservation of the capacity bound. People and material transfers execute its finite-state procedure.
A card can be Free, Reserved outside, Inside with its visitor, or Awaiting return after exit. Let F, R, I and E count those states. The fixed stock gives:
Issuing a card changes Free to Reserved. Entry changes Reserved to Inside. Exit changes Inside to Awaiting return. Return changes Awaiting return to Free. Cancellation can return a Reserved card to Free while its visitor remains outside.
Every operation moves one existing card between states. Entry requires its exclusive reservation. Return requires that its visitor is outside. These conditions preserve both the stock and the association between visitors inside and Inside cards. They establish the capacity bound throughout permitted histories, including overlapping admissions.
After two cards have been issued, one visitor may be inside and the other waiting outside. Then F = 1, R = 1, I = 1 and E = 0. The count of unavailable cards is 2, while occupancy is 1. Reading occupancy as 3 − F would be wrong. That count is an upper bound and equals occupancy when R = E = 0.
At the shared issuing point, a free card can be exclusively reserved for the next admission under this procedure. If all three cards are issued and none is free, another reservation must wait. The room itself could be empty with all three cardholders still outside, or full with all three inside. The free-stock count settles whether another card can be issued now. With all three cards free, the room is known to be empty. When several occupancies fit the available count and the exact number matters, count the visitors inside or reconcile entries and exits to the same instant.
A second entrance can share the same stock or receive a partition of it, such as two cards at one entrance and one at the other. Both preserve the total of three. Giving the second entrance three copied cards instead permits six simultaneous admissions, despite each attendant following their local procedure.
Partitioning can cause waiting at one entrance while a card is free at the other. Moving an existing free card preserves the total; creating another accepted card changes it. This separates the capacity property from the additional question of useful service across entrances.
The result is a conditional admission arrangement, its preserved bound and an occupancy bound derived from the free stock. The administrative method supplies controlled entry, exclusive possession and return. The mathematical construction makes explicit which physical and procedural properties the computation consumes.
Bias-Annotation
Conformance Checklist
Apply the checks to the realization and use. Their answers can remain in the working calculation, diagram or operating explanation.
Common Anti-Patterns and How to Avoid Them
Consequences
The practitioner can locate an implementation failure without replacing every part of the reasoning. The desired count, arithmetic and motion model can remain useful while input encoding changes. A readout can be repaired while the circuit remains intact. A shared-stock rule can preserve capacity while another method improves waiting time.
The comparison also clarifies the division of work. One contributor supplies the computation, another the executing mechanism, and another the needed measurement relation. Their results join through the preparation and interpretation that the receiving use consumes.
Constructing those connections can cost more than the displayed calculation. Reuse settled implementations and sufficient bounds. A new device, computational model or concurrency mechanism can require substantial specialist work beyond the common method.
Architectural Rationale
Why realization needs its own method
Mathematical result transfer compares accounts and their operations. Computational formulation constructs a procedure for the requested result. Realization begins with that procedure and asks how an executing arrangement supplies it. Its new work is input preparation, execution by the embodied arrangement, readout and comparison under the conditions of use.
A.6.1 provides the declaration and realization relation. It leaves the design and interpretation of an executing arrangement to the method that uses the relation. C.29.3 supplies that constructive work within the mathematical-use family. The method can be entered with an available computation.
Checking a program alone is useful when its execution environment is already established. It is insufficient for the unresolved interface, readout and shared-stock cases here. The present comparison follows only connections on which the receiving result depends.
Preparation, execution and interpretation
Horsman and colleagues (2014) distinguish abstract computation from its use through physical preparation, evolution and representation. Adopt their comparison as the basis for :4. The practical extension here is to return the equality, bound or behavior needed by the work, and to repair the connection that defeats it.
That account is one theory of physical computation. The method uses its constructive comparison. For the declared use, establish whether the interpreted execution supplies the required result under the stated conditions, using :4.5.
The examples explain why both directions matter. Input preparation asks what state can be made; output interpretation asks what result can be obtained from the state or indication. They can have different scales, ranges and physical means.
Realization and the surrounding physical activity
A computation can contribute to physical control while the controlled system continues to evolve. Horsman, Stepney, Clarke and Kendon (2026), §§3.3–3.4 and 5.3 distinguish the compute cycle within control from the broader physical control cycle. Adopt that distinction in :4.1 and the requirement to include consequential timing in :4.3.
The robot calculation supplies a command; the command's execution and its relation to travel supply further claims. In the card arrangement, the abstract admission rule is carried through reservation, crossing and return. The physical conditions that maintain those relations must remain understandable.
Changing model and means together
A fixed computation can be a useful constraint on the device design. A fixed device can instead suggest more suitable computational operations. Stepney (2019), §§4–5 proposes combining these directions in model-and-substrate co-design.
Kalita and colleagues (2026), §§1 and 3–4 develop the reverse direction through a bosonic-device example: capabilities of the physical system inform a computational model and language. Adapt this source contribution as the formulation return in :4.6. Developing such a language requires its own constructive repertoire; a return from realization identifies that work.
The analog example shows a small instance of the same design freedom. Keeping the averaging circuit and changing its encoding or decoding can supply a useful addition operation. The physical relation constrains which interpretation works.
Thermodynamic sampling and optical ML hardware extend this design choice to computations based on distributions or iterative physical evolution. Melanson et al. (2025) demonstrate sampling and matrix inversion on a small stochastic circuit. Kalinin et al. (2025) co-design an optical/electronic fixed-point computation with its learning model. Their different result and execution forms motivate the choices in :4.1–4.5. The detailed sampler, model-training and device-construction Methods supply the corresponding specialist work.
SoTA-Echoing
The selected answer combines a physical-realization comparison with scope-sensitive result use and a return to formulation when the available means suggest a different construction. The examples are constructed under their stated models.
Reconsider the realization when a required input, observation relation, execution condition or receiving tolerance changes. A new substrate or computational model can also change which preparation, operations and interpretation are worth developing.
Relations
C.29.3:End
Construct a Balance across a Boundary
Type: Method Status: Stable Normativity: Normative
Problem frame
Use this pattern when you need to relate a total's change over an interval to transfers across a boundary and creation or removal within it. The difficulty may be choosing what belongs inside, combining accounts for parts, or understanding why a balance changes after the boundary is redrawn.
Two tanks can gain less liquid than their external inflow minus outflow suggests because some liquid remains in their connecting line. A job moving between two stages changes their separate counts while leaving the total unfinished count unchanged. A computation can introduce a discrepancy if two cells use different amounts for their shared transfer.
The Method constructs a balance for an additive quantity over an interval and collection of parts. Its first result can be a predicted total, a bound, a missing-transfer question, or a computational update that preserves the supplied balance. The construction may describe a physical region, a defined population or a mathematical partition; identify which one answers the working question.
The reader needs the quantity's meaning and enough arithmetic to combine amounts with signs. A physical application also needs storage and transfer laws. A rate or field calculation needs the calculus used by that model.
If a balance already answers the question, use it. This pattern is useful when the balance itself must be built or revised. Detailed transport laws, reaction models and numerical solvers remain contributions of their respective subject Methods.
Problem
An equation such as “change equals input minus output” can leave its subject undecided. The parts may overlap, an intermediate store may be omitted, or two records may count the same transfer at different times. The quantity may also be created or removed within the boundary: the number of unfinished jobs changes when a job is completed, while the amount of a conserved material follows a different law.
Adding local equations works only after their quantities, intervals and shared transfers have been made compatible. Otherwise cancellation can hide a missing store or combine values that cannot be added.
Forces
Solution
Choose the quantity and boundary, account for changes in each included part, and combine only matching transfers. Use the resulting balance to answer the question at the needed resolution.
Choose an additive quantity and a common interval
Start from where and when the recipient needs the result. Identify what would count there and which things belong to that total. Choose the quantity Q, its unit or counting rule and the interval. When the boundary is yours to choose, try treating a consequential exchange as an internal transfer by including both sides. Can you then recover the requested total from the larger amount and what was added? At a smaller boundary, net transfers and internal production or removal can establish the change in stock. To obtain an ending stock from that change, also establish the starting stock or a sufficient bound. An observation of the ending stock can settle the question directly. Choose the boundary whose terms and return to the requested quantity can be established. Let S consist of disjoint included parts, or account for overlap before summing.
For disjoint parts with amounts q_i, additivity means that Q(S) = sum q_i. The rule needs a basis in the model or counting definition. For a population, assigning each included job to one stage can provide it. For physical storage, use the physical quantity and law.
For stored thermal energy, construct an additive quantity from the material and state relations. A question about the maximum temperature needs a measured maximum, the local temperatures or another relation from which that maximum follows. If the question asks only whether a temperature limit holds everywhere, a suitable upper bound can suffice. Use a sum when its relation to the requested result supports the inference.
Keep all amounts on the same interval. A transfer that leaves one store before the endpoint but reaches another afterward requires an in-transit store or separate boundary crossings. For snapshots at different times, first recover their relation to the requested interval.
Identify transfers and internal production or removal
For each included part, distinguish the amount crossing into it, the amount crossing out, and the change created or removed within it under the quantity definition.
Orient each transfer once. When a transfer from part i to part j removes amount f from i and adds the same amount to j over the common interval, use -f and +f in their accounts. If a connector stores some of the quantity, its incoming and outgoing amounts can differ; represent that store or retain both crossings.
Production and removal depend on the quantity. A reaction can change the amount of one chemical species within a vessel. Completing a job removes it from the population of unfinished jobs. Determine those terms from the physical or counting rule. Assign each contribution once: classify completion as a crossing out of the population or as removal under its counting rule, and use that convention throughout.
Do not set an unobserved term to zero merely because the first diagram omitted it. Use a supplied law, a justified approximation, a bound, or an unresolved term according to what the answer needs.
Combine the local balances
Write each included part's change using the terms. Sum those equations. Matching internal transfers cancel, leaving
The transfer terms are amounts over the interval. A rate must be integrated over that interval, or multiplied by its duration when the rate is constant. The signs follow the boundary.
For a differentiable rate model, the corresponding instantaneous relation is
Use the finite-interval form for discrete events or when a rate description adds no help. If the boundary moves, count crossings relative to that moving boundary using the transport relation.
A closed boundary eliminates crossing terms only when the model excludes those exchanges. Constancy of Q additionally requires net internal production and removal to cancel. Preserve that premise in any conservation conclusion.
Redraw the boundary by reclassifying the same exchanges
When enlarging S, include the added stores and their changes. Transfers between the old and added parts become internal only if both sides now refer to the same amount over the interval. The new exterior crossings remain.
When shrinking S, formerly internal transfers can become inputs or outputs. A direct balance around the smaller region uses those crossings. To recover its amount by subtraction, use the larger total and the amount in the removed region at the same time under compatible counting conventions. A direct reading or a justified bound can also settle what the recipient asks.
Check the two descriptions against each other: the larger total equals the sum of its disjoint parts, and their changes agree after shared transfers cancel. This comparison can expose an omitted connecting store without resolving every internal rate.
Changing the parts over time can also change membership. Include that effect through boundary crossings or a population rule; keep it visible when comparing successive totals.
Use uncertainty or discrepancy to choose the next move
Substitute the available amounts and compute the requested consequence. If some terms are bounded, propagate the bounds through the signed sum. Combining separate extremes can give an outer bound even when those extremes cannot occur together. Use it when it settles the requested threshold. For an attainable value or a guarantee that the outer bound leaves open, use the jointly allowed transfers and stores. A permitted case can refute the guarantee.
For observations, subtract the change predicted by the recorded transfers and internal terms from the observed change in storage. Propagate their uncertainties and shared constraints into a range for this discrepancy. If the range excludes zero, the records and balance cannot agree under those assumptions. If it contains zero, establish whether jointly admissible values actually give zero before claiming agreement is possible. Equal unrecorded inflow and outflow can cancel even when the discrepancy is exactly zero. To locate a fault, examine a term or an independent observation that could distinguish its possible causes.
Resolve a remaining term only when it can change the answer enough to justify the work. A bound that settles the decision is a usable result. An unresolved cause can remain open when the decision tolerates the resulting uncertainty.
Preserve the balance in a computation and return its result
When implementing a partitioned calculation, represent a shared transfer consistently on both sides. Use the same orientation, interval and quantity conversion. Recover total amounts with the cell volumes or population weights before comparing them.
Derive the aggregate change from the proposed updates. If a numerical approximation changes the balance, determine whether the discrepancy fits the receiving use or repair the update. Balance preservation addresses one property of the computation; accuracy of the local solution requires its own approximation argument.
Return the total, bound or identified missing contribution with the quantity, boundary and assumptions needed to use it. For a further question about a local maximum, delay or available action, retain or recover the component information or other grounds that the new question requires. A direct law or bound can already settle it.
Archetypal Grounding
A transfer line changes which balance answers the tank question
Use a constant-density liquid, measured by volume in litres. At the start, tank A holds 30 L and tank B holds 10 L. Over the interval, an external supply adds 12 L to A, and an external outlet removes 7 L from B.
First suppose the line transfers 15 L from A to B with no change in its liquid content. Then
The internal 15 L cancels. The larger balance gives the combined amount without needing that internal transfer, while the separate balances give each tank's amount.
Now change the condition: the line begins empty, receives 15 L from A, delivers 13 L to B, and retains 2 L. The results become
The two tanks contain 43 L. Their boundary excludes the line, so 15 L leaves that boundary and 13 L re-enters. The combined tank change is 12 - 7 - 15 + 13 = 3 L. Including the line gives 45 L and recovers the five-litre increase from the external exchanges alone.
If the line's increase is known only to lie between 0 and 2 L, the final tank total lies between 43 and 45 L. This bound settles a request for at least 42 L in the tanks. The admissible 2 L retained in the line leaves only 43 L in the tanks, which refutes a guarantee of 44 L for every filling allowed here. An empty line would leave 45 L. To decide whether one unknown filling delivered at least 44 L, determine whether the line retained at most 1 L. The 42 L guarantee already holds throughout.
Reverse the question. Suppose the tank measurements show 43 L at the end and the net recorded external inflow was 5 L. Starting from 40 L leaves 2 L to account for outside the tanks. If the line began empty and could have retained 0 to 2 L, unrecorded net outflow from tanks and line together could be anywhere from 0 to 2 L, assuming exact amounts and no production or removal inside. These data rule out unrecorded net outflow above 2 L. They still permit all 2 L to be in the line. To diagnose leakage, distinguish storage and other unrecorded exchanges.
Moving and reworking a job preserve one count but change its distribution
Count accepted jobs that have not yet been completed. Each job occupies one of two stages, including any processing position in that stage. Initially A contains 6 jobs and B contains 4.
During an interval, 12 new jobs arrive at A, 9 move from A to B, and 7 are completed at B. With no other arrivals, completions or membership changes,
A limit of 12 unfinished jobs is therefore exceeded at the endpoint. Determining whether the limit was exceeded earlier requires the event history or a bound on its prefixes.
If 2 of B's remaining jobs return to A for rework before the endpoint, the stage totals become 11 and 4; the unfinished total remains 15. The extra work is consequential, but moving the same jobs changes their distribution rather than their count.
If the question instead counts separately created work orders, specify that new population and its creation rule. A parent job and two newly opened orders cannot be mixed into the earlier job count without changing its meaning. The next operations-management question may need workload or completion-time modeling in addition to this balance.
One shared transfer preserves the computational total
Two computational cells store amounts 10 and 20. One step adds 5 from outside to cell A, removes 2 from cell B, and transfers 3 from A to B. Using one transfer amount gives
Suppose A subtracts 3 but B adds 2.9. The resulting total is 32.9; the inconsistent interface updates lose 0.1 in that step. Repairing the shared amount restores the aggregate balance. Whether the transfer approximates the intended transport well remains a separate question.
Now let both sides use 2.9. The result is A'=12.1, B'=20.9 and total 33. The aggregate balance holds, but the supplied transfer of 3 requires the local values 12 and 21. Agreement between the two interface amounts preserves the total; their accuracy must be established for the requested local result.
If the cells store average densities, first multiply by their volumes to obtain the amounts. Summing unequal-volume cells' densities would test another quantity.
Bias-Annotation
The storage examples make additivity easy to see. Apply it elsewhere only with a corresponding quantity definition. For a count of file copies inside the boundary, making one additional copy adds one through internal production.
The job example is a pathwise counting account. It introduces no arrival distribution, steady-state condition or claim that all jobs require equal effort. Those assumptions become relevant only for further questions that use them.
Conformance Checklist
- The quantity, units or counting rule, included parts and common interval are sufficient to identify the requested total.
- Additivity is justified for those parts, with overlap or in-transit storage accounted for where present.
- Each crossing and internal production or removal term has its meaning and sign; only matching internal transfers cancel.
- A changed boundary includes the added or removed stores and reclassifies the affected exchanges.
- The returned value or bound supports the stated question. Any unresolved discrepancy retains its uncertainty instead of being assigned an unsupported cause.
- When a computation is used, its aggregate update has been derived with the correct weights and shared transfers.
Common Anti-Patterns and How to Avoid Them
Consequences
A boundary can remove the need to know complicated internal exchanges. A second boundary can locate where missing storage or transfer information matters. The same construction supports physical balances, defined counts and numerical preservation tests.
The total compresses information. Local capacity, spatial extremes, causal response and completion time can remain unresolved even when the aggregate balance is sufficient for its own question. Retaining the component account makes those further inquiries possible.
Architectural Rationale
The difficulty is constructing what is added and what cancels. Once compatible local equations are supplied, summing them is elementary; selecting parts, quantities and shared exchanges is the work that makes the sum useful.
A.3.3.TR constructs state-change rules and demonstrates cancellation of internal forces. This pattern develops the choice and revision of a balance boundary across different quantities and descriptions. C.29.1 handles result transfer between mathematical accounts; the present construction explains why the aggregation preserves a total and what must be restored for a smaller-boundary question.
The finite-interval relation is the first construction because it works for discrete events as well as physical storage. Differential and field forms require the subject's regularity and transport assumptions. This separation keeps an elementary count accessible while leaving richer physical and numerical constructions available.
SoTA-Echoing
OpenStax, University Physics Volume 1, §14.5 derives fluid continuity from mass flow and uses constant density to pass to volume flow. Adopt the need for a physical quantity basis; adapt the equal-flow case by retaining changing storage in the region.
Sonin, On Choosing and Using Control Volumes (2001), Methods 1–6 compares fixed and moving boundaries for the same piston-driven flow. The constructions obtain the same exit speed while changing accumulation and crossing terms; a boundary through the piston makes one formulation require a density-discontinuity treatment. Use this control-volume Method when its continuum quantities and mathematical preparation fit the question. The finite construction in :4 also serves event counts and partitioned computational updates.
Ketcheson, LeVeque and del Razo, Riemann Problems and Jupyter Solutions, “Finite volume methods” obtains aggregate conservation by weighting cell averages and canceling neighboring fluxes. Adopt compatible interface transfer in a numerical update. The book's approximate-solver construction supplies transport and approximation methods beyond this balance test.
For the changed tank question, compare two accounts with the same supplied interval amounts and elementary arithmetic. Direct accounting at the tank boundary uses the line's 15 L departure and 13 L arrival to obtain 43 L. Accounting over tanks plus line gives 45 L; recovering the tank amount then needs the line's final 2 L store. Choose the account whose needed quantities are available. Applying only the outside exchange to A+B gives the wrong 45 L answer because it omits the line's increase. If a balance already answers the question, another account adds no benefit.
The boundary-revision procedure is a methodological synthesis of additive accounting, physical storage and computational conservation. The tank and job continuations are constructed cases under their stated models. Detailed continuum transport, chemical reaction accounting and stochastic queue models need their own subject Methods.
Reconsider the quantity, boundary or balance form when membership changes, needed storage information is unavailable, boundary motion changes the crossing law, or a cheaper adequate construction becomes available.
Relations
- C.29 - Mathematical Lens Use: identifies the subject correspondence and the consequences that may be carried back from the mathematical balance.
- A.3.3.TR - Construct a Rule for State Change: supplies the broader construction of state evolution; this pattern develops an additive change account.
- C.29.1 - Mathematical Result Transfer: preserves operations and requested results when changing mathematical accounts, including an aggregation that loses local values.
- C.29.2 - Computational Formulation: constructs the computation and its approximation conditions; the balance test here checks its aggregate update.
- B.5.MPC.R: coordinates a changed question across physical, mathematical and computational accounts when revising the balance requires more than an arithmetic update.
- E.18.2 - Transformation Flow Mathematical Description: applies when the balance mathematically describes a transformation-flow structure or network.
- C.11.DUA: helps decide whether resolving a remaining discrepancy or term can improve the available action enough to justify the effort.
C.29.BB:End
Architecture and Selected-Structure Adequacy
Type: Architectural pattern Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern to establish how a holon (U.Holon) is organized, explain what that organization makes possible, or identify what a design would change. A description may show an existing arrangement, propose a different one or specify conditions that an arrangement has to meet.
Begin with the work the architecture account must support. Which property, relationship or unrealized requirement does the recipient need to understand? For an actual arrangement, establish which subject relations hold and how the selected A.22 structure organizes the holon. For a proposal, show how its parts would work together under the stated conditions. Constitute a C.2.1 claim when its later use requires one.
Select the structures the question turns on. Follow the constituent relation for each relevant kind of holon far enough to account for the property at issue. Where the result depends on several structures, trace the connections through which they contribute to the result under the same admissible conditions. Some connect constituents of one holon. Others connect independent holons through an exchange, a shared rule or an interaction. Establish composition where the explanation relies on it. Choose the detail of each model by asking what it would treat alike that the requested explanation or design needs to distinguish. A request to identify the next architecture action can end when that action and its grounds are clear.
For example, a team wants to know what to investigate before replacing a module. Its diagram proposes a payment-system interface. Establishing whether the running payment service actually depends on that interface is a useful next inquiry. A commission to explain the existing system requires the answer to that inquiry and what the dependency means for service behavior. Use the card below when its fields help the recipient keep these facts together.
The card can stop before a durable claim. An actualRelationNamed result requires obtaining ArchitectureRelation occurrences and their A.22 structure participants. Candidate, planned, required, desired, expected, modeled, or diagrammed organization remains claim content. When that content is carried by a separate identified episteme, candidateOrExpectedStructureClaimRefs refers to it. Use selectedStructureRefs only for independently identified A.22 structures; neither designation makes an architecture relation obtain.
architectureConcernCue briefly describes the concern and helps identify the structures on which the question depends. When a controlled cue is useful, use changeLocalization, substitutionOrReplacement, flowBottleneck, controlOrRateMismatch, dataCustodyOrStateResidence, physicalSeparationOrPlacement, evidenceReuseOrAssuranceReuse, scaleWindowOrCoarseningLoss, runtimeFailureMode, crossScopeResidual, descriptionViewLoss, or otherDeclared. Local phrases such as change localization failure, hidden crossing, source return, generated-view loss, or state-residence uncertainty may remain in sourcePhrase? or Plain prose. If the described holon, distinction between actual and candidate structure, architecture concern, and first move cannot yet be named, set questionDisposition to concernCueOnly or problemCardReady.
ArchitectureQuestionCard@Project is a triage aid for choosing one architecture move. questionDisposition records whether to keep a concern cue, prepare a separate ProblemCard, constitute an ArchitectureClaim, or name the pattern for a non-architecture claim. architectureRelationDisposition separately records whether an actual relation has been recovered or the content is candidate or expected only. claimPatternRefs contains PatternIDs whose content defines, constrains, or tests any separate claim; it does not identify a pattern-application occurrence. The card is not an evidence record, gate or decision. Include non-admissible overread? or a nonAdmissibleUse? field only for a supported limitation that changes the receiving use; an unused field needs no invented exclusion.
Across C.30, @Project in a record name is a compatibility and retrieval cue only. It identifies neither a project entity nor a composite project U.Work, and it establishes no context, authority, viewpoint, or parthood. When the card is local to one actual project, projectWorkOccurrenceRef identifies the composite U.Work recovered under A.15.6. Include architectureQuestionProjectUseRelationRef only when a named pattern defines the relation by which this card use concerns that Work and an occurrence of that relation obtains. A Work reference alone does not establish project locality. If that locality matters but the relation is not yet defined, record missing-governor; if locality does not matter, omit both project-local fields. Description publication and other project-local uses follow the same rule.
Use the ArchitectureDescription bridge when the next architecture move depends on a description whose subject, correspondence or use conditions need recovery. Cross-team reuse or a regulated use can create that need; retaining an already sufficient explanation does not. Ordinary prose or a question card can finish the question. Use C.30.AD when the description itself needs construction or repair.
What goes wrong if C.30 is missed: the practitioner reasons from a document, module diagram, transformation-flow graph description, mathematical lens, benchmark, maturity score, or decision record instead of recovering the described holon, selected structures, first architecture move, and non-architecture claim kind.
What C.30 buys in practice: the practitioner can determine whether the next work concerns the actual organization, a proposed change, or the description through which it is understood. This selects an architecture action and the basis it still needs.
Do not use C.30 when the question does not concern the architecture of one holon, an obtaining ArchitectureRelation, a selected architecture-relevant structure, an ArchitectureClaim, or the thin architecture-description bridge needed for one architecture move. Use the pattern that defines or tests the source, description, view, publication-use, or other non-architecture relation. If one piece remains an architecture claim, use C.30 only for that piece. Common non-architecture claim boundaries are summarized in C.30:12.
Precision-restoration pointer: if the issue under repair is whether architecture, architecture description, structural view, module diagram, model, source material, functional architecture, or a source label such as layer, level, tier, stack, block, expert, cache, router, or gate names an architecture claim, description, view, representation, publication form, source relation, structure, or claim defined or tested by another pattern, use C.30.P or C.30.STRAT as triggered before applying C.30 to the recovered architecture portion. Use C.29 for an unresolved mathematical formulation, transfer or reliance question that changes the architecture work. An adequate domain calculation or ordinary explanation can finish without a separate lens-use note. Keep trigger tables in those patterns; C.30 is applied only after an ArchitectureClaim, selected architecture-relevant structure, conditional ArchitectureDescription bridge use, C.30.AD application, or other claim named by value is recoverable.
Problem
Engineering teams use "architecture" for several different things:
- the selected structure of a holon;
- a diagram, model, table, dashboard, generated relation graph, or document;
- a module layout;
- a selected transformation-flow structure, flow description, or mathematical graph description;
- a functional, control, information, deployment, logical, or physical structure view;
- an ADR-like publication;
- a project-side claim defined or tested by another FPF pattern.
These uses are all useful in ordinary engineering speech, but they cannot carry the same FPF claim. The core distinction is the one used across FPF: actual subject-relation occurrences; the A.22 structure selected from them; the ArchitectureRelation that may obtain between that structure and one holon; a C.2.1 claim about the holon, relation, or structure; the Description episteme or view; the representation and publication objects; and any project decision about changing architecture are different objects.
The first-minute practitioner asks four questions:
- Are we recovering an actual architecture relation, considering a candidate structure, or only reading a representation?
- For an actual architecture claim, which subject relations obtain and which A.22 structure is selected from them?
- Which structure kind is in view—function, flow, control, module, Work, system-role-kind or assignment, enactor, information, data, placement, deployment, scale, or a declared logical structure—and which adjacent interface, evidence, or assurance relations matter?
- How is the inspected material being used: as claim content, description, view, representation, publication form, decision, source relation, or mathematical lens?
The useful result preserves the distinctions needed by the architecture question. A module diagram may omit a control or flow relation that changes the next action. Conversely, a candidate statement may settle the question before an actual relation or full description is available.
Forces
Solution
Start with the architecture question and its subject. Establish the obtaining relation under :4.1 for a claim about actual organization. A candidate or expected arrangement can support a conditional explanation or a design. When the open question is what a description represents, use :4.2 or C.30.AD. When the question is which permitted arrangement operated, use evidence of that operation. The description itself can settle which arrangements its rules permit. For an explanation, follow the relations that account for the required property. When the recipient needs a design, carry the grounded requirements to C.32.P2S and develop the organization still missing. The account can end when it supplies the finding, explanation, design contribution or next action that the recipient asked for.
In C.30, the EntityOfConcern is one identified subject of the current claim: the described holon, an ArchitectureRelation occurrence, a selected structure, or another subject object. A claim episteme, description, diagram, or publication is not a proxy EntityOfConcern for a world-side relation or structure. Description hygiene supports this boundary but is not the center of C.30.
C.30 provides a bridge from the holon, architecture relation, or selected structure to a separately constituted architecture-description episteme only when durable description use changes the architecture move. C.30.AD carries the full general architecture-description EntityOfConcern: multi-view description sets, viewpoint-based views, correspondences, source return, freshness, specification use, and publication boundary. C.30.AD.BA carries built-asset architecture-description, asset-information, digital-twin, and reference-designation specialization. Generic episteme, view, viewpoint, publication, form, representation, and carrier machinery remains with C.2.1, E.17.0, E.17.1, E.17.2, E.17, E.24.PUB, and C.29. C.30.ASV carries the selected-structure-to-view branch; C.30.TFS-REL, C.30.LCA, and other named subpatterns carry their direct structure relations and claims.
C.30 does not mint U.Architecture and does not redefine U.Viewpoint. It defines ArchitectureRelation and the architecture claim form. It also supplies the question card and rules for using selected architecture-relevant A.22 structures in one architecture question, recovering structure kind, concern, admissible use, and inspected-material use, choosing the first move, routing characteristic claims, using small boundary notes, and opening the thin description bridge. It does not make descriptions or views conform merely by form and does not test every structure-specific view. Generic rules about publication, deontic permission, promise, evidence sufficiency, assurance, decision, gate passage, Work authorization, or release authorization remain in the patterns that define or test those claims.
Direct architecture relation and architecture claim
C.30 keeps one subject-side relation and one claim-bearing episteme distinct.
Direct relation kind. ArchitectureRelation is the direct dependent U.Relation defined here between exactly two actual participants:
architectureBearingHolonRef— theU.Holonwhose realized organization is at issue; andselectedArchitectureStructureRef— oneU.Structureselected under A.22 from declared constituents, obtaining subject-relation occurrences, applied constraints and invariants, and an admissible-use frame.
The relation is applicable only when the structure's constituents and selected subject-relation occurrences are recoverable for that holon or its admitted constituents and the structure is being used as architecture-relevant organization of that holon. Its obtaining predicate is satisfied only when the selected structure is constituted under A.22, every selected subject relation required by that structure passes the obtaining test defined for it, and those constituents and relations organize that holon in the declared way. A design, requirement or diagram can identify the structure to examine. An affirmative claim about the holon needs grounds that this structure actually organizes it.
Occurrence identity is the participant pair over one maximal continuous interval during which that predicate remains satisfied. A different holon, a differently identified A.22 structure, or cessation followed by later renewed obtaining yields another occurrence. A changed concern, claim scope, effective reference scheme, description, viewpoint, view, representation, publication, or carrier does not by itself reidentify or create the relation. Ordinary prose may state the readable relation and stop; use A.6.REL only when a later receiver must distinguish this occurrence from another one.
Architecture claim episteme. ArchitectureClaim is an ordinary C.2.1 U.Episteme, not the direct relation and not a new architecture kind:
The C.2.1 identity basis is the exact content, one entityOfConcernRef, and effective reference scheme. Each by-value scheme in this pattern uses the C.2.1:4.1.2 range EpistemeInterpretationScheme and must supply the rules for reading the episteme or claim content named by that form, including its designations, predicate meanings and polarity. claimScope qualifies what the claim covers. modelUseStructureRef appears only when one independently selected bounded-model-use structure changes structure interpretation or selection for this receiving use. empiricalGroundingRelationRef names a separately obtaining grounding relation; neither grounding nor the optional model-use structure is an ArchitectureRelation participant.
For an affirmative actual claim, every architectureRelationRef resolves to an obtaining occurrence whose participants and predicate satisfy the settlement above, and every selectedStructureRef is the structure participant of one of those occurrences. A negative, unresolved, candidate, required, desired, or expected claim can remain truthful claim content without an obtaining occurrence; it uses no invented positive reference. A description, diagram or decision record can carry that claim. An authorized decision can also have a constitutive effect under an applicable rule, for example when a delegation takes effect on adoption. Identify the act and rule that establish that relation. For a change requiring installation, establish the relevant installed relations.
Earlier consumers may still say “open the ArchitectureOf@Context form in the current C.30 edition.” In this edition that legacy retrieval instruction resolves to the ArchitectureClaim form plus the separate ArchitectureRelation settlement above. The suffix supplies no field, participant, scope, scheme, grounding, project identity, or relation fact, and new records use the current names.
EntityOfConcern bridge. C.30 may make the described holon, one ArchitectureRelation occurrence, or one selected structure the EntityOfConcern of a claim. A later architecture description independently chooses the holon, relation occurrence, or structure it describes under C.2.1; it does not use a claim record as a world-side proxy. Publication occurrences, forms, representations, and carriers remain separate.
Holonic architecture modes
Recover the holonic architecture mode before applying MHT, structure, description, or mathematical-lens language:
Evolutionary-engineering architecture candidate bridge
Use this bridge when an open-ended search, quality-diversity archive, current pool, front, or selected set contains possible architecture moves. The archive or front is not an actual architecture relation. It becomes C.30 material only when the claim names the described holon, the existing or candidate structure and structure kind, the affected architecture characteristic, and the next architecture move.
ArchitectureCandidateMove is a claim note about a possible structural change. It records why a generated, retained, front-member, or selected-set variant can be considered as architecture material; it is not an obtaining ArchitectureRelation, work plan, local choice result, declared selected-set result, publication occurrence, decision, performed Work, or new kind. Candidate structure content remains modal until the structure is constituted and the architecture predicate obtains. Recover the proposed organization from candidateMoveClaim or a named proposal episteme even when no selected structure can yet be identified. References to identified structures can supply comparison targets without establishing the proposed architecture.
For common exits from this architecture question, use C.18 for archive generation or front maintenance, C.19 for current-pool treatment, G.5 for selected-set result declaration, and C.11 for local choice. If that result is made available to an audience, use E.17 for a source-backed publication face and return to source, and E.24.PUB for the publication occurrence, particular expression, carrier, audience, bounded use, and availability; identify a reusable form separately when its rules are needed. Cite E.11.PUR for a recommended FPF pattern use. Use A.15.2, A.15.5, A.21, or A.15.1 when the move enters planning, work-entry readiness, a gate decision, or performed Work. Keep only the architecture claim here: which holon and current relation are at issue, which candidate structure matters, which characteristic may change, and which next use is admissible.
Architecture-move wording creates no root U.Move, structure, relation, WorkPlan, readiness relation, gate decision, performed work, decision, or source-use claim by itself. When source wording uses “move” outside this architecture-candidate use, restore the concern through E.10.MOVE and name the pattern that defines or tests the recovered claim.
When the useful next work is synthesizing candidate architecture variants rather than judging or repairing one actual relation, stop the C.30 question card after naming the described holon, the distinction between current and candidate structure, the structure kind, the concern, the admissible-use frame, and the next admissible use. Use C.32 only to build the candidate architecture palette. When another claim becomes current, use the pattern that defines and tests it. For example, use A.19.CPM for comparison, A.19.SelectorMechanism for selector-policy use, G.5 for selected-set result declaration, C.11 for final local choice, C.32.PAD for a project architecture decision, A.10 for evidence, B.3 for assurance, A.20 for an internal-constraint result, A.21 for a named gate decision, the direct domain pattern for the release claim, and A.15 for Work. For audience publication, use E.17 for the source-backed face and source return and E.24.PUB for the publication occurrence and audience availability.
Conditional architecture-description bridge
C.30 does not define a second local ArchitectureDescription record shape. C.30.AD:4.1 defines the architecture-use specialization of the canonical C.2.1 episteme. C.30 admits only a bridge when durable description use changes the first architecture move.
The minimum bridge recoverable in C.30 is:
This bridge does not mint another description definition, local view-membership relation, subject-side architecture relation, selected structure, or truth fact. It lets the C.30 reader say why a description episteme matters for the next architecture move. Apply C.30.AD whenever the description itself becomes the EntityOfConcern under repair or the full mechanism is needed: multi-view description-set use, viewpoint conformance, correspondence, source return, freshness, specification-use boundary, representation and publication boundaries, or reusable architecture-description use.
An architecture-description freshness claim is canonical in C.30.AD:4.4. C.30 may point to it only to bound admissible use of the first architecture move; it is not empirical grounding, publication currentness, evidence sufficiency, or assurance.
Publication-use boundary
An architecture description or its publication does not by itself establish deontic permission, promise, prescription, evidence sufficiency, assurance, decision, gate passage, work authorization, release authorization, source authority, or publication-use authority. Use C.30.AD for the description boundary and the pattern that defines or tests each separate claim.
ArchitectureDescriptionPublication@Project is subordinate to E.17 and MVPK machinery. It records the publication of one source episteme or episteme-lane view. publicationViewpointRef? names the publication-side viewpoint only when MVPK needs one; it is not an architecture viewpoint and not a TEVB viewpoint. mvpkFaceRef is a publication-lane face reference, not an alternative source episteme, source view, or source relation.
Model cards, system cards, and evaluation harness reports enter C.30 through the same publication boundary or source-relation boundary. They may describe a model, deployed AI system, architecture claim, evaluation harness, or policy, but the architecture move needs the actual or candidate structure distinction, an obtaining ArchitectureRelation only when its predicate is satisfied, a bounded ArchitectureClaim when claim content is needed, and the pattern for any proof, release, or gate claim.
For an architecture use of the card or harness, recover the structure kind at issue. For a separate evidence, measurement, causal, assurance or decision claim, use its governing pattern; that question may need no architecture claim.
Architecture name formation
The word architecture is shorthand only after the described holon, selected structures, structure kind, architecture concern and admissible-use frame, and use of inspected material as source, description, view, representation, or publication form are recoverable. Without those qualifiers, it is a recovery trigger, not a stable FPF term.
Architecture characteristic assignment
C.30 identifies the bearer before any quality, fitness, measure, metric, score, modularity, or ility wording carries an architecture-adequacy claim. Those words cue recovery of the bearer and the characteristic being claimed.
An ArchitectureClaim may state a characteristic claim, but the claim episteme is not automatically the characteristic bearer when its content names the holon, direct architecture relation, or selected structure. Select the bearer using the pattern that defines or tests that characteristic claim. Likewise, a diagram or publication cannot inherit the subject's quality by describing it.
C.30 keeps only a bridge from structural characteristics to Q-Bundle relevance. If the claim says architecture causes an outcome improvement, assign causal use to C.28. If a structural characteristic is used as a mechanism, constraint, predictor, proxy, evidence relation or causal hypothesis for a Q-Bundle slot, state that connection and its basis. An ordinary sentence may suffice; ArchitectureStructuralCharacteristicQBundleClaimLine organizes the needed distinctions. The formula low coupling = maintainability leaves that connection unexplained.
ArchitectureStructuralCharacteristicQBundleClaimLine is claim content for first contact, not a U.Relation occurrence or reusable relation declaration:
The line supports an inspectable next question without claiming measurement, modularity score, evidence sufficiency, assurance, or causal proof. admittedRelationAndOccurrence is available only when the direct characteristic, evidence, or causal rule defines the relation kind or declaration, participant meanings, obtaining predicate, applicability, and occurrence identity and the referenced occurrences obtain. missingGovernor instead names the actual participants, proposed predicate, affected use, and missing definition need. If no defining rule exists for a needed reusable relation, use A.6.RCD; neither a local token, PatternID locator, nor this line admits one.
Minimal structural-characteristic claim-line examples:
ArchitectureCharacteristicQBundleClaim is the triggered full claim episteme. Use it only when publication, comparison, causal use, evidence reliance, assurance, gate, decision, or reusable cross-case reliance needs a durable bounded claim and the thin line cannot keep the content inspectable.
The full claim keeps the proposal inspectable through assertion polarity, the structural-characteristic and Q-Bundle-slot referents, scope or scale window, viewpoint when it changes interpretation, qualifiers, witness expectations, admissible semantic change classes, and bridge or loss boundary. These are claim-content fields. They neither declare a reusable relation kind nor make an occurrence obtain; a direct relation needs an admitted kind, participants, a defining predicate and applicability rule, and occurrence identity.
Reusable product-quality vocabularies may supply candidate characteristic names. Claim content may connect bearers and Q-Bundle slots. A direct relation obtains only when its participants and predicate pass the test defined for it. Use the applicable patterns for measurement, modularity scoring, reusable-structure accounting, bespoke-residue accounting, evidence, assurance, gate, causal, and scale-audit claims.
Relation to structural views
Use C.30.ASV to test structural-view adequacy for an architecture-description episteme about one selected structure. E.17.0 separately admits that same episteme as U.View through independently obtaining conformance to a viewpoint. C.30 defines ArchitectureRelation occurrences, bounded ArchitectureClaim content, and, only for durable description use, how its ArchitectureDescription bridge cites structural views. Hidden or lost structure, correspondence, source or reliance relations, and source-return boundaries stay explicit when they affect action. C.30.AD defines the full description mechanism.
A diagram, model, table, selected transformation-flow diagram, mathematical graph description, LCA diagram, C.29 lens output, ADR, dashboard, generated explanation, or other publication face may carry an architecture description or an architecture structural view. It does not become the architecture, and it does not become a conforming view only because it looks like a view.
Name affected structures or view losses when they change the next architecture move. An ordinary account can suffice; this note organizes that content when needed:
This note only names affected architecture structure for the next architecture use. For a separate decision, ADR-publication, gate-passage, evidence-sufficiency, or release-authorization question, use the pattern that defines or tests that object or claim.
Minimal boundary notes
Use these notes when a common architecture phrase is close to a claim defined or tested by another pattern but full use of that pattern is not yet needed.
Use the smallest account that preserves the next architecture move. Use a fuller claim or relation record only when its content or independently admitted relation cannot otherwise be inspected, compared, refreshed or bounded. These notes can carry a mathematical-use limit, affected structure or characteristic connection. They are not mandatory preliminary records before using the pattern for a separate decision, measurement or causal question.
These notes are not substitutes for the module-and-interface repair pattern named by value, interface specifications, signature records, conformance evidence, or module-and-interface repair. An open or platform label is not substitutability proof, security proof, or assurance. A source label such as layer, stack, block, expert, cache, router, or gate enters this note only after C.30.STRAT recovers a module-interface or adjacent architecture-relevant item. It becomes architecture-relevant only through local structure, interface, variation, substitution, migration, update, and hardening boundaries. Relation-heavy wording inside these notes remains a Plain cue until the module or interface relation is identified, the relation establishing the asserted use is identified, or the pattern that defines or tests the non-architecture claim is named.
Architecture mathematical-lens boundary
State the architecture decision that the mathematical model can settle. A proof that every behavior of a feasible candidate satisfies a criterion can settle that part of a design choice. To rely on that proof for an installed system, establish that the relevant behaviors of the identified configuration, under the intended operating conditions, are covered by the proof. A behavior permitted by a simplified model may be unreachable in the installation. Check its realizability before using it to reject the installed design. C.29.1 supplies the transfer test.
Use an ordinary sentence or the one-line boundary when it states the mathematical contribution and its limits sufficiently. Apply C.29 when a representation or transfer question remains unresolved. Its :4.4 reliance rule selects the sufficient account; a changed architecture action alone does not require a Mini or Full card.
Lens use by architecture problem:
A proposed controller halves the deviation at each update in a model with a fixed reference and no disturbance. Repeated halving proves convergence for every initial deviation admitted by that model. A wiring diagram with the same arrows also permits a controller with the correction sign reversed. Both give zero deviation when started exactly at the reference. A run at the reference therefore leaves the sign open. The installed correction law, its timing and the plant response over the intended range determine whether the convergence proof applies. Use C.29 when the model or this inference still needs work.
Epiplexity-like use remains a C.29 bounded-observer structural-information lens. It may help recover learnable structure from traces, but it is not an architecture quality, task relevance proof, causal proof, assurance, or selector authority.
Boundary and repair table
Currentness and smallest reopen. When a decisive input changes, reopen only the C.30 object and use conclusion that depend on it. A changed holon or obtaining subject relation reopens the affected selected structure and, if asserted, the direct ArchitectureRelation predicate; a changed selected structure or predicate result reopens that relation occurrence and any affirmative ArchitectureClaim reference; a changed claim scheme or ClaimScope reopens only that claim; and a changed description, view, source edition, admissible-use boundary, or definition of a directly used relation reopens its reference and dependent ArchitectureQuestionCard@Project disposition. Admissible results are to update the affected reference or claim mode, narrow use, re-run the direct predicate, or reopen the card when its next architecture move is no longer supported; unrelated structures, descriptions, and claims stay closed.
Worked slices
"We have the architecture in this diagram." The diagram is a representation or publication form. It creates neither architecture nor U.View; recover an ArchitectureDescription episteme and, when view use is claimed, its independently obtaining E.17.0 conformance relation.
Continue the payment question. In a constructed case, take an independently identified payment-system holon, PaymentSystem1, and an A.22 module-interface structure, PaymentStructure1, selected for its latency and replacement concern. Its constituents are Orchestrator1 and Scorer1; supplied subject facts establish that Orchestrator1 synchronously invokes Scorer1 under interface contract K. The selected structure's constraints and invariants hold, and this dependency organizes PaymentSystem1 in the declared way. These premises establish the ArchitectureRelation between that holon and that selected structure. A diagram showing a direct bypass is only a candidate until its subject relations obtain.
A code optimization inside Scorer1 can reduce its response time while leaving these participants, selected relations and invariants unchanged. If the architecture predicate continues to hold, retain that relation occurrence; assess the latency change on its own measurement basis.
A proposal to replace Scorer1 by a distinct Scorer2 has different candidate constituents. Describe that candidate structure and work out the intended invocation and compatibility with K. This can support the next implementation decision without asserting that the new relation already obtains. A successful isolated Scorer2 benchmark supplies neither an actual invocation relation nor end-to-end latency. After a replacement, an affirmative architecture claim requires the resulting structure's A.22 identity, actual subject relations and architecture predicate; until those are established, retain the supported candidate or unresolved claim.
"Low coupling gives maintainability." State the proposed connection and its basis. The thin claim line can organize that explanation:
Use ArchitectureCharacteristicQBundleClaim only when publication, comparison, causal use, evidence reliance, assurance, gate, decision, or reusable cross-case claim reliance needs the fuller bounded claim. If repeated use needs an independently admitted direct characteristic, evidence, or causal relation, apply the relation's defining pattern to identify its participants, obtaining predicate, applicability, and occurrence identity and to verify that the occurrence obtains. Do not accept the slogan as architecture truth.
"The backup-pump architecture is safe because the loop is redundant." C.30 starts with the plant holon, operating claim scope, effective reference scheme when local terms need it, and selected structures: control loop, material-flow structure, placement structure, module-interface relation, and maintenance-work relation. The redundancy phrase may motivate an architecture move, but use the applicable patterns for safety proof, causal proof, evidence sufficiency, gate passage, and work authorization. The C.30 output is the selected structure and next architecture move, not a safety case by slogan.
"We replaced the neural-network block, so the architecture improved." Recover what block names through C.30.STRAT when unclear. Then identify the changed structure or relation and the characteristic for which improvement is claimed. C.30.ASV's neural-network replacement case shows the conditional recovery of flow, module-interface, preserved or lost structure, source use and decision grounds. A modification need not assert all those claims; the claimed improvement needs its comparison basis.
Archetypal Grounding
Bias-Annotation
Lenses: Arch, Onto, Epist, Prag, Did, Gov. Scope: FPF architecture-description use over holons.
Use the following checklist for the architecture claim or action being considered. Its conditions may already be established in the working account; completing a separate form adds no missing relation or argument.
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Consequences
Rationale
An architecture question concerns how a holon is organized and what a proposed change would affect. ArchitectureRelation connects the holon to a selected structure; ArchitectureClaim can assert, deny, leave unresolved or propose that organization. Keeping these separate lets a candidate guide work before realization, and lets an actual change be distinguished from a new description of the same organization.
C.30 and C.30.ASV establish an FPF architecture kernel: actual subject relations first; selected A.22 structure; ArchitectureRelation to the described holon; separately constituted claim, description, viewpoint, and view epistemes; structure-kind discipline; correspondence and source-return boundaries; and characteristic-claim applications. They do not by themselves provide full measurement, synthesis, decision, causal proof, safety proof, or assurance.
A short account can explain a structural relation or settle a specific architecture question. Its length follows from what the reader needs to recover. A question about the organization of a whole can require showing how several selected structures interact. A request to name the next useful action can be met by identifying that action and why it matters.
Episteme identity and direct view conformance also preserve plurality. The same holon, architecture-relation occurrence, or selected structure may be described by several independently identified epistemes; one episteme may conform to several viewpoints through distinct occurrences; several publications may render one description. C.30 keeps those variants usable without turning any publication form into architecture or any bundle or list into view membership.
SoTA-Echoing
Relations
Builds on: A.1, A.22, E.24.PUB, C.30.P, C.2.1, A.6.3, A.7, E.10.D2, E.17.0, E.17.1, E.17, E.17.2, A.6.P, F.18, E.10, and C.2.P.
Coordinates with: C.30.STRAT, C.30.ASV, A.6.F, C.30.TFS-REL, C.30.LCA, C.30.ILC, E.18, C.29, C.16, C.25, C.28, A.10, G.6, B.3, A.20, A.21, A.15, C.11, G.5, E.17, C.32.P2S, C.32, C.32.PAD, C.32.ADR, C.32.ADA, C.33, C.34, and C.35 when problem-to-structure carry-through, candidate-set, architecture-decision, ADR-projection, decision-adequacy, structure-capture, preservation, or discovery-adequacy claim kinds are being made.
For neighboring claims, use A.1 for the described holon, A.22 for a selected-structure EntityOfConcern, C.30.STRAT for stratification-wording and source-label repair, C.30.ASV for structural-view adequacy, C.33 for captured and lost selected-structure adequacy plus source return, C.34 for preservation or correspondence adequacy, C.35 for generated or discovered carrier adequacy before C.32 admission, E.18 for selected transformation-flow structure, path, and crossing discipline, E.18.2 and C.29 for mathematical graph descriptions and mathematical-lens use, C.16 for characterization, C.25 for Q-Bundles, C.28 for causal use, A.10 and G.6 for evidence, B.3 for assurance, A.20 for an internal-constraint result, A.21 for a named gate decision, the direct domain pattern for the release, admissibility, or approval claim, A.15.2 for a WorkPlan, A.15.5 for work-entry readiness, A.15.1 for performed Work, C.11 for decisions, E.11.PUR for pattern-use recommendation, E.10.MOVE for move-like wording outside C.30 architecture-candidate use, and C.32.P2S for the connected problem-to-structure architecturing flow. For publication, use E.17 for a source-backed face and return to source and E.24.PUB for the publication occurrence, form, carrier, audience, bounded use, and availability. Use C.30 to state and test actual ArchitectureRelation occurrences, bounded ArchitectureClaim content, selected structures, and the next admissible architecture candidate use.
C.30:End
Architecture Description Adequacy
Type: Architectural pattern Status: Stable Normativity: Normative unless marked informative
Plain-name. Architecture-description adequacy.
Intent. Make architecture descriptions useful for the intended work while keeping their claims and representations distinct from the architecture they describe.
Builds on. C.30, C.30.ASV, A.1, A.22, E.24.PUB, A.7, A.6.3, E.17.0, E.17.1, E.17.2, E.17, C.2.P, E.10, E.10.ARCH, and E.10.D2.
Coordinates with. C.30.AD.BA, C.30.P, C.30.TFS-REL, C.30.LCA, C.30.ILC, C.32.P2S, C.32, C.32.MLAO, C.32.PAD, C.32.ADR, C.32.ADA, A.6.3.NAR, A.19.CPM, A.19.SelectorMechanism, C.18, C.19, G.5, A.6.F, A.6.M, C.29, C.16, C.16.P, A.10, B.3, A.20, A.21, A.15, A.15.5, C.11, C.28, E.8, E.10.MOVE, E.11.PUR, E.24.CD, and F.18.
Use this when
Use this pattern when work must create, inspect, compare, reuse, or rely on an architecture description, a set of such descriptions, a generated view of architecture relations, or a description used as a specification. First name what the description is about: one holon, one ArchitectureRelation occurrence that obtains, or one selected U.Structure.
Use it to answer:
- what architecture-side thing each description is about: a holon, an obtaining architecture-relation occurrence, or a selected structure;
- which architecture claim the description carries or lets the practitioner inspect;
- which selected structures and structure kinds the description covers;
- whether a description qualifies as
U.View: name the viewpoint and show that the E.17.0 conformance relation holds; - how views correspond, which sources enter the use, when a stronger use must return to a source, how fresh the description is, and whether specification use is allowed;
- what the description may guide, what it may not be used for, and what architecture move comes next.
What goes wrong if missed. A diagram, documentation set, generated relation graph, model card, ADR publication set, file, or architecture model is treated as architecture, selected structure, U.View, proof, gate, assurance, decision, work authorization, or release authorization merely because it presents those claims.
What this buys. A reader can tell what each description is about, how its views correspond, where reused material came from, how fresh it is, what it may be used for, and which other claims need their own patterns.
First useful description-use output. For a new description, begin with what its audience must distinguish or explain. Choose the subject, structures and detail needed to express that difference, then construct the account and try it on the intended question. For an existing description, say what it describes, which reference scheme gives its terms meaning, why it is being used, which structure matters, what use is allowed, and whether this answers the question or leaves an architecture move to make. If you call it a U.View, also name the viewpoint and the conformance relation that holds. Stop if this answers the question. Ordinary wording can also be retained, compared, or handed on. Use ArchitectureDescriptionUseCard@Project when its organized fields help the receiving work:
@Project is only a retrieval cue. It creates no project, authority, context, viewpoint, parthood, or Work. When an actual project matters, projectWorkOccurrenceRef names the composite U.Work recovered under A.15.6. Include architectureDescriptionProjectUseRelationRef only when a named pattern defines how this description use concerns that Work and the relation holds. A Work reference alone is not project locality. If locality matters but the relation is not defined, return missing-governor; otherwise omit both project-local fields.
The card is optional and does not identify the description. For its declared use, it retains the described thing, reference scheme, purpose, selected structures and their kinds, allowed use and any material limitation, and the next architecture move or pattern needed for a separate claim. If it calls the description a U.View, it also retains the viewpoint and the conformance relation that holds. Use the fuller ArchitectureDescriptionUseAccount when its fields help keep the correspondence, source use or return, freshness, specification use, publication, representation, or project locality inspectable. An adequate existing account need not be reformatted. Keep any authority claim in its own pattern and relation.
For publication use, each carrier reference resolves to an independently identified physical or digital entity whose state or realization yields the expression. The carrier's domain supplies its identity rule. When the bearing claim needs inspection, recover its PublicationExpressionBearingRelation and exact expression under E.24.PUB. A reusable form has its separate conformance relation. A file object, immutable snapshot value and individual copy retain their different identity rules.
Not this pattern when.
- If the use is an architecture claim, an obtaining
ArchitectureRelation, or one first architecture question, useC.30. - If the use is a selected structure or structural description outside architecture, use
A.22. - If the use is one architecture structural view and its viewpoint-conformance test, use
C.30.ASV. - If the use is built-asset architecture-description, BIM, IFC, asset-information, digital-twin, or reference-designation specialization, use
C.30.AD.BA. - If architecture or structure wording is ambiguous, use
C.30.P. - If the use is only a representation, publication occurrence, publication face or form, report, dashboard, file, carrier, source-expression relation, or publication-currentness relation, use
C.2.P,E.17,E.24.PUB, or the pattern that defines or tests that representation, publication, or source-use claim. - If the description is being used as a pattern-use recommendation, work-entry readiness, evidence, assurance, gate passage, decision, work authorization, causal-use claim, release authorization, deontic permission, or mathematical-lens use, keep
C.30.ADonly for the description boundary and use the pattern that defines or tests the other claim.
Problem frame
Architecture practice needs descriptions that remain useful over time: multi-view documents, view models, generated relation graphs, transformation-flow views, control sketches, module or interface diagrams, deployment views, model cards, system cards, and architecture-decision description sets. Teams use them to compare, reuse, refresh, and inspect architecture claims. If a project also claims a system-role assignment, Work attribution, authority, or responsibility, keep that as a separate claim: use A.2.1 for assignment, A.15.1 for Work admission, and F.6 only for assignment-bound Work attribution; use an admitted domain relation for authority or responsibility, or return an A.6.RCD missing governor if a rule needed to state or test that claim is absent. VP.AllocationResponsibility is only a clue to the concern.
A description is not the architecture, an architecture relation that holds, or the selected structure. The same holon or relation occurrence can have several descriptions, and a description set can contain several separately identified epistemes. A description counts as U.View only while the E.17.0 conformance relation holds between that same episteme and one viewpoint episteme. Different views can hide, lose, coarsen, or emphasize different structures: for example functional, flow, control, module, interface, placement, information-custody, evidence-reuse, assurance, or scale structure.
The first-minute practitioner can ask:
- What holon, obtaining
ArchitectureRelationoccurrence, or selected structure is this description about? - Which ClaimGraph, EntityOfConcern, and reference scheme identify the description?
- Which structures and structure kinds does it describe?
- If it is called a
U.View, which viewpoint and which conformance relation make that true? - What claim or relation connects it to architecture claims and other views? Proximity alone does not establish correspondence.
- Which sources, representations, or publications enter this use, by what path, and when must stronger use return to a source?
- After using the description, what architecture move remains admissible?
Problem
A practitioner needs to know whether an architecture description supports the intended work: what it describes, which claims and selected structures it makes recoverable, how it corresponds to other descriptions, and where its use must stop or return to a source. A polished view set can leave these questions unanswered; requiring a full documentation apparatus for an adequate sentence creates a different failure. The task is to supply the missing description content and its grounds at the scale the use needs.
Forces
Solution
An ArchitectureDescription is the local name for a C.2.1 U.Episteme that describes one architecture-side EntityOfConcern: a holon, an obtaining ArchitectureRelation occurrence, or a selected U.Structure. Use the name only when its ClaimGraph makes that subject, the described structures, purpose, and use boundary recoverable. It remains an episteme identified by <ClaimGraph, EntityOfConcern, ReferenceScheme>; it is not a record or a new root kind. A cited ArchitectureClaim is content or trace, not automatically the thing described.
Keep ClaimScope, empirical grounding, concern, viewpoint, view membership, selected model-use structure, representation, publication occurrence, publication form, carrier, project Work, and project-use relation outside that identity triple. Add each only when it independently applies. modelUseStructureRef is optional and appears only when a selected DDD model-use structure changes interpretation or selection.
C.30.AD does not mint U.Architecture, redefine U.Viewpoint, or replace generic Description, view, representation, publication, or publication-form machinery. It defines their architecture-description use while keeping every selected architecture-relevant structure recoverable.
For built-asset architecture descriptions, BIM, IFC, asset information, digital twins, and ISO/IEC 81346 reference designation, use C.30.AD.BA. C.30.AD keeps the general architecture-description bridge and does not absorb that specialization.
Architecture-description use account
The account points to a constituted episteme; it is not the episteme and does not add slots to it. Its claimGraphRef resolves under C.2.1:4.2.6 to that episteme's complete semantic claim content; the designation and any graph representation remain distinct from the value. Together with entityOfConcernRef and effectiveReferenceScheme, it exposes the constituents and EntityOfConcern that identify the episteme. When the described thing is a relation occurrence or selected structure, the participant trace can recover its holon. architectureClaimRefs carries claim content or trace; selectedStructureRefs names the structures described, and structureKindRefs classifies them.
In the description-use card and account, effectiveReferenceScheme carries the rules of the selected architecture-description episteme. In the later C.2.1 constitution blocks, it carries the rules of the named view-use, correspondence or freshness claim itself. Each value satisfies C.2.1:4.1.2 EpistemeInterpretationScheme and must cover the claims and references being interpreted. Missing interpretation leaves the dependent description-use judgement unresolved; membership alone establishes neither view conformance, correspondence, freshness nor adequate support for the intended use.
Minimum conformance for a retained ArchitectureDescriptionUseAccount:
- the account resolves to one architecture-description episteme and exposes its exact ClaimGraph, one EntityOfConcern, and effective
ReferenceScheme; - actual architecture-relation references identify independently obtaining
ArchitectureRelationoccurrences; required, desired, expected, candidate, unresolved, or negative architecture content stays claim content; selectedStructureRefsnames the architecture-relevant structures being described, andstructureKindRefsclassifies those selected structures;- any cited
ArchitectureStructuralViewis the same description episteme admitted asU.Viewonly by a separately obtaining E.17.0 conformance relation to one viewpoint episteme; - cross-view composition uses explicit description-set use claims, correspondence claims, or independently obtaining relations; source use names source-to-use paths; a source-return condition appears only when stronger use requires return to a named source or exact defining or constraining ClaimGraph;
- representation and publication fields identify their own objects and occurrences; they do not establish the description, architecture, selected structure, view membership, empirical grounding, or truth;
admissibleUsestates what the description supports. IncludenonAdmissibleUse?when a supported limit changes the receiving use.
Traceable architecture multi-view description chain
When a full architecture-description use relies on a view, the reader must be able to recover the chain that makes that view useful. The chain is a trace requirement, not a prescribed method or work plan:
When allocation or responsibility is current, add the direct relation separately. A system-role kind or assignment can support the work context but does not establish responsibility; VP.AllocationResponsibility only helps recognize the concern. When a source episteme or source view is used, a source-to-use path joins it to the view or description. Representation adds its own representation relation or object. Publication adds a publication occurrence with its form and carrier kept distinct. Cross-view use adds a correspondence claim or a direct correspondence relation only when its predicate is satisfied. A source-return condition is added only when a stronger use must return from a derivative or reused expression to a named source or exact defining or constraining ClaimGraph.
E.17.0 tests whether the description is a U.View; C.30.ASV tests whether it carries the selected structure and structure kind needed for the use. C.30.AD explains how to compose and use the description: what it describes, which views and correspondence it uses, where source material enters, when stronger use must return to a source, and what architecture move or separate claim remains.
If a needed link is absent, do not substitute a label, query result, bundle, diagram, file, or publication. Add the missing reference or relation that holds, narrow the allowed use, or use the pattern that defines how to recover it.
View membership, viewpoint, and structure-kind binding
An architecture description episteme is not a U.View because it is put in a multi-view set, authored under a viewpoint label, constructed by A.6.3, returned by a query, selected, bundled, diagrammed, rendered, or published. First identify the candidate episteme by its C.2.1 identity. Then identify one viewpoint episteme and test the fixed five-part E.17.0 predicate. Only a separately obtaining EpistemeViewpointConformanceRelation(candidateEpisteme, exactViewpoint) admits that same episteme as U.View.
When a receiving use needs one multi-view description set, identify the collection by its independently identified description members and membership rule. Use C.13 when a construction trace is needed; set membership remains ordinary collection membership. A shared file, bundle, heading, graph, publication, or query result neither identifies that collection nor grants U.View membership. The collection keeps no second episteme identity for its members.
C.30.AD can record use of recoverable architecture structural views inside one description set without minting a local relation kind:
ArchitectureDescriptionViewUseClaim is a C.2.1 episteme about one description set. The block separates what the claim says from the objects that identify it; it does not add slots to the episteme. The claim cannot create a view, set, viewpoint, or structure or confer U.View membership. Each referenced view must already satisfy E.17.0. Use C.30.ASV to check viewpoint conformance and selected structure, A.22 for structure itself, and C.30 for an obtaining architecture relation or architecture claim. Use C.30.AD only for description identity and use, cross-view correspondence, source use or return, freshness, specification or publication use, and the remaining architecture move.
Common architecture-description views:
Cross-view correspondence, source use, and return conditions
Before combining two views, establish whether they describe the same holon, the same architecture-relation occurrence, the same selected structure, related structures, or different subjects. State that correspondence as a supported claim or cite a direct relation that holds. When source material enters the use, identify the source and its contribution. Use sourceToUsePathRefs when that provenance needs a citable G.6 path: PathCitationRecordRef here means a reference to one G.6 PathCitationRecord, resolving its EvidenceGraph, path and slice. That descriptive path cites independently established direct relations; it is not itself one of them. Add a return condition when a stronger use needs the named source or its defining claims.
Coarse-graining check. A coarser description groups, omits or summarizes distinctions. Identify what was grouped and how it corresponds to the finer description. For a positive conclusion shared by every admissible arrangement, establish that at least one is possible and that the conclusion holds throughout the admitted possibilities. When the question is whether any is feasible, a proof that the conditions are incompatible can settle it directly. If a binding or observation narrows the arrangements further, carry that source with the conclusion. If two give different answers, name the distinction. To establish the actual arrangement, find a binding or observation that resolves it. To show which arrangements are permitted, the supported alternatives may already answer the question. In a design commission, make or assign the open design choice on its stated grounds. Use A.22 or the governing subject pattern to establish any claimed levels or parts of the holon itself.
ArchitectureDescriptionCorrespondenceClaim is a C.2.1 episteme about one description set. The block separates claim content from its C.2.1 identity; it does not add slots or create a world-side relation. Cite a direct correspondence relation only when its predicate is defined, the facts satisfy it, and the relation holds. A warranted correspondence claim lets a reader combine views to the extent it supports: for example, a sameDescribedHolon claim establishes their common subject. It does not thereby establish the truth of their other claims, assurance, gate passage, or the identity of an architecture-relation occurrence.
Freshness and currentness boundary
Use a freshness claim only when the architecture description's admissible use depends on source edition, structure edition, model version, deployment state, or an external condition. Keep this bounded claim distinct from any publication-currentness relation:
ArchitectureDescriptionFreshnessClaim is a C.2.1 episteme about one architecture description. The block separates claim content from its C.2.1 identity. Add a source-return condition only when stronger use must go back to a named source or defining or constraining ClaimGraph. Freshness bounds current use; it does not make the description true, grounded, evidence-sufficient, or publication-current.
Specification-use and publication boundary
An architecture description can be used as a specification only when that use is declared. Specification use is not a new architecture kind; it is a bounded use of a description episteme or of one of its publications.
This account describes how an existing description or publication is used as a specification. Its claim-bearing content can constitute an episteme under C.2.1; the record form alone establishes neither that identity nor the described use. claimPatternRefs cites patterns that define or constrain separate claims. When project locality matters, name the composite U.Work and the obtaining project-use relation under its defining pattern. If that relation is needed but undefined, return missing-governor; otherwise omit the project fields. Claimed Methods, MethodDescriptions and performed Work retain their own admission grounds.
If specification use is also claimed to be a pattern-use recommendation, work-entry readiness, evidence, assurance, gate passage, performed work, work authorization, decision, causal use, or release authorization, use the pattern that defines or tests that other claim. C.30.AD governs only the description-use boundary.
Keep the description episteme, its possible U.View membership, diagram or other representation, publication occurrence, publication form, and carrier distinct. Authoring, construction, querying, selection, bundling, rendering, filing, or publication creates none of the subject-side architecture relation, selected structure, description truth, empirical grounding, project Work, or project-use relation by itself.
Other claims and applicable patterns
Candidate, front, and selected-set description boundary
An architecture description may also carry a project architecture decision or selected structures cited by an ADR-like publication. Use C.32.PAD for the decision relation, C.32.ADR for its publication projection, and C.32.ADA for decision adequacy. C.30.AD retains only description identity, E.17.0 view conformance, description-set use, correspondence claims or relations that hold, source paths and return conditions, freshness, representation, publication use, and specification use.
An architecture description may contain claims about an archive, front, selected set, candidate palette, local choice, or planned architecture move. That content does not turn the description into any of those things or establish recommendation, readiness, authorization, or permission. Use C.32.MLAO and C.32 for candidates, C.18 and C.19 for archives, fronts, and pools, G.5 for a selected-set result, C.11 for local choice, C.30 for the architecture move, C.30.ASV for the structural view, E.11.PUR for recommended pattern use, and A.15.5 or the A.15 family for readiness and Work. If the content is published, use E.17 for the source-backed face and source return, and E.24.PUB for the publication occurrence, form, carrier, audience, bounded use, and availability. C.30.AD records only the architecture description and its publication use.
For an architecture-description claim, record its C.2.1 identity and only the view conformance, set use, viewpoint, correspondence, source path or return condition, freshness, representation, publication use, and specification use that apply. For the first architecture move, use C.30; for synthesizing alternatives, use C.32 or C.32.MLAO. When using source content to change which variants are archived, pooled, compared, selected, published, locally chosen, or decided, use the pattern that defines or constrains that relation.
Archetypal Grounding (Worked Cases)
Worked use: one subject, two descriptions, one changed placement. Consider a constructed payment-system case. PaymentFunctions1 describes the authorization and capture functions of PaymentSystem1; PaymentDeployment1 describes its application instances and their host placements at time t1. Each description has its own ClaimGraph and effective reference scheme, with PaymentSystem1 as its EntityOfConcern. The common subject is supported by those identifications. Suppose authorization can run on either of two application instances. If the current placement inventory locates both on HostA and these are the only eligible instances, the host is known even with no record of which instance served a call. If one is on each host, both answers fit these descriptions. A binding in force at the time can settle which hosts are permitted. If both hosts are permitted for that call, use evidence that locates it when it ran. Each claimed U.View also needs conformance to its selected viewpoint under E.17.0.
For example, suppose the invocation record identifies instance Auth1 as the handler for call Q at t1 and the deployment record places Auth1 on HostA at that time. The records together locate Q on HostA. A short account citing both suffices. To move Auth1 to HostB, the designer can specify the new placement now. Its available capacity, achieved latency and actual deployment are separate decisions with their own grounds.
After an actual placement change at t2, the t1 deployment description can remain a true historical description but no longer answer the current-placement question. Return to the changed configuration and relevant facts to produce the revised deployment description and correspondence. Retain the functional description if its functions, claims and reference scheme are unaffected. A new rendering of the old deployment description would not supply this repair. The changed condition therefore selects the description and connection to revisit; it does not require rebuilding the whole view set or withdrawing the historical answer.
Bias-Annotation
Conformance checklist
Common Anti-Patterns and How to Avoid Them
Consequences
Positive consequences:
- Architecture descriptions become reusable while remaining distinct from the architecture, obtaining relation, or selected structure they describe.
- Multi-view work can keep each episteme identity, viewpoint conformance, selected structures, cross-view correspondence, source-to-use paths, source-return conditions, freshness, representation, publication, and specification use inspectable.
- Separate non-description claims remain identifiable and receive the tests needed for their own use.
- Architecture decisions can use descriptions whose purpose, correspondence and limits are recoverable.
Costs:
- A useful architecture document must make its description epistemes, EntitiesOfConcern, effective schemes, selected structures, and admissible use recoverable. Ordinary wording or existing links can do this.
- A claimed view additionally needs the viewpoint episteme and independently obtaining E.17.0 conformance relation.
- Reused or regulated descriptions may need correspondence refs, source-to-use paths, source and structure editions, source-return conditions, and freshness claims before they can be relied on.
- Familiar diagrams, files, and publication forms lose implicit authority; establish grounding, evidence, assurance, gate, decision, and release claims through their patterns.
Rationale
Architecture work needs descriptions, but a good description is not necessarily a good architecture. A description can guide work only when the reader can identify it, tell what it describes, recover the selected structures and any view conformance, see how it corresponds to other descriptions and sources, and know what use is allowed.
The pattern therefore specializes generic Description and publication machinery for architecture use. It does not mint a new architecture kind, direct subject relation, local view-membership relation, or second meaning of U.View; it does not replace C.30; and it does not let diagrams or documentation formats establish non-description claims by presentation alone.
SoTA-Echoing
Relations
- Use
C.2.1to identify every architecture-description episteme. - Use
C.30for obtaining architecture relations, selected structures, and bounded architecture claims. - Use
C.30.Pwhen architecture or structure wording remains overloaded; useC.30.ADwhen the architecture-description use is clear. - Use
C.30.ASVto test architecture structural-view adequacy; only E.17.0 conformance admits the same episteme asU.View. - Use
C.33to account for captured and lost structure when a description, generated relation graph, ADR-like record, or view set carries only part of the needed architecture content. - Use
C.34to test preservation or correspondence when comparing a description with another view, source model, generated output, candidate, or realized structure. - Use
C.29for a mathematical-lens or coarse-graining result,C.30.STRATfor level-word admission, andA.22orC.30for the separately supported subject structure or architecture relation. Description-side grouping establishes none of those subject claims by itself. - Use
A.6.3.NARfor a reader-facing narrative made from a description, view set, or decision route. C.30.AD tests description adequacy; A.6.3.NAR handles structure-to-sequence, source carry-through, lost structure, reader use, and return conditions. - Use
C.30.TFS-REL,C.30.LCA, andC.30.ILCfor their named architecture-relation subcases. - Use
C.32.P2Sfor a connected architecturing flow when the description carries only part of the selected structure, decision handoff, method expectation, source continuity, return condition, or feedback from actual structure. - Use
A.7,E.17.0,E.17.1,E.17.2,E.17, andE.24.PUBfor generic EntityOfConcern, view, viewpoint, representation, publication occurrence, form, carrier, and MVPK machinery. C.2.Pnormalizes source-expression, source-to-use, publication-form, and publication-currentness relation-set overreads.- Use
E.11.PURfor recommended FPF pattern use after reading a description; C.30.AD records only the description-use boundary. - Use
A.15.5for work-entry readiness and full-kit condition; use the A.15 family for Work and the pattern defining any needed project-use relation. C.30.AD records only descriptions and their view conformance, set use, correspondence, source paths and returns, freshness, representation, publication use, and specification use. E.10.MOVErestores move-like wording when source prose about an architecture description does not mean a C.30 architecture move or a C.30.AD remaining architecture candidate use.
C.30.AD:End
Built-Asset Architecture Description and Reference Designation
Type: Architecture-description subpattern under
C.30.ADStatus: Stable Normativity: Normative for built-asset architecture-description, reference-designation, model-exchange, and digital-twin use
Builds on. C.30, C.30.AD, C.30.ASV, A.1, A.22, C.2.1, E.17.0, E.17, E.24.PUB, and A.7.
Coordinates with. A.6.P, A.6.RCD, A.6.REL, A.6.F, A.6.M, C.30.TFS-REL, C.30.LCA, C.29, C.16, C.27, C.27.TA, C.28, A.3.4, A.10, G.11, A.15, A.15.1, A.15.5, A.21, A.2.8.PER, A.2.9, B.3, and F.18.
Use this when. Use this pattern when a BIM or IFC publication, asset-information description, reference designation, cost, schedule, operation, maintenance, sustainability, or energy view, or digital-twin description is being used to say something about the architecture of one built asset.
What goes wrong if missed. A rich model is treated as the asset or its architecture; a file or multi-view bundle is allowed to select structure or grant U.View membership; or a designation and a live data feed are asked to carry identity, occurrence, truth, and currentness claims that their direct relations do not establish.
What this buys. An engineer can retrieve the intended asset or component, combine the relevant descriptions, and determine which architecture use their contents and freshness support. Actual asset relations, candidate claims and the descriptions of them remain distinguishable.
Not this pattern when. Use C.30 when the object is the direct architecture relation or bounded architecture claim, C.30.AD when no built-asset specialization is needed, and C.30.ASV when one structural view is under repair. Use the direct designation/reference, evidence, currentness, Work, decision, transformation, or causal-use pattern when that relation rather than built-asset architecture-description use is current.
When auxiliary-view, telemetry, simulation, maintenance, or digital-twin material is used to claim that an intervention caused an effect, handle that causal use under C.28; C.27 remains the owner of temporal-claim adequacy.
When a twin, dashboard, exchange result, or release screen looks like gate passage, recover whether a named gate decision is current. Use A.21 when that gate must decide a bounded action or when its GateDecisionResult is being relied on; the display remains a cue until the required result is established. Return any separate evidence, work-entry readiness, assurance, Work, or other claim to its own governor.
Do not collapse release into gate passage: route a release action or other performed Work to the A.15.1 U.Work occurrence, work-entry readiness to A.15.5, a permission result or exercise to A.2.8.PER, an instituting or revoking grant act to A.2.9, and a claim that a subject was released to its named subject predicate and participants; if that predicate cannot be recovered, return A.6.RCD missing-governor. An authorization-looking label does not choose among these claims.
Problem Frame
Built-asset work joins descriptions made for design, fabrication, construction, commissioning, operation, maintenance, adaptation, and decommissioning. A hospital, bridge, plant, railway corridor, or campus can therefore have a geometry model, spatial decomposition, functional and flow descriptions, product and equipment structures, cost and schedule descriptions, operation and maintenance records, sustainability and energy views, asset registers, live telemetry, inspection histories, and several reference-designation schemes.
These descriptions select different structures of the same built asset, or sometimes describe different entities related to that asset. They are not interchangeable descriptions of one undifferentiated object. The architect needs to recover which EntityOfConcern each description has, which A.22 structure each claimed view describes, how the views correspond, which designation or reference relation makes an entity retrievable, and what currentness boundary permits the description to guide the next architecture move.
Problem
How can an engineer assemble a usable built-asset architecture description from model exchanges, designation systems, and operational information without making the exchange schema, designation code, dashboard, or digital-twin system stand in for the built asset, an obtaining architecture relation, a selected structure, or a conforming view?
Forces
Solution
Start with one intended architecture use, not with the available tool outputs. Name the built asset and selected A.22 structures that matter. Recover subject-relation occurrences when making actual-relation claims; a proposed arrangement can remain candidate content. Cite an obtaining ArchitectureRelation only when its C.30 predicate holds; otherwise keep required, desired, expected, candidate, negative, or unresolved content in a bounded ArchitectureClaim.
Constitute each architecture description under C.2.1 about exactly one EntityOfConcern: the built asset, one obtaining ArchitectureRelation occurrence, or one selected structure. Keep its ClaimGraph and effective reference scheme recoverable. The same episteme is a U.View only while a separately identified E.17.0 conformance relation to one viewpoint episteme obtains. Then recover reference designation, model exchange, source use, representation, publication, and currentness through their own objects and relations.
For a first use, make the relevant descriptions, structures, reference designations, sources and allowed use recoverable. Ordinary wording is sufficient when it answers the question; stop there. The following record can organize the references when that helps the receiving work:
Resolve descriptionContentDesignation under C.2.1:4.2.6 to the whole content of the same episteme selected by architectureDescriptionRef. Its by-value effective scheme must be the rules governing that content, not the rules of a document that merely describes the scheme. A missing or conflicting resolution leaves this use record unresolved; it does not reidentify the asset or establish view conformance.
This is a project-side use record, not a description identity constructor or a new root kind. @Project is a compatibility and retrieval cue and establishes no project entity, Work occurrence, authority, context, viewpoint, parthood, or use relation. When the use is local to one project, projectWorkOccurrenceRef identifies the composite U.Work, and builtAssetDescriptionProjectUseRelationRef identifies the separately governed obtaining relation by which this description use concerns that Work. Otherwise both remain absent.
architectureDescriptionRef resolves to one C.2.1 episteme whose identity is the cited ClaimGraph, one descriptionEntityOfConcernRef, and effective reference scheme. If that EntityOfConcern is the built asset, it is exactly builtAssetRef. If it is an architecture-relation occurrence or selected structure, its participants or selection trace recover builtAssetRef without deriving identity from an optional architecture claim. architectureClaimRefs carry bounded claim content or trace only.
Every value in architectureStructuralViewRefs identifies a description episteme admitted as U.View only through an independently obtaining EpistemeViewpointConformanceRelation to one viewpoint. It can be the same episteme as architectureDescriptionRef only when that description's one EntityOfConcern is the selected structure required by C.30.ASV; otherwise it is a separately identified description episteme connected through an explicit description-set use or correspondence claim or independently obtaining relation. A multi-view use can cite several such description/view epistemes; the use record, collection, file, bundle, list order, or publication creates neither their identities nor their conformance.
claimScope, architecture concern, empirical grounding, and modelUseStructureRef remain neighboring qualifiers or relations. A DDD-style bounded-model-use structure appears only when that independently selected structure changes interpretation or selection for this use. It replaces none of the asset, relation, structure, description, scheme, scope, grounding, viewpoint, Work, or project-use objects and is absent from base identity.
When a model publication, exchange, measurement description or other source enters the architecture use, identify its edition and contribution. Use sourceToUsePathRefs for citable G.6 paths as defined in C.30.AD:4.3, keeping their cited direct relations distinct. Add a source-return condition when stronger use needs the named source. Keep representation and publication occurrence, form and carrier separate. Use G.11 when description freshness, publication currentness, model decay or synchronization currentness changes the claim.
When ISO 19650 contributes to a use, cite the published part and edition and identify what it contributes. The cited ISO 19650-1:2018 edition supplies whole-life information-management concepts and principles; ISO 19650-3:2020 supplies operational-phase information management and exchanges. Keep the used information or model edition recoverable. Add a source-status date, validity window or refresh condition when the receiving use or governing regime depends on current status; a historical interpretation of a named edition needs no invented expiry. A draft or later edition does not silently replace the source used. These standards supply information-management discipline, not the asset's architecture facts or a separate authority claim.
Recover the selected structures before combining views
For every included architecture structural view, state the candidate description episteme, its selected-structure EntityOfConcern, the architecture concern for which it is used, the viewpoint episteme, and the obtaining E.17.0 conformance occurrence. A geometry or coordination model can expose several structures, but the file boundary does not select one structure or grant view membership.
The last row is a recognition branch, not a new auxiliary-view kind. An asserted architecture structural view must have the selected-structure EntityOfConcern, viewpoint and independently obtaining E.17.0 conformance required by C.30.ASV. Another U.View keeps its own EntityOfConcern and E.17.0 viewpoint conformance without thereby becoming an architecture structural view. Return each embedded claim to the applicable pattern named in the row.
A single IFC publication may carry source descriptions for several rows. Conversely, one selected structure may be the EntityOfConcern of several descriptions and may be represented or published several times. C.30.AD.BA therefore keys each use to description identity, selected structure, view conformance, built-asset trace, and declared use rather than to a file, platform, package, or view count.
Recover a reference designation as a relation
A reference designation makes information about an entity retrievable under stated structuring and designation rules. When comparing two identifiable systems of designations, E.5.2.L:4.1 supplies the tests for formation, reading and resolution. For one asset, give the rules by value when that already identifies its aspect, issuer and effective interval. A well-formed but unassigned code in a complete register, two possible assignments and an unavailable register give different answers about which asset the reader can locate. First recover the designation or reference relation through its direct representation/reference/naming owner. C.30.AD.BA records only its built-asset architecture-description use; a code, field, repeated string, list row, or the record below neither admits a relation kind nor makes an occurrence obtain. If no current subject pattern supplies the needed relation, keep the designation use as bounded C.2.1 claim content; apply A.6.RCD only when a named repeated receiving use needs a reusable predicate definition or admitted direct relation.
selectedAspectStructureRef names the structure in which the designation is interpreted, such as a functional, product, location, or declared local structure. It is not a free aspect label. designatedEntityRef names the entity designated in that structure. Include designationOrReferenceRelationRef only when the designation or reference relation kind is admitted and the occurrence independently obtains. If a design object and a realized component both need to be retrieved, name the two entities and a bounded correspondence claim or independently obtaining correspondence relation rather than letting one code silently collapse them.
The designation use permits retrieval and cross-description coordination. Part-whole, function, location, identity across aspects, and evidence claims still come from their direct relations or claim owners. Repeated appearance of the same designation expression is insufficient to merge referents when the scheme, selected structure, local sense, or qualification window differs.
Keep exchange checking distinct from architecture evaluation
An IFC exchange or another machine-readable model is a representation and publication of one or more epistemes. Its schema relations can preserve source structure. Before using it as architecture-description content:
- identify every source episteme used, its representation, and the publication occurrence, form, and carrier;
- recover the A.22 selected structures and the status of the relation content being used; identify subject-relation occurrences only when asserting that those relations obtain;
- record the source-to-use path into the architecture description or view episteme;
- state the admissible architecture use and any lost, inferred, unknown, stale, or unavailable relation content.
An exchange checker can establish which records satisfy a computer-interpretable information specification. State the population it actually checks. If each exported equipment occurrence carries a maintenance identifier, a receiver may accept that property of the delivered records. For coverage of the installation, use an established correspondence to all equipment in scope. A verified complete export of a current asset register may already provide it. Otherwise reconcile the export with a grounded inventory and check for omitted units or multiple records for one unit. A shared equipment type may suffice to select maintenance instructions. Establishing which unit was serviced requires a correspondence to the individual unit. Use the relevant architecture, characteristic, evidence or assurance method for conclusions about the asset.
Keep a digital-twin description coupled without merging its objects
The phrase digital twin can cover a model episteme, software system, sensor systems, telemetry epistemes, simulation methods, operational Work, interfaces, representations, and publications. Recover each object by its kind and relation. C.30.AD.BA uses only the descriptions and views that contribute to the built asset's architecture-description use.
When one architecture-description use crosses design-side and run-side material, identify the descriptions on each side and the correspondence needed for that use. Preserve an adequate existing explanation; the optional by-value designRunSeparationUse record can organize it:
This is a by-value classifier for one built-asset description use, not a new FPF kind, a generic tag, or a relation constructor. designSideDescriptionRefs cite epistemes used for intended, required, proposed, or design-state material; runSideDescriptionRefs cite as-built, observed, operating, inspection, or maintenance-state epistemes. The classification is local to the declared use, not intrinsic to an episteme. If one publication carries both, identify the description or ClaimGraph loci before classifying them.
Every referenced object and relation keeps its subject pattern. C.2.1 governs description identity; A.15.1 governs each U.Work; G.11 governs freshness, currentness, and decay; the source-use owner governs each source path; the direct correspondence or coupling owner governs an affirmative relation; and A.10 or B.3 governs reliance or assurance.
C.28 governs a causal use of telemetry, simulation, maintenance, or a claimed physical or energy change; C.27 continues to govern the temporal adequacy of the change statement. A.3.4 governs an actual physical transformation only when the changed referent, boundary, conditions, before/during/after facts, and continuity or reidentification basis are complete. A required or desired effect, live value, simulation result, control-view row, or local design/run classification is not that actual transformation.
The local classification states only which already identified references belong on each side of this one cross-lifecycle use. Its nested source and currentness refs must resolve to the same refs cited by the enclosing use record, not duplicate or replace them.
When a required side identity, source path, claimed Work occurrence, currentness boundary, or direct relation cannot be recovered, state the gap and narrow or block the dependent cross-lifecycle use. Use blockedMerge for that gap when keeping this record.
The twin's coupling to the asset establishes neither parthood nor identity between the digital and physical objects. Connection, synchronization, rendering, bundling, and publication also establish neither architecture relation, selected structure, description truth, empirical grounding, view membership, evidence sufficiency, assurance, gate passage, Work, nor project-use relation.
A green twin, dashboard, exchange result, or release screen is therefore a cue, not gate passage. Use A.21 when a named gate must decide a bounded action or when its decision result is being relied on. Establish or recover the GateDecisionResult episteme, its applicable profile application and check-application results. Publication and the optional DecisionLog follow A.21:4.10 when publication, audit, or reuse is needed; an ordinary one-time decision can stop at its result and rationale. If no named gate decision is current, keep the display as a cue and handle any evidence, work-entry readiness, assurance, Work, or neighboring claim under A.10, A.15.5, B.3, A.15.1, or its direct governor. Even a GateDecisionResult with decisionValue=pass establishes neither release, readiness, permission, authorization, nor performed Work.
Recover the release-looking claim before routing it. An actual release action is one A.15.1 U.Work occurrence. Use A.15.5 for work-entry readiness, A.2.8.PER for a non-prohibition, granted permission, permission exercise, non-violation, or permission conflict, and A.2.9 for an instituting or revoking grant act. A further claim that a subject was released needs its named subject predicate and participants. Keep the display as a cue while either is unresolved; return A.6.RCD missing-governor only when the predicate cannot be recovered, and state any missing participant information separately. No current model, source, gate result, dashboard, or the word authorized supplies one of these relations by appearance.
Currentness and smallest reopen. When a decisive input changes, reopen only the built-asset description-use locus and conclusion that depend on it:
- A changed asset or selected structure reopens the dependent description identity, built-asset trace, or structural-view use.
- Changed view conformance reopens that one view admission.
- A changed designation scheme, referent, or qualification window reopens only its
BuiltAssetReferenceDesignationUse. - A changed source, model, publication, or telemetry edition or freshness/fidelity boundary reopens its source-to-use or currentness locus.
- Changed design/run classification, cited Work, source/currentness ref, correspondence, coupling, or transformation reopens only the affected
BuiltAssetDesignRunSeparationUseand dependent admissible cross-lifecycle use. - A changed project-use relation or direct governor reopens only that relation reference and dependent admissible-use conclusion.
Update the affected description, designation, design/run, or currentness locus; when the required input cannot be recovered, narrow or block only that use while unrelated views, descriptions, designations, and uses stay closed.
Worked Cases
Hospital ventilation and fire compartmentation
A hospital renovation uses an IFC publication, a fire-compartment view, a ventilation flow view, an equipment register, an energy-use view, and live air-handling telemetry. The immediate architecture concern is whether the changed ventilation arrangement preserves smoke-control functions across compartment boundaries. A second intended use is a bounded comparison of air-handling energy consumption; it does not share the smoke-control verdict.
The engineer names the hospital facility as builtAssetRef, recovers the actual subject relations and fire-compartment, ventilation-flow, equipment-module, and control structures, and cites an obtaining ArchitectureRelation only if its C.30 predicate holds. Required or proposed content remains in a bounded architecture claim. Each used architecture description has its ClaimGraph, one EntityOfConcern, and effective reference scheme; each claimed structural view additionally has a viewpoint and independently obtaining E.17.0 conformance relation.
The IFC publication supplies a source-to-use path for the spatial and equipment descriptions while its representation, publication occurrence, form, and carrier stay separate. The telemetry episteme has a currentness boundary and can support separately governed operating-state claims; it is not itself the control structure, an architecture relation, an actual physical transformation, or architecture evaluation.
For the energy use, one current C.16 measurement-result episteme attributes 112 kWh ± 4 kWh electrical-energy consumption to air-handling unit AHU-3 over a declared 24-hour commissioning window. It names the measurand, Characteristic, Scale and unit, method and model, calibration basis, dated measurement Work, time stance, and uncertainty. Its source-to-use path cites the meter telemetry and source edition; its G.11 currentness condition limits use to the named sensor, calibration, model editions, and validity window and reopens that use when one changes.
If the engineer claims that the control revision will reduce the daily consumption rate, the action-guiding temporal claim enters C.27 with the intervention, window, resistance or cost, evidence or assumption relation, supported use, unsupported use, and reopen condition. For the causal claim that the control revision produces the reduction, C.28 additionally governs the causal-use class, support basis, supported use, unsupported use, and verdict. C.27 temporal adequacy does not establish a causal intervention effect, and C.28 does not replace the C.16 measurement result, C.27 temporal claim, or G.11 currentness boundary.
A.10 and B.3 still govern material reliance and assurance. With no comparable baseline given, the reading supplies the reported commissioning-window consumption, not a conclusion that consumption fell. Neither the energy view nor the recent reading proves sustainability, smoke-control adequacy, or architecture adequacy.
An air-handling unit has a product-aspect designation and a location-aspect designation. Each designation use names its scheme, selected structure, designated entity, qualification window, and direct relation when one obtains. A bounded correspondence claim or obtaining correspondence relation lets the maintenance team retrieve both descriptions without making the product structure identical to the location structure.
For the proposed arrangement, identify the compartment boundaries covered by the requirement and establish whether the proposed flow and control structures preserve that function in each required mode. For a decision about a mode, one may establish the same answer for every admissible arrangement, even when a different mechanism supplies the grounds in each. To identify the actual mechanism, distinguish the arrangements whose mechanisms differ. For a claim about the installed system’s capability, establish its actual configuration and grounds for its behavior in the required modes. For an account of a past event, establish which mode and paths were involved then. The separate energy comparison retains its own unit, period and measurement basis.
Bridge inspection twin
A bridge operator combines an as-maintained geometry model, structural-member view, inspection history, strain telemetry, and a simulation view. The bridge remains the built asset across model editions. The structural-member and sensor-placement structures are selected explicitly; each description keeps C.2.1 identity and each claimed view E.17.0 conformance. Inspection and telemetry claims retain their evidence, grounding, source-use, and currentness relations. A revised simulation model retains the same C.2.1 episteme identity when its claim content, EntityOfConcern, and effective reference scheme are unchanged. Otherwise identify another episteme; assert an EpistemeEditionRelation only when its historical-continuation predicate obtains. Assess any declared lineage or continuity claims separately for the intended reuse.
When the operator compares an original design description with the as-maintained geometry, inspection history, and live telemetry, designRunSeparationUse cites the design-side and run-side descriptions, the design, inspection or maintenance Work claimed, the sources and currentness conditions needed by the comparison, and any separately governed design-to-realization correspondence. actualTransformationRefs remains absent unless a repair or other physical change satisfies A.3.4. The architecture description can therefore support a decision to inspect or redesign a connection while retaining the route back to the geometry publication, measurement descriptions, and selected structures. It cannot treat a successful data-exchange check, recent sensor sample, simulation result, polished dashboard, or local design/run classification as proof that the bridge architecture is adequate or that design and realized objects are identical.
Conformance Checklist
Consequences
Assembling multi-view built-asset descriptions takes additional work because the engineer must recover the objects and relations represented in the tool's outputs. That cost is local and reviewable: each description gains C.2.1 identity; each asserted view gains a selected structure, viewpoint, and conformance relation; each designation use gains a referent, scheme, aspect structure, qualification window, and direct-relation disposition; and each live description gains an explicit currentness boundary.
The gain is stronger reuse. A changed model edition, replacement sensor system, revised designation scheme, new representation, or new publication can be incorporated without changing the built asset's identity, inventing an architecture occurrence, or silently reidentifying a description. Architecture evaluation can use the description while remaining distinct from exchange checking, information currentness, evidence sufficiency, assurance, Work, and project locality.
Rationale
Built-asset practice is exposed to semio-bias because one information environment often spans geometry, equipment, Work, sensor state, and maintenance history. The repair is a constructive chain from the physical asset and independently obtaining subject relations to selected A.22 structures, any actual ArchitectureRelation, bounded architecture claims, C.2.1 description epistemes, independently conforming views, and the source, representation, publication, designation, grounding, currentness, Work, and project-use relations that make those descriptions usable.
Reference designation demonstrates why this chain matters. Its engineering value is retrieval across heterogeneous descriptions. That value depends on the scheme, selected aspect structure, referent, qualification window, and direct-relation disposition being explicit. Treating the code as universal identity would remove the aspect discipline that makes the designation useful.
SoTA-Echoing
Relations
- Specializes:
C.30.ADfor built-asset architecture-description use. - Uses architecture and structure patterns:
C.30,C.30.ASV,A.22,A.6.F,A.6.M,C.30.TFS-REL, andC.30.LCA. - Uses description, view, representation, and publication patterns:
C.2.1,A.7,E.17.0,E.17,E.24.PUB, andC.29. - Uses relation, naming, and currentness patterns: the designation/reference predicate and assertion;
A.6.Pfor precision repair;A.6.RCDonly for a demonstrated missing reusable predicate;A.6.RELonly after the relation kind is admitted, current facts satisfy its obtaining predicate, and occurrence identity matters;F.18; andG.11. - Uses lifecycle information-management source discipline from: published
ISO 19650-1:2018andISO 19650-3:2020editions through source-to-use, edition, currentness, source-return, and admissible-use boundaries, without ontology or authority import. - Records design/run separation in: an ordinary account or local
BuiltAssetDesignRunSeparationUseover the descriptions and correspondence needed by the use. Include A.15.1 Work, source paths, G.11 currentness, coupling or A.3.4 transformation refs when those facts are claimed or consumed. - Use for auxiliary-view claims:
C.16for characteristic measurement,A.15.1for operation or maintenance Work,C.27.TAfor positive temporal aspects,C.27for action-guiding temporal-claim adequacy,C.28when an intervention, maintenance action, simulation, telemetry change, or claimed effect is used causally,A.10for evidence or material reliance,B.3for assurance, andG.11for currentness and reopen conditions. - Use for gate- and release-looking claims:
A.21when a named gate must decide a bounded action or itsGateDecisionResultis being relied on; the display remains a cue until the required result is established. Route a release action or other performed Work toA.15.1, work-entry readiness toA.15.5, a permission result or exercise toA.2.8.PER, an instituting or revoking grant act toA.2.9, and a subject-release claim to its named predicate and participants orA.6.RCD missing-governor; none of these conclusions follows merely from a display or an authorization-looking label. - Use for other claims:
A.3.4and the direct transformation, evaluation, evidence, assurance, Work, decision, acceptance, or project-use pattern named by the claim.
C.30.AD.BA:End
Clarify Architecture and Structure Wording (Precision Restoration)
Type: Architectural pattern Status: Stable Normativity: Normative unless marked informative
Plain-name. Architecture-structure wording repair.
Intent.
Recover architecture or structure wording whose selected structure, architecture relation, architecture-description use, structural-view use, source-return relation, or named C.30 subcase is hidden before a reader applies A.22, C.30, C.30.ASV, or a named C.30.* pattern.
This pattern repairs overloaded wording so the architecture, structure, description, view, publication, source, relation, characteristic, mathematical-lens, evidence, assurance, gate, work, decision, causal-use, release, or ordinary-prose use becomes recoverable by value. It does not mint U.Architecture, does not fuse architecture and structure into one kind, and does not replace grounded architecture adequacy or structural-view adequacy.
Builds on. E.10, E.10.ARCH, A.22, C.30, C.30.ASV, C.2.P, A.6.P, A.6.F, C.29, C.16.P, C.16, C.25, E.17, and E.8.
Coordinates with. C.30.TFS-REL, C.30.LCA, C.30.ILC, named C.30.* structure and view patterns, A.10, B.3, A.20, A.21, C.11, C.28, A.15, E.11, and work, release, and publication patterns defining or constraining those claims.
E.10.ARCH applicability rule. When E.10 encounters architecture or structure wording whose selected structure, architecture relation, architecture-description use, structural-view use, source-return relation, source label, or neighboring claim is hidden, E.10.ARCH selects C.30.P only until the use under repair and subject pattern are recovered. C.30.P then stops applying; it does not become a registry of architecture topics or a substitute for A.22, C.30, C.30.AD, or named C.30.* patterns.
Use this when
Use this pattern when architecture or structure wording hides which use is being made.
Typical triggers:
architecture,architecture description,architecture model,architecture diagram,architecture map,architecture dashboard,architecture score;structure,structural view,structural model,module layout,component structure,interface structure, or stratification wording or source-label wording such aslayer,level,tier,stack,ladder,rung,block,expert,cache,router, orgatewhose technical meaning must be recovered throughC.30.STRATbefore local architecture or structure assignment;graph,flow,transformation-flow graph expression,control sketch,LCA diagram,ADR,dashboard,benchmark,source, orviewbeing treated as architecture or structure by wording alone;- a function, module, interface, signature, flow, control, quality, score, evidence, assurance, gate, work, decision, causal-use, or release claim hidden by architecture or structure wording.
What goes wrong if missed. A diagram is treated as the architecture, a graph as proof, a view as the selected structure, a source document as an architecture decision, a score as architecture adequacy, or a function, module, or interface claim as architecture by default.
What this buys. The reader can state the intended claim, retain its useful content, and use the pattern that addresses the remaining question. Clear ordinary wording needs no repair record.
First useful move. Recover what the sentence says and what the reader needs to do with it. Keep adequate ordinary wording. Otherwise rewrite the ambiguous claim, then use its subject pattern for any remaining substantive question. A repair note is optional when the repair needs to remain inspectable.
Not this pattern when.
- If the use under repair is already a selected structure, use
A.22directly. - For an already recoverable
ArchitectureRelationor boundedArchitectureClaim, useC.30directly. For architecture-description identity, correspondence or use, useC.30.AD; useC.30for the initial description question tied to one architecture move. - If the use under repair is already an architecture structural view, use
C.30.ASVor a namedC.30.*view pattern directly. - If the claim being made is evidence, assurance, gate, work, decision, causal-use, release, mathematical-lens use, characteristic and scale construction, quality characterization, source-use, or relation construction, use the subject pattern for that claim after any architecture or structure wording is demoted or assigned.
Problem frame
Working engineers often say "architecture" or "structure" while pointing at a useful artifact: a diagram, model, graph, table, dashboard, ADR, code-agent relation graph, neural-network architecture-operation diagram, benchmark result, or source document. Ordinary shorthand is acceptable in FPF-governed prose when the intended use is recoverable. If the artifact is named by a source label such as block, layer, expert, cache, router, or gate whose technical meaning remains unclear, use C.30.STRAT before assigning the recovered use locally.
The repair question is:
Which selected structure, architecture relation, architecture-description use, structural-view use, source-return relation, or neighboring claim does the wording name, and which FPF pattern defines or constrains that claim?
The architecture or structure use under repair may be:
- selected structure under
A.22; - an
ArchitectureRelationor boundedArchitectureClaimunderC.30, its initial description question, or architecture-description use underC.30.AD; - an
ArchitectureStructuralView@Contextor namedC.30.*subcase; - a publication, view, face,
PublicationUnit, carrier, dashboard, ADR, source document, or source-return relation underC.2.PorE.17; - a relation construction under
A.6.P; - a function or functionality-kind use under
A.6.F; - a mathematical-lens use claim under
C.29; - a characteristic, scale, score, coordinate, threshold, or quality-coordinate claim under
C.16.PorC.16; - a Q-bundle or quality-characterization claim under
C.16.Q,C.25, orE.21; - an evidence, assurance, gate, work, decision, causal-use, release, or method claim under its subject pattern;
- ordinary prose with no FPF-governed use being made.
Problem
The same word can name a structure, a claim about it, or a description used to inspect it. A repair must recover the intended object and claim without losing what the reader can use. It should finish when the sentence and applicable subject rule are clear.
Forces
Solution
Repair architecture or structure wording with a local rewrite or an architecture-structure repair note.
An optional note can organize the following information; include only what the repair needs:
Use the note only when the repair must remain inspectable. A local rewrite is enough when one sentence clearly names the selected-structure claim being made, architecture relation, architecture-description use, structural-view use, source-return relation, or subject pattern.
Recovery sequence
- Read the span. Read the architecture or structure wording in its sentence and the nearby context needed to recover its meaning. Keep a copy only when the repair needs an inspectable record.
- Recover the object and claim. Identify the subject, predicate, participants and intended use. Distinguish, for example, a selected structure from its description, or an architecture claim from a score or decision. Use the assignment table below for the unresolved kind; keep ordinary prose when no formal distinction changes the claim.
- Recover source-publication relations before architecture assignment. If the wording relies on a source, publication, view, face,
PublicationUnit, dashboard, ADR, file, carrier, or source-return relation and a source-use, source-currentness, or publication distinction remains hidden, applyC.2.Pbefore assigning the architecture or structure claim. If that distinction and its subject pattern are already recoverable, use the subject pattern directly. - Choose the subject pattern for the architecture or structure use.
- selected structure ->
A.22; ArchitectureRelation, boundedArchitectureClaim, or its initial description question ->C.30;- architecture-description identity, correspondence or use ->
C.30.AD; - architecture structural view ->
C.30.ASV; - architecture transformation-flow relation ->
C.30.TFS-REL; - control-structure view ->
C.30.LCA; - cross-scope conflict or frustration triage ->
C.30.ILC; - stratification wording or source-label wording such as
layer,level,tier,stack,ladder,rung,block,expert,cache,router, orgate->C.30.STRATwhile the label's technical meaning remains unclear, before choosing the final subject pattern; - named C.30 subcase -> that subpattern.
- selected structure ->
- Assign non-architecture claims to their subject patterns. If the sentence uses architecture wording to carry relation, function or functionality, mathematical-lens, characteristic and scale, quality, evidence, assurance, gate, work, decision, causal-use, release, or method claim, apply the subject pattern for that claim and keep this pattern only for the architecture or structure wording repair.
- Preserve the useful conclusion. State what the repaired wording supports. Add a blocked interpretation only when a plausible reader could take it here and the distinction changes the conclusion or action.
- Stop C.30.P after assignment. Stop after the subject pattern or ordinary-prose demotion is named.
Subject pattern assignments
Architecture-synthesis routing note:
- Use
C.32,C.32.MLAO,C.32.CONWAY, orC.32.FAILwhen the recovered claim is a candidate palette, residual-reducing multilevel frame, transformer and transformed correspondence frame, or architecture-synthesis repair cue. - Use
A.19.CPM,A.19.SelectorMechanism,C.11, orG.5when the recovered claim is comparison-policy use, selector-policy use, local choice, or selected-set result declaration. When publication is current, useE.17for a source-backed face and source return andE.24.PUBfor the publication occurrence and audience availability. - Use
C.18orC.19when the recovered claim is archive, front, or pool policy. - For transformation-flow, function, module, transformer, mathematical-lens, relation-signature, affordance, architecture-use, or move-like wording, recover that claim kind first and use its subject pattern by value.
Refresh and reopen conditions
Reopen the wording when a changed source meaning, intended use, subject predicate or receiving rule changes what the sentence asserts or what the reader may do with it. Repair the affected claim and its direct use. Retain an adequate sentence when only its packaging changes; a new template or neighboring pattern does not by itself make its meaning unclear.
Worked cases
Worked repair. A review says, “The new module layer improved maintainability.” The available results report 80 seconds for revision R1 and 50 seconds for R2 on regression fixture F; the word layer merely labels the changed source grouping.
For the reported timing, one can write, “On fixture F, the reported regression-test runtime was 80 seconds for R1 and 50 seconds for R2.” For a claim about the contribution of the faster test, establish how it enters the maintenance work and whether it changes the required effort or elapsed time. For a claim about total maintenance time, a comparison of complete maintenance work under comparable conditions may already settle the question.
If the receiving task only needs that runtime comparison and its measurement basis is adequate, stop. If it needs a maintainability conclusion, establish how the measured runtime bears on that characteristic under C.25 and the domain method; if it needs a causal conclusion, recover the additional intervention and comparison basis under C.28. If the grouping itself must be changed, recover its constituents and relations before using A.22 or C.30. The stronger question selects the missing work; it does not invalidate the narrower measured result.
Reduced SoTA row
Architecture-description, model, view, and decision-record practice treats architecture as distinct from architecture descriptions, models, views, viewpoints, diagrams, and decision records. FPF adopts that line only where it changes action guidance: examples, non-use boundaries, subject-pattern assignments, source-return conditions, and conformance checks.
These distinctions block diagram-as-architecture, graph-as-proof, view-as-structure-kind, publication-as-claim, and ADR-as-decision overreads. This use does not import any external standard as FPF ontology.
Conformance checklist
Common anti-patterns
Related patterns
E.10catches architecture or structure wording and selects this pattern only when the selected structure, architecture relation, architecture-description use, structural-view use, source-return relation, or named C.30 subcase is hidden.E.10.ARCHdefines the shared wording-use recovery order and applicability row.A.22governs selected structure and structural views as structure.C.30governsArchitectureRelation, boundedArchitectureClaimand the initial architecture-description question.C.30.ADgoverns architecture-description identity, correspondence and use. The description retains its own EntityOfConcern; it need not itself be the subject of another description.C.30.ASVgoverns architecture structural views.C.30.STRATrecovers unclear stratification or source-label meaning before C.30.P assigns the recovered architecture or structure portion.- Named
C.30.*patterns define or constrain their own structure adequacy or view adequacy questions. C.2.Precovers source, publication, view, face,PublicationUnit, carrier, and source-use disposition.A.6.Prepairs relation construction;A.6.Frepairs function and functionality wording;A.6.Mrepairs module-relation and interface-specification wording.C.16.Prepairs unclear characteristic-and-scale wording, andC.16.Qrepairs unclear quality-term or evaluative characterization wording. Already recoverable claims use their subject rules directly.C.29governs mathematical-lens use and does not become architecture by analogy.
C.30.P:End
Clarify Stratification and Architecture Source Labels
Type: Architectural precision-restoration subpattern under
C.30Status: Stable Normativity: Normative unless marked informative
Plain-name. Stratification and architecture-operation source-label repair.
Intent. Help a reader decide what a source label such as layer, level, tier, stack, ladder, rung, block, expert, cache, router, or gate means in one sentence. Keep useful local language, but recover the object, relation, or claim before relying on it. This pattern introduces no universal kinds for these source labels or for stratification.
Builds on. E.10, E.10.ARCH, E.8, F.19, F.18, C.30.P, A.22, and C.30.
Coordinates with. C.30.ASV, C.30.LCA, C.30.TFS-REL, C.30.ILC, A.6.M, A.6.F, E.18, C.16.P, C.16, A.19.SPR, C.2.P, E.17, C.29, C.28, A.10, G.6, B.3, A.20, A.21, A.15, A.2, G.5, and C.11.
Result boundary. A practitioner receives the shortest sentence or note that names the recovered object, relation, or claim and its useful conclusion. Name a next action or stop or return condition when it changes the intended use. Include a blocked overread only when it passes F.19's plausible-reader test.
Use this when
Use this pattern when a source uses a compact architecture or stratification label and that word alone does not tell you what technical claim is being made.
Typical labels are layer, level, tier, stack, ladder, rung, structural uses of vertical or horizontal, and architecture-operation words such as block, expert, cache, router, and gate.
What goes wrong if missed. A useful local label is treated as ontology. A layer is assumed to be a holon level, control layer, publication layer, scale window, or module boundary without deciding which. A stack is treated as architecture; a block as a module; an expert as a system-role kind or performer; a cache as a state or memory relation; a router as a decision policy; a gate as a gate decision. Word shape establishes none of these.
What this buys. The reader can keep the source word while making its meaning and safe use explicit. Once the object, relation, or claim is clear, use the pattern that defines, constrains, or tests it.
First useful move. Read the sentence in context and ask: “What does this label name here, what may I infer from it, and what must I do next?” If it is ordinary wording, keep it and stop. If the answer is already clear, use the applicable pattern directly. Otherwise write one line: label -> recovered meaning and useful conclusion; any needed next action or return condition.
Not this pattern when. Do not detour through C.30.STRAT when the object, relation, or claim is already clear. Do not use it merely because a familiar word appears. Ordinary source prose with no FPF claim remains ordinary prose or a quotation.
Problem frame
Architecture and engineering sources use compact labels because they work in local practice. Neural-network prose says block, expert, cache, or router; control architecture says layer; organizations say level or tier; documentation says section, stack, or view; scale prose says level, resolution, or coarse-graining step.
These words are good recognition cues but poor stand-alone kinds. The same label can point to a selected structure, module relation, control relation, transformation-flow element, characteristic or scale, publication grouping, state, evidence claim, decision, or nothing beyond ordinary prose.
The repair question is: what does the label name in this sentence, what use does the recovered claim support, and which existing pattern supplies the needed definition, constraint, or test?
Problem
A compact label can carry a useful source-local meaning while leaving the intended technical claim unclear to its receiver. Recover the claim and its use without inferring a kind from word shape or replacing an already clear domain term.
Forces
Solution
Write the local repair first. For example: Here “gate” names the neural-network path selector; it selects which expert receives this token. That sentence can be the complete result. Use E.18 when the receiving question needs the selected transformation-flow organization, path or crossing.
A sufficient sentence can be compared, handed on, or revisited as it stands. Retain a compact note when organizing the repair helps that use:
groundedOverread? is an alias for the same optional blockedOverread? value, selected through F.19's plausible-reader test.
The note is neither the selected structure nor the relation, claim, publication, or pattern result it points to. Omit it when the sentence is enough.
Recovery sequence
- Read the sentence and label. Keep enough source context to tell what the sentence is doing; retain a copy when it is needed for the repair.
- Try the cheap exits. If the word carries no FPF claim, keep ordinary prose or quote it and stop. If the source already gives the technical term one clear local meaning, keep that term and use its rule directly. If one local rewrite makes the meaning clear, write it and stop. A controlled vocabulary or preferred-word list is not by itself a reason to replace useful domain language.
- Recover plausible meanings. Treat the label as a designation, not as the object, relation, or claim. If it belongs to a named model, viewpoint, standard, or local vocabulary, recover that source-local convention first. Then ask which object, relation, participants or bearer, claim, scope, time, and source use the sentence could be compressing. Include literal and metonymic readings when both are plausible.
- Choose by the recovered meaning. Use the row in C.30.STRAT:4.2 that matches the recovered claim; separate several claims when they need different rules. Do not choose from the label alone. A standard or model may settle the label inside its declared use, but neither its status nor its popularity extends that meaning to another subject or source.
- Open only the needed rule. Name the participants, relation, structure, characteristic, state, publication, evidence, work, decision, or other object that makes the claim true or false. Do not copy every possible field into the result.
- Return to ordinary wording. Write the shortest sentence that preserves the recovered claim and names the next pattern only when its contribution matters.
- State the result. Give the useful conclusion and any next action or return condition the receiving use needs. Stop if the question is answered. Include a blocked overread only when it passes F.19's plausible-reader test. If the needed meaning or use remains unresolved, state that limit through quote-only, reduced-use, blocked, or incomplete-rewrite disposition.
Recovered meanings and patterns to use
Functional layers and modular levels. A functional layer groups functions by what they contribute or are intended to achieve. A module interface states conditions under which another module can use or replace what it supplies. To speak of modular levels, name the order, such as enclosure by a composite module or an acyclic dependency among modules. Two peers may exchange data through a well-defined interface. Their exchange alone supplies no higher or lower module. Dependencies can leave some modules unordered. When a numbered level for each module matters, state how the levels are assigned. A functional layer can span several modules and a module can serve several functions. Use A.6.F for the functional contribution and A.6.M for the module and interface claims.
Composition, extent and model abstraction. A system-composition level locates an identified system along a stated chain of integrated components. Physical size requires a named quantity, such as length, and a basis for comparison. For a Method, Work, episteme or discipline, recover the composition relation for that kind. A system level can thus identify a scale of System composition. A stack of Methods can identify scales of Method composition when its Methods actually stand in those relations. Following several such chains in one inquiry requires the links that connect them. When a number is used for a level, recover the ordering and the reference whole. A composition with several branches need not give all participants one common ranking. Model abstraction compares what two models admit for the same inquiry under B.5.MR.
In a functional account, functional layer names the local kind of grouping defined by its organizing contribution. Identify a particular grouping through its organizing function, its membership criterion and the functions included in that design or operation. Different Methods can realise the same functional layer. When the source uses layer to designate the repeatable Method itself, identify that Method under A.3.1. Call the Method composite when its construction admits the specified constituent Methods and organizes their contributions under B.1.5.CM. If the layer denotes a functional grouping, the composition claim concerns its implementing Method. Membership of that Method in a further encompassing Method needs its own basis. A supervisor's command to a controller may establish a control relation across layers while both perform separate Methods. Use C.30.LCA or C.32.LCA for that relation. Library.Reference:6.2.2 collects the working terms for these different uses.
When a level claim matters. When an explanation, comparison or design relies on a sentence such as “X is at level L” or “A is above B,” name the subject at stake (the EntityOfConcern), what is being ordered, compared, grouped, or mapped, the relation or scale mapping that gives the claim its meaning, when it applies, and whether the sentence asserts, proposes, assumes, or illustrates the claim. Apply the same test when layer, tier, band, scale, or stage carries the stronger claim.
A named model or standard may provide this mapping within its declared use; its status does not extend the claim beyond that use. The source word may remain, but these facts—not the label—carry the claim.
A list, diagram row, first-then order, carrier section, curriculum, scale label, stage sequence, or coarse-grained description does not establish a subject level by form. If the facts are missing, keep the wording local or illustrative and block reliance on the stronger level claim.
When orientation carries the claim. Recover the structure's subject, the represented relation and the convention that gives vertical or horizontal its meaning through A.22:4.3a. Keep literal spatial directions under their reference frame. For a Method-architecture question, C.30.ASV:4.5a helps choose the structure; the axis does not choose it.
Same-sentence claim boundary
One sentence may use a source label while making several claims. Split them instead of adding a local catalogue of everything the label does not prove. C.30.STRAT repairs the label; the applicable pattern defines, constrains, or tests each separate claim. The table above lists common destinations, not a mandatory reading list.
For a dated Work claim, recover the performer basis under A.13 and independently admit the Work under A.15.1. Use F.6 afterward only for assignment-bound attribution. The source word expert, ordinary practitioner action or instrumental use of a pattern does not require inventing those identities.
Source-label cue table
Use the source context and receiving question to select plausible readings from this cue table. It is not a required sweep of every meaning; stop when the intended claim is recoverable.
Choose by recovered meaning
C.30 placement does not decide the recovered meaning. After recovery, use the rule that defines, constrains, or tests the object or claim.
Worked cases
Filled repair note
For The cache proves the architecture scales, do not hide the split inside one formal record. Read it as three candidate claims:
cachemay name a state-bearing module, interface arrangement, flow buffer, or ordinary source label; the sentence does not yet decide which;provesneeds grounds appropriate to the claim: a formal derivation under stated assumptions, support for an empirical claim, or the argument required by a named assurance use. Retain proof wording only within the established scope;scalesrequires a characteristic and bearer, comparison or scale construction, architecture scale-preference claim, or mathematical-lens use.
The note can remain compact:
The questions can be resolved separately. Recovering the cache referent does not establish the scaling claim; a sufficient answer to one question remains usable within its boundary.
For a formal result, suppose a memoizer processes requests one at a time, starts with an empty cache, uses exact input keys for a pure deterministic function f, stores each computed result, and retains all entries. On a hit it returns the cached value without calling f; on a miss it calls f once and stores the result.
These steps preserve the invariant that, after each request, the cache contains f(x) for every input seen so far. Therefore N requests with k distinct inputs cause exactly k calls to f. This derivation can finish that mathematical question without measurement or an assurance record.
Eviction or state-dependent outputs change the premises and require a different argument. The result alone predicts neither wall-clock speed nor the behavior of an implementation that has not been shown to satisfy the premises.
Lowering and reopen conditions
A repair remains usable while its source meaning, applicable rule and conclusion remain clear for the intended use. Reopen or narrow it when the label begins carrying another relation or claim, the object becomes clear and makes this detour unnecessary, the interpretation was chosen from word similarity rather than evidence, or the repair still fails to answer the receiving question.
Use ordinary wording or quotation when that is the intended result. Narrow or block a stronger use when its needed object, rule or support cannot be recovered; an adequate answer needs no invented next action or return condition. Include a blocked overread only when it passes F.19's plausible-reader test.
Archetypal Grounding
Bias-Annotation
Recover the claim from its source use instead of word shape. Keep ordinary prose and already clear domain terms; a known meaning uses its rule directly.
Conformance checklist
Common Anti-Patterns and How to Avoid Them
Consequences
Rationale
Stratification words compress local practice. That compression is useful for recognition and unsafe as a substitute for an object, relation, or claim. C.30.STRAT therefore keeps the source word, recovers what it means in the sentence, and returns the reader to the technical rule.
SoTA-Echoing
The sources below contribute designation discipline, the architecture-description distinction and bounded use of controlled language. Their combination helps recover the claim while preserving a sufficient ordinary sentence. A standard's scope still bounds the distinctions it supplies.
Reopen only the affected cue, worked case, check, or pointer when a source changes. Reopen the wider pattern when a current approach demonstrates a meaning-recovery method that is more reliable at comparable reader effort, or when repeated use shows that the present cue table either misses consequential readings or sends clear ordinary language through unnecessary formal repair.
Relations
- E.10 supplies wording cues; E.10.ARCH helps recover a still-unresolved FPF meaning or relation.
- C.30.P is the broader architecture and structure wording repair; C.30.STRAT is the narrow recurring source-label case.
- Use A.22, C.30, C.30.ASV, C.30.LCA, C.30.TFS-REL, or C.30.ILC only for the selected structure, architecture relation, view, control, flow, or conflict question each pattern defines or tests.
- Use A.6.M for recovered module and interface relations; A.6.F for recovered function claims; E.18 when the selected transformation-flow organization, path or crossing is needed; C.16.P, C.16, C.29, C.31, or C.31.RSA for recovered characteristic, scale, mathematical-lens, reusable-locus, bespoke-residue, or report-only-share claims.
- Use the episteme, publication, and source-use routes in C.30.STRAT:4.2; A.19.SPR for unresolved state wording; the applicable state rule, A.3.3, C.27.TA, or C.27 for recovered state, dynamics, temporal, or rate claims; C.28 for causal use; A.10 and G.6 for evidence; B.3 for assurance; A.20 for internal-constraint results and A.21 for gate decisions; A.15.1 for Work; A.2 for system-role kinds; G.5 for selector-facing set-result declarations and C.11 for local choice among already available options.
- C.33, C.34, and C.35 handle captured, lost, preserved, generated-carrier, or discovered-carrier structure when those claims are at issue.
- A recovered level claim returns to the pattern that defines or tests its subject relation or mapping. C.30.STRAT establishes no level by itself.
Stop when the repaired source label supplies the needed meaning and conclusion. Use the defining or testing pattern for each recovered object or claim, and keep its rules there.
C.30.STRAT:End
Architecture Structural View Adequacy (ASV)
Type: Architectural pattern Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when you have a structural description of one selected architecture-relevant U.Structure and need to decide whether that same description is also a U.View under one viewpoint.
Name the described holon or ArchitectureRelation when known and the structures whose relations matter to the audience. A question about information lost in a diagram can end with the consequential omission. To establish that the description conforms to a particular viewpoint, apply its coverage and omission rules under §4.1 and report the outcome.
For example: “The failover diagram describes runtime-interaction and control structures, but it hides the failover relation. Recover that relation before relying on the diagram to decide whether failover works.” This is a usable triage result. An adequate sentence can be retained, compared, or handed on. Use ArchitectureStructureKindTriage@Project when its fields help that work.
@Project is a compatibility and retrieval cue for a project-side use record. It supplies no project identity, authority, context, viewpoint, parthood, or Work occurrence.
When one actual project matters to this triage, projectWorkOccurrenceRef identifies the composite U.Work recovered under A.15.6. Include architectureStructuralViewProjectUseRelationRef only when a named pattern defines the relation by which this use concerns that Work and an occurrence of the relation obtains. A Work reference alone does not establish project locality. If locality matters but that relation is not yet defined, use A.6.RCD before filling the field; otherwise omit both project-local fields.
claimPatternRefs points only to patterns that define or test separate claims; it does not identify a pattern-application occurrence.
Start with C.30 when the architecture relation, selected structure, or architecture claim is unclear. Use C.30.ASV only when a structural description over selected architecture-relevant structure changes the next architecture use. Claim ArchitectureStructuralView only when the description episteme passes E.17.0 conformance to the viewpoint. Use the fuller record when its organized fields help the intended action, correspondence, source return, comparison or other use; adequate existing wording remains sufficient.
What goes wrong if C.30.ASV is missed: one favored diagram, module view, TEVB viewpoint, generated relation graph, control sketch, or neural-network block diagram is treated as the architecture, selected structure, U.View, or proof without naming the description episteme, selected structure kind, viewpoint-conformance occurrence, hidden or lost structure, correspondence, and next architecture use.
What C.30.ASV buys in practice: the practitioner can distinguish description identity, selected structure and structure kind, viewpoint conformance, representation, and publication before relying on a view. Construction history, hidden or lost structure, correspondence, source return, and admissible use remain separately inspectable.
Do not use C.30.ASV when the question is only about a general architecture claim, subject-side ArchitectureRelation, structure as such, selected transformation-flow relation, mathematical graph description, transformation-flow path relation, or crossing relation. Use C.30, A.22, E.18, E.18.2, C.29, or C.30.TFS-REL as appropriate. If the view is used for another claim, use the pattern for that claim and keep C.30.ASV only for the view portion.
If the issue under repair is still whether view, architecture view, architecture structural view, diagram, model, graph, layer, or functional architecture names a structural description, a U.View, an architecture description, a representation, a publication occurrence, a publication form, a source relation, or another claim or relation named by value, use C.30.P first. Apply C.30.ASV only after the architecture structural-view claim or non-ASV claim named by value is recoverable.
Problem
Architecture structural-view work is selected-structure triage: which architecture-relevant structure is described, which structure kind is under consideration, which viewpoint's fixed rules the description satisfies, and what relation, constraint, invariant, operation, dynamics description, hidden or lost structure, correspondence, source-to-use path or work-reliance relation, and source-return condition changes the next architecture move.
The candidate is first one C.2.1 description episteme. That same episteme is a U.View only while an EpistemeViewpointConformanceRelation to an independently identified U.Viewpoint episteme obtains. Diagram, representation, publication occurrence, form, carrier, and rendering remain separate.
The problem includes these misreadings:
- a module-interface view is treated as all architecture;
- a selected transformation-flow structure, mathematical graph description, or control diagram is treated as proof;
- a structure kind is treated as a
U.Viewpoint; - a viewpoint label, query, authoring route, family-declaration membership, diagram, or publication is treated as enough for
U.View; - E.17.2's TEVB template or a project-local TEVB declaration is treated as a global bundle and mutated to carry architecture-specific structure kinds;
- a diagram, table, dashboard, generated relation graph, or ADR is treated as the view episteme itself;
- functional architecture is treated as a peer ontology rather than a structure-kind interpretation under C.30;
- cross-view consistency is asserted without a supported correspondence claim or obtaining relation;
- omitted structure is relied on in subsequent work without a source-return condition.
Forces
Solution
For an architecture structural view, first identify one candidate description episteme, its selected U.Structure EntityOfConcern, effective ReferenceScheme, structure kind, viewpoint episteme, and the conformance occurrence. Then add construction history, correspondence, hidden or lost structure, source-to-use path or work-reliance relation when at issue, source-return condition when needed, admissible use, and next architecture move. Use ArchitectureStructuralView only for the same episteme whose E.17.0 conformance obtains.
A conforming ArchitectureStructuralView is not a second individual beside its description. The candidate retains one C.2.1 identity <ClaimGraph, one selected-structure EntityOfConcern, effective ReferenceScheme>. The viewpoint and EpistemeViewpointConformanceRelation qualify that same episteme as U.View; they do not enter its C.2.1 identity.
The full view's claimGraph and effectiveReferenceScheme use the C.2.1:4.1.2 ranges and retain the narrower condition that they are the complete content and operative interpretation rules of the viewEpistemeRef. A graph representation is obtained through its semantic correspondence; its nodes and edges do not supply that content by shape alone. A partial scheme can belong to the declared range yet leave a required structure or conformance claim uninterpreted. Resolve that gap before the dependent judgement.
C.30.ASV is the selected-structure structural-view adequacy pattern for architecture work. It explains how descriptions of different selected structure kinds may satisfy declared viewpoints and concerns. It is not a complete architecture-description pattern; C.30.AD composes separately identified descriptions, view-use claims, and correspondence only when that broader description use is being made.
C.30.ASV does not extend any TEVB family by implication. It defines architecture structure kinds and architecture-specific bindings to viewpoint epistemes. A one-off architecture viewpoint can be identified and used directly under E.17.0 without a family catalogue. Reuse of a declared TEVB or architecture viewpoint family resolves its U.ViewpointRef members through the materialized declaration in an E.17.1 catalogue. E.17.2 and C.30.ASV otherwise provide templates, not family values. Source-to-use, work-reliance, project-use, representation, and publication relations remain separate from viewpoint conformance.
Architecture structural view record
StructuralAspectDescription describes one selected structural aspect under A.22. It is not an ArchitectureStructureKindRef or U.View by itself. ArchitectureStructuralView names the same architecture-description episteme only after E.17.0 conformance obtains.
recordPatternLocator identifies the pattern whose record form is being used. It is a locator, not a substitute for any rule needed by a claim at issue. When source-to-use citation is needed, PathCitationRecordRef has the meaning given in C.30.AD:4.3: it resolves the G.6 graph, path and applicable slice together with citations of the independently established direct relations.
For publication use, each carrier reference resolves to an independently identified physical or digital entity whose actual state or realization yields the relevant expression. The carrier's domain supplies its identity rule. When the bearing claim needs inspection, recover its PublicationExpressionBearingRelation and expression under E.24.PUB. A reusable form has its separate conformance relation. A file object, immutable snapshot value and individual copy retain their different identity rules.
The selected U.Structure is the one EntityOfConcern. Before that field can be filled, A.22 identifies the structure from constituents, selected independently obtaining relation occurrences, applied constraint claims, and one receiving-use frame.
A description, query, diagram, family declaration, file, representation, or publication creates none of those discriminators. relatedStructureRefs may name structures needed to interpret correspondence, allocation, or crossing, but they do not create a union-valued EntityOfConcern.
When another selected structure becomes primary, identify another description episteme under C.2.1. Assert edition continuity only when its historical-continuation predicate obtains, and apply A.6.4 separately when retargeting is claimed; do not overwrite identity through a view field.
The conformance occurrence has exactly two participants: viewEpistemeRef as candidate E and viewpointRef as P. Its fixed E.17.0 predicate requires:
- Independently identified E and admitted P;
EntityOfConcern(E)recovered;- P's fixed EntityOfConcern-kind criterion succeeds for that object;
- E has an independently admitted episteme kind accepted by P without circular
U.Viewuse; and - E's fixed claim content under its effective scheme satisfies P's concern-coverage, semantic-form, consistency, completeness, and admitted-omission rules.
The occurrence is participant-determined by <E,P>.
An architecture claim can carry positive, negative, unresolved, required, desired, expected, or candidate content. architectureRelationOccurrenceRefs is affirmative only for independently obtaining ArchitectureRelation occurrences. describedHolonRef and participant traces keep the subject recoverable; neither an optional claim nor a diagram derives the subject-side occurrence.
ClaimScope, concern, model-use structure, and empirical grounding remain optional neighboring qualifiers or relations. modelUseStructureRef appears only when an independently selected DDD-style bounded-model-use structure changes interpretation or selection for this use. None enters base episteme or selected-structure identity.
viewConstruction? records provenance when that history is known and needed. It is separate from the E.17.0 conformance test: a candidate can satisfy the fixed viewpoint rules even when its generation history is unknown. If the receiving use requires that history, the gap limits that use. Direct authoring, A.6.3 construction, query, extraction, selection or publication does not by itself establish conformance or the selected structure.
structureKnowledgeState? states how the selected structure is known when partial knowledge matters: declared, observed, inferred, generated, simulated, extracted, hypothesized, or with an unknown region present. Unknown or inferred structure may guide inspection or source return; it cannot by itself supply architecture truth, assurance, gate, release, causal proof, or architecture decision.
Architecture structure-kind classifier
ArchitectureStructureKindRef is a C.30-local DiscriminatorToken enumeration over architecture-relevant U.Structure references selected under A.22 and used by C.30. It is not U.Kind, U.Viewpoint, U.ViewpointBundle, StructuralAspectDescription, ArchitectureStructuralView, or a root U.* kind. An ArchitectureStructuralView uses structureKindRef to state which kind of selected structure its claim graph describes; that token neither identifies the structure nor grants U.View membership.
The first group is the seed classifier set for ordinary architecture structural-view use. SecurityTrustBoundaryStructure, WorkMethodStructure, AllocationResponsibilityStructure, EvidenceAssuranceStructure, and ScaleEvolutionStructure are classifier values over selected U.Structure references, not new root kinds. ASV may use them to name the selected architecture-relevant structure, but their full semantics stay in the named security, work and method, allocation-responsibility, evidence and assurance, scale, characterization, or mathematical-lens patterns.
Select the structure kinds by the distinction the recipient needs to explain or act on. For a claim about their interaction, establish which connections make them interact under compatible conditions. One view can describe a selected interaction structure that connects several systems or practices. When the contributing views describe different selected structures, relate their findings through C.30.AD:4.3. Ordinary names suffice where identifying a specific ArchitectureStructureKindRef would add nothing to the result.
OtherDeclaredStructureKind is admissible only when the local text names:
declaredStructureKindName;declaredStructureKindDefinition;- allowed relation families;
- locally triggered overreads;
- applicable patterns for non-ASV claims;
- a selected-structure admission test, plus the effective
ReferenceSchemewhen the local classifier name depends on one.
Each structure kind needs a short definition, allowed relation families, locally triggered overreads, applicable patterns for its non-ASV claims, and example architecture structural-view records. This is a controlled classifier set over U.Structure values, not a new root-kind set.
Small triage output
Retain an adequate triage sentence for comparison or handoff. ArchitectureStructureKindTriage@Project can organize the result when the practitioner needs its fields; it is not a required step before a full view record. In either triage form, include effectiveReferenceScheme only when the selected classification or description claim needs its interpretation rules stated separately, and name the claim those rules interpret. The value is an EpistemeInterpretationScheme; satisfying that range does not establish the selected-structure admission test or viewpoint conformance.
architectureConcernCue? and sourcePhrase? are recognition wording. claimPatternRefs? names only patterns that define or test separate claims; it does not assert that a pattern-application occurrence exists. inspectedDescriptionOrViewRef? and candidateViewEpistemeRef? name epistemes only when identified under C.2.1. None creates an ArchitectureStructureKindRef, selected structure, architecture relation, or U.View.
Fill exactViewpointRef when the independently admitted viewpoint is identified; fill viewpointConformanceRelationRef only when the same candidate episteme conforms to that viewpoint.
When architectureClaimRef is absent, describedHolonRef, architectureRelationOccurrenceRef, or at least one candidate selected structure must keep the subject recoverable for the intended triage. claimScope, effectiveReferenceScheme, and modelUseStructureRef are present only when they change this use. The card publishes no architecture claim and creates no subject relation. A full ArchitectureStructuralView requires the candidate episteme's C.2.1 identity plus obtaining E.17.0 conformance; it does not require an architecture-claim record when the selected-structure EntityOfConcern and subject trace are otherwise recoverable.
Practitioner prompt labels are first-entry cues, not ArchitectureStructureKindRef values. When the triage is retained as a record, use the technical values below:
Evolutionary-engineering candidate structural view
Use this branch when a retained variant, front member, selected set, or architecture-candidate palette needs structural-description triage. The claim is "this candidate makes a selected structure or structure kind relevant to an architecture claim or possible architecture move." It is not "this archive is architecture" or "this record is already a U.View."
If the candidate cannot name a selected structure or structure kind, keep it in C.18, C.19, or G.5. If the description is used only to publish, compare, or explain the candidate, use C.30.AD, the comparison pattern, or the publication-use pattern named by value. It is an ArchitectureStructuralView only when the same candidate description episteme independently conforms to the viewpoint under E.17.0.
Project-local architecture viewpoint-family template and binding rows
C.30.ASV ships no architecture viewpoint catalogue, family value, reference, or viewpoint episteme edition. VF.ARCH.STRUCTURE, VF.TEVB.ENG, and VP.Architecture* are therefore not current global values. To claim reuse of such a family, identify the catalogue episteme L and its reference bindings. A local viewpoint itself can be admitted and used independently under E.17.0.
When a recurring viewpoint family is needed, reuse its admitted catalogue and declaration while their edition, scheme, bindings and relied-on premises remain unchanged; resolve only the members needed for this use. For a new family, or when its constitution or binding is missing or disputed, perform these steps:
- constitute or verify L through obtaining
EpistemeConstitutionRelation(G_L, K_L, R_L); - identify or declare the local claim block inside
G_L, retrieved by ordinaryfamilyDesignator = f_archunderR_L; - state or verify the target-kind compatibility condition and a finite non-empty set of
U.ViewpointRefmembers; - resolve every retained reference under
R_Lto one P already admitted under E.17.0; and - preserve catalogue and member provenance for any imported project-local TEVB reference rather than importing a family label.
Until those bindings exist, use this only as an authoring template:
Every r_*, P_*, d_*, L, and f_arch above is a variable until one project supplies the binding. A d_* value is only P's ordinary designator; it is neither a reference nor P. Another project with matching spellings has not reused this family unless it resolves the same L, declaration, and members.
Project-local TEVB reuse follows the same rule. A project may retain one or more references from a materialized local TEVB declaration when their P rules fit. It preserves each <editionDesignator(L_source), sourceFamilyDesignator, sourceViewpointRef> tuple and any omission decision. It does not expand a global TEVB core, infer a cross-family import relation from labels, or claim that E.17.2's four-position template is a materialized family.
candidateViewRecordSetRef names a collection of permitted description or specification-use record forms for one structure-kind binding. Use C.13 when a construction trace of that collection is needed.
The binding and family declaration may help retrieve candidate E and resolve P, but neither makes the fixed E.17.0 predicate true, identifies an EpistemeViewpointConformanceRelation occurrence, grants U.View membership, or creates the selected structure.
The collection is not a publication face or package grouping, and it neither supplies a ViewFamilyId nor adds a viewpoint; publication forms, carriers, catalogue locators, and family designators remain separate.
Structural-view publication-use boundary
C.30.ASV covers the description episteme's identity, selected architecture-relevant structure, structure kind, E.17.0 viewpoint conformance, construction history, correspondence, hidden and lost structure, source return, and the next architecture move. When a view, diagram, graph, card, benchmark, probe output, model publication, or architecture note is used for evidence sufficiency, safety assurance, gate passage, release permission, work record, or decision authority, use the pattern for that claim; keep only the structural-view record and next architecture move in C.30.ASV.
Initial architecture structure kinds and view records
Use the table to find a structure kind that bears on the question. When the explanation needs several structures, establish their contribution and correspondence. Follow a further pattern where its subject claim is needed.
Classifier values defined outside C.30.ASV remain admissible when they are the architecture-relevant structure under consideration, but C.30.ASV does not define their full record families:
Minimum useful seed examples:
Minimal SecurityTrustBoundaryStructureView fields:
SecurityTrustBoundaryStructure carries adversarial-boundary interpretation: which protected assets or effects are under consideration, who or what is trusted, where untrusted input crosses, what authority or privilege is exposed, which adversarial paths and attack exposures matter, which data-flow or control-flow security boundaries matter, and where secure defaults, hardening, update or supply-chain channels, detection, or response boundaries change the next architecture move.
Apply evidence, assurance, gate, or compliance patterns only when the architecture move relies on evidence sufficiency, assurance verdict, gate passage, regulatory acceptance, or release authority. If the selected move is structural, first recover the structure: trust boundary, loss-control relation, control relation, evidence reuse structure, or affected structure or affected view.
Use a SafetyLossControlStructureNote when a safety-architecture concern first needs the architecture-side loss-control structure rather than a safety-case verdict:
The note gives a first architecture move: find the loss-control structure, controlled process or plant, constraint, foreseeable misuse, operational design scope, and action-relevant boundary. It does not replace evidence, assurance, gate, causal, dynamics, or temporal claims.
Select structures for a Method and its uses
Use this profile when “the architecture of the method” is too vague to decide what can change, what must remain compatible, or what a drawing leaves out. Start from one identified Method and the decision at hand. A.3.1 supplies that non-agentive holon's identity; C.30 relates a holon to a selected A.22 structure when its architecture predicate obtains. For several independently usable Methods, retain that plurality until an encompassing Method is established. A useful account of their connections can remain ordinary prose.
Start with the Method concern and intended effect or preserved condition under A.3.1:4.1 and :4.4. They explain what future enactments are for. A proposed receiving use, an actual use of a recording, a statement in a description and a binding in a declaration answer different questions; state the one needed by the architecture decision. E.10:0.2c.7–7a helps when result, purpose or work-product wording hides that choice.
Choose questions that change the decision. The following questions are a first-use repertoire, not a mandatory collection of views or new universal structure kinds. Answer in the language of the work. Select an A.22 structure only when reliance on its organization matters, and apply the rule that defines each relation. A proposed dependency or correspondence remains a claim until its obtaining relation is established; a convenient name or a local graph edge does not admit it.
Connect the answers through one changed condition. Suppose a studio proposes a recording-preparation Method for correcting relative timing while preserving the recorded movements. Its construction requires corresponding time marks and observed offsets, a calculation of the correction, a domain basis for applying it between marks, a suitable editing operation, and a check against the receiving-use criterion.
B.1.5 supplies the question for admitting the constituent calculation as part of this preparation Method under its stable construction rule. Another calibration Method could include that same calculation under its own rule; the shared part would not merge the two wholes.
At three marks the observed offsets are 40, 42 and 45 ms. Subtracting 42.5 ms gives residuals −2.5, −0.5 and 2.5 ms; no constant correction can lower the largest absolute residual at these marks below 2.5 ms.
Separately, stipulate that the recording engineer supplies a justified domain claim that the relative offset remains within 40–45 ms throughout the stated interval, and an editing operation qualified to apply the chosen constant correction without adding timing distortion. That timing envelope is a supplied premise of this hypothetical case, not an inference from three observations.
Applying this 42.5 ms correction to the identified recording with the qualified editor would bound its absolute timing error by 2.5 ms throughout the interval. If the interval envelope loses its basis, the three marks still settle a narrower mathematical question about constant corrections. If the edit itself is unqualified, the calculated residuals describe the specified operation. For a delivered file, establish which edit actually produced that version or check the file directly for the receiving use.
Now select only the structures needed by the studio's choice:
- The composition account explains why the calculation is a constituent of preparation. An interface account says whether the receiving Method gets the corrected recording alone or also the timing qualification and its limits.
- The result-use account connects preparation to movement analysis and demonstration viewing. If the specified edit is applied under the interval and editing premises, its 2.5 ms bound meets the 3 ms analysis and 8 ms viewing tolerances. The studio can select that preparation. To accept a delivered file under the same argument, establish that this correction was applied to the right recording. Consuming the file makes neither receiving Method a constituent of preparation.
- Analysis tightens its requirement to 2 ms. The two extreme marks alone rule out every constant correction for this recording: their 5 ms separation exceeds the 4 ms span the analysis tolerance allows. That rejection still holds if the interval envelope is withdrawn, provided the extreme readings and constant-correction model remain valid. The proposed preparation still meets the 8 ms viewing criterion under the interval and editing premises. If an interval-wide viewing decision is still needed after losing those grounds, obtain a sufficient bound or check the actual recording. The analysis team can instead try a different correction, another recording or a justified change of requirement.
- If the team develops a replacement correction rule, compare it against the affected preparation and receiving uses before introduction. If only the editor workstation moves to another room, inspect the changed access, timing or resource conditions; Method identity changes only if its own defining conditions change.
C.39:5.3 develops the same calculation and changed-condition case for ordinary method development; E.4.CM:5.4 applies it to framework authorship. The interval claim here depends on the separately stipulated domain envelope and editing qualification; neither the numbers at the marks nor a diagram supplies those premises. This is a conditional construction, not evidence of practitioner success. Performing the preparation later supplies Work occurrences and evidence under their own rules.
Choose the picture last. A composition drawing may place the preparation above its constituents, with each edge naming Method parthood. A result-use drawing may place preparation to the left of analysis and viewing, with each edge naming the used result and its acceptance condition. These are the vertical and horizontal directions of those representations, as A.22:4.3a explains. Neither axis gives a universal Method level, a lattice, a performer hierarchy or a place in a room.
The first useful result is a short explanation of what changes and what remains usable. A typed structural view adds value when the decision depends on exact correspondence, hidden structure or viewpoint conformance. If several structures expose a conflict or move burden between participants, C.32.MWA supplies the synthesis and comparison; a clear single connection can stop here.
Functional structure view boundary
A FunctionalStructureView under C.30.ASV does not mint U.Function, U.Transformation, or a bearer relation. It is the same ArchitectureStructuralView episteme whose EntityOfConcern is one selected functional U.Structure and whose viewpoint conformance obtains. It may carry FunctionalElementClaim epistemes when the claim graph relates that selected functional structure to required or desired behavior or effect content and to a bearer or candidate-bearer locus. The claim is not identical with any actual behavior occurrence, bearer, capability, port, allocation, or relation.
Keep three branches explicit:
- required or desired behavior or effect: a C.2.1 claim; use the applicable requirement, architecture, capability, method, or functional-view pattern for that claim;
- actual transformation: an independently identified
U.Transformationonly after A.3.4 recovers the changed referent, extent or boundary, boundary conditions, before, during, and after facts, and continuity or reidentification basis; - compound flow organization: a selected
TransformationFlowStructureunder E.18, whose constituents and selected obtaining relations are independently identified; the structure is not itself an actual transformation.
FunctionalElementClaim has ordinary C.2.1 identity <ClaimGraph, one EntityOfConcern, effective ReferenceScheme>. For this use its EntityOfConcern is the selected functional structure. Its claim content may name:
- one or more required or desired behavior or effect claim refs;
- actual transformation designations only when the complete A.3.4 basis independently obtains;
- selected transformation-flow structure refs for compound flow organization;
- a bearer or candidate-bearer locus, normally a
U.Systemor candidate system for a separately established transformer system-role-kind claim; - capability, input and output condition, functional-port, dependency, allocation, and correspondence refs only when the applicable pattern defines or tests that relation or claim.
If no bearer or candidate allocation is current, do not claim a filled functional element. Record a required-behavior gap, required-effect gap, capability gap, functional-behavior slot, or candidate allocation question. This preserves the practical architecture move. A module, component, diagram row, function word, requirement, or selected flow structure does not establish the bearer or actual change by itself.
actualTransformations and transformerSideSystems designate the identified Transformation and System individuals directly. When an external format stores addresses, declare its correspondence to these values and use its effective scheme to resolve each address to one individual of the stated kind. No target, several targets, or an unsettled identity leaves that dependent use unresolved. A report or record may identify the individual but does not substitute for it. Retarget an address explicitly and recheck the target and receiving conditions. Keep proposed bearers in candidateBearerRefs; a System designation alone establishes neither transformer-role membership nor actual participation in a transformation.
Required cooling effect followed by actual cooling. Requirement episteme RequiredCoolingEffect-1 says that Rack 7 should be brought below 30 °C during declared operation. Before the rack or cooling loop has changed, that is required effect claim content: there is no U.Transformation, even if a functional-view row, flow diagram, or selected TransformationFlowStructure cites it.
Later, Rack7CoolingTransformation-42 may be identified under A.3.4 when the changed referent and boundary are fixed, operating and ambient boundary conditions are stated, before facts show 38 °C, during facts recover heat removal, after facts show 27 °C, and continuity or reidentification keeps the same referent recoverable.
Read the requirement at its stated time and for its stated effect. Reliable before and after temperatures on this rack, with the identified cooling during the interval, can establish that the rack was brought below 30 °C during that operation. A requirement to stay below 30 °C for the whole run would be refuted by a verified 38 °C reading within that run. A reading before the run cannot settle that question. Showing that the named path carried heat from the rack establishes its participation. To credit that path with achieving the threshold, account for other heat gains and losses and test the stated causal claim under C.28.
A selected transformation-flow structure, mathematical graph description, transformation-flow path slice, crossing, or flow valuation is not a functional element or actual transformation by default. When a transformation-flow relation is being used, connect the functional view to the TransformationFlowStructure through C.30.TFS-REL. When a mathematical graph description is being used, connect it through E.18.2; when math-lens use is being claimed, connect it through C.29.
When module allocation is being claimed, use A.6.M to repair the module claim and identify the admitted allocation or interface relation separately rather than treating function and module as one kind.
Functional ports and module interfaces can both use U.Signature discipline, but functional ports specify behavior input and output slots while module interfaces specify substitution, compatibility, boundary, and change-policy claims.
Composability and quality compositionality are separate claims. If the view says parts can be assembled, keep that structure or use claim. If it says a quality of the whole follows from the parts, establish the quality and its composition rule under the applicable domain pattern or C.25. Use C.16 for measurements consumed by that argument; formal or qualitative grounds need no invented measurement chain.
Compositional formalisms may express composition structures, view relations, and model relations. They do not make required behavior actual, create transformations, or make safety, latency, reliability, or another quality propagate automatically.
Correspondence and source return
When a view relates structures, state the correspondence and its basis. An ordinary sentence can carry a warranted correspondence claim; use a record when it helps retain or compare that content. Shared packaging or graph adjacency alone supplies no correspondence.
Correspondence examples:
Use SourceReturnCondition when compression, extraction, coarsening, evidence reuse, ML evaluation, bounded exception, many-to-many allocation, publication, or decision claim hides a distinction needed for action, assurance, causal use, law-domain review, regulatory review, comparison, or reopening.
If viewConstruction is query, extraction, coarsening, correspondenceSlice, or sourceReturnSlice, and omitted structure changes action, assurance, causal use, law-domain or regulatory review, or subsequent decision reopening, SourceReturnCondition is needed.
When the view names affected structures for an architecture use, recover their kinds and references, any relevant ASV references or view loss, and the allowed use or needed source return. C.30 AffectedArchitectureStructureNote can organize this content; an adequate ordinary account is sufficient. The note is not an architecture decision, ADR, gate passage, evidence sufficiency, or release authority.
Use the thinnest source or reliance relation that preserves the next architecture move. Use fuller source, evidence, assurance, or claim-kind relation only when the source or reliance relation being used cannot be inspected, used, compared, refreshed, or bounded without it. A ControlStructureViewNote may precede full C.30.LCA use or use of the applicable proof or assurance pattern when one control relation and its boundary are enough for the architecture move being made.
Treat source return as a user action, not only a metadata field:
Require source return when a hidden distinction carries a selected reliance relation, assurance, law-domain, comparison, causal, gate, or decision commitment. The condition is needed only when the repaired text still relies on the hidden source-side distinction. Keep source return optional for ordinary local recognition when no hidden distinction is being used for action.
Model cards, system cards, and evaluation harness reports may publish or substantiate an architecture structural view only when the structural-view claim is recoverable. The view must name the relevant structure kind, such as RuntimeInteractionStructure, InformationDataStructure, SecurityTrustBoundaryStructure, EvidenceAssuranceStructure, ModuleInterfaceStructure, or another declared structure kind.
It must also state:
- intended-use scope;
- evaluation scope and known loss when evaluation is used;
- deployment-context mismatch when that mismatch is being claimed; and
- the applicable evidence or assurance pattern when the publication is used beyond transparency.
A card or harness is not architecture adequacy, safety proof, or release claim or gate claim by publication alone.
Currentness and smallest reopen. When a decisive input changes, reopen only the ArchitectureStructuralView locus and use conclusions that depend on it.
- A changed selected structure or structure kind reopens its structure fields and, when another structure becomes the EntityOfConcern, requires a separately identified description episteme.
- A changed description identity reopens only that episteme's view admission.
- A changed viewpoint or conformance occurrence reopens only the E.17.0 predicate.
- Changed construction or knowledge state, correspondence, source edition or lost structure reopens the matching provenance, correspondence, source-to-use, hidden or lost structure, or source-return locus.
- A changed admissible-use boundary or applicable rule reopens only its dependent use or claim.
Update that locus, narrow use, return to the named source, or stop. If conformance to the current viewpoint is not established, use a structural description or ArchitectureStructureKindTriage@Project for the next move.
Changed use alone leaves fixed <E,P> conformance unchanged; unrelated structures, views, and claims stay closed.
Worked slices
Runtime degradation. A team wants to explain why service sometimes becomes unavailable during failover. It follows the promised service across the handover and asks which participants can still serve the same request with the required state. The view below is useful where runtime interactions, control, placement and the available observations can distinguish rival accounts. An account of an actual outage also needs grounds that identify the conditions on that occasion.
Use C.24 when a fixed action or option needs budgeted tool-call planning or replanning. Use A.15.1 for Work and identify any call-graph structure under its applicable pattern. Keep those claims and uses separate from the architecture structural-view claim.
CPS or plant architecture. Suppose an engineer must explain how separation of plant sections can be maintained during a control change. Select the paths that could connect the sections and the conditions under which the control can operate them. The views below illustrate what may be needed when a drawing omits one of those relations.
Chiplet or device architecture. A team wants to know whether a replacement can meet the agreed interface and timing conditions in its assembled package. A packaging diagram or interconnect sketch may describe several relevant structures.
Organization or operating-model architecture. Suppose colleagues need to explain how an approved handover keeps a service available when participants change. Follow the work that the recipient needs across the handover. Identify which responsibility, enactor, information and evidence relations account for that continuity. A question about one assignment can use just its relevant relations.
Evidence reuse across product variants. A certification or test package reused across module variants may be architecture-relevant as an evidence-and-assurance structure view, but it is not an assurance verdict:
Organization service architecture. A service organization sketch that shows teams, handoffs, escalation points, and dashboards is not the organization architecture by itself. First recovery can require:
AI agent diagram. A "planner-memory-tools" diagram is not the agent's architecture by itself. It may start first recovery as a structure-kind set, without minting an AI-domain ontology:
Structural AI-agent security is architecture structure when these structure kinds change the next architecture move. When the claim is instead about latent representation, decoding, or effect adequacy, keep the phrase as a reduced-use source cue and use the applicable representation, decoding, or effect-adequacy pattern.
Generated code-agent relation graph. A probe output can supply a diagnostic description or representation. To claim an architecture structural view, establish its description identity, selected-structure subject and E.17.0 conformance. Recover relation meanings and the source basis for the claims being made; distinguish observed, inferred or unknown relations and unexplored regions where that distinction changes the use. Add a source-return condition when hidden distinctions matter. Claims about an agent's internal belief or the safety of a code change need their own grounds.
Neural-network block replacement. For a neural-network modification, such as substituting an attention block or changing MoE path selection, recover which selected structure or relation changes and use C.30 for its architecture relevance. Identify flow, module-interface, preserved or lost structure, characteristic and source-use claims when they matter to that change. Establish any separate decision or evidence conclusion under its own rule; a modification need not assert every one of these claims.
Archetypal Grounding
The Method case in §4.5a shows a different use: composition, result acceptance, method development and performer placement answer different change questions. The practitioner chooses the needed structures before drawing their representations.
Bias-Annotation
The following risks concern structural descriptions and their use in architecture decisions.
This checklist verifies the preceding guidance after the practitioner has chosen the repair action; it is not a required project control form and not a substitute for the card, note, description, conformance relation, or repair guidance above.
Conformance Checklist
For a Method-architecture use, check that the selected questions in §4.5a recover the affected Method, its relations and the consequence of a changed condition. A calculation and a domain claim used to extend its application retain their separate premises and limits; state which uses must be reconsidered if either loses its basis. Keep allocation and Work facts under their own rules. When orientation carries a claim, apply A.22:4.3a to its named structure and representation convention.
Common Anti-Patterns and How to Avoid Them
Consequences
Rationale
Architecture descriptions are commonly multi-view, but "view" cannot stand in for every architecture claim. A structure kind and a viewpoint are different. A structure kind says what kind of selected structure is described; a viewpoint is one episteme whose fixed rules the candidate description must satisfy. The conformance occurrence, not a label or bundle, makes the same episteme a U.View.
The first useful result is an explanation of the selected structure, its kind and the intended use. An adequate ordinary sentence can finish triage. ArchitectureStructureKindTriage@Project and the fuller view record organize content when that helps a receiving use; neither is required merely because the explanation is retained or published. A claimed architecture structural view still requires description identity and E.17.0 conformance.
E.17.2 supplies a four-position project-local authoring template, not a current family or importable bundle. To reuse a declared family, resolve U.ViewpointRef values from its materialized local declaration and preserve catalogue and member provenance. Direct use of an independently admitted viewpoint does not require forming that family. Architecture-specific structure kinds and candidate-record bindings are defined beside those local references rather than mutating their resolved viewpoint epistemes or treating declaration membership as conformance.
SoTA-Echoing
Which structures help a Method change without inventing a universal Method diagram? The answer in §4.5a is an FPF construction from A.3.1, B.1.5 and A.22: select by the change question, then connect composition, interfaces and result use to separately grounded performance facts. A generic view catalogue is a useful starting alternative, but its names leave the studio's failed 2 ms connection unresolved.
Essence 2.0 beta 2, §§7.3–7.4 and 9.3–9.4 is the method-engineering comparator: it supports composable and adaptable Practices in a common descriptive language. Adapt its attention to coherent composition; recover the FPF participants and relations instead of transferring its Method/Practice metaclasses by name.
The studio case demonstrates the narrower gain: under the separately supplied interval basis, preserve viewing while reopening the stricter analysis condition, without first imposing a common method-description metamodel. It does not establish superior expressiveness, lower measured effort or a general ranking of the frameworks.
The cost is relation recovery; an existing suitable domain model can reduce that cost. Reopen the profile when a case needs a decision-changing structure it cannot recover, or another explanation supplies the same distinctions with less reconstruction.
Relations
Builds on: C.30.P, C.30, A.1, A.22, C.2.1, E.24.PUB, A.6.3, E.17.0, E.17.1, E.17.2, A.7, E.10.D2, E.10, C.2.P, and F.18.
Coordinates with: A.6.F, A.6.M, C.30.TFS-REL, C.30.LCA, C.30.ILC, E.18, C.29, C.16, C.25, C.28, A.10, G.6, B.3, A.20, A.21, A.15, C.11, C.32.P2S, C.32, C.32.PAD, C.32.ADR, C.32.ADA, C.33, C.34, and C.35 when problem-to-structure carry-through, candidate-set, architecture-decision, ADR-projection, decision-adequacy, capture, preservation, or generated-carrier claim kinds are being made. Use A.6.M when a module-interface claim is being made, and separately identify the admitted module or allocation relation required by that claim.
Use these patterns for the other claims:
C.30.ASV covers structural-view adequacy for the selected structure being viewed.
C.30.ASV:End
Control Structure View Adequacy (LCA)
Type: Architectural subpattern under
C.30Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when a control diagram, control-language source, or selected control structure changes the next architecture move. Start by asking: what controls what, through which observation, command, reference, supervision, or feedback relation? A label such as controller, plant, observer, planner, supervisor, or policy loop is only a cue; identify the relation and what each participant does in it before relying on the diagram.
A participating System, local system-role kind, System-classification judgment, assignment, Method, or Work is a separate fact. Add it only when it independently obtains and changes the use of the control-structure result.
A first result can state what the drawing specifies: “Supervisor S is to send allowed-mode commands to controller C and receive its status.” To describe the installed arrangement, establish the inputs the enabled rule uses in the mode of interest and where its commands can reach. For a particular update, trace the input used and the command that followed. Establish the physical response when the question concerns what that command achieved. A failed actuator can receive a command without acting. A short architecture claim can retain the proposal. Use the note below for an identified control relation and its participants. If the source gives only a label such as layer, level, tier or stack, C.30.STRAT helps recover the intended relation.
What goes wrong if C.30.LCA is missed: a control diagram is treated as the control structure, U.View, or proof; stratification labels bypass C.30.STRAT and carry undeclared scope; and B.2.5, E.18 transformation-flow prose, or Layered Control Architecture prose is overread as control adequacy.
What C.30.LCA buys in practice: the practitioner can keep useful controller, plant, observer, regulator, supervisor, feedback, rate, and control-layer language while recovering a selected control structure, one description episteme, its possible E.17.0 view conformance, and the pattern used to state or test each proof or claim.
When the question is how to construct or change the cooperation of observations, decisions and interventions, use C.32.LCA. It develops alternatives with their information, timing, authority and continuation conditions. Supply the recovered control account to that construction; return here when a candidate's description or view needs qualification. An adequate arrangement needing only a clearer account can finish with C.30.LCA.
Do not use C.30.LCA when the issue under repair is generic stratification or source-label repair, only an E.18 transformation-flow path slice, function description, module boundary, measurement head, causal intervention, or safety case. Use C.30.STRAT, C.30.TFS-REL, A.6.F, A.6.M, C.16, C.28, or the assurance or evidence pattern to state or test the claim at issue.
The primary EntityOfConcern for a full C.30.LCA description or view is one selected control U.Structure. The description, selected structure, controlled holon, architecture relation, architecture claim, viewpoint, conformance occurrence, control relations and their participants, any participating Systems, classifications, assignments, Methods or Work, diagram, representation, proof claims, and publication remain separate. Start with the smallest useful note:
Use either selectedControlStructureRef or a structureGap. A positive control claim also names at least one obtaining control relation and its participants. This note is enough when those values make the next action clear; its fields do not turn it into a C.2.1 episteme or U.View.
Add a described holon, an architecture-relation occurrence or claim, rate bands, control-layer relations, boundaries, view and viewpoint-conformance facts, source return, representation, or publication only when they change the intended use. Add participating Systems, local classifications, assignments, Methods, Work, and F.6 attribution only when those neighboring facts are independently current.
When either form includes control Work, each Work ref names an occurrence independently admitted under A.15.1 after every performer is recovered through A.13.
assignmentRows and actualControlWorkAttributionRefs remain optional: include them only when the note, view, or receiving use expressly represents precise assignment-bound attribution. Any present attribution ref resolves through F.6 to the same obtaining A.13 assignment; absence or failure of attribution alone leaves independently admitted Work intact.
If the same finding invalidates an admission premise, recheck that premise and the conclusions that used it. The note or view creates none of these facts.
Use full ControlStructureView only when an independently identified architecture-description episteme about the selected control structure satisfies the fixed E.17.0 predicate for one viewpoint. Full use is justified when control-participant meanings, direct relations, rates, recovered control-layer labels, boundary refs, source return, representation or publication, or the patterns used for claims matter beyond the note.
An obtaining relation can be ineffective or harmful. Describe the command, observation and response, including their delay, rather than supplying missing ideal connections. Distinguish an established absence from a relation not yet known: oneWay and unclear support different next inquiries. A controller whose feedback amplifies a deviation still has describable feedback; adequacy is a judgement on its behavior for the receiving use. C.43:4.1.1–4.1.3 connects that description to explanation and a possible change. C.32.LCA develops the new arrangement when that is required.
Problem
Control diagrams are persuasive because they look operational: arrows imply feedback, boxes imply responsibility, and recovered control-layer labels imply separation. In practice that is often enough for orientation, but not enough to identify selected structure, make direct relations obtain, admit the description as U.View, or establish architecture adequacy. A control-stack description can overclaim stability, safety, evidence sufficiency, gate validity, assurance, or causality; a non-control layer, level, tier, or stack label belongs first to C.30.STRAT.
Keep control terminology useful for recognition while distinguishing its cues from structure, relation, and proof. Direct control relations, their participant meanings, feedback relations, externality boundaries, and rate separations can enter an architecture structural description. The same episteme is a view only through viewpoint conformance. Systems, local kinds, separate System-classification judgments, assignments, Methods, and Work are optional neighboring facts; use the relevant patterns to state or test authority, responsibility, safety, stability, gates, evidence, assurance, and causal effects.
Forces
- Control terminology helps engineers identify observations, commands, feedback and their participants; preserve it when its local meaning is clear.
- The same source labels can name different things. C.30.LCA applies after a direct control relation, rate-band relation, control-layer relation, or
B.2.5supervisor-subholon relation is recovered. An assignment is neither required nor sufficient for control; include it only when it independently obtains. A model-use structure is cited only when that independently selected structure changes interpretation. - Layered and multi-rate control descriptions often need timing and dynamics claims before they can carry stability or safety claims.
B.2.5supplies a compound supervisor-subholon feedback claim grounded in observation, return and coupling facts. A feedback diagram alone establishes none of those facts.- Mathematical graph descriptions of E.18
TransformationFlowStructurevalues can describe flow, path, crossing, or transformation-flow relations relevant to control, but the selected flow structure, graph expression, and control structure remain distinct. - Practitioners need one small first output; viewpoint conformance, dynamics, C.29, evidence, assurance, and gate records are used only when the question calls for them.
Solution
Start with the controlled holon and the decision at hand. For design work, recover the observation, command, reference or supervision relations that the proposed structure specifies, including the meaning of each participant. For a claim about a running loop, establish which of those relations obtained in the named configuration and period. A full structural view also requires the selected control structure, its described-holon trace and the E.17.0 viewpoint conformance. Include an architecture relation only on its C.30 basis. For B.2.5 feedback, retain the compound assertion alongside its base relations.
Add participating Systems, local kinds, separate System-classification judgments, assignment species and obtaining occurrences, Methods, and Work only when the corresponding fact is independently established. Use A.22 to identify the selected structure from its constituents, selected obtaining relation occurrences, applied constraint claims, and receiving-use frame; a note, diagram, list, description, kind, or assignment creates none of them. If a source label is not yet control-specific, apply C.30.STRAT first. Then state the supported conclusion and any further pattern needed by the question.
Retain any needed boundary, admissible-use or handoff detail in the existing account. This expanded ControlStructureViewNote can organize it:
Use rateBandRef?, controlLayerRelationRef?, and externalityBoundaryRef? only when that object or relation changes the control-structure use. Read the feedback-closure state for the selected return path, operating configuration and interval. One control relation and its feedback-closure state can answer a bounded question; name a further pattern only if its contribution is needed. Generic stratification labels stay with C.30.STRAT until a control-specific relation is recovered.
For a decomposition or replacement, identify the behavior the enclosing arrangement requires across the layers and establish how the proposed combination preserves it. A qualified assumption–guarantee contract, a common constraint on the reachable joint states, or evidence that covers the combined behavior can supply the ground. Use interLayerControlRelationRefs? for an identified control relation on which the design or assurance claim relies.
Use this note only when a recovered control-layer relation is used for decomposition, substitution, a safety or stability claim, or an architecture decision. It is not proof and does not make the relation obtain. Otherwise keep C.30.LCA at the small note or ordinary description form, or use C.30.STRAT to recover the source label.
A full ControlStructureView carries the common ArchitectureStructuralView content from C.30.ASV:4.1 plus the control-specific fields below. Here selectedControlStructureRef names the same structure as the common selectedStructureRef; the full view retains hiddenOrLostStructure and includes viewConstruction? when the receiving use needs that provenance even though this form does not repeat them.
The full view is the same C.2.1 episteme identified by its exact claim graph, selected-control-structure EntityOfConcern, and effective scheme. Its E.17.0 conformance occurrence has exactly that candidate episteme and one viewpoint episteme as participants. It obtains only when the fixed five-part predicate is true, and those participants determine its identity.
Authoring, A.6.3 construction, a viewpointRef, query, selection, bundle membership, diagramming, rendering, publication, or current use does not make it obtain.
controlledHolonRef names the holon whose state is observed or changed by independently obtaining control relations and may be the described holon or one of its parts. Architecture claims, ClaimScope, model-use structure, concern, and empirical grounding remain optional neighboring objects or relations. modelUseStructureRef appears only when an independently selected DDD-style bounded-model-use structure changes interpretation or selection.
For publication use, each carrier reference resolves to an independently identified physical or digital entity whose actual state or realization yields the relevant expression. The carrier's domain supplies its identity rule. When the bearing claim needs inspection, recover its PublicationExpressionBearingRelation and exact expression under E.24.PUB. A reusable form has its separate conformance relation. A file object, immutable snapshot value and individual copy retain their different identity rules.
For every positive control-relation reference, identify the occurrence and use the relevant pattern to recover what its participants mean. Identify any participating System, local classification, assignment, Method, Work, or F.6 attribution independently on its own admitted basis. A classification or assignment establishes neither control nor action.
The description, control note, view record, and diagram create none of these occurrences. Representation, publication occurrence, form, and carrier likewise remain separate from the selected structure and view episteme.
Safety-loss control-structure note
Use a SafetyLossControlStructureNote only when safety wording is being used for a loss-control claim and the practitioner first needs the architecture-side loss-control structure, not a safety-case verdict:
The note gives a safety-triggered architecture move: find the loss-control structure, controlled process or plant, constraint, foreseeable misuse, operational design scope, and action-relevant boundary. It does not replace the generic control-structure view and does not replace evidence, assurance, gate, causal, dynamics, or temporal claims.
Interpret control relations and the claims they support
Control-participant interpretation.
Control-specific stratification gate. Layer, level, tier, and stack enter C.30.LCA only after C.30.STRAT or the local sentence recovers a direct control relation, inter-layer control relation, rate band, or B.2.5 supervisor-subholon relation. An assignment alone is insufficient, and the label by itself establishes neither control structure nor separation.
B.2.5 boundary. Use B.2.5 for the compound supervisor-subholon feedback assertion. The description cites its observation and returned-influence base relations in feedbackRelationRefs and the assertion in feedbackClaimRefs; it does not coerce an episteme into U.Relation. Keep the coupling rule and facts recoverable from the assertion. Use the relevant patterns for stability, safety, causality, evidence, gate and assurance claims.
If a dated Work occurrence involving an episteme is claimed, recover the performing System through A.13 and admit the Work and enacted Method independently through A.15.1. Add an assignment occurrence and F.6 only when the account expressly consumes precise assignment-bound attribution; keep publication, source-to-use, and work-reliance relations separate.
Transformation-flow boundary. An E.18 transformation-flow path slice may supply flow-structure, path, crossing, or transformation-flow-structure input to the control view when that relation is being used. The transformation-flow description may use a mathematical graph expression under E.18.2; the description is a U.View only when E.17.0 conformance obtains. Neither the expression nor the description becomes the functional architecture, the control structure, or proof of control adequacy.
Use of a mathematical model. A proof can settle the response of the modeled loop for all disturbances and operating conditions covered by its assumptions. To apply that proof to the installed loop, identify the operating mode and check the feedback path, plant response and timing on which the conclusion depends. A successful trial can establish the response in that trial. C.29 handles the transfer, A.3.3 the dynamics, C.27.TA and C.27 the temporal claims, and C.28 a claim about the effect of changing the controller.
Nesting and scale. Identify the property at issue and which control relations could produce it. A regulator can use feedback while a supervisor supplies a fixed reference. For a claim of supervisory control, show how the reference, limit or operating mode set by the supervisor governs the regulator’s action. For supervisory feedback, also trace the observation returned to the supervisor and how it can affect that setting. Several regulators can produce a joint response through their interactions and shared limits. Trace these when they account for the behavior of the whole. Recover the bounds needed to judge latency, stability, observability, accountability or assurance, and keep a scale-audit reference when the receiving rule calls for it.
Worked slice A — a diagram used as proof. A safety note claims that a plant is safe because the supervisor monitors the lower controller. A display can show measurements even while commands run on a timer. For feedback between the supervisor and the lower controller in one episode, establish what the supervisor observed about the lower controller, how the rule or judgement in use depended on that observation, and which resulting influence reached the lower controller. A timestamp can fit both a fixed schedule and an observation-dependent rule. Records of the input and decision can distinguish them even when both would issue the same command. A claim about the physical effect of that command needs evidence of the plant response. To extend the feedback claim to another operating mode, check the signals available in time, the rule enabled in that mode and the return to the controller. To assess safety, identify the hazard and establish how the response addresses it under the conditions at issue.
Use the applicable safety pattern for the safety claim, B.3 only for a named assurance claim, A.10 for bounded source-to-use reliance, G.6 for citable provenance paths, C.27.TA for temporal aspects and rate bands, C.27 for authored temporal-claim adequacy, and A.3.3 or the applicable dynamics pattern for dynamics or stability.
Worked slice B - multi-rate controller. A source says a control stack has a slow planner, a faster regulator, and an observer with a different update period. Apply C.30.LCA only after the stack label has been recovered as reference-provision, regulation, observation, or other control relations with their participant meanings and rate bands; otherwise use C.30.STRAT first.
Systems, classifications, assignments, Methods, and Work are added only where independently current. A C.30.LCA description establishes no rate adequacy. If the rate relation matters for oscillation, latency, stability, or safety, next use C.27.TA for temporal aspect or rate-band structure, C.27 when an authored temporal-claim adequacy question is under repair, and the dynamics or assurance pattern named by value when that claim kind is being made.
Worked slice C - supervisor-subholon loop. A subsystem is supervised by an external controller System. The C.30.LCA note cites the observation and returned-influence relations together with B.2.5's compound assertion and its coupling basis.
If that System performs mode-change Work, recover it through A.13 and admit the Work and enacted Method independently through A.15.1. Add an assignment occurrence and F.6 only when this slice also expressly represents precise assignment-bound attribution; missing or failed F.6 attribution alone leaves the mode-change Work intact while its independent A.13/A.15.1 grounds still hold.
Authority, responsibility, gate passage, safety, stability, and policy-constraint results remain separate claims under their own patterns; the supervisor relation establishes none of them.
Currentness and smallest reopen. When a decisive input changes, reopen only the control-structure locus and the use conclusions that depend on it.
- A changed selected control structure or controlled holon reopens the affected
ControlStructureViewNoteor full description and view. - A changed direct control relation or participant meaning reopens that occurrence and its dependent structure selection.
- A changed classification, assignment, Method, Work, or F.6 attribution reopens only that neighboring fact and any view use that relied on it.
- A changed B.2.5 coupling rule, supervised set or report/return fact reopens the affected pair and interval assertion, its cited base relations where changed, and dependent structure uses.
- Changed rate or control-layer relations reopen only their matching relation or boundary fields.
- Changed view conformance reopens only the E.17.0 admission.
- A changed source edition reopens its source-to-use and source-return locus.
- A changed authority, responsibility, safety, proof, evidence, assurance, or gate claim reopens only that neighboring claim unless a control-structure input also changed.
A communication break need not end an assignment; reconnection requires new supporting facts.
Update the affected locus, demote full view use to a note or orientation, narrow use, or reopen the control-structure question; unrelated structures and claims stay closed.
Archetypal Grounding
Bias-Annotation
- Diagram authority bias. A neat feedback diagram can look more persuasive than the structure, source-to-use path, work-reliance relation, or claim it supports. Repair by naming each object or relation and the pattern used to state or test the claim.
- Stratification-label bias. A
layer,level,tier, orstacklabel can hide whether it names a control relation, rate band, aggregation, scale, organization, Work scope, evidence scope, deployment, or publication section. Repair withC.30.STRAT; C.30.LCA applies only to the recovered control-specific case. - Supervisor anthropomorphism. A supervisor label can make an episteme, policy, assignment, or dashboard sound agentive. Repair by recovering the supervision relation first. If a dated Work occurrence is claimed, recover the performer System through A.13 and admit the Work and enacted Method independently through A.15.1. Add assignment and F.6 only for an expressly consumed precise assignment-bound attribution; recover authority, responsibility, gate, safety, and evidence separately.
- Transformation-flow and LCA conflation. A transformation-flow graph expression and a control description or view can inform each other, but neither replaces the other. Repair by naming the EntityOfConcern, structure kind, and direct relations for each.
This checklist verifies the preceding guidance after the practitioner has chosen the repair action; it is not a required project control form and not a substitute for the note, description episteme, conformance occurrence, direct control relation, or repair guidance above.
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Consequences
The gain is a small control-structure output that preserves common architecture language while blocking structure, view, and proof overread. Practitioners can still say controller, plant, supervisor, feedback, and control layer, but the record shows the selected structure, the boundary between description and view, and the direct relations those words carry; generic stratification labels use C.30.STRAT first.
The cost is recovering missing control facts or checking the conformance or property on which the use depends. Reuse an adequate existing account. Recognition can finish with an ordinary sentence; view membership, safety, stability, evidence, assurance or gate passage requires its own basis.
Rationale
Keep the practice cue and recover control-structure content first: selected structure, controlled holon, architecture relation when current, direct control relations and participant meanings, recovered rate or control-layer labels, observation and actuation boundaries, externality boundaries, and next admissible move. Systems, classifications, assignments, Methods, Work, and F.6 attribution are added independently; use their own patterns for authority, responsibility, gate, safety, stability, and evidence.
The full view is one description episteme admitted as U.View through E.17.0 conformance; its record represents the relevant content and references. It can cite C.30.STRAT, B.2.5, E.18 transformation-flow structure, dynamics, C.27.TA, C.27, C.28, evidence, assurance, gates, and C.29, but does not absorb their claim kinds.
This protects subject, structure, episteme, View, representation, and publication boundaries. Several descriptions may have the same selected control structure as EntityOfConcern, and one description may be published repeatedly without changing identity, creating the structure, granting view membership, or making direct relations obtain.
SoTA-Echoing
Relations
- Builds on
C.30for direct architecture relations and selected-structure adequacy,C.30.ADfor description identity and use,E.17.0for viewpoint conformance, andC.30.ASVfor structural-view adequacy. - Uses
A.22for structure identity and structure-kind discipline. - Coordinates with
C.30.STRATwhen layer, level, tier, stack, ladder, rung, block, expert, cache, router, gate, or similar source labels must be recovered before control-specific use. - Coordinates with
B.2.5for supervisor-subholon feedback relation recognition. - Coordinates with
C.32.LCAfor constructing and comparing control arrangements. C.30.LCA supplies the recovered control account and its use limits; C.32.LCA returns explained alternatives whose descriptions can be examined here. Construction and view qualification retain their different results. - Coordinates with E.18 and C.30.TFS-REL when transformation-flow path slices supply structure input to the control view; use E.18.2 for a mathematical description when that description is at issue.
For neighboring claims, use these patterns:
Use C.30.LCA only for the control-structure description or view-adequacy question at issue.
C.30.LCA:End
Cross-Scope Architecture Residual Triage
Type: Architectural subpattern under C.30 Status: Stable Normativity: Normative for FPF pattern, companion-text, and project-description use that claims architecture-specific residuals across declared scopes.
Problem frame
Use this pattern when a project situation contains a cross-scope architecture residual for a described holon, often described in project speech as:
First-minute use slice. After an assembly team retunes its arm controllers, individual arms finish faster. The team also reports more cell stoppages and more audit exceptions. The engineer takes the assembly cell as the holon of interest and separates the two decisions.
- For the stoppages, compare the time required for an arm to acknowledge a request with the interval the cell controller allows. Use the active configuration and the recorded reason for a stop to distinguish a timeout from a shortage of material. To explain an increase across jobs, also check which jobs and operating conditions were compared.
- For the audit exceptions, compare the records the retuned controllers supply with the information the audit now requires. Changed audit criteria can produce more exceptions even if the controllers supply the same records.
A documented timing incompatibility rules out timely acknowledgement through that interface. If timely acknowledgement is required, the specified exchange cannot meet that requirement. Carry this finding into redesign when a different arrangement is commissioned. If timeout with a fallback is permitted, check whether that alternative meets the cell’s need. Evidence that the interface caused a particular stop warrants addressing that occurrence. The aggregate counts alone leave both diagnoses open. A mediator or changed control layer becomes a candidate when the diagnosed coupling calls for it. Use C.30.ASV when its structural view is needed.
The first result names the affected holon levels or scopes, the structure that carries them, the residual-bearing locus, any attempted local repair or reason for its insufficiency, and the first warranted architecture move. An adequate ordinary explanation can finish the question. CrossScopeArchitectureResidualTriageRecord@Context can organize the result when its fields help the receiving use.
The primary EntityOfConcern is the cross-scope or interlevel architecture residual in the described holon or holon family for a named architecture concern and intended use. The described holon may be an admitted system, organization-as-system, episteme, work occurrence, discipline, or another admitted holon kind. Publication-family material uses the episteme and publication patterns. A MethodDescription is an episteme; a Method uses A.3.1 and the relations claimed for it.
A phrase in a description, a diagram label, or a mathematical-lens output may make the residual visible, but it is not the residual itself and does not become the center of this pattern.
InterlevelConflict@Context applies when two or more declared holon levels, declared scopes, or level-bearing structure relations of the same described holon or holon family impose incompatible or tensioned constraints, objectives, admissibility conditions, tempos, resource allocations, information-transfer relations, or assurance requirements.
Examples include declared system levels, declared episteme levels, aggregation scopes, typed control layers, declared organizational scopes, work scopes, evidence scopes, system scopes, environment scopes, description-use scopes, publication-use scopes, or other declared scopes.
A selected structure matters here only when it carries, separates, or relates the declared levels or scopes. A conflict between structures belongs in C.30.ILC only when those structures are assigned to different declared holon levels, declared scopes, scale windows, or coarse-graining steps; a conflict confined to the same level and scope, or one whose relevant assignment remains unknown, belongs elsewhere until that cross-level or cross-scope basis is recovered.
FrustrationResidual applies when a persistent cross-scope or interlevel residual remains after local repairs have been attempted or deemed insufficient: local optimization in one declared holon level or declared scope improves local fit while degrading, blocking, or destabilizing another declared holon level, declared scope, or level-bearing structure relation.
ComplexityGrowthPressure is admitted only as conditional architecture pressure: reducing an interlevel residual may require adding, splitting, mediating, or stabilizing a declared holon level, declared scope, aggregation scope, interface grammar, control loop, evidence scope, work-method scope, abstraction scope, source-return scope, or declared system level when that special case is being claimed. It is not a claim that complexity is good or that complexity necessarily grows.
Entry condition. For the described holon in context, name:
- declared holon levels or declared scopes;
- the selected structure or architecture structure kind that carries them;
- one residual-bearing locus;
- one first admissible architecture move or a supported
noArchitectureMoveresult.
If any cannot be named, keep the issue at ordinary problem framing or ProblemCard@Context; do not claim CrossScopeArchitectureResidualTriageRecord@Context yet.
What goes wrong if C.30.ILC is missed: a local improvement, control layer, scale label, interface grammar, or evidence reuse is treated as whole-holon architecture adequacy while the residual moves into another declared holon level or declared scope.
What this buys: the practitioner can keep conflict or frustration language as an entry label while naming the residual-bearing locus, the affected declared holon levels or scopes, the selected structure that carries them, any local repair or reason for its insufficiency, and one first architecture move or sufficient noArchitectureMove result.
Interlevel conflict and frustration may appear in ordinary project descriptions, but the conforming record describes the residual through declared holon levels or declared scopes, the selected structure that carries them, and a residual-bearing locus. The pattern does not create a generic level scale or U.Frustration.
It asks which declared holon level, declared scope, aggregation scope, control layer, organizational scope, work scope, evidence scope, system scope, environment scope, scale window, interface grammar, allocation boundary, publication section, or source-return condition bears the residual. A system level or episteme level is a special case of a declared holon level.
Use the subject pattern when the issue under repair is only ethical value framing, interlevel ethical conflict structure, ethical mediation or decision use, measurement, scale relation, coarse-graining relation, mathematical-lens validation, candidate generation, residual-reducing candidate-set work, final selection, causal outcome, evidence, or assurance. Use C.30.ILC only for the architecture residual-bearing locus.
Problem
Architecture work often starts from a residual: a local fix works in one declared holon level or declared scope and fails in another. Component optimization increases whole-holon or product-line integration cost. A new module boundary reduces local complexity and increases exceptions at the product-line scope. A control layer improves local safety and creates accountability or latency claims elsewhere. A reusable evidence set reduces repeated work and hides a new source-return condition.
The architecture intuition concerns how local optimization at one declared holon level or declared scope can create a persistent residual in another declared holon level, declared scope, or level-bearing structure relation. It does not introduce a new Frustration kind.
When a recoverable multilevel, scale, or coarse-graining mapping is claimed, use C.29 to state and test that lens use. When the claim compares architecture alternatives over a declared scale window, use C.31.ASAP.
Use C.32.MLAO and C.32 for residual-reducing candidate work, and use G.5 only to declare a selected-set result. If that result is made available to an audience, use E.17 for a source-backed publication face and return to source and E.24.PUB for the publication occurrence, form, carrier, audience, bounded use, and availability.
An ordinary conflict between structures is not enough for the RG lens or frustration mathematical lens, but a conflict between structures assigned to different declared holon levels or scale windows may be enough when the mapping, preserved-structure line, and lost-structure line are recoverable. The first C.30.ILC output is the grounded residual account and first warranted move.
Without a pattern, teams either discuss the residual as vague complexity, treat it as an ordinary negotiation problem, jump into measurement, use mathematical frustration language as proof, or jump to candidate generation too early. C.30.ILC keeps the first move small: identify whether the residual is architecture-shaping and name the first admissible architecture move or subject-pattern application.
The practical work is often not to draw another view. It is to assign the residual to the locus named by value that can bear it: declared holon level, declared scope, level-bearing selected structure, structure kind, constraint, characteristic or Q-bundle, evidence-reuse boundary, source-return condition, or non-architecture claim kind.
Forces
- Local optimization can be real and still be harmful at another declared holon level or declared scope.
Level,layer,scope,scale,abstraction,organization,system, andenvironmentlabels can sound precise while naming different project-side entities, relations, scopes, or claim kinds.- Frustration language can be useful because it points to incompatible constraints or fitness contributions, but it can also implicitly introduce a physics, biology, psychology, or global-optimizer ontology into architecture prose.
- Measurement is tempting because the residual feels numeric, but a measure before declared-scope and structure-kind recovery can hide the conflict.
- Ethical conflict or mediation may be present, but not every cross-scope residual is an ethical conflict or mediation problem.
- Architecture synthesis may be needed, but a small triage output often identifies a narrower move: split scope, add mediator, add interface grammar, change allocation, expose coupling, add evidence scope, accept bounded exception, or return to source.
- The pattern is not a prescribed sequence of moves; architecture work is often case-managed through loops, checks, and dead ends.
Solution
When an architecture concern involves residuals across declared levels or scopes, identify the related structures, the conflicting conditions and the locus where the loss remains. Preserve an adequate ordinary account; use CrossScopeArchitectureResidualTriageRecord@Context when its fields help retain or compare those distinctions. The record does not itself establish the conflict.
Layer, level, tier, stack, and declared-scope labels. Declared holon level is the general level-bearing recovery field for this pattern; system level and episteme level are special cases when the described holon or selected structure makes them relevant to the claim. System level may remain as ordinary recognition language when a practitioner would naturally use it, but the record recovers the project-side scope references through declaredHolonLevelRefs or declaredScopeRefs; a system level is not the default architecture level.
When the source says layer, level, tier, or stack, recover one or more of: declaredHolonLevelRef, controlLayerRef, declaredSystemLevelRef, aggregationScopeRef, rateBandRef, organizationLevelRef, workEvidenceScopeRef, scaleWindowRef, or publicationSectionRef when the wording only names a document layer.
A move such as splitDeclaredHolonLevel, splitDeclaredScope, or, in the special system-level case, splitDeclaredSystemLevel is admissible only when the affected declared holon level, declared scope, selected structure, declared system level, aggregation scope, control layer, organization scope, work scope, evidence scope, system scope, environment scope, rate band, scale window, publication section, or source-return condition is named.
A layer label is not a structure kind, not a system level, not a rate band, and not evidence of separation by itself.
Interlevel conflict, frustration residual, and complexity-growth recovery. A conflict is architecture-shaping only when the account identifies the declared holon levels or declared scopes, the selected structure or structure kind that carries them, and the conflict carrier: constraint, objective, admissibility condition, tempo, resource allocation, information-transfer relation, or assurance requirement. A frustration residual is architecture-shaping when the cross-scope loss persists after attempted local repair, or when a stated reason shows why that repair would leave the loss unresolved. Complexity-growth pressure is only a candidate reason to add, split, mediate, or stabilize structure when that change is expected to reduce the residual enough to justify the new cost and its own residue.
Identify the constraints whose combination matters, their connections and what their sources establish. When different models group the arrangements differently, check whether the apparently conflicting conditions can occur together in the use under inquiry. If separate summaries lost their pairing, restore the joint constraint before attributing a conflict to the architecture. The loss of that relation can itself answer a question about the descriptions. If an exchange can return an acknowledgement no earlier than 10 seconds after a request and the receiver accepts it only within 5 seconds of that same request, the design cannot complete that exchange as specified. The two bounds already settle the design decision. To attribute a particular cell stop to this conflict, establish which request was outstanding at the deadline, whether a reply had arrived by then and which rule actually stopped the cell. A competing explanation matters when it changes the causal account or the action the recipient needs. Apply C.28 when attributing a change in the rate of stops across jobs to the retuning.
Multilevel optimization boundary. First establish whether local optimization in one declared holon level or scope degrades another declared holon level or scope. This triage neither optimizes the architecture nor proves that one global function exists. A mathematical treatment needs a recoverable mapping between declared levels, scopes, scale windows or coarse-graining steps, with the distinctions preserved and lost for its use. Retain an adequate domain treatment directly. Use C.29 when formulation, transfer or reliance remains unresolved; its MLU.Description@MultilevelLearningFrustration form is optional unless the receiving interface requires it.
Conflicting structures can enter this lens only when each structure is assigned to a declared holon level, scope, scale window, or coarse-graining step and the mapping shows why the conflict is interlevel. If scale window, RG relation, coarse-graining relation, preserved structure, lost structure, or conflict residual slope becomes an architecture scale-preference claim, use C.31.ASAP and use C.29 for any remaining mathematical formulation, transfer or reliance question.
If the practitioner needs to generate residual-reducing candidate architecture moves, apply C.32.MLAO for the residual-reducing multilevel candidate frame and C.32 for the candidate palette. Use A.19.CPM for comparison under a declared comparator, G.5 only when selected-set result declaration is current, and C.11 when final local choice is current. For audience publication, use E.17 for a source-backed face and return to source and E.24.PUB for the publication occurrence and audience availability. Use C.32.PAD when a project architecture decision is current.
Stop the C.30.ILC use at the residual and first admissible move. If the case is only a conflict between two selected structures with no recoverable assignment to different declared levels or scopes, keep it in C.30, C.30.ASV, D.3, D.4, C.28, evidence, assurance, or decision patterns as applicable.
Stop condition. Stop when the account names the residual and the first warranted move, including a supported conclusion that no architecture move is needed. The triage does not measure scale preference, generate candidate architectures, mediate ethical conflict, or select a decision. Apply a subject pattern only when making a claim of its kind:
D.3 and D.4 boundary. D.3 connects each ethical claim to the affected entity, declared level relation or scope, value premise and any frame edition used, expected consequence, horizon, and evidence use or uncertainty when current, then states the tension among the sides. Identify the conflict-description episteme when the receiving use needs to cite or revise it. A D.4 use takes that description into mediation, refusal, evidence demand, causal return, assurance return, architecture return, accepted residual, or bounded decision use.
C.30.ILC handles architecture-specific recognition: whether the conflict or residual is borne by declared holon levels or declared scopes within a selected structure. Such a residual may concern allocation, interfaces, control rates, work reuse, evidence reuse, scale windows, or coarse-graining loss. Structural views may help describe the selected structure and locate the residual. C.30.ILC is a triage and architecture-move pattern, not an ethical mediation pattern.
Architecture-move examples.
Worked slice A — clean module layout, bad flow. A team redraws its module boundaries and reports more order-to-cash exceptions. For the proposed layout, trace one eligible order through the required handoffs and identify any new transfer the design imposes. For the current workflow, compare the handoffs actually used for comparable orders before and after the change. The redrawn boundaries may simply expose transfers that already occurred. A new transfer can justify inspecting its interface. Attributing the rise in exceptions to that transfer requires comparing its effects under the relevant conditions. Use C.30.TFS-REL for the flow and C.16 for the exception measure.
Worked slice B - AI agent control conflict. A local agent optimizes its local objective and violates a supervisor's allowed-mode constraint. C.30.ILC names the agent scope, supervisor scope or control scope, control relation, local optimization claim, residual-bearing locus, and local repair attempted. The first move may be add control layer, change allocation, or apply C.30.LCA. Safety, causality, and gate claims use their subject patterns.
Worked slice C - evidence scope residue. A reusable certification evidence set removes repeated evidence work for several product variants, but one variant has a hidden environment difference. C.30.ILC names the work scope or evidence scope and source-return condition. The practitioner uses A.10 for source recovery and bounded reliance, and G.6 when the provenance path must remain addressable. Whether the evidence meets the certification requirements for that variant is tested under the applicable certification rule.
Worked slice D - frustration residual before synthesis. Several decompositions reduce local module work but each creates a different integration, control-rate, or evidence-reuse residual in another declared scope. C.30.ILC records the residuals and first architecture moves. If the team needs a residual-reducing candidate palette, stop the C.30.ILC use and apply C.32.MLAO for the residual-reducing frame and C.32 for the candidate palette. Use G.5 only when the palette or retained set must be declared as a selected-set result for downstream use.
A multilevel-learning or frustration treatment retains the relevant level, scope, scale-window or coarse-graining mapping and the distinctions it preserves or loses. Use C.29 for a remaining formulation, transfer or reliance question; an adequate domain result can finish without its record form.
Archetypal Grounding
Bias-Annotation
- Local-success bias. A local improvement is treated as whole-architecture improvement. Repair by naming the wider declared holon level or declared scope and the residual.
- Pseudo-level bias.
Level,layer, orscopesounds precise but no declared holon level or declared scope exists. Repair throughdeclaredHolonLevelRefsordeclaredScopeRefs. - Generic-structure-conflict bias. A conflict between selected structures is treated as cross-scope or interlevel conflict although the relevant levels or scopes and their relation are unrecovered. Identify those levels or scopes and the structure that carries them; otherwise keep the case in
C.30,C.30.ASV, or its subject pattern. Different declared scopes can matter at the same level. - Frustration-ontology bias. A useful conflict or frustration entry label becomes a new first-order architecture kind, physics ontology, biology ontology, or psychology claim. Repair by recovering declared holon levels or declared scopes, the level-bearing structure, conflict carriers, residual-bearing locus, and the subject pattern for any lens, proof, evidence, mediation, or synthesis claim.
- Global-optimizer bias. Local optimization in one declared holon level or declared scope is used as if the architecture optimizes one global function. Repair by keeping the local optimization claim as a triage input unless
C.29supplies an admissible mathematical-lens use with recoverable level mapping or scale mapping and the candidate-set or decision pattern carries any synthesis or selection claim. - Measurement-first bias. A number is used as a cross-scope residual without identifying its characteristic, bearer or comparison basis. Use
C.16or the applicable characteristic rule to establish that measurement; it may help discover the residual before architecture triage is complete. The architecture conclusion still needs the affected scopes and structural relation. - Mediation-default bias. Every conflict is treated as ethical mediation or negotiation. Repair by checking whether the use under repair is architecture structure, allocation, interface grammar, control, work or evidence scope, source-return, or another declared level-bearing architecture relation.
- Synthesis-jump bias. A local residual immediately triggers candidate generation. Repair by identifying the first admissible architecture move before applying
C.32.MLAOandC.32; useG.5only when selected-set result declaration is current.
The conformance checklist below verifies the preceding guidance after the practitioner has chosen the repair action; it is not a required project control form and not a substitute for the card, note, view, relation, or repair guidance above.
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Consequences
The gain is an early architecture move that is small and precise. The practitioner can preserve useful problem language such as conflict, frustration, level, layer, or local optimization while making the affected scopes, relations and remaining loss inspectable.
C.30.ILC supplies the residual account and first warranted move, which can complete the receiving question. A further construction or evaluation uses the applicable subject pattern. Measurement, scale relation, RG relation, coarse-graining relation, mathematical lens use, ethical mediation, candidate generation, evidence, assurance, and final choice remain outside until those claims are being made.
Use C.30.ILC to identify the residual that makes optimization relevant. Use C.29 for a remaining mathematical formulation, transfer or reliance question, retaining the applicable level or scale mapping and preservation/loss boundary; an adequate domain result can be used directly. C.32.MLAO and C.32 supply residual-reducing candidate frames and palettes, G.5 supplies selected-set result declaration, and C.32.PAD governs a project architecture decision when those results are needed. For publication, use E.17 for a source-backed face and source return and E.24.PUB for the publication occurrence and audience availability.
Rationale
Interlevel conflict and frustration serve as Plain entry labels by pointing to a recurrent architecture failure: local repair in one declared holon level or declared scope leaves a residual in another. Used as generic labels, they can hide which level, scope, level-bearing structure, relation, conflict carrier, or source-return condition bears the residual.
A local optimum or successful local repair is therefore not treated as whole-architecture adequacy. It becomes relevant to C.30.ILC triage only when the residual-bearing locus is recoverable and the next architecture use can be named.
C.30.ILC keeps the entry label but recovers the architecture relation or structure claim. It treats conflict or frustration as architecture-shaping only when declared holon levels or declared scopes, the level-bearing structure, conflict carriers, and residual-bearing loci are named. This lets FPF preserve the practical intuition without introducing a second ontology of levels, a hidden measurement pattern, a physics or biology transfer, a global optimizer proof, or a prescribed architecture work order.
SoTA-Echoing
Relations
- Builds on
C.30andC.30.ASVfor grounded architecture, selected-structure, and structural-view adequacy. - Uses
A.22for selected-structure grounding. - When the residual concerns flow, control, function, allocation, module, or interface structure, use the applicable pattern:
C.30.TFS-RELfor the bounded architecture-use record connecting an architecture locus to a transformation-flow structure or network;C.30.LCAfor the selected control structure, its description, and view conformance when current;A.6.Ffor function-use repair; andA.6.Mfor module-interface claim repair. - Applies
C.16or the characteristic pattern that defines or constrains the characteristic under evaluation for measurement or characteristic claims. - Uses
C.29for a remaining question about mathematical formulation, transfer or reliance. For a multilevel-learning or frustration lens, retain the applicable level or scale mapping and preservation/loss boundary;MLU.Description@MultilevelLearningFrustrationis an available form for that account. ApplyC.31.ASAPfor a separate architecture scale-preference claim. - Applies
C.32.P2Swhen residual triage must continue through problem-to-structure architecturing; appliesC.32.MLAOfor residual-reducing multilevel candidate frames andC.32for candidate palettes; appliesG.5only when selected-set result declaration is current.
For these claims and uses, apply:
When audience availability is current, use E.17 for a source-backed face and return to source and E.24.PUB for the publication occurrence, form, carrier, audience, bounded use, and availability. Use C.30.ILC only to triage a cross-scope architecture residual.
C.30.ILC:End
C.30.TFS-REL - Use Transformation-Flow Structures and Networks in Architecture
Type: Architectural pattern Status: Stable Normativity: Normative unless marked informative Tech-name:
ArchitectureTransformationFlowStructureRelation(relation record) Plain-name: architecture transformation-flow structure relation Governed object: the bounded architecture-use record connecting an actualArchitectureRelation, selected architecture structure, structural-view or description episteme, or bounded architecture claim to one selectedTransformationFlowStructureunderE.18or one selectedTransformationFlowStructureNetworkunderE.18.NET; the record does not itself instantiate a direct relation.
C.30.TFS-REL:1 - Problem frame
Use this pattern when an architecture discussion depends on one selected TransformationFlowStructure, one selected TransformationFlowStructureNetwork, or a current path, path slice, crossing, flow valuation, edition pin, plane pin, context pin, no-hidden-scalarization claim, or mathematical description of the selected flow structure.
First identify the architecture locus, the selected E.18 TFS or E.18.NET network, and what this use lets the practitioner inspect or decide. An adequate existing account can finish the question. ArchitectureTransformationFlowStructureRelation is an optional record for retaining those distinctions.
The locus can be an ArchitectureRelation occurrence, selected architecture structure, architecture-description or structural-view episteme, or bounded architecture claim. For a network, distinguish an actual containing-holon relation, a bounded claim about one named holon, and a use spanning several named holons.
When the optional scheme field is used, supply the rule collection by value in the C.2.1:4.1.2 range EpistemeInterpretationScheme. It must interpret the designations and claims used by this architecture-flow record.
This is a use/trace record, not a universal direct U.Relation declaration and not an obtaining-condition shortcut. Establish each positive architectureRelationOccurrenceRef, flow relation, cross-member relation, correspondence relation, empirical-grounding relation, publication occurrence, or project/work relation under its direct pattern before asserting it. Keep a bounded claim or description separate from an obtaining relation occurrence.
groundedNonAdmissibleUse? is an alias for the optional nonAdmissibleUse? value, used only for a supported restriction that matters to the receiving use.
Ordinary minimum: name at least one architecture-side reference (architectureRelationOccurrenceRefs, selectedArchitectureStructureRefs, architectureStructuralViewRefs, architectureDescriptionRefs, or a bounded architectureClaimRefs entry) and at least one flow-structure reference (transformationFlowStructureRef, transformationFlowStructureNetworkRef, transformationFlowUnfoldingStructureRef, selectedPathOrSliceRefs, crossingBundleRefs, or flowValuationRefs), the admissible use, and one stop or governing-pattern application. A network use also selects exactly one network architecture-use branch and supplies its required holon and relation/claim refs.
Keep another trace field only when it changes the next architecture move or receiving use; unused optional fields may be omitted. State non-use, uncertainty, or missing information when that distinction changes the operation.
Include an explanatory nonAdmissibleUse guard only when an independently grounded alternative is plausible for the intended reader here and distinguishing it changes understanding or use. An unused guard may be omitted without an absence entry unless a concrete receiving use needs to distinguish absence from missing information.
Use this record only when an actual architecture relation, bounded architecture claim, selected architecture-relevant structure, structural-view episteme, functional-structure view, transformation-flow-structure claim, or conditional architecture-description use depends on an E.18 TFS, an E.18.NET network, or one of the selected TFS's paths, crossings, or valuations. Stop when that architecture-to-flow-structure use and its stop or return condition are clear. If another claim is being made, apply its governing pattern and keep this record to the architecture/flow boundary.
What goes wrong if this pattern is missed: a transformation-flow diagram, graph-shaped mathematical description, path slice, flow valuation, requirement, or functional-view row becomes functional architecture, whole architecture ontology, actual U.Transformation, performed Work, work-result record, evidence, gate passage, or project decision by appearance.
What this buys in practice: the practitioner can use E.18 for one TFS or E.18.NET for one network while C.30 remains the direct architecture-relation and selected-structure adequacy locus and C.30.ASV remains the architecture-structural-view locus.
Not this pattern when the question is only the TFS or network, a mathematical description, path, crossing, or flow valuation and no architecture use is being made. Use E.18 for one TFS, E.18.NET for one network, E.18.2 for its mathematical description, and C.29 when mathematical-lens use is current. Use C.30 for a direct architecture relation or architecture claim without this flow-structure use; use C.30.AD for a durable architecture description and C.30.ASV/A.6.F for a functional view without it. Apply any other claim's governing pattern and keep C.30.TFS-REL only for the architecture/flow relation.
C.30.TFS-REL:2 - Problem
Using actual architecture relations, selected architecture-relevant structures, architecture structural views, and conditional architecture descriptions often requires E.18 TFS objects or one E.18.NET network when the question concerns transformation-flow structure, required functional dependencies, actual change, data movement, control paths, evidence-flow descriptions, neural-network dataflow, or code-agent relation graphs.
C.30.TFS-REL prevents collapse by requiring the architecture-side reference before any E.18 TFS, E.18.NET network, path, slice, crossing, or valuation receives architecture use. It also keeps required or desired behavior/effect claims distinct from actual A.3.4 transformations. A network use chooses the branch in :4.4a from its named holons and actual relation or bounded claim; its graph, description, publication, or record cannot supply those facts.
C.30.TFS-REL:3 - Forces
C.30.TFS-REL:4 - Solution
C.30.TFS-REL is the C.30 entry record to E.18 and E.18.NET when an actual architecture relation, bounded architecture claim, selected architecture-relevant structure, architecture structural view, or conditional architecture description uses one selected TransformationFlowStructure, one selected TransformationFlowStructureNetwork, or a current path, crossing, or flow valuation.
It supplies only the architecture-to-transformation-flow use boundary. The field set in section 1 names the contributions that a use may need. An existing account can carry these contributions if it is sufficient and qualified and has accessible exact returns. No separate form or entries for unused optional fields are required, and no populated field makes a direct relation obtain.
At least one architecture-side field and at least one E.18 or E.18.NET field must be named by value. Network branch fields obey C.30.TFS-REL:4.4a; other optional trace fields follow the ordinary minimum in section 1. The explanatory nonAdmissibleUse guard follows the ordinary minimum in section 1 and needs no absence entry unless a concrete receiving use requires that distinction.
C.30.TFS-REL:4.1 - Use trigger
Use this pattern only when an ArchitectureRelation occurrence, bounded architecture claim, selected architecture-relevant structure, architecture structural view, functional-structure view, transformation-flow-structure claim, or conditional ArchitectureDescription use depends on one or more of the following:
TransformationFlowStructureRef;TransformationFlowStructureNetworkRef, when architecture use selects an E.18.NET-conforming network;PathIdorPathSliceId;CrossingBundleRef;- a flow valuation on the E.18:5.2 connection/cut domain, with its channels and regime and its resolution to the same selected structure;
- edition, plane, or context pin;
- no-hidden-scalarization or set-return discipline;
- a correspondence claim or independently governed relation between functional structure and transformation-flow structure;
- a generated or extracted relation graph used as candidate input for the architecture-to-transformation-flow use.
For a dated Work claim, recover its A.13 basis and independent A.15.1 admission; use F.6 only for precise assignment-bound attribution. Ordinary practitioner action wording needs no invented occurrence or assignment. If it says that a referent changed, use A.3.4 before citing a U.Transformation. If it only says that one selected TFS exists, use E.18; if it only says that one independently identified E.18.NET-conforming TFS network is selected, use E.18.NET. If the sentence uses a graph-shaped expression as mathematical description, use E.18.2. If it relies on a mathematical lens, use C.29.
Use transformationFlowUnfoldingStructureRef? only when the architecture use depends on one A.22-selected CGUS qualified under E.18.3. The ref names that selected CGUS; its E.18.3 account separately names one independently identified E.18 substrate branch and the positions, bindings, and already-obtaining occurrences the CGUS uses.
Architecture, decision, work, feedback, narrative, or refresh values connect only through already-obtaining supporting relations, with predicate-definition content and current facts when the claim needs them; the pattern reference adds no connection relation.
Generic architecture use of a constraint-governed unfolding structure belongs in C.32.P2S or the direct C.30 architecture governing pattern; this pattern keeps only the architecture-to-transformation-flow trace.
C.30.TFS-REL:4.2 - Relation to functional structure
A FunctionalStructureView under C.30.ASV may cite ArchitectureTransformationFlowStructureRelation when a transformation-flow use is current. That record does not make the selected E.18 structure a functional element or actual transformation, and does not make a functional-element claim identical with the system, module, method, bearer, or flow. It states a bounded claim or trace that functional-view content corresponds to, or is declared relative to, one E.18 selected structure, member-local path, crossing, or valuation. Positive co-reference states that the functional-structure and flow-structure designations refer to the same selected U.Structure.
actualTransformations, transformerSideSystems, and the single transformerSideSystem in the note below use the direct-value designation and external-address resolution rules in C.30.ASV:4.6.
Keep the same three branches used by C.30.ASV:
functionalBehaviorClaimRefsandrequiredOrDesiredEffectClaimRefsremain C.2.1 claim content under their requirement, architecture, capability, method, functional-view, or other direct owner;actualTransformationscite only independently identified A.3.4 occurrences with changed referent, boundary or extent, boundary conditions, before/during/after facts, and continuity or reidentification basis;selectedTransformationFlowStructureRefscite E.18 structures, which may organize several independently identified transformations and transfers but are not themselves required effects or actual transformations.
A FunctionalElementClaim is a bounded C.2.1 claim about one selected functional structure. Its bearer or candidate-bearer locus, capability, port, allocation, transformation, and correspondence refs retain their direct owners. A graph-shaped expression, path, valuation, required-effect statement, or flow packet is therefore not the functional element by default.
Required-cooling-effect / later-actual-cooling countercase. RequiredCoolingEffect-1 can require Rack 7 to be below 30 °C and can correspond to a selected cooling-flow structure before any change occurs. In that first use, fill requiredOrDesiredEffectClaimRefs and the selected TFS fields; leave actualTransformations empty.
A later Rack7CoolingTransformation-42 is actual only when A.3.4 fixes Rack 7 as the changed referent, its thermal boundary and operating/ambient conditions, 38 °C before facts, heat-removal during facts, 27 °C after facts, and continuity or reidentification of Rack 7.
Apply the requirement to the rack and period it names. A reliable reading of 27 °C establishes “below 30 °C at the specified instant” when it belongs to that rack and instant. For “below 30 °C throughout the shift,” one reliable 38 °C reading within the shift already establishes a violation. A reading before the shift leaves that guarantee open. Temperature and net heat removal can agree under two different cooling routes. Observed transfer through the named route can establish its participation. To decide whether that route was necessary for bringing the rack below 30 °C, compare what would have happened with the other heat exchanges but without that transfer.
Use this note when the practitioner needs to see whether the function-to-transformation-flow relation changes inspection, split, relation-making, downgrade, selection of the governing pattern for the claim named by value, candidate generation, or stop.
Use C.30.ASV for the functional structure view, A.6.F for function-like wording recovery, A.3.4 for an actual transformation, A.6.M for module-claim repair and the direct allocation/interface owner, and E.18 for selected transformation-flow structure.
FunctionTransformationFlowRelationNote is the one-TFS form. When architecture use selects a network, use the top-level ArchitectureTransformationFlowStructureRelation and the branch in C.30.TFS-REL:4.4a. Name a member TFS in this note only when the function correspondence is to that member; membership in the selected network alone does not create a function correspondence.
When several transformation-flow variants are kept or compared as candidate architecture inputs, keep each selected transformation-flow structure, path, crossing, valuation, graph-shaped expression, or mathematical description under E.18, E.18.2, and this record. Apply C.32 only to the architecture candidate palette that uses those selected structures. The graph, path, and flow description do not become architecture adequacy, evidence, assurance, gate passage, selected-set result declaration, publication occurrence, or decision by serving as a candidate input.
C.30.TFS-REL:4.3 - Claim-kind applications named by value
Apply a row when its claim is made; a sufficient architecture-to-flow account can finish without those further inquiries. For another claim kind, use the rule that governs it.
C.30.TFS-REL:4.4 - E.18 selected-structure boundary statement
For an E.18-governed selected TransformationFlowStructure used by an ArchitectureRelation occurrence, selected architecture structure, ArchitectureStructuralView episteme, or conditional ArchitectureDescription episteme, the architecture-use record may cite that E.18 structure plus MVPK faces and correspondence claims or independently governed relations.
Grounded architecture adequacy and bounded architecture claims are governed by C.30; description identity by C.30.AD; view conformance by E.17.0 and C.30.ASV. E.18 supplies selected transformation-flow structures and relations; it does not define all architecture structure kinds, create an architecture relation, or turn required flow content into actual change.
C.30.TFS-REL:4.4a - Architecture use of a transformation-flow structure network
First recover the independently selected E.18.NET network and the architecture question. For one named holon, distinguish an obtaining architecture relation whose selected structure is that network from a bounded claim that does not assert such an occurrence. Use the inter-holon branch when the claim concerns a connection among participants. Use the containing-holon branch when the question depends on the network actually organizing that whole. A connection between two participants can be the subject of inquiry even when a containing holon is known. Select the branch for the particular claim.
-
Named containing-holon use. Set
networkArchitectureUseBranch=namedContainingHolon. Name exactly onecontainingHolonRefand one actualcontainingArchitectureRelationRefwhose selected structure is the same network.containingArchitectureClaimRefis optional claim/trace content. Keep all participating arrays andnoNetworkBearerHolonAssertedabsent. Member TFS values and their Work, valuations, boundaries, actual transformations, and direct relations remain independently governed. -
Explicit inter-holon use. Set
networkArchitectureUseBranch=explicitInterHolon. Put at least two distinct holons inparticipatingHolonRefs[]. Add exactly the actualparticipatingArchitectureRelationRefs[]and boundedparticipatingArchitectureClaimRefs[]on which this question relies; a network member whose architecture is not used by the question stays outside those arrays. Keep all containing fields absent and setnoNetworkBearerHolonAsserted=true. This states one architecture-use question spanning named holons; it does not invent a containing holon, architecture relation, or characteristic bearer whose identity is the network. -
Single-holon claim use. Set
networkArchitectureUseBranch=singleHolonClaim. Name one C.30ArchitectureClaimQ inarchitectureClaimRefs; its content names the described holon H and the same selected network N. Preserve Q'sdoesNotObtain,unresolved, orcandidateOrExpectedOnlydisposition and its scope, reference scheme, and supporting or missing basis. Keep the containing fields and participating arrays absent; setnoNetworkBearerHolonAsserted=true.Use N's admitted positions and relations to inspect the bounded architecture question. Keep the resulting claim or question with its basis. Use
admissibleUseandstopOrReturnConditionto state what may be done with it and what must be recovered before stronger reliance. An unavailable fact remains unknown; a negative claim needs its own C.30 basis.An architecture use that relies on an actual R(H,N) belongs in the containing-holon branch; retaining Q does not supply that occurrence.
Every other populated architecture-side reference must agree with the selected branch.
- In
namedContainingHolon, each value inselectedArchitectureStructureRefsbelongs to the containing architecture relation's selected structure route, and each structural view, architecture description, functional structure view, or architecture claim used by this record traces to the same containing holon and relation. - In
singleHolonClaim, each reference traces to H or Q and retains its own actual or modal status; an independently grounded current relation R(H,S) used for comparison does not establish R(H,N). - In
explicitInterHolon, each such reference traces to one named participating holon and, when actual, its architecture relation; a singular reference names only that participant and does not imply a containing architecture.
If a reference depends on another holon or architecture relation, add it only when the current question relies on it, or use a separate record.
The branches are mutually exclusive. When transformationFlowStructureNetworkRef is absent, networkCrossFlowRelationRowRefs[] and all network branch fields are absent. A network ref without one complete branch is not ready for architecture use.
When the record also names a path, slice, crossing, valuation, required effect, or actual transformation, bind it to the member TFS that owns the value and to the local positions, participants, or bindings that identify that value. When it names a network-aware unfolding, the E.18.3 substrate branch must name the same network and preserve its admitted position mappings, while selectedCGUSRef continues to name the separate A.22-selected CGUS. The network ref does not lift member-local values into network-global state.
Use networkCrossFlowRelationRowRefs[] only for E.18.NET-owned composite locators. Each locator's current containing record must describe the same selected network, and the direct occurrence plus complete ordered endpoint-binding identity must resolve exactly one nested row. Zero matches, several matches, or a record for a different network stop this architecture use. The locator identifies the row; it neither creates the relation occurrence nor changes its direct governor.
For every maintainability, capability, responsibility, production, safety, or other architecture-characteristic claim made or used by this record, name the holon, actual architecture relation, selected structure, description/view episteme, bounded claim, or other bearer governed by C.30 or the characteristic's direct owner. A network may have selected structural facts—members, relations, recursion, or exposed positions—but those facts do not make an unnamed network the bearer of holon characteristics, agency, Work, production, required effects, or actual transformations.
A network diagram, member graph, mathematical description, publication, or TransformationFlowStructureNetworkRecord may supply facts or claims used in the account. Its form alone establishes neither an architecture relation nor the applicable branch. Keep its identity distinct from the network and any architecture relation, and qualify its representation, description or publication use under the applicable pattern.
Named containing-holon case. The holon ManufacturingPlatform-7 has one obtaining architecture relation whose selected structure is the product-development/production-system-change network. C.30.TFS-REL may use that network to localize an architecture change while each member TFS, production relation, Work occurrence, and actual transformation keeps its own owner.
Explicit inter-holon case. A supplier holon and a plant holon use one selected E.18.NET-conforming supply-linked TFS network to inspect a cross-company dependency. Both appear in participatingHolonRefs[], with only the actual architecture relations and claims the question uses in their corresponding arrays. No containing supply-chain holon has been grounded, so noNetworkBearerHolonAsserted=true. The network is not called the architecture of an unnamed enterprise.
Single-holon claim case. N is an independently selected development/production network. Q asks whether N organizes ManufacturingPlatform-7; the available account leaves one constituent's inclusion in that holon unknown. Keep Q unresolved and inspect N's admitted cross-member dependency to identify which allocation or boundary fact must be recovered. A proposal to use N instead keeps candidateOrExpectedOnly and may support a bounded comparison. If a sufficient C.30 basis establishes that N does not organize this holon, retain doesNotObtain; missing facts alone do not support that result. These uses retain a useful architecture question without an actual R(H,N).
If N's own cross-member relation or endpoint binding is missing, stop at E.18.NET's proposed description or selection blocker instead of naming a selected N.
A detached account carries Q's disposition, H, N, basis and usable return; the bare phrase “N is the platform architecture” would lose the distinction on which this use depends.
C.30.TFS-REL:4.5 - Worked slices
Functional architecture with a transformation-flow relation being claimed. A team says, "The functional architecture is this flow diagram." The repair is:
This required-effect account asserts no actual transformation.
For the filled record below, suppose the current checkout configuration and its verified connections independently establish the cited CheckoutService architecture relation. With only the proposed diagram, retain the bounded architecture claim and omit the occurrence. For the designer’s question, identify the effect the path is meant to support and which dependency the drawing preserves. An operator asking how a particular request was authorized needs evidence of the decision actually returned. The route matters to that question to the extent that it distinguishes the answer.
For the Rack 7 case in §4.2, the flow design supplies a candidate route for the required cooling. A dated temperature reading can settle the specified end condition. Before, during and after facts about Rack 7 can establish the cooling occurrence. Identified heat transfer through the selected route establishes participation. A second route might already have been sufficient to reach the target. Use the heat balance and a warranted comparison for a claim about the first route’s effect on the outcome. These uses can be decided separately.
Near miss: if the selected transformation-flow structure has no C.30-side architecture reference named by value, the case stays in E.18. If the same sentence is a mathematical description, use E.18.2; if it is a math-lens-use claim, use C.29. If it is a Work log, evidence claim, gate decision, or benchmark result, apply the governing pattern to that non-flow claim and keep this record only for the architecture-to-transformation-flow use.
Pump-station flow relation. A plant team says, "the safety architecture is the bypass flow." C.30.TFS-REL applies only if the plant holon, its actual architecture relation or bounded architecture claim as current, selected control or material-flow structure, and E.18 selected bypass-flow structure are named.
The bypass path may be architecture-relevant, but it is not an actual cooling/pumping transformation, safety proof, performed maintenance Work, gate passage, or release permission.
The account names the plant architecture locus, selected E.18 path or crossing, any action-changing hidden relation and return condition, and what the bypass relation lets the team conclude or change. An adequate conclusion need not propose an architecture change.
Supply-chain transformation-flow relation. A logistics architecture view may use an E.18 selected flow structure for supplier handoff, transport crossing, freshness window, and valuation. The subject holons, actual architecture relations when claimed, and selected supply-chain structures remain named; Work occurrences, contractual commitments, evidence, and gate decisions stay with their governing patterns.
Neural-network dataflow change. Source labels such as attention block, SSM block, convolution block, memory mechanism, cache mechanism, and MoE expert-selection go through C.30.STRAT unless the changed value is already recovered. C.30.TFS-REL applies only when the changed structure kind and transformation-flow relation are named. A benchmark, ablation, or pruning result may bear on a non-architecture claim named by value, but it does not make the flow relation an architecture decision, actual transformation, or evidence sufficiency by itself.
Extracted relation graph. In a generated graph distinguish an import present in a source edition, a call permitted in a configuration and a call actually observed. An allowed path with jointly compatible conditions establishes a possible route in the model. A response marked authorized could have come from a cached decision or an invoked payment service. Matching send and receive records can establish contact with the handler. To identify the source of the issued response, establish which decision the responding branch used. A background refresh can contact the service while the caller receives the cached decision. A verified bypass refutes a claim that every request took the service route. To conclude that no bypass exists, account for the permitted alternatives, including any paths outside the extraction. Retain the edition, extraction locus and relation semantics needed to repeat that judgement.
C.30.TFS-REL:4.6 - Lowering and currentness conditions
Lower, narrow, or reopen the relation at the smallest changed locus when:
- E.18 one-TFS structure, path, crossing, or flow-valuation semantics change;
- E.18.NET network identity, direct membership, exposed positions, cross-member relations, or nested-row locator resolution changes;
- the selected network architecture branch or any architecture claim used by that branch changes;
- edition, plane, context pin, set-return, or no-hidden-scalarization discipline changes;
- source publication or graph edition, path slice, relation observation class, edition or context pin, unexplored region, or hidden relation-structure return condition changes;
- the C.30 architecture locus, selected architecture-relevant structure, architecture structural view, conditional architecture description, or C.30.ASV relation changes;
- functional-to-transformation-flow correspondence changes;
- a non-flow claim is being made and requires the direct governing pattern identified in
C.30.TFS-REL:4.3rather than this architecture-to-flow record; - C.29, C.16, C.28, A.10, G.6, B.3, A.20, A.21, A.15.1, A.15, C.30, C.30.AD, C.30.ASV, A.6.F, C.30.STRAT, E.18, or E.18.NET changes the governing boundary used by the relation.
Admissible repair results are the following alternatives:
- update the affected TFS or network reference, network branch, or row locator;
- add or change correspondence or the hidden relation-structure return condition;
- narrow admissible use;
- keep the one-TFS claim inside E.18 and the network claim inside E.18.NET;
- keep the mathematical-description claim inside E.18.2;
- keep the math-lens-use claim inside C.29;
- apply the governing pattern to a non-flow claim;
- lower to quote-only or reduced-use cue;
- block the architecture-to-transformation-flow use.
C.30.TFS-REL:5 - Archetypal Grounding
C.30.TFS-REL:6 - Bias-Annotation
This checklist verifies the preceding guidance after the practitioner has chosen the repair action; it is not a required project control form and not a substitute for the card, note, use record, direct relations, or repair guidance above.
C.30.TFS-REL:7 - Conformance Checklist
C.30.TFS-REL:8 - Common Anti-Patterns and How to Avoid Them
C.30.TFS-REL:9 - Consequences
C.30.TFS-REL:10 - Rationale
E.18 governs one selected TFS, its paths, crossings, valuations, and pins; E.18.NET governs one selected network and its cross-member relations. Architecture needs to use either object without taking over its ontology or inventing an unnamed architecture bearer. The smallest stable result is therefore one qualified C.30-side account identifying the architecture and flow objects it uses and preserving the applicable architecture-use branch from :4.4a when a network is selected.
This pattern also protects functional architecture and actual-change semantics. A functional structure may correspond to a transformation-flow structure, and in some cases both views may designate the same selected U.Structure; that identity is not automatic. Required or desired effect remains claim content, while an actual U.Transformation requires the independent A.3.4 basis.
C.30.TFS-REL:11 - SoTA-Echoing
Currentness boundary. Inputs are E.18 TFS semantics and pins; E.18.NET network identity, cross-member relations, and row-locator resolution when selected; the chosen network architecture branch and its named holons, relation refs and claim content; C.30/C.30.AD/C.30.ASV architecture-side rules; observation class; required-versus-actual status; and non-flow governors named in C.30.TFS-REL:4.3. When one changes, the record changes only at the affected reference, branch, row locator, correspondence, hidden relation-structure return condition, admissible-use boundary, or governing-pattern assignment.
C.30.TFS-REL:12 - Relations
Builds on: C.30, C.30.AD, C.30.ASV, A.22, A.6.F, A.3.4, E.18 for one TFS, E.18.NET for one selected conforming network and obtaining cross-member relations, E.17.0, E.24.PUB, A.7, E.10, C.2.P, and F.18.
Coordinates with:
C.30.STRAT;C.32.P2Swhen architecture-to-transformation-flow grounding is one stage of problem-to-structure architecturing;C.32when selected transformation-flow variants become candidate architecture inputs;C.33when transformation-flow relation descriptions capture or lose selected architecture structure;C.34when transformation-flow claims must be preserved across a mapping, model, generated output, or realization;C.35when a generated or discovered carrier may seed synthesis;A.15.1for Work occurrence;A.15for Method/Work alignment;A.20,A.21,A.10,G.6,B.3,C.28,C.29,C.16, admitted measurement, selection, or candidate-set governors;A.6.Mmodule-claim repair and direct interface owner;A.6.5when reusable relation-participant typing is needed;A.6.0when a signature declaration is being made.
Related claims stay with their governing patterns:
C.30.TFS-REL governs only the bounded architecture use of the selected TFS or TFS network.
C.30.TFS-REL:End
Modularity and Reusable Structure Characteristics
Type: Characterization pattern Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when modularity or reusable-structure language is doing work in an architecture discussion and the practitioner needs to know which few characteristics matter, what repair follows, and whether any measurement or comparison is admissible.
First identify the few characteristics that matter to the architecture question, the observed difficulty and any needed repair. An adequate ordinary account can finish; ModularityVectorLite offers a compact way to retain it:
For the first diagnostic move, select the characteristics needed to explain the observed problem and choose the next warranted repair. Keep this first inquiry small while preserving every distinction that changes that move. A bounded report-only proxy can be a sufficient result.
Claim-use boundary: before relying on a characteristic for comparison, causal explanation, assurance, selection or another further use, establish the grounds required by that claim. Making a qualified local account available to an audience preserves its limits; publication alone does not turn it into a measurement or a selection result. Add only the information on which the receiving use depends.
What goes wrong if C.31 is missed:
- "modular" becomes a binary label;
- a single modularity score hides incompatible characteristics;
- interface publication is confused with substitutability;
- internal cohesion is improved while evidence reuse gets worse;
- bespoke residue moves from templates into work or assurance; and
- complexity language becomes a commensurable score without a declared characteristic, scale, measurement basis, comparison basis, or admissible-use boundary.
What C.31 buys in practice: the practitioner can choose the modularity characteristic that matters, identify a warranted repair, and determine what further measurement, evidence, causal, scale, selection or accounting grounds the intended use requires.
For a question limited to one of the following, use the indicated pattern instead of C.31:
Problem
Architecture work often asks for more modularity, better reuse, lower coupling, more explicit interfaces, cleaner allocation, or less bespoke residue. These phrases point to engineering work, but they do not name one scalar property.
Modularity can improve along one characteristic while worsening another. A design can reduce external coupling and increase interface alphabet size. A platform can standardize interfaces and create unhelpful rigidity. Evidence reuse can improve while functional-to-module allocation remains tangled. A report can show a high reuse share while hiding source-return work or residual uncertainty.
C.31 turns modularity talk into a small characteristic vector plus repair direction. It does not promise that every characteristic is measurable now, comparable across cases, causal, admissible for decision use, or evidence-backed.
Forces
Solution
C.31 defines and constrains modularity and reusable-structure characteristic assertions as C.16-compatible characteristic heads, composite descriptions, lens-backed characteristic interpretations, temporal or scale-sensitive characteristic interpretations, causal-use-sensitive characteristic interpretations, or report-only proxies. It starts from action guidance and adds fields for beyond-local-repair use only when a use being made requires them.
Ordinary output: ModularityVectorLite
Start with the change someone has to make, such as replacing a module, qualifying an interface or servicing a variant. Identify the characteristics that could alter that action. For a substitution, specify what the replacement has to do with the receiving equipment under its operating conditions. Compare the proposed replacements against those conditions. A shared interface name can help find candidates. A check that covers the required fit and behavior can admit one directly. To claim that changing the interface would save work, compare the installation, migration and qualification work under the alternatives over the same planned uses. ModularityVectorLite can retain the account.
A diagnosis can finish a request to locate a modularity problem. If the commission asks which change would reduce service or qualification work, compare the work required with and without the change for the same required service under comparable external conditions. Keep a fixed installed fleet when the commission fixes it. Where an alternative changes the fleet or the number of service calls, include the work that actually follows from each arrangement. An identified repair supplies a candidate for that comparison.
relatedClaimPatternLocatorsIfClaimed locates a governing pattern; it supplies neither the claim nor its grounds. Further use needs the subject, assertion and applicable rule. Rule identity is required only under section 4.3. Architecture scale-preference claims use C.31.ASAP.
Filled ModularityVectorLite
The engineer can now name the eligible board among the two examined. To decide whether an interface change would reduce the cost of field service, compare the adapter and qualification work for the installed variants with the migration and service work after the change. An increased standardization share can coexist with higher service cost.
Characteristic classes
Identify the classes that apply to the characteristic use. They distinguish requirements and can overlap: a measured temporal relationship used causally needs its measurement, time-window and causal grounds.
In C.31, declared basis and comparability basis name C.16-compatible measurement or comparison fields. They are not generic reason words and are not substitutes for evidence, assurance, cause, source, decision, or architecture-description relations.
Measurement-head mapping
When a proposed comparison or decision relies on a measurement head whose meaning or admissible operations remain unresolved, recover the following mapping. Reuse an adequate existing account; the form is optional.
Use an ordinary PatternID to locate the rule.
Under E.10 L-EPI-PUB, relationFunctionClaimRef identifies the defining or constraining ClaimGraph only when interpretation, comparison, migration, publication or reuse depends on that rule identity. It is not a demand to establish a separate relation-function individual.
When an external standard, policy or other non-pattern source carries the relevant authority, use authoritySourceRef for that source instead.
For example, an ordinary InterfaceStandardizationShare account states its subject, the counted interface population, conformance specification, ratio Scale, source/test basis and limitations. It needs no extra rule-identity claim merely to describe that share.
If a comparison combines two editions that count partial conformance differently, cite the defining predicates through relationFunctionClaimRef before treating their shares as comparable. The comparison still needs its measurement and evidence basis; citing the predicate supplies neither test results nor assurance.
This mapping prepares a C.16-compatible measurement or a report-only interpretation; it does not supply observations or a measurement procedure. Recover the evidence or source relations that the claim relies on. Use evidenceClaimAbsentBecause only when the receiving use needs an explanation of that absence; a local cue needs no invented evidence claim. Sharing a bounded account preserves its existing qualification.
C.31 characteristic card
Use the full card when retaining or comparing the characteristic requires these distinctions. An existing adequate account can carry them:
A measurement or comparison states the characteristic, bearer, scale, unit interpretation, measurement or comparison basis, and evidence on which it relies under C.16. Retain relevant limits and any warranted repair. When source material is used as evidence, identify that source relation. A checklist, dashboard label or inherited score supplies no missing basis. The optional card organizes this content; it does not make a qualitative diagnostic cue a measurement.
Seed characteristic heads and repair actions
These heads are seeds, not an exhaustive taxonomy. Use only the heads that change the next action.
Claim-scoped residual heads
C.31 uses residual heads only as qualitative repair cues. These heads do not create one complexity characteristic.
Proxy-risk discipline
When a decision relies on a proxy, identify how it could diverge from the intended value and how that divergence would be detected. The card can retain these as proxyRisk and auditQuestion. An observed divergence blocks the unsupported value inference; the measured proxy can remain a valid report within its scope.
Rejected shortcut
The expression ModularityScore = average(all measures) is not admissible as a C.31 result. A local score is admissible only when the scoring method, codomain, polarity, characteristic basis, comparability basis, and use boundary are disclosed through the governing scoring or comparator pattern. Without that, keep the result as report-only or return to ModularityVectorLite.
Lowering and currentness conditions
Lower or reopen a ModularityVectorLite, ModularityCharacteristicCard, or report-only proxy when any of these conditions changes the characteristic use:
- proxy audit worsens, such as more integration failures, workarounds, source-return events, stale evidence reuse, or bounded exceptions;
- measurement basis, comparability basis, scoring method, codomain, polarity, unit policy, or declared characteristic basis changes;
- evidence relation, evidence-provenance relation, source relation, evidence-claim-absent reason, or source-return condition changes;
- described holon, architecture question or intended use, ClaimScope or qualification window, architecture claim, structure kind, characteristic head, or repair direction changes;
- a report-only proxy is used to support a stronger comparison, assurance, causal, selection or scale-preference claim; publishing its qualified local meaning does not create that stronger use;
C.31.RSA,C.31.ASAP,C.16,C.25,C.29,C.30.STRAT,A.6.M,C.30,C.30.ASV,A.10,B.3,A.20,A.21,G.5, orC.11changes the boundary for the neighboring claim being made.
Admissible repair results are:
- keep the result report-only;
- split or rename the characteristic head;
- update basis or evidence fields;
- revise the repair direction;
- change
relatedClaimPatternLocatorsIfClaimed; - lower a score to a local proxy; or
- stop the unsupported use and establish the required subject claim under its applicable rule.
Archetypal Grounding
Modularity has distinct action-guiding characteristics. Use them to identify the next repair.
A product architecture can have high interface standardization and still poor substitutability. A software-system architecture can reduce external coupling while increasing hidden data custody. A safety-case architecture can reuse evidence while increasing regulatory bespoke residue. Each case needs a different characteristic and a different repair.
A DSM or dependency graph can substantiate a modularity interpretation, but the graph does not by itself say which dependency kind matters, what scale applies, whether the interpretation is comparable, or what action follows.
Holon and episteme: C.31 characterizes selected module-interface and other architecture-relevant structures of described holons for named architecture questions and uses; the described holon may be an admitted system, organization-as-system, episteme, work occurrence, discipline, or another admitted holon kind.
Publication-family material uses the episteme and publication patterns. A MethodDescription is an episteme; a Method uses A.3.1, and any relation asserted for it uses the pattern that defines that relation.
C.31 heads, cards, vectors, and report-only proxies are characteristic records, declared-measurement-basis records, comparability-basis records, or report-only records about those structures.
Bias-Annotation
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Consequences
Benefits:
- Modularity findings guide the next repair.
- Cheap repair remains possible before measurement.
- Characteristic, declared-measurement-basis, comparability-basis, proxy-risk, and subject-pattern boundaries are visible.
- DSM, MOSA, platform, product-line, and architecture-operation sources can inform practice without importing their ontology wholesale.
Costs:
- Familiar score language often has to be downgraded to report-only use.
- Cross-case comparison needs the applicable measurement and comparison grounds, with quality, source, evidence or decision work when that claim requires it.
- Some "complexity" statements need a characteristic, a lens-backed interpretation, or a residual cue.
Rationale
C.31 is a characterization pattern because modularity and reusable-structure talk changes engineering action through characteristics: coupling, cohesion, interface variation, substitutability, reuse, evidence reuse, hidden coupling, source-return cost, and residual growth. Their use depends on recovering the subject and relevant structural relation, then adding scale, measurement or comparison grounds only for the claims that need them. The account should explain the answer or warranted repair and any action-changing limit.
The small first inquiry preserves affordability. A practitioner may need one relation to repair, one interface grammar to tighten or one residue to account for. These distinctions determine whether measurement could improve the next move; a complete diagnostic answer can stop without a full card.
The pattern rejects a single complexity or modularity score because selected heads are not automatically commensurable. When a local score is useful, it belongs under disclosed scoring, comparator, characteristic, declared-measurement-basis, and subject-pattern discipline.
SoTA-Echoing
Source-use and refresh. A source row identifies a contribution and its limits. A historical method or result can support a use within its established premises; it does not establish the current comparative standing of an approach. When the question depends on present acquisition rules, a changed interface standard or newer comparative evidence, recover that basis through G.2 or the applicable source rule before relying on it. A source citation supplies no missing measurement, conformance test or decision premise.
Reopen only the dependent characteristic, comparison or repair when its source meaning, assumptions or applicable rule changes. For example, a changed substitution specification can alter InterfaceStandardizationShare and SubstitutabilityWidth while leaving an earlier counted population intact. A queueing-model change reopens the dependent bottleneck inference; it does not erase an independently observed dependency. Preserve unchanged source uses and bounded historical results.
Relations
C.31:End
Reusable Structure Accounting
Type: Characterization pattern Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when planning another delivery or reviewing a claim of reuse. Work out which part came from a shared design, procedure or physical article, which work was specific to the delivery, and what the next use requires. A team can use the same procedure twice and still need new measurements each time.
Claim-use boundary: any use that relies on the RSA account to make a claim beyond reusable-structure accounting is outside RSA. Examples include comparison, publication, evidence validity, assurance or safety-case reliance, gate use, architecture scale preference, causal use, selected-set result declaration, candidate synthesis, and local decision.
Record with C.31.RSA only the reusable locus, bespoke residue, accounting basis, report-only share, repair direction, and source-return condition. Add another claim only after naming and applying the pattern that defines and tests it.
Start from the receiving reuse question and an adequate existing account, or recover reuse and residue from relevant instances as in :4.0. Explain any warranted repair or bounded exception. An adequate ordinary account can finish. ReusableStructureTriage can retain these distinctions:
Use a fuller accounting description when the receiving question needs its structure references and accounting rule. Ordinary use stops when the practitioner knows where reusable structure lives, where bespoke residue grows, what can be refactored and what remains a bounded exception. Retaining or sharing that answer need not create another form.
What goes wrong if C.31.RSA is missed:
- a reusable share is treated as a proof of modularity;
- one-off work is hidden under a reuse label;
- evidence reuse is counted without validity context;
- hidden residual uncertainty is averaged with reusable templates; and
- "more reusable structure" is treated as always better.
What C.31.RSA buys in practice: the practitioner can state where structure is reusable, where it is bespoke, what source-side distinctions must remain reachable, and when the result is report-only accounting.
Not this pattern when the question under repair is source-label recovery, module-interface relation repair, modularity-characteristic selection, measurement or comparability admissibility, architecture scale preference, mathematical-lens use, or any outside-RSA use named above.
Use C.30.STRAT, A.6.M, C.31, C.16, A.10, B.3, G.6, C.31.ASAP, C.29, G.5, or C.11 as appropriate; do not treat C.31.RSA as the synthesis or selector pattern.
Problem
Architecture teams often say that structure is reusable, repeated, templated, common, standardized, or bespoke. Those phrases are useful, but they do not say what is being counted, described, or compared.
Structure can be selected from functions, flows, control relations, module interfaces, work methods, evidence packages, regulatory arguments, data schemas, deployment constraints, or exception networks.
Functional, flow, control, module-interface, work, evidence, and assurance structures may be included only under their declared accountingBasisRef. Name any evidence, assurance, or source relation being claimed by value, and state any source-return condition relied on by the account.
The practical question is: which reusable loci matter, which bespoke residue remains, what source distinctions are lost by accounting, and what repair or source return follows?
Forces
Solution
C.31.RSA governs reusable-structure accounting over declared structures and structural aspects. Use an adequate existing account directly. When reuse and residue have not yet been distinguished, recover them from the available instances below. The triage and accounting forms can retain the answer; a quantitative account still needs its declared basis.
Recover reuse and residue from instances
Start with the decision. To establish past use, trace what an identified delivery took from the shared design, procedure, evidence or physical article. Name what was carried over and which changes the delivery team made. Two independently made products can both fit the same specification. That can establish conformity even when the origin of one product is unknown. To identify the design choices a person or team made, compare the instructions, designs, parameter settings and help available with the choices made during the work. A worker can produce the correct adapter from a supplier's drawing or choose and justify its geometry independently. A cutting or assembly record can establish execution in either case. Attribute design of the part to the person or team when the evidence establishes the design choices they made. To assess their ability to justify an option, also establish the reasoning they supplied for it. Work completed before one particular template arrived rules out the use of that template but can still have followed another drawing. To commission a further design, compare the choices it requires and the help that will be available with those the person actually made under comparable conditions. One independently designed flange can support a further flange assignment while leaving design of an entire adapter open.
For a proposed use, compare what the recipient needs with what the shared design or procedure supplies in the intended conditions. Check that the chosen settings work together, including their interfaces and any local adaptation. A supported design rule, qualification or test that already covers this combination can answer the question before the first delivery there. If a necessary condition fails, identify what has to change or which alternative fits. If its value is unknown and it could change the answer, identify what would settle it. Actual use at the new site will have its own production or work history.
For a claim about a whole family, explain why the common rule covers all the admitted combinations. One admitted combination that needs an extra operation defeats a promise of use without that operation throughout the family. Preserve the supported uses elsewhere. A known count of template uses can finish a request for that count.
- Follow the same required result through the instances. Use the available products, configurations, instructions, test and approval material, or actual work records to recover what supplies that result, what it receives, and the conditions of each connection. C.30.ASV helps recover the selected structure; C.32.MWA.PP:4.1 and :4.4 help trace performance and provision when those relations matter. Compare the relation or way of doing.
- Explain the differences that matter. Identify the requirement, input, environment or local arrangement that accounts for each consequential difference. A contract can already settle whether the customer needs an on-site witness. When the reason is unclear, use a contrasting delivery, a work record, the provider's account or a bounded trial that can distinguish the plausible explanations. Preserve an established shared part while investigating a difference that could change its use.
- State the reusable rule and its permitted variation. Identify what can remain the same, which values or steps change, and under which conditions the resulting instance still supplies the required result. A necessary customer-specific approval or local adaptation may remain bespoke. A repeated workaround may instead expose a missing parameter, interface condition or Method step. Actual substitution, evidence validity and assurance claims retain their subject rules; resemblance between instances does not establish those claims.
- Choose the next sufficient result. The shared rule, its permitted changes and the remaining local work may finish the account. When considering a repair under :4.5, trace its consequences through an affected delivery, including who would take on the work moved elsewhere. C.32 can use the proposed change as a candidate. Record a relocation as accomplished when it has taken effect.
- Count only when the question needs a share. Select the population, unit and reuse test from that question, then apply :4.1–:4.4. A count of template uses, a claim of valid evidence reuse and a claim that whole deliveries can share a Method answer different questions. Preserve the conditions and uncertainty of the one counted.
Typed accounting description
accountingBasisRef identifies the accounting rule. The rule may concern description length, dependency edges, work items, evidence package count, cost share, template instances, interface variants, regulatory case sections, or another declared basis. The accounting rule is not implied by the word "reuse".
Well-formedness: every slot is over declared structureRefs, declared structuralAspectRefs, and one declared accounting basis. Slot labels are explanatory; they are not root kinds and are not automatically commensurable.
Explanatory slot labels
A local accounting description may use explanatory slot labels such as:
These labels are local slots, not FPF ontology. H_residual is residual uncertainty or unmodelled variance under the accounting basis. Its value is not automatically commensurable with counts or other measures of interface grammars, work templates, evidence packages, or regulatory arguments.
Report-only shares
Every numeric share needs its accounting rule, counted population and scale or unitless-value interpretation, with a unit when relevant.
A comparison or ranking additionally needs aligned bases, preference direction when relevant, and its applicable comparator, such as CG-Spec, ComparatorSetRef or a governing comparator reference. A gate or decision needs that comparator only if its rule relies on such a comparison; it retains its own other requirements.
A share with an established counting basis can remain a bounded local report when stronger reliance lacks grounds.
Pseudo-sum boundary
An explanatory decomposition may be useful:
This is a decomposition of one declared accounting description, not an equation defining an architecture amount. If the slots do not share a declared accounting basis and comparability rule, they cannot be summed or ranked.
Structure-relocation actions
RSA is useful because it points to relocation and repair actions:
A high reusable-structure share is not always good. The architecture question is where structure lives—for example, in reusable templates, interfaces, flows, control relations, work methods, evidence packages, or unique exception networks and hidden coupling—and what action follows.
After a relocation or reuse move, ask what got worse:
An RSA move identifies the reusable and residual loci, the basis needed for its accounting claim, and the warranted repair or retention. Inspect the possible costs and losses above; state those supported by the case and any condition that could change admissibility. Preserve an unresolved consequence as a question rather than inventing a loss or return condition.
Triage and accounting use boundary
A qualitative account can be sufficient when the practitioner needs to locate reuse, residue and the next warranted repair. Use the triage form if it helps retain that answer.
Develop the accounting rule and structure references when the question depends on a numeric share, comparison of aspects or distinctions hidden by summarization. The fuller description can organize that content. Preserve source-return conditions when they change use. Publication of an adequate account does not require a new form; assurance, scale preference and decisions need the additional grounds of their own claims.
Reopen and lowering conditions
An RSA result retains its declared scope, structure edition and accounting basis, together with source-return and comparator conditions on which its use depends. Reopen the triage or lower the admissible use when any of the following changes:
- a hidden source distinction becomes action-relevant;
- the accounting basis changes or proves heterogeneous;
- the selected structure, structural aspect, interface grammar, evidence package, work method, or assurance argument changes edition;
- a comparator set, CG-Spec, or outside-RSA use is added after a report-only share was recorded;
- downstream reliance uses the RSA result for outside-RSA evidence, assurance, gate, causal-use, scale-preference, or decision work that the RSA note did not admit;
- evidence validity, assurance window, or source-return condition decays;
- recurrence changes a bounded exception's cost, conditions or justification;
- a reuse move improves one locus while worsening interface cost, variation loss, evidence decay, assurance work, source-return cost, or hidden bespoke residue.
Block the dependent comparison or ranking when its required comparator or aligned basis is missing; preserve a valid bounded report. For another gate or decision, recover the missing premise under its own rule.
If the accounting basis is unknown, retain only the qualitative or source-only meaning supported. If reuse and residue cannot be distinguished at one locus, mark that part unresolved while preserving independently established parts.
Archetypal Grounding
Account for structure at declared loci under a declared accounting basis.
In one architecture, reusable structure may be located in a template and interface grammar. In another, it may be located in a test package, regulatory argument, work method, or flow pattern. In a third, the reusable part may be small, but the bounded exception is what preserves safety or local fit.
A share can be useful as a local report. It becomes misleading when it hides which structure was counted, which structure was not counted, and when the reader must return to source records.
Holon and episteme: the structures being accounted over are selected architecture-relevant structures in context. The RSA description describes those structures; it records the local accounting slots, report-only share values, and any source-return condition.
Worked case: reusable evidence package, bespoke delivery work
Situation:
Start with the delivery instructions, completed approval material and integration exceptions. The ordinary variants use the same component interfaces and test procedure within the package's stated operating envelope. One customer's approval adds an on-site witness required by its contract; another delivery repeats an adapter workaround because its input format falls outside the shared interface grammar.
Comparing those cases with their requirements separates three contributions:
- a reusable test procedure within its conditions;
- a necessary customer-specific witness; and
- an unresolved design choice about accommodating the input format.
Where the record does not explain a difference, recover that condition from the provider or governing requirement before deciding whether to remove it.
The next account can therefore retain the common test and approval preparation, preserve the contractual witness, and propose an interface or adapter change for the recurring format exception. Replaying that proposed change through input, integration and test shows whether it preserves the required result and whether it merely relocates work.
For the next delivery, the contract requires an on-site witness. A prior test may cover the unchanged component, while the witness clause still calls for work at this site. If the new adapter changes a signal used in the component test, inspect the qualification boundary. Evidence covering that adapter and signal can be used directly. Otherwise qualify the affected path or choose an already covered configuration. The purchase or acceptance decision can use these findings under its own rule.
ReusableStructureTriage:
Publish the account of shared and local work. The interface change is available for design and testing. Once it is implemented, use the affected deliveries to establish where the adapter work now occurs and what its cost is.
Anti-case: high share hides a bad architecture move
Situation:
For an agreement requiring at least 85 percent of the released screens to use the template, the 17 traced uses suffice. Between 17 and 18 screens used it. An exact percentage, or a decision at a 90 percent threshold, depends on the origin of the remaining screen. Comparing screen appearance could establish conformity with a visual specification. It cannot recover that missing build history.
The same account leaves open whether to fund the template for the next release. Compare the cost of maintaining and adapting it with the cost of developing and maintaining the independent designs for the planned assortment. Include the local work on both alternatives. A common template can save production work while moving more work into integration.
Compare applicable offers, engineering estimates or work records for the planned assortment, including interface maintenance and local adaptations. Use observed costs when reporting work already done. An offer covering the full intended service can settle its quoted price while actual future expenditure is still open. Use A.6.M for a disputed interface interpretation, C.16 for a cost measure and C.11 for the local funding decision.
These grounds can support comparison of the existing variants before any interface is changed. A chosen repair changes the candidate and reopens the affected accounting and comparison.
Keep the release count when the next release changes the planned assortment or the cost comparison. Reopen only the answer that depends on the changed facts.
Transfer case: neural-network block replacement
Situation:
To transfer RSA from product-line architecture to neural-network architecture, use the already recovered reusable locus or first apply C.30.STRAT to treat block, cache, and related terms as source labels. Then RSA names the declared structures and accounting basis:
- trace which earlier topology, dataflow, cache rule or evaluation harness the team used. To credit a new design, establish which choice the team made beyond implementing the supplied architecture;
- check the target's memory layout, input distribution and tuning needs before offering the same change for another model;
- benchmark gain is not reusable-structure accounting by itself;
- establish evidence and causal claims under their direct rules; use
A.10for a needed replayable source-to-use account,G.6for a needed citable provenance path, andC.28for causal-use recovery. UseC.29for an unresolved mathematical representation choice, transfer or reliance question.
Admissible move: record which recovered structural locus was reused, what changed, what source distinctions must remain reachable, and which pattern defines or tests each benchmark, evidence, causal-use, or mathematical-lens claim. Non-admissible move: treat "block replacement improved the architecture" as RSA proof.
Bias-Annotation
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Consequences
Benefits:
- Reusable structure and bespoke residue become visible without a false architecture amount.
- Practitioners get a cheap triage before accounting.
- Report-only shares can guide repair without becoming proof.
- Evidence reuse, work repeatability, interface grammar, and bounded exceptions can be separated instead of averaged.
Costs:
- Some reuse reports remain report-only.
- Numeric shares require an explicit counting and scale basis. Comparison needs aligned bases and its comparator; each further claim adds only its own required grounds.
- The pattern raises a source-return question whenever accounting hides distinctions needed by downstream action.
Rationale
C.31.RSA is separate from C.31 because accounting over reusable structure answers a different question from choosing modularity characteristics. C.31 asks which characteristic changes action. C.31.RSA asks where reusable structure and bespoke residue are located under a declared accounting basis.
C.30 distinguishes an architecture relation from its selected structure. RSA does not define StructureAmount. An accounting description can still report shares, but only under declared structure refs, structural aspects, and an accounting basis.
The pattern also keeps residue ethically and practically neutral until interpreted. Bespoke residue can be a defect, a local necessity, a safety boundary, a regulatory constraint, or a deliberately accepted exception. The repair is to name which one.
SoTA-Echoing
Source use and refresh. Preserve each adopted contribution's scope and assumptions. A dated standard or historical method can support a claim within those conditions; a claim about present requirements or comparative standing needs the applicable current basis. A source row supplies neither conformance nor the evidence-validity, assurance or decision grounds of a reuse.
Reopen the affected RSA use when a changed source or rule alters its reusable locus, residue, accounting basis, source-return condition or relied-on comparison. For example, a new evidence-validity window can limit reuse of a test package without erasing the count of earlier uses. Preserve unaffected historical accounts, and recover new grounds before stronger reliance.
Relations
C.31.RSA:End
Which Architecture Is Preferable Under Scale? (Scale Amenability)
Type: Characterization pattern Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when a practitioner compares architecture alternatives under a declared scale variable or scale window and needs to avoid treating a modularity label, platform label, product-line label, reusable-structure share, or coarse-graining metaphor as an automatic scale-preference claim.
First identify the alternatives, the scale variable and its window, and the grounds and limits of the proposed preference. An adequate ordinary account can finish; ScaleClaimTriage can retain it:
Ordinary use starts by naming:
- the alternatives and described holon;
- the
U.ClaimScopeand relevant A.2.6U.ContextSlicemembership; - the scale variable and scale window;
- the claimed preference under scale;
- the observations or derivation supporting the proposed scale behavior, or what remains open in their absence;
- the structural feature or exception behavior on which that preference depends, when it calls for either.
Name modelUseStructureRef only when one independently selected BoundedModelUseStructure changes this receiving interpretation; it never replaces the claim scope.
Use ArchitectureScaleAuditRecord@Project when those distinctions help retain or examine the preference for its receiving use. Sharing an already adequate account needs no new form.
What goes wrong if C.31.ASAP is missed:
- "modular", "platform", "product line", "reusable", "general", "open", "coarse-grained", or "RG-like" becomes a shortcut for a scale-preference claim;
- a locally hand-engineered solution is called debt even when safety, law-domain, or mission constraints justify it;
- exception growth is hidden until the architecture is already expensive to change; and
- coarse descriptions keep losing lower-scope safety or semantic distinctions without a source-return condition.
A practitioner can use the resulting account to say which architecture is preferable for the planned scale and criterion, which conditions support that judgement, and what could reverse it.
For questions limited to the following concerns, use the corresponding patterns instead of C.31.ASAP:
C.31.ASAP governs architecture scale-preference claims; it does not select the architecture, prove the scale law, or certify the project.
Problem
Architecture work often asks whether one structure will scale better than another: a product-line architecture rather than bespoke variants, an open interface rather than a closed integration, a stable flow topology rather than repeated project exceptions, or a reusable evidence package rather than repeated assurance work.
These claims are useful only after the scale relation is declared. "Scales better" can mean fewer interface grammar variants, lower exception growth, better learning transfer, more reusable templates, more stable control relations, less source-return cost, fewer regulatory exceptions, or more freedom of action. Those are not the same claim.
C.31.ASAP connects a declared scale variable and window to a preference claim, its evidence or no-probe reason, the behavior or exception risk on which it depends, and any needed source return. It does not turn scale preference into proof, selection, assurance, causal use, or a universal law.
Forces
Solution
C.31.ASAP specializes scale-amenability preference for architecture alternatives. It applies when an architecture alternative set, scale variable with its window, and claimed preference under scale are named.
Applicability fields
C.31.ASAP applies only when all of the following are present:
- a declared architecture alternative set, described holon,
U.ClaimScope, and relevant A.2.6U.ContextSlicemembership; - a declared scale variable and its window;
- a claimed preference under scale;
- the observed scale behavior, a derivation from applicable conditions, evidence from a scale probe, or a reason why the claim remains hypothetical;
- the grounds relating the architecture alternatives to that comparison. These may include an applicable quotation for the required service, or differences in structure and local exceptions where those affect the claimed advantage.
If those fields are absent, keep the claim in C.31 as a temporal or scale-sensitive characteristic cue, in C.31.RSA as report-only accounting, in C.29 as a bounded lens-use output, or in ordinary architecture prose.
ScaleClaimTriage
Recover the following content before a heavier scale audit. Reuse an adequate account; the triage form is optional:
A supported preference can finish a comparison. If the evidence still permits different rankings, identify the condition that changes the ordering. An inquiry to resolve it can itself be the deliverable when that inquiry was commissioned.
Include nonAdmissibleUse? only for an independently grounded alternative that is plausible here and whose rejection changes understanding or action; groundedScaleShortcut? is an alias for that optional value.
Retaining or publishing the bounded answer preserves its limits. A stronger comparison, assurance, selected-set or decision claim needs its own grounds.
claimScopeRef designates one U.ClaimScope; selectedContextSliceRefs records the A.2.6 membership relevant to this use. A scale window is the range of the scale variable for which the preference is claimed, not a substitute for either scope object.
modelUseStructureRef is optional and is filled only when an independently selected A.1.1 BoundedModelUseStructure changes the interpretation of this preference use.
Architecture scale-preference rule
When the commission values stable reusable structure, compare how well the alternatives preserve it over the declared scale window, subject to safety, law-domain and assurance requirements. If the question concerns another characteristic, carry the alternatives through the permitted operating conditions and compare them on that characteristic. Select the functional, flow, control, interface, work-template or evidence structure and any learning-transfer characteristic relevant to that comparison.
Define the advantage that would change the recommendation. Compare both alternatives for the same deliveries under the conditions that affect that advantage. Separate a one-time investment from work repeated at each installation or each new interface type. When deciding whether to commission a shared design, count the preparation it requires. When deciding whether to keep using one already built, compare future costs and commitments that the choice can still change. The original expense remains part of the investment history. More deployments at an established site may require none of the qualification work that a first deployment in a new jurisdiction requires. Check where the two alternatives tie or change order. When different admissible mixes of sites give different recommendations, state which mix supports each and what is known about the proposed rollout.
Observations can support a comparison for the conditions examined. A binding rate, an engineering bound or a supported construction can also establish an answer for the entire stated range under its assumptions. Record the observations or the applicable derivation and its assumptions. A quotation or bound that orders the alternatives throughout a window can be cited in comparisonBasisRef even when neither slope is known. An advantage in price throughout a window establishes which quote is cheaper. To claim its price grows more slowly, compare the changes in price. When an untested premise could reverse the preference, state what depends on it. A reason for skipping a probe supplies no evidence about its outcome. An applicable derivation can already settle the question. Conflicting characteristics need the preference or decision rule for the work in question.
C.31.ASAP defines the scale-preference claim and its boundary. Use G.9 for a parity claim and A.21 for a gate claim; the candidate, comparison, set-declaration, and choice uses are distinguished below.
A scale-preference claim may inform C.32 candidate generation or supply one input to an A.19.CPM comparison by naming the scale variable and window, the grounds for ordering the alternatives, and any structural uncertainty or exception that could reverse that order.
Use:
C.32to construct the candidate architecture palette;A.19.CPMto compare alternatives;G.5to declare a selected-set result;C.11to make a final local choice; andC.32.PADto record a project architecture decision.
For audience availability, use E.24.PUB for the publication occurrence, form, carrier, audience and bounded use; use E.17 when a source-backed face and return are needed.
Apply the relevant evidence, assurance, gate, or release definition and test only when that claim is current.
When the same scale-sensitive pressure must also become a project criterion, C.32.ACS creates a separate row for the characteristic or Q-Bundle slot, bearer, scale form, and use class. That row may supply declared input to an ASAP preference. Use C.31.ASAP to state which alternative is preferable under the scale window; use C.32.ACS to admit and classify the criterion row as an optimization indicator, guardrail, or context-only row. Keep the preference record and criterion row separately referenced.
Scale variables
Typical architecture scale variables include:
Name the scale variable, its window and the characteristic whose behavior matters to the comparison. Add a source-return condition when omitted distinctions could change the receiving action. Increasing the number of sites changes a quantity in the model. Abstracting the model can leave the jurisdiction of a site open. The same disciplinary model can compare five and forty sites. Use B.5.MR:4.2 when asking which conclusions still follow after leaving a distinction open.
Scale audit outputs
Use a fuller audit when unresolved behavior, exceptions or comparison grounds could change the receiving use. An existing adequate account can carry the necessary content:
For ArchitectureScaleAuditRecord@Project and BespokeResidueRegister@Project, @Project is a compatibility and retrieval cue.
An audit local to one project names both the composite U.Work in projectWorkOccurrenceRef and the obtaining direct audit-use relation in architectureScaleAuditProjectUseRelationRef. BespokeResidueRegisterRef may cite the register episteme. Assert that register's project locality only after a separate direct register-to-work relation is governed, and cite that occurrence; the audit-use relation remains about the audit. Otherwise retain only retrieval use and make no audit or residue-register project-locality claim.
ArchitectureScaleAuditRecord@Project records the triage of an architecture scale-preference claim. An assurance proof, gate record, selected-set result declaration, publication occurrence, local decision, or work plan remains a separate result under its own pattern; use A.15.2 for the work plan.
Waiver discipline
A deontic constraint, safety boundary, law-domain boundary, mission constraint, assurance infeasibility, or scale-probe overturn can justify a bounded exception without creating ArchitectureHeuristicDebt.
ArchitectureHeuristicDebt remains report-only unless tied to a decision, risk, work, evidence, assurance, or selected-set record through its governing pattern.
Scale-refactoring moves
Name at least one possible repair or stop before using the scale preference to guide action:
C.29 lens relation
A scale preference may use an RG, coarse-graining, epiplexity, graph, multilevel-learning or frustration model. Establish the assumptions and correspondence needed by that inference; an adequate domain treatment can suffice. Use C.29 when choosing or transferring the representation, or relying on its result, leaves an unresolved question. Keep the architecture scale-preference claim in C.31.ASAP and use C.18.1 for a scale-law claim.
The account retains its scale variable, window and the basis for the comparison, including any variation in exceptions that can change the answer. State where to recover an omitted fact when the answer depends on it. When a C.29 account is needed, it can use MLU.Description@RGArchitecture, MLU.Description@MultilevelLearningFrustration or another local account to explain the representation, correspondence, preserved and lost distinctions, payoff and limits. A separately named output is needed only when the receiving use requires it.
Archetypal Grounding
Filled triage slice
The six platform deliveries show which interfaces and exceptions occurred there. Predicting a lower growth rate than the bespoke alternative calls for comparable grounds for that alternative and for the future mix of sites. Identify the change in configuration or jurisdiction that could reverse the preference.
Suppose both suppliers offer the same required service for five to forty sites during the quotation period. The shared platform costs 18 units to prepare the common interface and 2 per ordinary site. The bespoke offer costs 5 per site. At five ordinary sites the quoted totals are 28 and 25. At six they tie. From seven through forty ordinary sites the platform costs less. This follows from the two binding offers for that range. The six installations alone would not establish those prices.
At eight sites the ordinary prices are 34 and 40. Under the same offers one site needing a different connector adds a one-time charge of 12 to the platform total and is included in the bespoke price. The comparison then favors the bespoke offer at 40 against 46. If the eighth site's connector is unknown and the other seven are ordinary, its interface specification settles which offer is less expensive. An assertion about actual expenditure still needs the charges incurred. An architecture choice can also depend on service or assurance conditions beyond price. At twelve sites with at most one such surcharge, the platform quote is either 42 or 54 against 60. Its price advantage then holds with either connector.
Near-miss lowering slice
Near-miss: a team says the platform architecture should win because it has an 82 percent reusable-structure share, an RG-like coarse description, and "less bespoke debt" than the local variant.
Lowering replay:
- Identify the feature of the proposed platform on which the preference depends. For a preference based on exception-handling effort, recover which variants require an exception and how much work it adds.
- The team's reported 82 percent gives a starting point for checking the claim. Recover the counting unit, population and records to establish the share. Even an established share can correspond to 41 of 50 or 82 of 100. It supplies an absolute count only with an adequate denominator and precision. For a scale preference, establish how the counted reuse affects the characteristic compared over the proposed deployments.
- An RG-like phrase alone supplies no mathematical correspondence or scale-preference evidence. Recover the model and its relevant preservation and loss; use C.29 if representation choice, transfer or reliance remains unresolved.
- The "bespoke debt" label is lowered to waiver review when safety, law-domain, mission, assurance, or scale-probe overturn reasons may justify the local variant.
Report the 82 percent as the team's statement until its basis is established. Preserve a verified accounting result within that basis. If compatible rollout plans favor different alternatives, identify the conditions for each preference and what the present decision requires.
Bias-Annotation
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Consequences
Rationale
C.31 and C.31.RSA can expose scale-sensitive characteristics and reusable-structure residue, but they should not themselves decide which architecture alternative is preferable under scale. C.31.ASAP governs this architecture scale-preference claim family; it is narrower than general BLP and broader than one measurement card.
The pattern adapts BLP-style scale-amenability to architecture: when continued reuse is the criterion, prefer an alternative with a demonstrated advantage over the declared scale window under comparable safety, law-domain and assurance conditions. Treat modularity, reuse, platform practice, and mathematical coarse-graining as candidate cues. Base the preference on the named scale variable and window, relevant structural behavior, exception-growth risk and the reach of available evidence. A comparison can follow from applicable rates or constraints throughout a range. A preference that still depends on an untested extrapolation remains provisional.
SoTA-Echoing
Relations
- Builds on:
C.31,C.31.RSA,C.16,A.2.6,A.17,A.18,A.19,C.18.1,C.19.1, andC.29; uses A.1.1 only when a selectedBoundedModelUseStructurechanges the receiving interpretation.
Coordinates with:
Claim boundaries:
- Precision-restoration relation: source wording recovered by
E.10,E.10.ARCH, orC.30.STRATis governed by C.31.ASAP only when the recovered claim being made is architecture scale preference over a declared alternative set, scale variable, and scale window.
C.31.ASAP:End
Architecture Candidate Synthesis
Type: Architectural pattern Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when a practitioner has a C.30-grounded architecture question for one described holon and needs to synthesize several candidate architecture configurations across selected structures before comparison, archive or front-policy work, selected-set result declaration, publication, or decision. Keep any obtaining C.30 ArchitectureRelation occurrences, the selected U.Structure values they relate to the holon, and any candidate, required, desired, or expected structures named only in an ArchitectureClaim distinct throughout the synthesis.
Primary working reader: an architect or architecture-responsible practitioner preparing alternatives for one described holon before comparison, selection, selected-set result declaration, publication, local choice, or project decision.
Typical entry phrases:
First-minute use slice. A regulated product-family team has a C.30-grounded architecture question for one field-device-family holon. The question names its current obtaining ArchitectureRelation occurrences and their selected structures separately from candidate or expected structures stated only in the current ArchitectureClaim.
The work question is synthesis: how should required functions, constructive modules, field placement, control responsibility, and certification evidence be coordinated so maintainability, substitutability, latency, and evidence reuse stay acceptable?
Follow one field replacement through removal, fitting and acceptance. Keeping a common module can save the field team a replacement procedure if the common qualification covers the installed interfaces. Splitting the family can simplify each interface while requiring two sets of spares and evidence. Keeping the existing split and securing qualified local approval can suffice for an exceptional site. These are different configurations because they change what must fit, who can accept the replacement or what the field team must stock.
The team can describe these candidates before any field unit has been changed. To decide which to carry forward, establish which sites each can serve and compare the work left to the field team.
The primary EntityOfConcern is the candidate architecture palette for one C.30-grounded synthesis question. Its inputs are the described holon, any obtaining ArchitectureRelation occurrences and their selected U.Structure participants, and any candidate, required, desired, or expected structures stated only in an ArchitectureClaim.
The described holon may be a system, product family, organization-as-system, discipline, AI-agent setup, built asset, episteme, Work occurrence, or another admitted holon kind. Do not admit a source label as a holon.
For example, practice, culture, tradition, style, Method, or role may refer to a Method, Method relation structure, relation among local system-role kinds, classification, assignment, Work structure, episteme, source-local meaning, or C.36 cultural-evolution relation. Recover the object and claim through its subject pattern; route unresolved claim-bearing role wording through E.10.ROLE.
ClaimScope and a bounded model-use structure qualify the named use; neither becomes the holon. Architecture pressure may concern Method-family structures, relations among local kinds, classifications, or assignments. Keep each as a selected structure or separate input for the named architecture use, not as a holon kind or function bearer by label. C.32 is not software-system architecture by default; software-system sources are one source family and one domain example.
What goes wrong if C.32 is missed: the team optimizes one visible structure, such as modules, placement, team responsibility, control relation, or evidence package, and then treats that local improvement as architecture synthesis. The competing structures, architecture characteristics, losses, and alternatives disappear before they can be compared.
What C.32 buys in practice: a practitioner can build a small set of candidate architecture configurations, each grounded in selected structure changes, architecture characteristics, known losses, and patterns for the next questions.
Ordinary working move: name the selected structures and architecture characteristics that make the question consequential, then construct the structurally different configurations needed by the comparison, with expected gains, relevant preservation or loss, and next receiving use. A supported account that already settles the question can finish without constructing another option.
A completed comparative palette makes at least two structurally different configurations recoverable, with their selected-structure changes, relevant characteristics, expected gains, known losses and next use. A single configuration can be the starting point for that construction; it is not yet the comparative palette. Preserve a source-return condition when omitted detail can change the receiving use.
Use C.32 only for candidate palette construction. Do not use it to ground the architecture claim, recover one structure, build characteristic criteria rows, design eval programs, handle architecture-influence correspondence, run archive or front-policy work, declare a selected-set result, publish it to an audience, choose locally, or decide the project architecture.
Use C.32.MWA instead when several structures of Methods, Work, subjects and their descriptions, capabilities and providers, and cultural change do not line up one-for-one and the needed result is one usable practice architecture. Keep C.32 for a palette of candidate configurations for one grounded architecture question.
For control alternatives whose observations, decisions and interventions must work together under information, time, resource and authority constraints, use C.32.LCA to construct and compare that connection. Bring its explained alternatives into the wider palette when other selected structures also change. C.30.LCA addresses the separate question of which control relations a description represents and what use that description or view supports.
Common exits by claim kind:
C.30grounds the described holon, any obtainingArchitectureRelation, its selectedU.Structure, and any separateArchitectureClaim;C.30.ASV,A.6.F, andA.6.Mrecover structural views, function wording, and module-interface relations.C.32.HCS,C.32.ACS,C.32.ACE,C.25,C.31,C.31.ASAP, andC.16govern starter heads, project criteria rows, eval programs, Q-Bundles, modularity or scale-preference claims, and measurement.C.32.MLAO,C.32.CONWAY,C.32.FAIL, andC.29govern residual-reducing frames, architecture-influence and transformed-architecture correspondence, candidate repair, and mathematical-lens use.- Use
A.19.CPMfor comparison,A.19.SelectorMechanismfor set-returning selection,C.18andC.19for archive, front, and current-pool treatment,G.5for selected-set result declaration,C.11for local choice, andC.32.PADfor a project architecture decision. For audience availability, useE.24.PUBfor the publication occurrence, form, carrier and bounded use, withE.17when a source-backed face and return are needed. C.30.AD,E.17,E.24.PUB,A.10, andB.3govern architecture-description, publication-face, publication-occurrence, evidence, and assurance claims, respectively.
The first useful output is the small candidate palette. An adequate ordinary account can carry it; CandidateArchitecturePalette@Project is an optional working record. Its name introduces no new U.* kind, and the record does not establish selection, publication, evidence, assurance or decision authority.
For a first pass, identify:
- the described holon, synthesis question and intended palette use;
- relevant actual and proposed structures and what they contribute;
- the architecture characteristics under pressure;
- the configurations; and
- the sufficient stop.
Add ClaimScope or a bounded model-use structure only when it changes synthesis; add optional references only when they change the next use:
Across C.32, @Project is a compatibility and retrieval cue, not a project kind or relation assertion. CandidateArchitecturePalette@Project, ArchitectureSynthesisFrame@Project, and ArchitectureCharacteristicImprovementLoop@Project establish no composite project work, context, authority, viewpoint, or parthood by name.
When one of these records is local to one project, identify the composite U.Work and the direct relation by which synthesis framing, palette construction, or improvement feedback concerns that work. Otherwise no project-work relation is implied.
A cited ArchitectureQuestionCard@Project transfers neither project locality nor architecture truth: each affirmative currentArchitectureRelationRef must already resolve to one obtaining C.30 occurrence, while candidate, required, desired, or expected structure remains claim content until the C.30 predicate is independently satisfied.
Problem
Architecture synthesis is the constructive middle of architecture work. A practitioner may already know the described holon, architecture question and intended use, some obtaining architecture relations and selected structures, and some concerns, but still need to configure those structures together into alternatives for later comparison or decision.
The typical synthesis problem is multi-structure. State each required function or functioning claim through the predicate and bearer recovered with A.6.F. Candidate module, placement, control, transformation-flow, information, evidence, Method, Work, local-kind, classification, or assignment structures may constrain or help explain a candidate, but a kind or assignment establishes no functioning, participation, capability, function bearing, or Work.
A control relation can improve supervision while increasing timing or responsibility burden; an information structure can improve maintenance access when exposed through a digital-twin view while still hiding source-return loss; a team structure can improve flow while failing to match module or deployment structure.
Every positive responsibility claim still needs its direct domain predicate, actual participants, applicability, and occurrence identity, or the A.6.RCD missing governor.
A functional description contributes only part of the synthesis. Recover the function or functioning claim and possible bearer, using A.6.F when that meaning remains unresolved. Module, Method, resource, placement, control, Work, evidence and assignment conditions can constrain its feasibility.
If no bearer can satisfy the function under the applicable constraints, mark the candidate infeasible: repair or reject it before treating it as a feasible option for choice. It can remain an explicit countercase for comparison or diagnosis. Missing feasibility evidence instead leaves the claim unresolved. Neighboring structures constrain the candidate; their labels do not establish function bearing.
The typical synthesis problem is also multi-characteristic. Architecture characteristics such as cohesion, coupling, substitutability, evidence reuse, work repeatability, latency, locality, control separation, source-return cost, and composite quality families often compete. Functional demands describe what the holon is to do; architecture characteristics describe whether the selected structures make those demands maintainable, controllable, evolvable, replaceable, inspectable, and otherwise acceptable in the current context.
One recurring candidate-generation heuristic is idealization: ask whether an existing bearer or resource can carry an additional required function under the selected-structure constraints, whether a support bearer can disappear, or whether a more general scale-amenable bearer can replace several special bearers. Admit that heuristic only as a candidate. The candidate must name the functions transferred to a bearer, the bearer removed or generalized, the architecture characteristics improved and worsened, and any BLP scale window or waiver when scale advantage is claimed.
Use C.32 to make the constructive translation explicit. Build a small palette whose candidates answer: which selected structures are configured together, which architecture characteristics improve or worsen, which constraints remain admissible, what source detail must remain recoverable, and which pattern supplies the next claim or test.
Forces
Solution
Use an adequate existing candidate account directly. To start from one current or proposed configuration, trace a required result or constraint through the selected structures: which bearer and relation supply it, and where does the arrangement fail, impose a material burden or leave feasibility unknown? Use the applicable change in the table below to construct a variant, then follow its changed inputs and connections through to the receiving result. C.32.MWA.PP supplies this tracing and construction for performance and provision; C.32.LCA supplies it for control, and C.32.CONWAY:4.1.1 for a constraining production or organization arrangement. Keep the incumbent when it is adequate, including as a comparison candidate.
Construct each alternative far enough to explain how it could deliver the same required result. Identify what changes hands, who or what performs the changed operation, and which other work depends on its output. Two drawings of the same connections can support different inquiries while describing one arrangement. A different assignment, interface or order of work counts as an architectural alternative when it changes an operative connection. When assessing an architect's work, identify which connections the brief already specified and which ones the architect constructed and justified. Selecting from supplied designs can answer a selection assignment.
The palette retains the configurations needed by the receiving question. ArchitectureSynthesisFrame@Project can expose their connections and CandidateArchitecturePalette@Project can retain the result. Selection, publication and decision remain separately governed uses.
Work in seven steps:
-
Anchor the palette to one described holon or holon family, synthesis question, and intended next use. Name any current C.30 architecture relations and selected structures that can change that question.
-
Explain the contributions of the smallest useful set of selected structures. Start with the recoverable functional demand and candidate bearer, using
A.6.Ffor unresolved function meaning, and constructive module/manufacture or placement/deployment structures when they shape the question. Add control, flow, Method, Work, local-kind or classification, assignment, information, evidence or scale structures only when they change it. Send unresolved claim-bearing role wording throughE.10.ROLE.For each candidate treated as feasible, establish which bearers can supply the required functions. A design that fixes an arrangement still has to meet the conditions of its commission. Distinguish a demonstration in one allowed mode from a guarantee over operating conditions or choices that remain open. For a guarantee, establish that the arrangement or a timely way of adapting it covers the required combinations, using the information and authority actually available. Retain unknown or failed feasibility for investigation or repair.
-
State the characteristics and any Q-Bundle slots that define the trade-off, reusing existing criteria rows when available. Use C.32.ACS when a criteria-row construction is needed. Keep functional demand, characteristic, criterion, evaluation result and decision distinct.
-
Generate candidate architecture configurations. A candidate claim may propose, for example, changed decomposition, allocation, A.6.F function bearing, bearer count, placement, interface grammar, a control or transformation-flow relation, Method use, future assignment conditions, an independently established responsibility relation, evidence scope, information structure, or a bounded exception.
Modal candidate wording creates no assignment occurrence and proves no Work occurred. Use a WorkPlan, policy, commitment, permission, decision, or other truthful prospective object when one applies. For actual precise Work, recover each actual performer System through A.13 and let A.15.1 independently admit the dated Work and enacted Method; add an assignment occurrence, its declared species, and F.6 only when the candidate account or its receiving use expressly consumes precise assignment-bound attribution through the same obtaining A.13 assignment.
F.6 identifies neither assignment nor performer, missing or failed F.6 attribution alone leaves the Work intact while its independent A.13/A.15.1 grounds still hold, and an assignment never carries responsibility by itself.
-
For each candidate, state its selected-structure changes, expected gains, known losses, constraint fit and preserved or hidden structure. Identify material unknowns rather than inventing a gain or loss. Add a source-return condition when omitted distinctions affect the next use.
Identify the event the recipient needs. The work can finish, be accepted and become available to its user at different times. Follow each alternative to that event. A later distribution batch can deliver two accepted versions at the same time. Keep the operating conditions common to the alternatives, including shared resources and any binding acceptance criterion. A gain on one operation can trade against a loss on another. Compare the candidates under each operating condition that the evidence still allows. When the same choice holds throughout, carry it forward with that range of use. If the choice reverses, locate the condition that distinguishes the alternatives. A preference among the candidates examined extends to all eligible configurations only when the comparison covers them or an applicable bound rules the others out.
-
When a front, archive, search result, or pool-treatment policy is being used, cite
C.18,C.19, or NQD and OEE support as generation or retention support only. Keep the C.32 candidate content separate from archive work, front membership, pool treatment, selected-set result declaration, publication, and local choice. -
Stop when the palette supplies the candidate content needed by the receiving question, such as comparison, archive use, selected-set declaration, publication, choice or repair. That content need not perform the receiving operation in advance.
Together these contributions explain how the proposed relations make a configuration possible and what they constrain. Ordinary prose or rows can carry that account. Cite a defining rule through relationFunctionClaimRef only when the interpretation or receiving use depends on its identity; a pattern citation normally suffices to locate the rule. Name an evolution window when the expected use or comparison assumptions depend on that range of change.
Architecture-characteristic improvement loop. A completed palette can finish the current question. Reopen synthesis when changed characteristic pressure or feedback warrants another configuration. The practitioner uses characteristic pressure or criteria from C.32.ACS, C.31, C.25, C.16, C.16.P, C.31.ASAP or a local Q-Bundle to construct selected-structure changes, stating which characteristics should improve and which protected conditions may worsen.
ArchitectureCharacteristicImprovementLoop@Project is a local feedback record for reopening C.32 synthesis when characteristic pressure changes.
Keep each receiving claim with its subject pattern:
- Criteria rows stay with
C.32.ACS; - Q-Bundles with
C.25; - scale preference with
C.31.ASAP; - measurement with
C.16; and - eval programs and eval results with
C.32.ACE.
Improvement-question framing and repeated-improvement method stay with E.22 or E.23.
Use A.19.CPM for comparison, A.19.SelectorMechanism for set-returning selection, G.5 for selected-set result declaration, C.11 for local choice, and C.32.PAD for a project architecture decision. For publication, use E.24.PUB for occurrence and audience availability, with E.17 when a source-backed face and return are needed.
For this loop, bring only the changed characteristic pressure into C.32 and return the next candidate palette. Open another synthesis question only when an evaluation result, front relation, retained alternative, rejected candidate or source-return trigger changes what must be configured.
An evaluation result that cohesion improved, evidence reuse decayed, coupling changed, latency worsened or exceptions grew does not choose an architecture. To use it as feedback, recover the bearer, characteristic or criterion, scale or qualitative reading frame, affected structures and the inference it supports. A parity frame is needed when that inference relies on parity.
Candidate architecture changes are local C.32 entries for candidate configurations. They are not FPF work occurrences, method steps, or receiving-pattern claims. A change is admissible only when the selected structure being changed is named.
Didactic mini-slices. Use these as examples of the kind of work C.32 expects, not as domain-specific templates.
When one independently typed architecture-side or other source constrains transformed-side architecture content for a changed referent, use C.32.CONWAY before using Conway, mirroring, or inverse-Conway language in candidate synthesis. Use :4.1.1 there to obtain an unexplained constraint and candidate change from available work, test and approval material.
The practitioner names:
- the changed referent and any actual A.3.4 transformation separately;
- each influence source by kind and its direct relation only when that occurrence is asserted; and
- for each actual architecture side, the C.30 described holon, obtaining
ArchitectureRelation, and selectedU.Structure.
Modal architecture content stays in an ArchitectureClaim.
Without an admitted and satisfied direct influence predicate, the pressure stays synthesis-local in the C.32.CONWAY frame with its missing-governor, unresolved-grounding, or false-predicate disposition and no pair row.
Candidate work then names influence-source-side, transformed-side, joint, or bounded-mismatch changes, architecture characteristics under pressure, expected gains, known losses, and any source-return conditions needed by the use.
Keep the candidate palette as the C.32 result. C.32.CONWAY carries the architecture-influence correspondence frame or one reusable pair-row episteme.
Influence alone supplies no acting System, local system-role kind, System-classification judgment, assignment, Work, changed-referent identity, or transformation participation. Transformation, acting and Work attribution, exact influence, transformation-flow, and module-interface claims belong to A.3.4, A.12, A.2.1, A.15.1, F.6, the direct influence pattern, E.18, C.30.TFS-REL, or A.6.M when current.
Establish preservation and correspondence under their applicable subject rules, using C.34 for adequacy questions. Use C.29 when mathematical representation choice, transfer or reliance is unresolved.
A richer dossier is optional. Open it when source views, relation notes, measurements, C.29 lens outputs, evidence or failure repairs affect the next architecture use. Ordinary use needs one recoverable account per configuration; a row is one convenient form.
Downstream use. The C.32 result is architecture-specific candidate content. Use G.5 to declare a selected-set result, C.11 for a fixed local choice, and C.32.PAD for a project architecture decision. Use C.18 or C.19 for archive, front, pool-treatment, or generation policy when that claim is current. Use C.30.AD for architecture-description work. For publication, use E.24.PUB for occurrence and audience availability, with E.17 when a source-backed face and return are needed.
Stop condition. Stop when the palette supplies the selected structures, proposed changes, gains, losses, constraint fit and relevant limits needed for the next use. Retain source return where it protects that use.
Lowering and return. A single visible configuration can be the starting point for synthesis; use :4 to construct another configuration only when the question needs it. Recover a missing architecture claim through C.30 or candidate content through its available source or description. Repair an unqualified feasibility, gain or loss under C.32.FAIL, retaining the supported content and identifying what remains unknown.
When an adequate palette or retained configuration answers the synthesis question, stop; continue to comparison, selection, publication, choice, decision, evidence or assurance only for the receiving question that remains.
Reopen C.32 when a criteria row, eval result, retained alternative, front relation, source-return trigger or source-currentness change alters the structures under pressure or acceptable loss.
Archetypal Grounding
Choose an arrangement for an accepted document
A team must deliver a document after an independent review. Under one design the author drafts for 50 minutes and the reviewer needs 20 minutes. A shared template changes those times to 35 and 30 minutes. Both designs preserve the same review of the finished document. For this comparison assume those durations, a common start, an immediately available author, review after drafting and no rework. The reviewer has time to finish once available, is authorized to accept the document and accepts it when the review ends. In this agreement, delivery occurs when the accepted document becomes available to the reader. Acceptance gives the reader immediate access.
If the reviewer is ready for either handover, the template delivers at minute 65 and the existing arrangement at minute 70. If the reviewer first becomes available at minute 60, both drafts wait. The existing arrangement delivers at minute 80 and the template at minute 90. The template saves five minutes of combined effort in both comparisons. The earliest acceptable delivery changes with the appointment. Even an inexact appointment can settle the choice. Availability by minute 39 puts the template ahead throughout that range.
A third design could qualify common material once and review its use in each finished document. To compare it, identify the common checks, the authority that accepts their reuse and the local checks each document still needs. A new heading in the template leaves this work in place unless a check or its input actually changes.
These times are consequences of the stated designs. Suppose instead that accepted documents are delivered together at minute 100. Both designs then meet the same delivery time under either appointment. If the recipient chooses the lower effort at equal delivery, the template can be selected without first resolving the appointment. To establish how long a particular release took, use its actual start and the event promised to the recipient. Use the delivery terms to identify the event and its evidence. A portal access record can establish when the document became available under this agreement. If delivery is complete when a named recipient receives the document, establish when it reached them. If the agreement also requires their acknowledgement to complete delivery, use the time of that act. A receipt at minute 100 followed by a required acknowledgement at minute 120 completes delivery at minute 120. To establish what passed review, use the applicable criterion and the review result for that document. A comparison across releases also depends on what they had to deliver, reviewer availability and any rework. A shorter observed release can raise a question about the design without identifying what caused the difference.
Bias-Annotation
Scope: Limited to constructing a small candidate architecture palette for one C.30-grounded architecture question about one described holon or holon family. C.32 is not a universal architecting Method, a software-architecture default, a comparison or selection rule, a publication route, or a decision procedure.
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Architecture trade-off failures
More repair cues
Consequences
Rationale
Architecture practice needs a method between a grounded architecture question and an architecture decision. Use C.30 to ground the question over selected structures of a described holon. Use C.30.ASV, A.6.F, A.6.M, C.30.LCA, C.30.TFS-REL, C.25, and C.31 to recover the structures and characteristics. Later, use C.18 or C.19 for front, archive, or pool treatment, G.5 for selected-set result declaration, E.17 and E.24.PUB for their distinct publication jobs, C.11 for local choice, and the applicable decision pattern for a project decision.
Use C.32 for the constructive middle: building a small set of candidate architecture configurations whose selected structures, allocations, characteristic trade-offs, known losses, source-return conditions, and patterns for the next questions are explicit.
The same middle repeats during improvement. A later criteria-row change, scale-row change, C.16 reading, C.25 or C.31 pressure change, C.31.ASAP scale-preference change, or C.18 or C.19 front, archive, or retained-alternative relation can reopen C.32 when it changes the architecture-characteristic pressure, the selected structures under stress, or the acceptable loss profile. The practitioner then synthesizes another candidate palette; the trigger does not decide the architecture.
The work is to make architecture content available for a later front, comparison, selected-set result declaration, publication, or decision.
SoTA-Echoing
These rows show how source practice contributes to C.32. The opening of each second-column entry classifies the source use; the opening of each transfer states its disposition. The blocked-overread column gives the use limit, and the source-currentness boundary below gives the reopen rule. Software-system sources are comparison inputs, examples, or lineage only; they do not narrow C.32 to IT architecture.
Source-currentness boundary. Use each source row only for the C.32 candidate-generation move that the row transfers. Recheck the row before using it again when:
- a named standard, guide, book edition, survey, or research line changes that move; or
- a receiving FPF pattern named in the row changes how it handles the source family.
If the project needs comparison, selection, selected-set result declaration, publication, local choice, decision, evidence, or assurance, leave C.32 and open the pattern for the next question.
Rows named as lineage, such as TRIZ ideality, information hiding, or mature DSM lineage, stay lineage until a current source relation is recovered.
Relations
- Builds on:
C.30for the described holon, obtainingArchitectureRelationoccurrences, their selectedU.Structureparticipants, and separately identifiedArchitectureClaimcontent;C.30.P,C.30.ASV,A.22,A.6.F,A.6.M,C.32.HCS,C.32.ACS,C.32.ACE,C.25,C.31,C.31.ASAP,C.16,C.16.P,E.22,E.23,C.19.1,C.30.LCA,C.30.TFS-REL,E.18,A.3.4,A.15, and local patterns for recovering source-side architecture referents.
Uses:
Patterns for the next questions:
- P2S docking:
C.32.P2Suses C.32 for the candidate-synthesis stages after problem pressure, selected structures, architecture characteristics, and structural uncertainty have been recovered; C.32 continues to define the candidate palette. - Routes to:
C.32.MWAwhen one usable practice-architecture answer must be synthesized from several structures that do not line up one-for-one; C.32 retains general candidate-palette construction. - Boundary: Use C.32 to construct a candidate architecture palette for one grounded architecture question over selected structures of a described holon. C.35 may support C.32 by recovering a generated or discovered result's kind and assessing its adequacy for candidate-palette use, but C.35 does not select candidates, publish sets, or decide the project architecture. Evidence, assurance, gate, release, work authorization, Method rules, ethical mediation, and causal claims use their own patterns when those claims are being made.
C.32:End
Construct and Compare a Control Architecture
Type: Method pattern in the C.32 architecture-synthesis family Status: Stable Normativity: Guidance within the stated applicability
Problem frame
Use this pattern when a result depends on decisions and actions that use different information, take different amounts of time, or belong to participants with different authority, and you need to construct or change how those contributions work together.
A robot can follow a target accurately and still receive a target that exceeds its actuator limit. A production team can respond quickly and still wait for a decision that nobody present may make. In both situations, improving one participant is insufficient: the arrangement must connect what can be observed, decided and done in time.
First useful move. Choose one consequential demand or disturbance. Follow it from the information available at the time, through the decision and permitted action, to the receiving result. Check whether the arrangement supplies the conditions along that path. A missing condition directs construction of an alternative; an adequate path can be retained while examining another consequential case or a justified improvement. Follow each proposed change through the same whole.
What this buys. The reader obtains comparable control arrangements, the mechanism by which each could obtain the intended result, their conditions and costs, and a reason to retain, select or revise an arrangement. A result can be a simpler incumbent, a bounded proposal, a qualified technical construction, or a finding that the requested outcome is unavailable under the supplied conditions.
The broad subject is functional organization: how contributions combine to obtain a whole result. This pattern develops its control-architecture branch. It constructs the organization of observation, decision and intervention, including coordination needed between controllers or participants. It does not presume a hierarchy, three layers, different update rates, or a correspondence between one function and one physical part. Layered and multirate control are important alternatives within this question.
The primary working reader is an architect, engineer or operations practitioner who can explain the relevant work and obtain the subject knowledge needed for its consequential claims. The common method requires reasoning about dependencies, time and conditions. A numerical stability or safety claim additionally requires the appropriate control-engineering competence and model.
Ordinary boundaries.
- If an available controller or working arrangement already answers the question, use it. A new architecture exercise needs a consequential reason.
- If the question is what control structure a description represents, use C.30.LCA. Its result can supply the present arrangement here.
- If the difficulty is functional construction without a control dependence, use C.32 for the candidate-configuration question or C.39 for a missing way to obtain the required contribution. Return here when the observation, decision and intervention dependence is the unresolved part.
- If the task is to establish the purpose or legitimate priority itself, obtain that decision from its governing practice and participants. Here the question is how that decision affects feasible control and when feedback requires its reconsideration.
- If a project needs a controller, estimator or safety proof, construct or obtain that technical result. The common architecture method below identifies and uses its necessary contribution; it does not replace the subject mathematics.
A proposed arrangement and an operating arrangement have different evidence. The explanation can justify a candidate before implementation. Claims about its actual effectiveness require observations of its use under the conditions that matter.
Problem
Separately reasonable choices can be incompatible when connected. A planner assumes immediate execution; an executor receives a stale state; two units allocate the same scarce resource; a requested exception requires authority outside the decision window. Each local description can be correct while the intended result remains unobtainable.
A familiar layered drawing does not resolve these dependencies. Nor does naming “strategy, planning and execution”: the designer still needs to determine which choices belong together, what information crosses a boundary, how long each contribution may take, and what remains possible when a premise fails.
The practical problem is to construct and compare those relations without losing the purpose, physical or organizational constraints, or the costs introduced by separation. The result must explain both how the arrangement can continue and which feedback calls for a different action, goal or architecture.
Forces
Solution
Construct the arrangement from the result and its causal dependencies, then compare the complete alternatives. The following sections explain connected operations; they do not prescribe a stack of controllers. Begin at the first missing contribution; retain an adequate model, decision or component result.
The architectural result changes the organization through which control occurs. C.43 supplies the broader connection between the current situation, possible continuation, choice, realized influence and revision; use it directly when those answers can be joined within an adequate existing arrangement. The present Method constructs an arrangement when that joining requires different information, timing, allocation, means or interaction. Its result then returns to the ongoing control question.
Recover the controlled situation and the consequence that matters
State what is meant to change, for whom, by when, and within which protected conditions. “Make the system faster” leaves the decision unresolved. “Reach the receiving position within two updates without exceeding the actuator limit” identifies a result and a constraint. “Have the approved panel installed by 16:00 without spending an unauthorized budget” identifies a different result and different conditions.
Identify the acting participants, the subject affected, the available interventions, and the observations that can inform their selection. Recover what an existing arrangement does from an actual consequential episode or an adequate model. For a proposed arrangement, explain the intended dependence without describing it as already operating.
For a proposed feedback arrangement, specify which later intervention can depend on an observation returned from the controlled system, and how the intervention can affect that system. For an operating arrangement, establish when that dependence was in use. B.2.5 develops the supervisor–subholon claim when an acting system supervises identified holons. It recovers the observation, the influence returned to each holon and the facts coupling that return to its observation. A dashboard can supply a useful reading before anyone has decided how to act on it.
Keep the selected structure and its representation distinct. C.30.LCA helps recover the observation, command, reference, feedback and timing relations represented in a view. Use those relations as inputs, then examine their consequences below. An organizational chart may omit an external supplier who controls a needed release; including that relationship does not make the supplier part of the enterprise.
Separate the goal, its realization and the right to revise it
Distinguish the intended outcome, the constraints that remain binding, and the choices available within them. A destination, an allowed operating region, a customer commitment and a preferred cost are different inputs to a decision. Some are negotiable by the present participant; others must be retained or returned to someone with the relevant authority.
Three changes often require different treatment:
- Change the immediate action while preserving the goal. Correct a disturbance, choose another route, reschedule an available resource or use an already qualified fallback.
- Change an operational goal within the current mandate. Select the next job, change a waypoint or choose among already authorized service priorities. Recompute the consequences for the work that depends on the earlier choice.
- Change the mandate, promised result or architecture. Renegotiate the deadline, remove a required output, obtain a new allocation right or change the available means. Include the time and conditions needed to make that change effective.
These are choices, not necessarily three organizational levels or three people. One actor can perform several of them; different actors can participate in one choice.
Construct the return from infeasibility to goal revision. Identify the observation or forecast that makes the present goal doubtful, the alternative outcomes still available, and who can select among them. Include the receiving work affected by that revision. A planner's failure to find a plan is evidence about that search under its conditions; it is not, by itself, proof that no feasible plan exists. A physical bound or an exhaustive argument can establish a stronger impossibility result.
While the revised decision is unavailable, specify the continuation that remains permitted. Continuing, holding a command, reducing demand, switching to another qualified method or stopping can each be right under different subject conditions. Stopping is not automatically safe: an aircraft, a chemical process and an unfinished organizational obligation have different consequences.
Construct the information that makes the decision possible
Begin with the future distinction the decision needs. To decide whether work can finish today, a count of unfinished jobs may be insufficient if their remaining effort differs. To constrain a moving system, the last reported position may be insufficient if motion since that observation is unknown.
Specify the state, transition account and future input assumptions used by the prediction. Retain pending actions, occupied resources or accepted commitments when their differences change the continuation. Separate the subject's state from the information available about it.
Use A.3.3.PI when the retained information needs construction or repair. Compare two admissible situations that the proposed description merges, under the same future inputs and change rule. If their decision-relevant continuations differ, preserve the missing distinction, use sufficient history, or propagate a set or distribution of possibilities. A bound can settle the decision without identifying the exact state.
Date an observation by when it describes the subject, not only when its message arrived. Bring it to the decision time using the known intervening actions and dynamics, or retain the uncertainty those missing facts create. If a model requires information that cannot arrive in time, compare a different observation arrangement, a more conservative action, another model or a different decision allocation.
Construct the forecast for the use. A one-step predictor repeatedly applied, a direct longer-horizon predictor and an upper bound can support different conclusions. The model used for planning may differ from the model needed to qualify rapid correction, but the correspondence between their inputs, outputs and retained distinctions must be explained.
Internal simulation and actual intervention supply different returns. A simulated failure can change the proposed action before execution. An observed result after execution updates the state estimate and may challenge the model. Do not count repeated replanning as progress in the controlled subject.
Derive a division of decisions from the whole dependence
Start with a feasible incumbent or a whole-problem account, not a fixed number of layers. Ask which decisions need which observations, available actions, resources and time. Group decisions when their strong interaction makes separate choice costly or unreliable. Separate them when different information or response demands make one undivided computation or authorization path impractical.
For every proposed separation, recover the dependence it removes from local consideration. Suppose two devices share a power supply limited to 10 kW. Each local plan can meet its device's 8 kW demand while their simultaneous demands total 16 kW. Giving both plans a copy of the resource limit does not preserve the joint constraint.
Where the work permits different allocations or times, possible constructions include a shared allocation decision, fixed compatible reservations, a request-and-grant interaction, or staggered work. They have different waiting, utilization and change costs.
If both devices must draw 8 kW simultaneously, no allocation fits the 10 kW supply; change the available supply, the demand or the simultaneity requirement.
The same question applies to planning and tracking. A planner that ignores actuator effort or response delay can select references that its tracker cannot realize. The joining contribution must restrict the references, supply a feasible trajectory, change the tracking arrangement, or revise the demand. Merely adding a feedback arrow leaves that work undone.
Recover functional contributions before assigning their bearers. Estimating state, choosing among goals, constructing a trajectory or work plan, correcting a deviation and handling an exception can each contribute without requiring a separate module, department or person. State the effect each must obtain, the inputs and initial conditions it needs, and the conditions it must preserve.
Then associate each contribution with a possible System or participant and the means through which that effect could be obtained. One bearer can carry several contributions; several bearers can cooperate in one contribution. Keep existence, capability, access and proposed allocation as separate claims.
Construct the cooperation in both directions. Work back from the required intervention to the decision, information and means that make it possible. Follow each missing intermediate result to a possible producer.
Then follow a consequential demand forward from the available observations and initial conditions: what changes, what becomes usable next, and how those changes obtain the receiving effect. Explain where contributions must coincide, wait or retain a result between updates. A state estimate that expires before authorization, or a command whose execution has no available bearer, defeats that candidate even if every named function is sensible.
C.32.MWA.PP:4.3–4.7 supplies the general construction across Method, Work, performer, means and allocation structures, including their joint demands. Use that account where the proposed division leaves performance or support unresolved.
For control, connect the source of the reference or input to the decision, the permitted influence and the means by which it reaches the subject. For a proposed feedback path, also trace how an observation returned from the subject can affect the next decision. Change the division or its realization where it leaves a needed connection open.
For each candidate, explain how a local result becomes usable by the next participant and how a consequential limitation returns. The result is a candidate arrangement with an operating mechanism, not a collection of independently desirable boxes.
Compare architectural alternatives
Retain alternatives that change the mechanism, not just the labels. The following are recurring choices, not an exhaustive classification.
A command or reference governor adjusts the requested input to an already qualified closed loop using its current condition and applicable constraints. It is an alternative to redesigning that loop.
Use C.32 to develop the candidate palette when several structures change together. For an identified conflict across scopes, C.32.MLAO helps generate alternatives that reduce it while exposing the burden moved elsewhere. Bring those candidates back to the information, timing and joint-condition construction here.
A policy in a lower control layer need not be less accurate than a decision in an upper control layer. Rapid tracking accuracy, predictive accuracy, semantic understanding, search quality and adaptability are different characteristics. Compare the required characteristics under the resource conditions. A rule such as “strategy is slow” is a possible arrangement, not a substitute for that comparison.
Construct the exchange that makes the parts work together
For each consequential connection, explain what the sender supplies, how the receiver interprets it, which action it can change, and which limitation or result returns. Specify the semantics before choosing a message format or transport.
A target value, a trajectory segment and an instruction to achieve an outcome impose different work on the receiver. A target may be held until replaced; a segment can specify its evolution; an outcome request can leave the means open. Determine which interpretation the receiver can realize and which uncertainty remains with the sender. In a technical realization, hold, reset, interpolation and delay rules can change the closed-loop dynamics.
Keep the following meanings distinct when the receiving action depends on them:
- an observation or estimate of what is the case;
- a prediction of what could or is expected to happen under stated inputs;
- a decision selecting a continuation;
- a command or request addressed to a participant;
- an accepted commitment to supply a result under conditions;
- a report of the action or result that occurred.
A decision does not necessarily authorize its recipient, and a request does not establish acceptance. An organization may require an explicit commitment; a technical protocol may instead have a specified execution rule. Construct the connection that governs the case.
Early communication is useful when a receiver can do worthwhile preparation before the final result exists. Name that preparatory action, the information it needs and its reversibility. Reserving a test slot can reduce later waiting while displacing other work. Preparing material can save time while creating waste if the intended design changes. Compare those consequences with waiting for confirmation; do not treat earlier messages as a free acceleration.
Include the response to change. If a plan, estimate or commitment is replaced, identify which pending action must be cancelled, revised or allowed to finish. Distinguish a message received from a message interpreted and used. Where conflicting editions would change execution, establish which instruction applies and how the recipient learns of its replacement.
Connect the proposed exchange to effective action. Identify who can request the result, who can commit the affected means, who can perform the work, and who can accept or return its result. For the needed contribution, state the supplied result, usable inputs, scope, time window, acceptance conditions and response to refusal or failure. A participant's right to choose an action and another participant's commitment to carry it out are distinct conditions; either can be missing.
Use C.32.MWA.PP:4.7–4.9 to construct and compare the joint workload, commitments and transition. Reuse existing effective arrangements whose scope and conditions fit. When a new permission, reservation or agreement is needed, identify its owner and the obtainable act that could establish it, including the time and participation required. Keep the candidate conditional until that act and its conditions are established. A design specification explains the intended relation; the actual authorization, acceptance or provision supplies a different fact. Return those effective conditions, or their unresolved remainder, to the timing analysis and continuation choice.
Derive timing from the action window
Start from the last time at which an action can still obtain the required consequence. Work backward through the necessary detection, communication, decision and execution. Include waiting for an available processor, person, authorization or resource, not only active processing time.
For a serial path with applicable upper duration bounds, a sufficient timing condition is
$$ t_{\mathrm{change}}+ L_{\mathrm{detect}}+ L_{\mathrm{deliver}}+ L_{\mathrm{decide}}+ L_{\mathrm{execute}} \leq t_{\mathrm{needed}}. $$
Each bound must cover its whole contribution under the compared conditions. Periodic observation contributes the worst wait until the next observation as well as the observation work. A computation bound must include the resource arrangement under which it can be achieved. A decision bound must include the availability of a participant with the required authority.
This sum applies to the stated serial path. For parallel or returning work, construct the timing from its dependencies and resource occupations. Identify when each required input and permission can be available, which resource each action occupies, and when it releases that resource. Reconcile different names for the same person, device or capacity before assigning intervals. A machine may continue a cycle after its operator is free; an allocation must reflect that occupation rather than charging both for the whole elapsed cycle.
Construct a candidate schedule by placing each action after its required inputs and authorizations, in intervals where all the resources it needs are jointly available. Preserve any required uninterrupted duration, overlap, synchronization and recurring demand. At a join requiring all outputs, the receiving action can start only after the latest required output is available; its own duration still contributes to final completion. Use C.32.MWA.PP:4.7 for the joint workload and commitment construction.
A failed placement rejects that schedule; it does not prove that a different order or allocation is impossible. A necessary bound can establish a stronger rejection: required occupation exceeding the available qualified capacity in the same window cannot be repaired by rescheduling within that window.
For example, two otherwise parallel branches each require ten minutes from the same editor before their outputs can be used together. With one continuously available editor and no other required work, serializing the two edits gives a twenty-minute completion. Both editing jobs become ready to start at the same time. A deadline fifteen minutes later is impossible under those conditions, although either branch alone fits. A second qualified and authorized editor, a changed operation or a later deadline changes a different premise of the construction.
Retain a timing summary only when the receiving answer follows from what it keeps. Total free time can conceal incompatible windows; an average rate can conceal a late individual response. Derive the response time or useful bound from the retained dependencies, availability and occupation conditions. If two allowed arrangements share the proposed summary but give different answers, retain the distinguishing condition, derive a bound that covers both, or restrict the claim.
Identify the end event the recipient relies on. Completion of an intervention and receipt of its confirmation may have different deadlines. If the receiving decision needs confirmation, include how the observer can establish completion and when the report reaches the person who must act. If only the physical completion has a deadline, later evidence can establish that the intervention met it.
Failure to find such a pair does not establish sufficiency. The timing-summary test specializes the information question in :4.3 to the time and capacity needed by the control decision. An average duration does not establish a worst-case deadline.
Choose an observation or decision rhythm from how rapidly a consequential difference can develop and how long a useful response takes. Periodic checking, event-triggered messages and prediction of an approaching limit are alternatives. An event trigger needs an obtainable detector and a response route; otherwise “notify on exception” supplies no timing result.
Separate update period, end-to-end latency and prediction horizon. A planner can update slowly while predicting far ahead; a fast controller can use a delayed estimate. Changing one quantity may change model error, computation, communication load or the available response. Recompute the affected dependence rather than assuming that more frequent updates are better.
Before the next update, specify what the current action may do and why that continuation remains acceptable. If a premise can fail sooner than a response can arrive, obtain a local protective action, shorten the exposure, change the observation or reduce the promised operating region. A monitoring arrangement cannot protect a condition it notices only after the last useful action.
Establish joint conditions and a usable continuation
For each relied-on component or participant result, state its inputs, operating conditions, output and limits. Then check the connection in both directions: does the supplier's result satisfy what the receiver needs, and can the rest of the arrangement keep supplying the conditions on which that result depends?
Begin with a jointly possible situation. Two controllers can each require the other to have already restored the shared condition. Two teams can each promise to start when the other releases the only specialist. Such mutual expectations supply neither an initial action nor a continuation. Construct an initial allocation, a compatible joint condition or another arrangement.
Next follow a consequential action or change through the whole. Establish what remains true, what must be updated and what is available next. For a promise about recurring work, start the next job from a state in which the preceding job can leave the shared means. A rule or bound can establish continued feasibility if the work can start under its stated conditions and every permitted transition leaves the next job with what it needs. A record of one execution can settle what happened on that occasion.
For a modeled controller interconnection, derive or obtain the joint dynamics rather than combining separate stability labels. Include reference changes, state estimation, hold rules, actuator limits and computational delay. A technical result can establish an admissible region, an error bound, recursive feasibility, stability or goal attainment; retain the property established. These results are not interchangeable.
For example, an invariant error bound can justify a margin in a planner constraint while leaving the question of reaching its goal unresolved. A continuous inequality proved only in a region applies until the first exit unless remaining in that region is established separately. Under mode switching, one invariant set respected by every allowed transition is a sufficient condition; different invariant sets for individual modes do not establish safe transitions between them.
Organizational continuation needs a different construction. An accepted commitment must have a capable participant, available means, compatible other commitments and the right to perform the promised action. Establish how an unavailable input, refusal, changed priority or missed window is handled. A promise can be part of the plan's basis without becoming a mathematical guarantee of performance. If the plan needs a deterministic deadline, justify the duration bounds and availability it relies on; otherwise retain the appropriate scenario, uncertainty or conditional promise.
Choose the fallback for the protected result. Keeping a previous command can work only while its qualifying conditions remain true. Escalation can work only if the recipient can decide in time. A replacement performer can work only with the required competence, access and capacity. Compare these as continuations, not names in an exception list.
If the joint conditions cannot be established, return the unresolved or conflicting dependence. It may justify a different interface, smaller demand, different allocation, additional observation or a revised goal. Keep independently usable component results.
Qualify learned contributions by their function
Learning is one way to obtain a contribution to the arrangement. Identify which contribution it supplies. A learned state estimate, forecast, action policy, representation or optimization surrogate has different inputs, outputs and failure consequences.
For a predictor, qualify the output, horizon, available observations, future-action assumptions and error relevant to the receiving decision. For a policy, qualify the actions it can select, the states and instructions it can receive, and the constraints of the actual subject. Agreement with recorded demonstrations and a useful closed-loop result are different tests.
Keep training, inference and online adaptation separate. A policy trained slowly can execute rapidly. An online update can change the policy while its input and output formats remain unchanged. Determine whether the relied-on error, timing or constraint result continues to apply to the updated component.
Compare the learned contribution with an adequate programmed or conventional controller at the same receiving use. Include data acquisition, training or calibration, inference resources, monitoring and recovery in the whole cost. Prefer the learned option when its supported capability changes the result enough to justify those burdens, not because the architecture contains a neural network.
If a separate filter or supervisor modifies a learned command, qualify the resulting combination. The modified command can preserve one constraint while changing a tracking or progress result. A human fallback also needs detection, sufficient time, usable information and authority.
The output of this operation is a qualified component use or a named missing result. It is not a general method for training every controller. When the application needs a bound that the learned supplier has not established, compare a narrower operating region, an applicable protective construction, a different supplier or a weaker warranted claim. Do not silently replace the missing bound with average benchmark performance.
Compare the whole arrangements and the work of changing them
Compare candidates for the same required outcome, external demand and disturbances over comparable horizons. Hold baseline resources and information conditions comparable. Where a candidate changes sensing, provision, authority or timing, include the work of obtaining that change and the conditions under which it can operate.
Separate an unavailable candidate from a costly feasible one, and both from a candidate whose consequential premise remains unknown. Compare the coordinates that can change the decision: useful result, delay, constraint exposure, effort, computation, scarce capacity, participant burden, flexibility and recovery. A mathematical optimum remains relative to its model and objective; a project choice also has to respect its obligations.
Include the transition. A new controller needs initialization and an admissible switch. A new organizational allocation needs effective authority, informed participants and preserved obligations. Existing work and performed history do not disappear when the new arrangement is selected.
C.11 supplies a reasoned local choice among the explained alternatives. The right to commit resources or direct action remains separately established under the permission and agreement relations. Use C.11.DUA when the value of another observation, calculation or trial is itself unresolved.
A constructed case can justify a conditional proposal. An effectiveness claim requires evidence from its intended use. Do not require a new trial when existing adequate evidence settles the decision, and do not count a simulated or taught example as operational evidence.
Return the chosen alternative, accepted trade-off and conditions of use to the receiving engineering or organizational decision. If only a candidate comparison was requested, it can finish here without authorizing implementation.
Make the result usable and revise the affected dependence
Explain the result in the ordinary form needed by its users. It must make recoverable:
- the intended outcome and protected conditions;
- the participants, decisions and available interventions;
- the information, exchange and timing that connect them;
- the jointly feasible conditions, continuation and exception response;
- the alternative, accepted cost and grounds of the choice when a choice was requested and made;
- the changed observation or condition that requires a different decision.
These are content requirements, not a mandatory record or a demand to model every relationship in the enterprise.
On new evidence, first identify the failed dependence. A fresh observation may only require a new state estimate. A different response can challenge the dynamics or capability. A blocked authorization can require a different allocation or commitment. A changed use can require a revised goal. Alter the architecture when the needed continuation cannot be obtained adequately within the existing arrangement.
Carry the change through its direct consumers. A new estimate can change the admissible command; that command can change arrival time; the changed arrival can defeat an organizational promise. Preserve unaffected results, and requalify the connections that changed.
Archetypal Grounding
Retain a qualified loop and change how it receives targets
Consider a constructed discrete model, in normalized units,
$$ x_{k+1}=x_k+u_k,\qquad u_k=\tfrac12(r_k-x_k). $$
The actuator condition is $|u_k|\leq0.06$, and the state must remain in $[-1,1]$. The state is available without delay or error in this first case. The controller computation fits each update. These are premises of the example, not claims about a physical device.
For a fixed target, the error contracts by one half per update. Nevertheless, if $x_0=0.5$ and a planner immediately applies $r_0=0.7$, the commanded input is $0.1$, outside the actuator limit. The separate tracking result did not qualify that reference change.
One alternative is to redesign the controller and planner together. Another candidate retains the controller and limits the applied target from the desired goal $g_k\in[-1,1]$:
$$ r_k=x_k+\operatorname{clip}(g_k-x_k,-0.12,0.12). $$
Here $\operatorname{clip}(a,l,h)=\min(\max(a,l),h)$. Substitution gives $$ u_k=\tfrac12\operatorname{clip}(g_k-x_k,-0.12,0.12). $$ Thus the actuator condition holds. The next state lies between $x_k$ and $g_k$, so an initial state in $[-1,1]$ stays there. This establishes the stated discrete constraints for this model. It does not establish intersample physical safety or a prescribed arrival time.
For the constant goal $0.7$, the first three applied targets are $0.62,0.68,0.7$; the corresponding inputs are $0.06,0.06,0.04$, and the resulting states are $0.56,0.62,0.66$. The limiter uses the loop's remaining action capability and returns a different progress forecast to the planner. It is a state-dependent governor construction, not a reason to add a new tracking layer.
Now request $x_2\geq0.68$ from $x_0=0.5$. Under the actuator bound, even an arbitrary control law can increase the state by at most $2(0.06)=0.12$, giving $x_2\leq0.62$. That requirement is impossible under the stated model and limit. The next choice concerns the deadline, attained position or available actuation; faster planning alone cannot obtain it. The participant authorized to change the demanded result must receive this bound.
Change the governor's observation condition instead. Suppose the governor receives only $\hat x_k$, with a qualified current bound $|x_k-\hat x_k|\leq0.02$, while the local tracker retains the original state-based feedback law. Limiting $|r_k-\hat x_k|$ to $0.12$ would allow $|r_k-x_k|=0.14$ and input $0.07$. A sufficient repair is to limit $|r_k-\hat x_k|\leq0.10$ and require $r_k\in[-1,1]$. Then $|r_k-x_k|\leq0.12$, while the original convex update preserves the state interval.
The set of targets satisfying both conditions must be nonempty; it is nonempty here when the state lies in the interval and the stated estimate bound holds.
The error bound must describe the current state. A bound for a stale measurement cannot be reused without accounting for intervening motion. If the tracker also loses its state measurement, its response equation and the convexity argument need reconsideration. Replacing it with a learned policy likewise reopens the calculation: the old response equation no longer supplies its proof.
This example derives the governor interface, proves the stated bound and identifies the conditions on which continued operation depends. A choice between the governor and joint redesign also needs the response, cost and changeover conditions of the redesigned arrangement.
Obtain a timely organizational decision
An exhibition panel must be installed by 16:00. Under a constructed operating arrangement, the following bounds are available:
These contributions are serial. For a change at 15:00, the full bounds put detection and dispatch to the editor by 15:20, receipt by 15:25, approval and dispatch for installation by 15:35, and physical installation by 16:00. A later change can still succeed when some operations finish earlier. The hour-long bound alone leaves that case open.
The useful state includes the current content, its approval status, the required materials and the availability of the people and means. “The file was sent” does not distinguish an approved file from an unapproved proposal, or a delivered message from one awaiting a decision.
Compare three changes to the arrangement.
- Shorten detection to five minutes. If every other premise remains true, the bound becomes 45 minutes. This helps changes that occur by 15:15. It adds observation effort and does not repair a missing decision right.
- Authorize a defined local exception in advance. If the current change needs an extra expenditure outside the editor's authority, the ten-minute decision premise fails. An effective limited authorization, available reserve and competent decision maker can restore a bounded path. A proposal to delegate does not do so until those conditions hold.
- Retain central authority and change the commitment or work. Obtain an available central decision window, agree to freeze content earlier, use an already accepted substitute panel, or omit the affected panel where the receiving commitment permits it. Compare the changed result, delay and cost; these are not equivalent ways to meet the unchanged promise.
A forecast of a likely change can allow reversible preparation before approval. Establish which materials or capacity may be reserved, what another job loses, and how the reservation is released if the forecast changes. Printing an unapproved final panel is a different action with different consequences.
The installer checks the panel in place and reports completion or a problem to the coordinator. The table bounds physical installation. If the venue also needs confirmation by 16:00, include time for this check and report or establish that the approved release already covers them. If the report is missing, examine any other evidence of installation. Receipt of a confirmation is a separate result when the venue has commissioned it. One person may perform several contributions when the demands still fit. Their number does not determine a management hierarchy.
The timing implication is mathematical under its premises. The premises about people, authority, resources and duration need organizational support and observation; they are not supplied by the equation. In actual use, distinguish a committed service bound, an empirical estimate and a planning assumption.
The same hour-long bound can cover another panel presented on its own if the path from the content change to physical installation includes all its work and waiting, with the same upper bounds, authority and available capacity. An added wait for material to set can break the bound even if the four operations take the same time. A promise to confirm completion needs a bound through confirmation. To promise two installations with overlapping windows, account for the shared editor and printer across both. One approval can serve both panels only when it covers both and arrives in time for each. Place the shared work and the remaining work on the joint schedule.
The smaller sufficient result
A machine already has an applicable controller, command range, state estimate, timing qualification and recovery procedure. The proposed change only redraws their representation for a maintenance team. Use C.30.LCA to make the obtaining relations and limits recoverable. Do not repeat architecture synthesis unless that reading exposes a consequential missing or changed relation.
Similarly, a team can retain an adequate allocation when a new report does not change any action, deadline or protected condition. More communication is not a required improvement.
Bias-Annotation
The method can overfavor engineered, measurable interactions. Human participants can contest the goal, decline a request or hold interests that the model omitted. Recover what they have committed to and their protected conditions before treating a proposed arrangement as available.
A layered picture can overfavor hierarchy. Shared resources, peers, external suppliers and overlapping authority may require coordination without one superior controller. Examine the actual influence and decision rights.
A successful demonstration can overfavor its observed conditions. Keep the operating region, information quality, available support and receiving criterion with the result. Compare the whole arrangement, including work displaced elsewhere.
Conformance Checklist
- The intended result, protected conditions and authority to revise them are recoverable.
- The candidate explains observation, decision and intervention dependencies, including the consequential return.
- The retained information supports the needed prediction or bound at the decision time.
- The proposed separation preserves shared constraints through a constructed interaction or allocation.
- Alternatives include retaining an adequate arrangement or changing the dependency itself.
- Timing accounts for waiting, resources, authority and the last useful action, not only message or computation speed.
- Initial conditions and continuation are jointly possible; component claims are not silently promoted to whole guarantees.
- A learned contribution is qualified for its function and its changed use is followed through the arrangement.
- The comparison preserves the same demand and includes transition and displaced burden.
- The result distinguishes a proposal, a conditional derivation and evidence of actual operation, and names its relevant return condition.
Common Anti-Patterns and How to Avoid Them
Consequences
The reader can change the relation that prevents a result: the available information, allocation of a decision, command semantics, timing, shared capacity or the demanded outcome. An adequate incumbent can survive the comparison.
The method makes the price of separation visible. Layers can reduce immediate computation or authorization demand while adding models, interfaces and coordination. Delegation can remove waiting while adding qualification, communication and exposure. A simpler arrangement can be preferable.
Construction itself has a cost. Stop with sufficient grounds for the receiving decision; pursue additional modeling or evidence when a plausible result can change that decision or when a binding obligation requires it.
Architectural Rationale
The arrangement is constructed from joint use because the consequence belongs to the connected system or work. Local correctness is useful only when the receiver can use the result under conditions the rest of the arrangement can supply.
Different time demands justify considering a division, not prescribing one. Separation is advantageous when its saved computation, waiting or coordination exceeds the burden of preserving the lost dependence. The governor case shows that a smaller change can preserve a qualified component. The exhibition case shows that changing a decision right can matter more than increasing message speed.
Goal revision belongs in the architecture because feedback can make the selected outcome unattainable or no longer useful. Its authority remains distinct from the ability to compute another plan. This prevents a fast local optimization from silently changing the promise made to another participant.
Model-based guarantees and organizational commitments remain different contributions. Both can constrain a design, but their grounds and failure responses differ. That distinction makes the common method transferable without importing a control theorem into an enterprise through matching labels.
The current arrangement can itself become the subject of a slower development activity. That does not identify every operational goal revision with architecture development: choosing another authorized destination can preserve the arrangement, whereas moving decision rights or changing the available interventions changes it.
SoTA-Echoing
Which line answers the working question? For constructing joint control under different information, timing and authority, this pattern selects a synthesis of whole-arrangement comparison, explicit dynamic and temporal connections, and domain-specific qualification. A fixed three-layer template is cheaper to draw but leaves those connections to the reader.
The additional work here is justified when a consequential dependency remains unresolved; :5 shows a changed command interface and an authority-dependent deadline. These are conditional constructions, not measured enterprise benefits. Retain a sufficient incumbent and reopen this choice if its extra analysis cannot change the receiving decision.
Adopt joint design rather than independent layer tuning. Matni, Ames and Doyle's 2024 treatment, §§2–3 and 4.5 supports reasoning from a whole control problem to interacting layers. This changes :4.4 and :4.8: a removed consistency condition becomes an obligation to construct the connection. The source develops a research programme, not a universal optimal layer count. The common method adapts that question beyond its technical constructions; a claimed transfer still needs the receiving dynamics, conditions and evidence. Reopen when the proposed separation has no supported interface or a simpler controller answers the same demand.
Adapt quantitative architectural co-design within its class. Verhoek et al., 2026, §§2, 5 and 7 make model order, sampling ratio and controller parameters joint design variables for a two-layer LTI construction with compute constraints. Compared with fixed rate heuristics, this supports the whole comparison in :4.7–:4.10. The accepted cost is a demanding optimization and qualification problem. This is a bounded preprint result; general composition and several extensions remain open. Use a simpler qualified design when sufficient, and reopen on a changed plant, observation regime or implementation premise.
Retain a governor as an alternative. The Garone, Di Cairano and Kolmanovsky survey, §§2 and 6 supplies the established alternative of managing commands over a pre-stabilized system. Castroviejo-Fernandez and colleagues, 2024, develop a disturbance-observer-based continuation. Their conditions motivate :4.5 and the observation change in :5.1. Adopt the architecture when its qualified loop and admissible-set construction answer the demand; reject treating it as automatic mission achievement. Its price is altered progress and model-dependent qualification. Reopen if those costs or assumptions defeat the receiving use.
Adapt organizational design and commitment-based coordination. Joseph and Sengul's 2025 review compares organizational-design approaches rather than making decentralization a universal answer. Guerreiro and Dietz, 2024, preserve request, promise, declaration, acceptance and exceptional continuations in business-process descriptions. These lines change :4.2, :4.6 and :4.8: the organization must supply decision rights and workable commitments, not merely fast commands. Their use can add negotiation and coordination cost. Reopen when the decisive dependency is an unavailable capability, resource or conflicting interest rather than the proposed information or authority arrangement.
Use safety analysis for its distinct result. The STPA Handbook, Chapters 2–3, is an established approach to unsafe control actions and inadequate process models. Adapt its attention to timing, missing feedback and conflicting control in :4.3, :4.7 and :4.8. It is an alternative to checking only nominal performance when harmful loss is possible; it does not replace performance synthesis or prove absence of loss. Reopen the relevant safety analysis when a control action, model, responsibility or operating condition changes.
Learned execution does not settle architecture qualification. Figure's Helix 02 description gives a comparator: semantic reasoning, 200 Hz joint targets and 1 kHz whole-body execution with a learned lower component. It supports considering different functional and timing demands, not the superiority of that allocation in every system. Adapt the distinction in :4.9; retain a programmed qualified component as a rival. Reopen on a changed policy, observation distribution, latency or relied-on constraint. The vendor demonstration supplies no general enterprise or closed-loop safety theorem.
Relations
- C.30.LCA recovers the obtaining or proposed control structure represented in a view and its claim limits. This method consumes that account and returns constructed alternatives whose descriptions can be examined there.
- B.2.5 establishes the observation, returned influence and coupling needed for a feedback claim. A new architecture proposal still needs evidence of those relations when put into use.
- A.3.3, A.3.3.TR and A.3.3.PI supply the dynamics, transition rule and predictive information. Their results enter the decisions and timing here; a changed input policy can reopen them.
- C.32 and C.32.MLAO supply candidate generation and changes that address a recovered cross-scope residual. This pattern adds the control-specific construction of information, intervention, timing and continuation.
- C.32.MWA.PP:4.3–4.9 supplies construction and comparison of joint performance and provision, including shared means, workload, commitments and transition conditions. This pattern connects those conditions to the observation, decision and intervention that must occur in time. A proposed relation remains prospective until its obtaining conditions are met; changed authority or resource demand returns to the affected comparison.
- C.11 and C.11.DUA supply the local choice and the worth of further inquiry. The architecture explanation is their input; neither that explanation nor the choice establishes permission to act. C.43 returns the selected arrangement to the connected continuation it must make possible.
- C.28 governs reliance on a causal result when an empirical effect is used to justify the change. A conditional calculation retains its model premises.
- C.39 and B.1.5.RS support construction of a missing contribution and preservation of receiving uses when one contribution is replaced. A replacement that changes observations, timing or actions returns to the affected joint-condition analysis here.
C.32.LCA:End
Problem-to-Structure Architecturing Unfolding
Type: Architectural process pattern under C.32 Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when an architect or architecture-responsible practitioner has a stated receiving-use basis and must carry its architecture pressure through selected structures, candidate synthesis, project architecture decision, realization Work, actual-structure feedback, and the next action selected for the current question.
In a systemic product undertaking, recover the product SoI, the extended enterprise organized for it and the intended use or withdrawal from use that justifies the architecture. In work on its creator, keep the organization as the architectural subject and connect its needed capability to that product use. An independently warranted non-system result retains its own receiving question. When a needed use, boundary or functioning hypothesis remains unclear, recover that basis before relying on the dependent architecture choice.
Common starting situations include:
- A required function has no recoverable bearer.
- An architecture characteristic is failing.
- A cross-scope residual survives local repair.
- A modularity, reuse, interface, scale, or description-loss problem blocks action.
- One typed Work, communication, tool, method, deployment, evidence, selected-structure, or architecture-side source cannot yet sustain the transformed-side architecture content needed for the changed referent.
- Operation shows that expected structures and actual structures diverge.
The first useful result explains how the architecture pressure bears on the structures needed for the outside use and what the receiving work needs next. An ordinary account can be sufficient. Use ProblemToStructureArchitecturingFlowCard@Project when a reusable record helps keep those connections recoverable. The card is a working U.Episteme about one project-local P2S architecturing transformation flow; the flow, Method, descriptions, decisions and Work remain independently identified objects.
For the first pass, make the following content recoverable, using the card only when needed:
- problem claim or pressure;
- described holon;
- architecture question and intended use;
- ClaimScope or qualification window when material;
- first pattern to use;
- one unknown or selected structure;
- selected transformation-flow structure when one is relied on;
- pattern for the next claim.
Add decision, work, eval, publication, and feedback refs only when the corresponding question becomes current.
For ProblemToStructureArchitecturingFlowCard@Project, flowId designates the project-local P2S architecturing transformation flow that is the card's C.2.1 EntityOfConcern. The claims carried by the filled card and the effective reference scheme for its designations remain recoverable; changed claim content, changed flow EntityOfConcern, or changed effective reference scheme identifies another card episteme.
@Project is a compatibility and retrieval cue only. It establishes no project entity, composite-work identity, context, authority, viewpoint, or parthood.
When the card is used in one project, projectWorkOccurrenceRef identifies the composite Work occurrence admitted under U.Work and architecturingFlowCardProjectUseRelationRef identifies the direct relation by which architecturing work uses the card. The card, the architecturing work it helps coordinate, and the larger project work remain distinct.
The problem, architecting-side, realization, and architecture refs point to independently established objects; they are not P2S relation kinds. acceptedProblemCardRef resolves to one C.22.2 C.2.1 episteme; the nested signal and pressure fields neither constitute that card nor make an actual Problem obtain.
An assignment claim uses architectingAssignmentSpeciesRef and architectingSystemRoleAssignmentRef for its separately declared species and obtaining occurrence; it supplies neither classification nor performance.
When performance is current, each performer first has its A.13 core and each performedWorkRef names Work independently admitted through A.15.1. performedWorkAttributionRefs are optional and appear only when the flow or receiving use expressly represents precise assignment-bound attribution through the same obtaining A.13 assignment; missing or failed F.6 attribution alone leaves the Work refs intact while the independent A.13/A.15.1 grounds for each referenced Work still hold. Authority, responsibility, capability, Work, assignment, attribution, and result remain separate.
actualTransformationRefs name independently identified actual changes under A.3.4. The Work-to-change fields pair each relation that obtains with the cited pattern or local claim used to check it. The three production refs resolve to separate local A.15.PROD claims and remain absent when their question is not current.
actualStructureRefs name subject-side U.Structure values whose declared substrate and selected relation organization are recovered under A.22 from direct facts that obtain; they introduce neither an ActualStructure kind nor an actualization relation.
Use C.30 to keep the described holon, obtaining ArchitectureRelation occurrences, selected structures, and any affirmative, negative, unresolved, candidate, required, desired, or expected ArchitectureClaim content separate. actualStructureDescriptionRefs name later descriptions of those structures and do not make them actual.
For work limited to a problem card, a grounded architecture claim, a structural view, a candidate palette, a project architecture decision, an ADR-like publication, work planning, performed work, measurement, a mathematical lens, or G.11 currentness, freshness, telemetry, edition, or decay orchestration, use the pattern named in Relations for that narrower claim.
Problem
FPF has patterns for problem records, grounded architecture, structural views, candidate palettes, architecture characteristics, eval programs, decisions, ADR-like projections, methods, Work occurrences, separate method descriptions and work-record epistemes, measurements, mathematical lenses, improvement loops, and currentness or decay orchestration. A practitioner still needs one readable pattern for the architecture work that connects them.
Architecture work can fail in two opposite ways.
First, the flow collapses into a description or decision artifact: a diagram, view set, ADR, memo, dashboard, score, or publication record is treated as if it carried the architecture, the decision, and the realized structure. The project then loses the distinction between selected structure, description, decision, method expectation, performed work, and actual structure.
Second, the flow starts inside the boundary or disappears into relation rows. An architect selects modules, interfaces, or an attractive configuration before stating the external change, relying use, project system-of-interest, and functioning hypothesis that would justify them; or every local pattern is correct but no readable flow carries the pressure through Work and feedback. In either branch, the user can name patterns but cannot explain why this internal structure serves the outside use.
Forces
Solution
Apply the smallest useful part of the action sequence below. Create or update ProblemToStructureArchitecturingFlowCard@Project only when its record is useful for that work. The numbered presentation guides attention; it is not by itself a claim about the order, identity, or occurrence of project Work or actual transformations. When the applicable rule in another pattern fully answers the current question, use that pattern and stop there. Continue P2S only while the connected project-local architecturing flow remains the object of the question.
Before selecting internal structure, state the expected receiving change, the beneficiary or relying use, its operating time and relevant encompassing functioning, and the hypothesis by which the proposed result could support that use. In a systemic product account, recover the SoI, its extended enterprise and agreements under A.1.STM:4.1, and its actual or intended boundary. Move toward use and upward through the relevant part–whole relations. Explain how the parts of the encompassing whole function together and derive the system's required contribution before moving inward to construction or backward to creation.
A creator, owner or larger project is not a containing System by that fact. A merely intended system stays in plan or description content. A.1 and A.1.SCR govern actual-system recognition; A.15.6 recovers the project Work, designation and local subject.
The architectural subject can be the product, a component or a creator being developed. For the creator, state the product uses that justify its capability, then investigate the organization's own structures and the finite local result to be accepted under A.1.STM:4.1.1. Here the organization is the system-in-hand; the product remains the SoI.
A distinct usable machine or component supply can have its own product SoI under A.1.STM:4.1; changing a contract or assigning the same organizational work to an external provider leaves its product-use basis in place.
Several products can constrain one creator decision. Preserve the relevant demands, shared capacity and authorized limits instead of choosing whichever product makes the preferred option look best.
Compare creator-architecture alternatives by the capability they can provide under those conditions, including the required Methods, people, means, access, cooperation and authority. Carry material gains and losses through to C.32.PAD and to the acceptance of the promised local result.
For example, acquiring equipment, preparing its operators and establishing a workshop's production capability have different completion grounds. Suitable evidence can establish a bounded capability before every future product is used; uncertainty about the later product effect remains explicit. Research may instead finish with a supported answer or a reason to stop the proposed development.
Use A.1.STM:4.1–4.3 when the common use basis or the local subject's contribution remains unclear. Preserve an adequate shared account and independently useful local answers. When only a dependency is unsupported, A.1.STM:4.4 locates that gap. A known local architecture question does not require repeating the whole account; a missing purpose prevents only a choice that depends on that purpose from being justified.
Use the analogy with E.18.1 P2W narrowly. A P2W use carries an accepted problem-side record or accepted C.22.2 ProblemCard episteme plus the carried distinction into the next FPF use. A C.32.P2S use carries architecture-relevant pressure and structural uncertainty into candidate structures, selected structures, project architecture decision, realization work, and actual-structure feedback. The practitioner then uses the pattern for the next question. The analogy ends when that question is method, work, telemetry, publication, or improvement-loop use; use its pattern rather than stretching P2S into generic process management.
-
Recover the problem pressure or architecture concern together with its receiving-use basis. B.5.PI can expose it before a complaint or actual Problem. Name the expected environmental or relying-use change, beneficiary or user, operating time and relevant encompassing relations, product SoI, extended enterprise, agreement basis and boundary hypothesis where a systemic product account is current, required functioning, pressure signals, source-use records, affected holon, and first pattern to use. For the systemic product account, derive required functioning from how the system contributes with the other parts of its encompassing whole.
If the pressure is still only a cue, use C.22.2 and cite the resulting
ProblemCardepisteme when it becomes the accepted input. If an actual Problem is claimed, cite an independently obtaining C.22.PFRProblematicForRelation; the card and its fields create no such occurrence.If the outside-use or system basis is absent, recover the missing contribution before continuing with P2S: use A.1.STM:4.1–4.3 for an unsettled common SoI and the local subject's connection to its use, and the applicable recognition or problem pattern for its own question. Preserve an adequate shared basis and the independently useful local answers.
-
Recover the described holon after the outside-use and boundary hypotheses are visible. State the architecture question and intended use, plus ClaimScope or qualification window when either changes the answer. Then recover candidate or selected structure kinds, selected structures when available, and architecture characteristics. Use C.30 for the grounded architecture claim, C.32.HCS for starter characteristic heads, C.32.ACS for project criteria rows, and C.25 when a composite quality family is current.
-
Represent the structural uncertainty that can change the next action. Identify the unknown structure kinds, internal composition, bearers, interfaces, allocations, variation points or constraints, and distinguish them from expected structures. Explain which information is available or lost at the receiving use. Add a return to stronger inspection when an omitted distinction can change that use.
-
Generate architecture ideas, principles, constraints, and candidate structure changes from the recovered pressure and outside-use basis. A problem account or source contribution can help reveal a missing structure as well as change an identified one. State the affected or still unknown structure, relevant characteristic, intended gain, known loss and receiving claim; resolve remaining uncertainty through candidate construction and comparison.
-
Synthesize candidate architecture configurations and candidate sets through
C.32. Keep function-bearing feasibility, ordinary work organization or actual Work, and candidate-relevant selected structures visible when they change the candidate. Such structures include constructive-module, placement, control, transformation-flow, information, evidence, and scale structures. Route every unresolved claim-bearing role cue throughE.10.ROLE; carry the recovered local system-role kind, separate System-classification judgment, assignment, direct-relation position, function claim, organization or representation position, or ordinary wording rather than a generic role structure. -
Compare, retain, declare a selected-set result, publish it, or return alternatives only under the predicate that defines or constrains that relation. Use
A.19.CPMfor comparison,A.19.SelectorMechanismfor set-returning selection,C.18andC.19for archive, front, and pool policy,G.5for selected-set result declaration, andC.11for a fixed local choice. For publication, useE.24.PUBfor the publication occurrence and audience availability, andE.17when a source-backed face and source return are needed. -
Make a project architecture decision through
C.32.PADwhen implementation commitment is current. The decision relation names the selected architecture option, affected structures, trade-off, accepted losses, method and work consequences, return for action-changing structural loss, and decision repair or supersession condition. For a shared creator, carry the product demands affected by the choice and the actual provision and decision rights on which the capability depends. -
Publish descriptions, views, ADR-like records, narrative renderings, or other records only as descriptions, structure-to-narrative renderings, or publication forms of claims. Those claims concern structures, decision relations, method expectations, description or view loss repair, and reader use. Use
C.30.AD,C.30.ASV,C.32.ADR,A.6.3.NAR,E.17, andE.24.PUBas applicable. -
Name the intended architecting or realizing Systems first, then make the needed Method descriptions, constraints, readiness expectations, work expectations, and structure-use return conditions available to them under their own access or use relations. If the claim is that a System holds an assignment, identify that separately obtaining assignment; assignment does not establish classification. Assess the System's capability and fit through A.2.2 independently. Keep availability or access, local kind, separate classification, assignment, readiness, permission, authority, responsibility, and plan or commitment independent.
When performance is claimed, recover each performer through A.13 and let A.15.1 independently admit the
U.Work. AddperformedWorkAttributionRefsonly when the flow or receiving use expressly represents precise assignment-bound attribution. Add a Work-to-change relation only when that direct predicate obtains. Return the missing governor for an unsupported direct claim. -
Carry the selected arrangement into domain work when the commission includes realization. For retention or withdrawal, identify the use that must continue or cease, the time and conditions, and which connections must remain, change or end. An existing arrangement can already supply the means. Establish the promised effect in use when it falls due. Use
A.15.1for dated work andA.3.4for an independently identified actual bounded change.If work is said to cause or realize a change, name the Work-to-change relation and use the cited pattern to check the applicable rule. If no reusable relation is available, cite a local claim selected under
A.6.RCDdisposition 2.Use separate local
A.15.PRODclaims only when production-work participation, entity-identity inception, or historically indexed production completion is current. The P2S card records refs to the independently established Work occurrences, transformations, Work-to-change relations, and local production claims. -
Observe, inspect, measure, and evaluate subject-side
U.Structurevalues whose declared substrate and selected relation organization are recovered underA.22from relation occurrences, applied constraints, invariants, or other facts that obtain. Include architecture-characteristic results and functional-characteristic or capability implications in operation or use. Ask whether those actual structures enable or block the functions and effects they were meant to bear, and ask what selected structure, accepted loss, counter-characteristic, or functional implication got worse when a visible metric improved.When a measured characteristic misses its target, first identify the promised value, the observed population and the operating conditions. To establish a structural discrepancy, compare the intended connection with the one in place. The same layout can deliver less under a different input mix or an unavailable operator. A verified discrepancy can justify correcting a binding design even while its contribution to the performance shortfall remains open.
A description, measurement, evaluation result, publication, or resemblance to the selected structure does not itself make a structure actual or establish conformance. Use
C.30to name an obtainingArchitectureRelationonly when its holon, selected-structure participant, and predicate are satisfied; keep candidate, required, desired, expected, negative, or unresolved architecture content in anArchitectureClaim.Use
C.30.ADorC.30.ASVfor actual-structure descriptions or views,C.32.ACEfor eval programs and eval results,C.16for measurement, andC.25for Q-bundles.Use
E.23when repeated improvement method is current,G.11when currentness, telemetry, edition, freshness, or decay orchestration is current,E.18for transformation-flow slice-local refresh,C.18orC.19for archive, front, and pool updates,C.32.PADorC.32.ADAfor decision repair or supersession,C.32for new synthesis, andC.30.ADorC.30.ASVfor architecture-description or structural-view loss repair. Use the pattern for the next question to select the return or repair action from actual-structure divergence, eval results, functional implications, freshness loss, description or view loss, and new constraints.
At the realization boundary, keep selected structure, expected structure, actual subject-side structure, dated work, independently identified actual transformations, local production claims, actual-structure description, and evaluation result as different objects. Shared assembly work, temporal adjacency, common affected referents, one flow, or one selected configuration establishes neither one composite transformation nor absence of finer transformation parts. If a receiving claim requires transformation composition, return the missing-governor blocker; do not use that blocker to stop independent work, inception, completion, actual-structure, description, evaluation, or return claims.
When architecture-influence correspondence constrains the candidate set, add the C.32.CONWAY branch before synthesis becomes narrow. Name the changed referent and any independently grounded A.3.4 transformation separately. Name every influence source by kind. Put only asserted influence facts with an obtaining direct relation in influenceSourceRows[]; otherwise keep the pressure synthesis-local in the C.32.CONWAY frame with its missing-governor, unresolved-grounding, or false-predicate disposition.
For each actual architecture side, name the C.30 holon, obtaining ArchitectureRelation, and selected U.Structure; keep modal content in an ArchitectureClaim. Then frame candidate families that change the influence-source side, the transformed side, both, or declare a bounded mismatch with the named correspondence or decision-repair return condition.
P2S Unfolding Structure Block
Use this local block when the architecture question requires the connections among pressure, structure, decision, work and feedback to be represented as one constraint-governed unfolding structure. P2SUnfoldingStructureBlock is an architecture-facing local A.22.CGUS U.Structure specialization block defined here for problem-to-structure architecturing use. That U.Structure, the project-local P2S architecturing transformation flow, the P2S method, and the flow-card episteme remain distinct.
A pre-admission ProvisionalUnfoldingDemonstrationDescription or post-admission DemonstrativeUnfoldingSlice is a separate episteme whose displayed order is not the CGUS, flow, method, or Work. The block is not a root U-kind, not an architecture decision, not an ADR, not an architecture description, and not a work plan by itself.
The block is useful when the architecture work has to show how problem pressure constrains candidate, selected, expected, or actual structures without hiding the rule for the next claim or the pattern that contains it. unfoldingStructureRef names the selected CGUS or local architecture-facing U.Structure block. When a narrower-specialization relation is needed, name and test it under the pattern that defines its predicate; use A.6.RCD if that predicate is missing.
decisionLinkageRef names the project architecture decision relation defined in C.32.PAD only when that decision is current. For the claims associated with descriptionRefs[], use C.30.AD, C.30.ASV, C.32.ADR, A.6.3.NAR, or publication patterns only when a description, view, ADR projection, narrative rendering, or publication claim is current.
realizationWorkLinkageRef points to the A.15-family work relation; use A.3.4 for actual transformations. The Work-to-change fields keep the obtaining relation separate from the cited pattern or local claim used to check it. The three production-ref groups point only to separate local A.15.PROD claims.
Establish any Work-authorization, performed-Work, or selected/expected-to-actual claim through its direct pattern. Otherwise stop with the problem-to-structure result, next receiving action, and stop or return condition.
groundedArtifactConfusion? is an alias for the same optional blockedOverread? value.
Use e18TransformationFlowUnfoldingRefs[] only for slices whose substrate is transformation-flow structure. P2S itself is broader: it can carry, for example, module, functional, placement, control, method, evidence, scale, and information structures through architecture synthesis and feedback. Any role-like source wording first resolves to the independently relevant local kind, classification judgment, assignment occurrence, direct-relation or representation position, function claim, organization position, responsibility or authority relation, or ordinary label; P2S has no generic role-structure carrier.
Architecture Unfolding Structure Use
Use ArchitectureUnfoldingStructureUse@Project when a named constraint-governed unfolding structure is being used as architecture-relevant structure inside problem-to-structure architecturing. This dependent architecture-use relation record is defined here; use it with the relevant C.30 or C.32 pattern. Record any architecture decision, description, ADR projection, or realization Work through its direct pattern when that claim is current.
For ArchitectureUnfoldingStructureUse@Project, the suffix remains a compatibility and retrieval cue until a use is asserted. Every asserted occurrence includes projectWorkOccurrenceRef as the composite U.Work participant; without that Work, keep only the retrieval cue and do not assert the use relation.
architectureQuestionCardRef may cite the C.30 triage episteme, while architectureBearingHolonRef names its independently identified subject. architectureRelationRefs[] contain only independently obtaining C.30 occurrences; modal architecture content stays in architectureClaimRefs[]. unfoldingStructureRef names the admitted CGUS or local block being used.
affectedSelectedStructures[], architectureCharacteristicRefs[], and acceptedLosses[] state why the unfolding structure matters for architecture rather than for a generic route. Method and work refs point to the A.15 family only as realization or feedback linkage.
For decisions, descriptions, ADR-like projections, measurements, evals, evidence, gates, publication, and performed work, use the subject predicates, consulting their pattern explanations when needed to interpret or apply them.
Stop conditions:
- stop at
C.22.2when the signal still needs problem-side recovery; - stop at
C.30orC.30.ASVwhen the current need is only architecture claim or structural-view adequacy; - return the candidate palette under
C.32when that is the commissioned result. A later design, decision or realization has its own remaining work; - return the adopted project architecture decision under
C.32.PADfor a decision commission. When the project owes a change in use, carry that decision into the work and evidence the change requires; - stop at
A.3.1for Method identity orA.3.2forMethodDescriptionmembership, or at the A.15 family when the current question is method-work alignment, work planning, readiness, or performed work; - use
C.16,C.25,C.32.ACE,E.23orG.11for a current measurement, quality-bundle, eval, improvement or currentness-refresh question; useC.29when mathematical representation choice, transfer or reliance remains unresolved; - reconsider P2S only when a later subject assertion establishes architecture pressure that changes candidate structures, expected structures, actual structures, selected structures, or the stronger-structure inspection condition.
Archetypal Grounding
Carry the outside need through to what the commission requires. An explanation may establish how existing relations produce the relevant property. A design shows how the proposed relations would do so. For installation, establish the work agreed with the recipient, including qualification when it belongs to acceptance. For a capability, establish what the participants can now do under the agreed conditions. When the commission calls for use to start, continue or cease, establish the required start, continuation or cessation in practice. An actual transformation has its own A.3.4 basis.
A throughput shortfall after a layout change. A plant architect finds that a production cell delivered fewer accepted units per shift than the layout design predicted. The count establishes the shortfall for the observed shifts. It leaves open whether a buffer was installed on a different route, the input mix changed or the cell spent longer waiting for release. Identify which of those differences would change the proposed action.
The approved layout requires output from station B to bypass the shared fixture. Inspection shows that B's conveyor actually feeds the fixture. This establishes a discrepancy in the installed connection. The architect can call for conformity to the approved layout where that decision is binding. To claim that rerouting will restore the missing output, trace which operations waited for the fixture and what would still limit accepted output after the change. If inspection instead confirms the bypass, keep the throughput shortfall and investigate the constraint that remains.
A short P2S account can therefore record a production-cell performance question first. Record actualStructureDivergesFromExpectedStructure for the inspected conveyor once its actual connection has been compared with the binding layout. Keep the shift data and any still unidentified bottleneck separately recoverable. Use C.32 to develop changes to the connections implicated in the shortfall. The output record alone supplies no particular buffer design.
Lens-use slice. If the plant team builds a DSM or epiplexity-style model over stations, buffers and routing events, P2S retains the architecture inference and its grounds: the dependency or learnable structural content preserved, any consequential flow distinction omitted, and the structures the result can inform. Use C.29 when choosing or transferring that representation, or relying on its result, remains unresolved; an adequate domain treatment needs no additional lens account. Establish any architecture-adequacy, eval or decision conclusion separately.
Built asset and technical system. A clinic has rising instrument-turnaround delays and infection-control pressure. The first P2S move does not ask for a better diagram. It names the clinic service system as described holon, the turnaround-and-contamination architecture question, intended operating use, applicable scope and window, candidate structure kinds, architecture characteristics, and uncertainty: room layout, sterile and contaminated flows, equipment modules, tray interface, maintenance work, throughput, contamination isolation, maintainability, and surge adaptability.
Candidate synthesis develops these alternatives: a centralized autoclave bay, distributed sterilization modules, and a reusable tray-interface change. Use C.32.PAD for the decision selecting one configuration, C.30.AD and C.32.ADR for its descriptions and publication, and A.15-family patterns for construction and operating work.
During operation, measure turnaround, contamination events, maintenance burden, and actual-structure feedback triggers.
Method, organization and Work. Inspection work catches ontological errors late. Recover what the source's "review practice" names before choosing the described holon. When the question concerns how source recovery, judgement and repair connect, the bearer can be the Method itself under A.3.1; candidate alternatives can change that sequence or its feedback.
When staffing or coordination among performers is the problem, recover the review organization-as-system and its relevant structures.
If a performed review is the subject, recover each performer through A.13 and identify the U.Work occurrence independently through A.15.1.
Use C.32.HCS and C.32.ACS to bind characteristics to the bearer recovered for the question and the required result. Use E.10.ROLE to separate any local checker kind, classification, current assignment, direct review-relation position, function claim, organization position, and ordinary audience label. Candidate synthesis develops modal alternatives. Establish any claimed actual subject relations independently. For later actual inspection Work, add F.6 only when the flow or receiving use expressly represents precise assignment-bound attribution. Compare the phase at which the same kinds of errors become detectable under each arrangement. Records of detection can establish when particular errors were found. A rise in recorded early findings may also reflect more errors in the incoming material or a change in recording.
Architecture-influence and transformed-side co-synthesis. A team wants a modular product architecture but its toolchain, team communication, release method, and evidence workflow only support one tightly coupled build. The team uses C.32.CONWAY within the P2S flow: those typed influence sources and their direct relations remain separate from both C.30 architecture sides, the changed referent, any actual transformation, acting Systems, assignments, and Work.
Candidate families include changing the product modules only, changing the influence-source structures only, changing both, or accepting a bounded mismatch while retaining a named correspondence-frame return condition. The decision states which side changes now, what architecture characteristics are protected, what Work and work-to-change relations realize the change, and what operation or delivery feedback can return to the C.32.CONWAY correspondence frame or to decision repair.
PumpSkid 7 realization and return. Architecture pressure calls for a modular pump-skid with selected module and placement structures; use C.32.PAD for the project architecture decision.
Assembly work W-PS7-ASSEMBLY and later commissioning work W-PS7-COMMISSION are separate Work occurrences admitted under U.Work by A.15.1. Mounting change T-PS7-MOUNT, wiring change T-PS7-WIRE, fluid-connection change T-PS7-CONNECT, and commissioning-related change T-PS7-COMMISSION are each independently identified under A.3.4. Where the assembly or commissioning work is asserted to cause a change, cite the work-to-change relation and its declared predicate. Shared work, temporal adjacency, and one selected pump-skid configuration do not establish one composite transformation.
The local A.15.PROD entity-identity-inception claim uses the applicable PumpSkid 7 identity-specification edition, its applicability basis, work-to-change and change-to-identity facts, and inceptionBoundary; it need not assert a composite transformation. Later commissioning can remain production work until subject-state facts at completionBoundary satisfy the applicable production-completion-criterion edition; only then can a separate historically indexed completion claim be written.
A later C.30.AD or C.30.ASV record describes the actual structure; use C.32.ACE to evaluate service-access coupling. When coupling is worse than the selected expectation, the eval result returns to decision repair or new synthesis. Neither the description, evaluation, identity-inception claim, completion claim, nor shared assembly work proves conformance to the selected architecture or transformation composition.
Workshop capability with finite acceptance. A workshop-development project must enable the delivery of machined components with an agreed surface finish under the expected order flow. The workshop and receiving organizations have agreed supply and acceptance obligations for those components. The components are the SoIs of their extended enterprise; the workshop is the architectural subject. One candidate adds a machine, another combines that machine with a revised production Method, qualified operators, suitable measurement and secured material access, and a third obtains suitable external machining. The architect compares their actual capacity, handovers, cost and provision conditions.
A machine-only candidate cannot establish the workshop capability while the measurement Method or skilled operator is absent. A bounded acquisition project may still deliver and accept the machine. For the capability project, choose a supported combination or a bounded investigation of the unresolved fit; carry that distinction to PAD.
Acceptance can demonstrate the agreed production capability under representative conditions without waiting for every future component to operate. Later use can reopen the product or capability hypothesis without erasing completed acquisition Work.
Bias-Annotation
Use these rows as repair cues for problem pressure, source-practice transfer, or observed signals.
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Consequences
The project gains one replayable architecturing flow from pressure to actual-structure feedback. Practitioners can see where the Work currently stands and which assertion is next, with the relevant pattern as a locator, without treating descriptions, decisions, eval results, Work occurrences, or separate records about them as interchangeable.
The cost is maintaining the connections among structural uncertainty, alternatives, accepted losses, receiving work and needed inspection. Reuse an adequate account rather than creating another record. If one narrower pattern already answers the question, use it directly and do not open P2S.
The pattern improves cross-holon and adjacent-structure reuse. Practitioners may apply the same P2S method and Plain action sequence to different project-local flows in system, built-asset, product-family, organization-as-system, episteme, AI-agent-setup, discipline, and C.36-recovered cultural-evolution cases, after identifying the described holon in each case. Sharing that guidance does not give the flows one cross-holon identity or turn the examples into performed-work order.
When architecture pressure concerns source wording such as roles, methods, practices, cultures, traditions, or styles, name the described holon and locality separately. Keep each recovered local kind, classification judgment, assignment, direct-relation or representation position, Method, Method relation structure, MethodDescription, Work claim, canon or memory episteme, recognition or selection regime, and mediation-system claim with the pattern for that claim.
The pattern does not guarantee adequacy. It makes the architecturing flow inspectable. For candidate quality, decision adequacy, evidence, assurance, gate passage, release, measurement validity, and G.11 currentness refresh, use the relevant patterns.
Rationale
The connected architecturing work includes recovering problem pressure and structural uncertainty, generating candidate selected-structure changes, preserving alternatives, making decision-ready content, and returning actual-structure feedback to the next synthesis question. During realization, identify and test every actual bounded change under A.3.4.
The next action depends on what is known about the structure and what the receiving use still needs. Candidate construction can propose a different arrangement or constrain choices a design left open. For a claim of strict model refinement, compare the admissible alternatives for the same inquiry under B.5.MR. A decision selects what to pursue, and domain work can make the selected structure obtain. Descriptions and records retain only part of that content. Explain which distinctions the receiving description, decision, Method or Work needs and whether its available information preserves them. If an omitted distinction can change that use, recover it through stronger inspection, description or view repair, decision repair, or a C.29 mathematical lens as appropriate. An adequate account can be retained without another inspection.
SoTA-Echoing
These rows document transfers from source practice into C.32.P2S. Software-system sources are used as source families and examples only; they do not narrow P2S to IT architecture.
SoTA-anchor currentness boundary. Use each source contribution within its stated domain and assumptions. When a source result, receiving use or relied-on condition changes, reconsider the affected construction or inference; retain unrelated results whose grounds still hold.
Relations
Builds on:
Uses:
Patterns for the next questions:
Boundary: Use C.32.P2S for the connected project-local architecturing transformation flow from architecture-relevant pressure to subject-side actual structures recovered under A.22 from obtaining facts, and then to feedback.
C.33, C.34, and C.35 deepen the structural-information slot group already present in P2S; they do not move the whole connected flow out of P2S.
C.32.P2S does not define or test architecture claims, architecture descriptions, structural views, candidate palettes, comparison, selected-set result declarations, decisions, ADR-like publications, publication forms, publication-use claims, methods, Work, measurement, eval, evidence, assurance, gate, release, improvement, currentness refresh, or formal structural-information theory.
C.32.P2S:End
Architecture-Bearing Family Characteristic Starter Packs
Type: Architectural characterization subpattern under C.32 Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when a practitioner must choose a few architecture-characteristic starter heads for a described holon or for method, system-role-assignment, work, evidence, or cultural-evolution structures recovered from a source label, and the available catalogues are too broad to choose the first project criteria rows.
Primary working reader: an architect or architecture-responsible practitioner choosing a small first set of architecture-characteristic heads for an admitted holon family or another recovered architecture-bearing family, after naming the described holon or source-bearing episteme or publication context and any recovery patterns used.
Typical entry phrases:
First-minute use slice. A review lead must choose qualities for a reusable review method. Recovering the source word "practice" through A.3.1 identifies the Method itself as the described non-agentive holon. The lead inspects how its parts connect evidence, judgement and repair, then asks which of repeatability, transferability, evidence reuse and exception growth matters for the receiving work. A particular completed review is a separate A.15.1 Work occurrence; its actual coverage, elapsed time and rework require their own observations. The organization that performs reviews is another possible System, with different coordination and staffing questions. Teachability is a likely C.25 Q-Bundle. HCS returns these distinct starter questions to C.32.ACS; it does not replace the Method with the performing organization.
The primary EntityOfConcern is one architecture-bearing family starter pack for beginning to turn broad architecture-characteristic names into project criteria rows. A starter head is only a possible characteristic head before project bearer, scale, use class, proxy risk, and protected counter-characteristics are bound. Carry admitted starter heads to ACS. Keep Q-Bundles, measurements, eval programs, candidate palettes, comparison rules, G.5 result declarations, actual publications, and architecture decisions as separate objects handled by their applicable patterns.
Ordinary working move: choose the starter pack for the admitted holon family or recovered architecture-bearing family, keep only the heads that plausibly fit the project, ask the first project question for each head, then hand those heads to C.32.ACS for bearer, scale, and use-class binding.
The first useful result is a small set of plausible characteristic heads and the questions needed to bind them to the project. An ordinary account can suffice. Use the following optional ArchitectureBearingFamilyCharacteristicStarterPack@FPF record when a reusable family guide helps. Its suggestions remain candidate heads; project criteria and stronger uses need their own grounds:
Use describedHolonRef when the starter heads concern a particular holon. Use presentationCarrierRef only when the carrier itself changes how the starter pack is presented or used; do not fill it as a substitute for the described holon.
A copied quality catalogue can hide what bears the characteristic, especially when software-module names are applied to a Method, organization, built asset or evidence workflow. Recover that object before narrowing the heads.
The practitioner gains a small starting set for C.32.ACS to bind to project use. The later choice of optimization indicators and guardrails depends on that use; mandatory constraints remain binding.
The result is a short starter set with first project questions. Likely bearers and scales remain proposals until their project binding is established.
Not this pattern when the project already has admitted architecture-characteristic rows with bearers, scales, and use classes. Also not this pattern when the current work is composite-quality modeling, measurement, eval design, candidate synthesis, comparison, selected-set result declaration, actual publication, local choice, or project architecture decision.
Common exits by claim kind:
C.32.ACSfor project criteria rows.C.25for Q-Bundles and composite quality families.C.16for measurement andC.32.ACEfor eval programs or eval results.E.13when a source-looking cue, score, benchmark, or dashboard starts replacing the architecture concern.C.32for candidate synthesis when the relevant characteristic pressure is recoverable; use existing criteria or develop missing ones through ACS.A.19.CPMfor explicit comparison andA.19.SelectorMechanismfor set-returning selection.G.5for selected-set result declaration,C.11for local choice, andC.32.PADfor a project decision. For publication, useE.24.PUBfor the publication occurrence and audience availability, andE.17when a source-backed face and source return are needed.
Problem
Characteristic heads such as reliability, substitutability, change reach, evidence reuse, control separation or coordination load recur across admitted holons and across method, system-role-assignment, work, evidence, or cultural-evolution structures. The same head may require a different bearer, scale, use, or pattern definition or test. Recurrence alone neither establishes that a source-labelled method-like object satisfies U.Method, nor turns a local system-role kind into a holon, nor admits practice or culture as holon kinds.
A function depends on what performs it, while a functional demand may also depend on the holon that needs that performance. A saw-as-system can cut; a system may satisfy a local system-role-kind criterion; an obtaining system-role assignment may place that system in work-facing use; and a responsibility relation may hold independently. A method description can guide work, an enacted work family can be repeatable, and an organization-as-system can coordinate. A culture or practice label must first be resolved into systems, disciplines, method and work families, local system-role kinds and assignments, canon or memory epistemes, recognition and selection regimes, or mediation systems. A project therefore needs starter packs that suggest common heads while requiring the practitioner to re-identify the bearer, any separate demand holder, and the scale before optimization.
Forces
Solution
Choose a starter pack by the admitted holon family or recovered architecture-bearing family. Use the pack only to start narrowing starter heads into project criteria rows; then hand the result to C.32.ACS for the project criteria set.
Starter pack construction
Build or use a starter pack in this order:
- Name the admitted holon family or recovered architecture-bearing family. If the source label is method, role, practice, culture, tradition, style, or evidence practice, recover the bearer before choosing heads. A Method admitted by A.3.1 is itself a non-agentive holon; its description, a Work occurrence under A.15.1, and the performing System remain separate. When the label is only a description-side family, name the source-bearing episteme or publication context. Record only the recovery patterns used.
- For a single project, propose characteristics from the work the recipient needs and the relation that could change it. To recommend a question across a family, establish the conditions under which it helps, using a common requirement or evidence covering the intended projects. A locally useful characteristic can remain a candidate for wider use.
- For each head, name likely bearers or selected structures, not only a quality word.
- Record likely C.25 Q-Bundle boundaries when a head is usually composite.
- State a first project question that helps the practitioner decide whether the head belongs as a draft row in the project criteria set.
- Hand the resulting starter heads to
C.32.ACS; do not optimize or measure inside HCS.
Choose with four distinct considerations: the common characterization method, the bearer kind, the domain or practice, and the concrete use and interests. These considerations intersect. A software-development Method and an instructional Method share a bearer kind but serve different work; a biological system and Work performed in the same project share a domain but have different bearers. Neither domain membership nor a shared quality name supplies the missing structure, value meaning or observation.
Built-in starter packs
In HCS, source-return cost is a starter head for a holon family when recovering the source could impose decision-relevant effort, delay or exposure, once or repeatedly. The return is from a derivative, coarsened, extracted, rendered, or reused publication or evidence carrier back to the named source expression, selected source U.Episteme, EpistemePublicationRelation occurrence when availability matters, source-bearing relation, evidence-provenance entry, evidence relation, transform record, or defining ClaimGraph needed for stronger reliance. It is not a generic source-quality name. If the project is only asking whether a catalogue term is useful, keep the wording as source catalogue wording; if recoverability itself is the concern, carry source-return cost to C.32.ACS and bind its bearer, scale, and use.
A commissioning team may need a single archived reading before releasing an installation. If recovering it would stop work for two days, its cost already matters. Frequency matters when the project decision depends on cumulative burden.
Rebinding rule
When a starter head is reused at another bearer kind, holon level, domain, or concrete use, rebind it. Identify the target bearer, relevant relations and receiving result. State what is known about the conditions, quality meaning and attainable observation; carry unanswered binding questions to ACS. State which part of the source construction survives and which part needs replacement. The reusable item is a candidate head, not an already valid project row.
Example: availability for an engineered service may use time-window and service-scope measures. A method-side case may ask whether a particular System can access a Method description and evidence relation in the working situation. A kind case may ask whether A.2.7 admits substitution; an assignment case may ask whether holder replacement or assignment continuity obtains; a responsibility case needs its own relation. These are different bearers, predicates, and scales.
Refresh the starter pack when a change in the family, source, bearer or use changes which heads help a project establish its criteria. One project can expose a missing head or refute a claim that every eligible project needs a particular one. Several projects in one domain can justify narrowing the guidance for that domain. To move a question from the usual starting set to guidance for specialist projects, establish where asking it first helps and where it can wait. Keep the question available wherever it still matters. A claim that no eligible project needs the characteristic requires grounds covering the whole family. Retain project-specific criteria in C.32.ACS, measurements in C.16 and evaluation programs in C.32.ACE.
ACS Criteria-Row Use
HCS stops with starter heads and first project questions. The next C.32.ACS use governs:
- whether C.32.ACS admits the head as a draft project criteria row;
- whether it is one characteristic or a C.25 Q-Bundle;
- whether the project uses it as an optimization indicator, monitored guardrail, or context-only row;
- which scale, reading, and pattern for the next question apply.
Before ACS criteria-row use, ask whether improving a visible catalogue, benchmark or dashboard value could hide a loss in the architecture concern. Retain a plausible action-changing risk; do not invent one to qualify the head. Carry a head when the architecture-bearing family, likely bearer, concern and first project question are recoverable. A likely scale or composite-quality boundary can remain an explicit question for ACS or C.25. Preserve the source item and relevant source-to-use connection for a borrowed head, or the inspection basis for a directly recovered one. Name formal source identity or publication availability when the receiving use depends on it. If no architecture concern can be recovered, keep the term as source vocabulary or remove it from the starter set.
Stop condition. Stop C.32.HCS when the family, plausible starter heads, likely bearers or selected structures, and first ACS questions are recoverable. Carry a suspected composite boundary or unresolved scale as a question for C.25 or ACS, and retain a caution only when it changes use. The next project criteria-row work belongs to C.32.ACS.
Lowering condition. Keep a term as source vocabulary, or remove it from this starter set, when its architecture-bearing family, likely bearer, concern or first project question cannot be recovered. Repeated rejection during ACS binding can justify narrowing or revising the family guidance; inspect why the head failed before excluding another use. An unresolved scale or composite boundary is a question to resolve before stronger reliance, not evidence that the head is irrelevant. Use C.25 for composite quality and ACS for project binding; measurement, evaluation, comparison and decisions retain their own conditions.
Worked slices
Engineered-system family. Reliability, maintainability, substitutability, evidence reuse, locality and source-return cost are possible starting questions for a field device. When a particular operator requires continuity during replacement and prefers less service work, C.32.ACS develops those two criteria and their distinct roles. Other projects can impose different requirements.
Method and Work. A reusable review Method connects source recovery, comparison, judgement and correction. Evidence reuse and change reach are plausible starter heads because changing how a dependent claim is found can alter both missed defects and repeated reading. A completed review has a separate Work question: were the required comparisons performed, and what delay or rework occurred? Improving the next Method version cannot retroactively change that Work. If the issue is reviewer staffing, select the performing organization and its contributions instead. Local system-role-kind substitution uses A.2.7; assignment continuity and Work coverage retain their own bearers.
Same bearer, different domain. An instructional Method also has parts, order and feedback, but its receiving result concerns what a learner can do under specified support. Removing corrective feedback may simplify the description while weakening the Method. Carry the learner result and help conditions into ACS; do not reuse software throughput or readability as if either measured learning. Teachability is composite and uses C.25.
For example, a correct architecture diagram may reproduce the solution included in an exercise. To assess the learner’s design work, recover the decisions the task left open, the help available, what the learner actually designed and the grounds they supplied. A learner who set and justified a timeout can receive credit for that choice. Credit for the complete design also requires the structural decisions the assignment placed with its designer. Earlier project work can already provide that evidence. If the purpose is to apply an approved design, supplying it is appropriate.
AI-agent workflow. A retrieval-action setup starts from evidence refresh, policy controllability, latency, observability, and rollback. Benchmark performance stays a benchmark signal or comparison input until an architecture bearer and scale row are named.
Starter-pack proxy near-miss. A review team copies availability, throughput, and testability from a software quality catalogue because the list looks mature. The copied heads make the starter pack look complete, but they hide exception growth, evidence reuse, kind substitution, assignment continuity, and Work coverage, which have different bearers and predicates. C.32.HCS keeps the catalogue terms as source wording; A.3.1 and A.15 separate the Method, descriptions, assignments, and Work, while A.2.7 supplies kind-relation structure and substitution only for kinds. HCS carries only the recovered questions to C.32.ACS.
Receiving-Claim Boundary
Use C.32.HCS only to build architecture-bearing family starter packs. For work beyond starter-pack construction, use the exits in §1. C.32.HCS neither establishes a source-labelled object as U.Method, nor turns a local system-role kind into a holon, nor admits practice, culture, tradition, or style as holon kinds.
Conformance checklist
Common failures and repairs
Consequences
Rationale
The 300-to-3 problem needs a middle step. A project cannot optimize from a catalogue, but it also should not invent criteria from scratch. Architecture-bearing starter packs give a small, recognizable entry.
SoTA-Echoing
These sources inform the starter-pack fields, ACS criteria-use conditions, and blocked overreads described below.
The bearer/domain distinction in §4.1 also uses C.30.ASV:4.5a and the developed instructional-design example in van Merriënboer, 4C/ID: An Overview of Main Design Principles (2019; author-hosted copy uploaded in 2021), its explanation of whole tasks, support and feedback. This is retained design lineage: it shows why a Method’s parts and their relations can matter to a learner result. HCS adopts that question, not a claim of instructional effectiveness for an untested population. Reopen the transfer when the target Method needs a relation or outcome the recovered structure does not represent.
Source-currentness boundary. Use ISO/IEC 25010:2023 as ICT product-quality vocabulary, not as holon-family ontology. Use the O'Reilly architecture-characteristic and evolutionary-architecture sources for recurring starter heads and for later metric or eval work after the heads are named. Use an FPF row only for the claim it defines, constrains, or tests. Reopen HCS when a named source edition changes starter-head guidance, when the pattern for a next question changes how it handles that source family, when an applicable project requires a missing head or shows that a head has a different range of use, or when source-label recovery changes the recovered family or bearer.
Relations
- Receiving use:
C.32.ACSproject criteria-set construction, including scale rows and use classes when the project later needs them;C.32.P2Swhen starter heads are needed before the architecturing flow can bind project criteria, candidate synthesis, eval, and refresh. - Uses:
C.25when a starter head is composite;C.30andC.30.ASVwhen the selected structures are not yet recoverable. - Boundary: HCS is not a catalogue, measurement pattern, Q-Bundle pattern, optimization method, or architecture decision pattern.
C.32.HCS:End
Architecture Characteristic Criteria Set
Type: Architecture characterization pattern under C.32 Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when a project must turn architecture-characteristic pressure into a small project criteria set for architecture improvement, candidate synthesis, residual optimization, and later eval work.
Primary working reader: an architect or architecture-responsible practitioner turning broad quality names into project criteria rows for the next improvement cycle.
Typical entry phrases:
First-minute use slice. A field-service operator requires continuity during replacement and prefers less service work among designs that meet this requirement. Use continuity in the permitted replacement conditions as an admission criterion and field work as a preference. Two designs can give the same downtime for each unit but different service availability when their outages can coincide. If the commission concerns one unit, retain that boundary. For each criterion identify the bearer, applicable conditions, scale or comparison, consequential limitations of any proxy, and grounds for returning to the source. A scale-amenability preference across a defined window belongs to C.31.ASAP.
This pattern constructs a project architecture-characteristic criteria set for the receiving improvement or evaluation question. The optional records below can keep its bindings recoverable. Its rows can supply C.32 synthesis, C.32.MLAO residual work, C.32.ACE eval programs, and later patterns for the next questions. The named set and row records are C.32.ACS-local forms, not new U.* kinds; starter packs, U.Characteristic values, Q-Bundles, measurement methods and results, eval programs and results, candidate palettes, comparison rules, selection results, G.5 result declarations, actual publications, local choices, and architecture decisions remain separate objects.
Ordinary working move: identify the bearer and scale of the characteristic, then state the decision each use of its value supports. A binding limit determines admission. A preference can use the same quantity to rank the admitted candidates. Record the uses separately when their decision rules differ.
The first useful result is a set of criteria with their meanings, use classes and necessary bindings. Use ArchitectureCharacteristicCriteriaSet@Project when a record helps:
For a first pass, identify the described holon, architecture use, a small set of draft criteria and their possible bearers. Keep missing bindings as explicit questions. Before admitting a criterion for its receiving use, establish its bearer or selected structure, claim scope and relevant context slices, reference scheme, applicable plane, scale form and use class, with the temporal qualification, protected losses and reopen condition needed by that use. An incomplete draft does not yet support that use. Add readings, target bands, and eval-program references only when the current receiving use needs them; add a selected BoundedModelUseStructure only when it independently changes interpretation of the row use.
For ArchitectureCharacteristicCriteriaSet@Project and ArchitectureCharacteristicImprovementRow@Project, @Project is a compatibility and retrieval cue only; it establishes no project entity, composite-work identity, context, authority, viewpoint, or parthood. A criteria set or improvement row local to one project names both the exact composite U.Work in projectWorkOccurrenceRef and the obtaining direct use relation for that exact record in architectureCriteriaProjectUseRelationRef; either field alone is insufficient, and a relation occurrence about the set is not reused for a distinct row. Otherwise the record remains retrieval-only and no project locality is asserted.
If precise Work is asserted, recover each performer System's A.13 core, then independently admit the dated occurrence under A.15.1 from its performance history, Method followed, temporal extent and required containing-System relation. Add F.6 only when the receiving use consumes precise assignment-bound attribution through the same obtaining A.13 assignment and its declared species. F.6 identifies neither assignment nor performer; missing or failed F.6 attribution alone leaves the Work intact while its independent A.13/A.15.1 grounds hold.
draftProjectCriteriaRows are draft project criteria rows. They are not candidate architectures, selected architectures, or a selected set returned by A.19.SelectorMechanism.
What goes wrong if C.32.ACS is missed: the team says that the architecture should be more maintainable, scalable, modular, safe, or evolvable, but no one can say which selected structures carry the characteristic, which few rows are criteria for the next optimization, which rows only guard against loss, which C.25 Q-Bundle is involved, or which eval result can reopen synthesis.
The practitioner can narrow catalogue or starter material to criteria that change the next action, with optimization indicators distinguished from binding constraints and safeguards against proxy loss.
Adoption test: after using C.32.ACS, the project can name the few rows that drive optimization, the guardrail rows that protect against loss, and the bearer, scale, proxy risk, receiving use, and reopen condition for each live row.
Not this pattern when the current work is choosing the holon-family starter pack, modeling a Q-Bundle, validating a measurement method, designing an eval program, synthesizing candidates, comparing or selecting candidates, choosing locally, declaring a selected-set result, publishing it to an audience, or deciding the project architecture.
Common exits by claim kind:
C.32.HCSfor holon-family starter packs.C.25for Q-Bundles and composite quality families.C.16for measurement templates, readings, units, thresholds, or comparability claims.C.32.ACEfor eval-program framing and typed-result classification over declared rows; obtain each result separately under the pattern that defines and tests its kind and exact predicate or constraint.E.13when an indicator, score, or dashboard starts replacing the declared architecture concern.E.22andE.23for improvement-question framing and repeated improvement method.C.32for candidate synthesis andC.32.MLAOfor residual-reducing candidates.A.19.CPMfor explicit comparison,A.19.SelectorMechanismfor set-returning selection,C.11for local choice, andG.5for selected-set result declaration. For publication, useE.24.PUBfor the publication occurrence and audience availability, and E.17 when a source-backed face and source return are needed.A.10for source recovery and bounded reliance,G.6when an addressable provenance path is needed, andB.3for a named assurance claim; use the pattern that defines or tests any additional evidence or validity claim.C.32.PADfor project decision.
Problem
Architecture synthesis needs criteria. A multi-criteria or multilevel optimization phrase is empty until the criteria are named. In C.32-family work, those criteria are admitted architecture-characteristic rows or declared C.25 Q-Bundle slots of the described holon, each bound to its exact bearer, U.ClaimScope, relevant A.2.6 U.ContextSlice membership, effective reference scheme, applicable plane and any qualification or evaluation window required by the use, and receiving use. A broad domain or bounded-context label supplies none of those bindings.
Architecture characteristics are not the same as user functions. Functional demand says what the holon must do. An architecture characteristic says whether the selected structures make that demand maintainable, controllable, replaceable, observable, evolvable, scalable, affordable, safe enough, or otherwise acceptable.
Source catalogues and textbooks can offer hundreds of possible quality or architecture terms. A project may inspect dozens. The optimization loop should normally use only a few indicatorized rows, often three to five. Other important rows remain monitored guardrails or context-only rows so that optimizing one visible measure does not damage functional adequacy, safety, evidence, maintainability, or another protected architecture concern.
C.32.ACS supplies the project criteria set and scale rows. It does not create the holon-family starter pack, define a Q-Bundle, validate a measurement method, run an eval, compare candidates, choose an architecture, or decide the project architecture.
Forces
Solution
Construct or reuse criteria from the project improvement question, architecture constraints and relevant starter or source contributions. Use the criteria-set record when it helps keep their bindings and use classes recoverable.
Kind settlement
ArchitectureCharacteristicCriteriaSet@Project is a C.32.ACS-local project working record: it holds criteria-row references and use classifications for improvement work. Each draftProjectCriteriaRows entry is another local record form, not the referenced U.Characteristic, Q-Bundle slot, scale, predicate, measurement result, eval program, or eval result. The set and rows create no new U.* kind and replace none of those objects.
A characteristic matters to an architecture decision when it can admit or exclude a design, distinguish eligible designs, or protect a result at risk. The same constraint can reject every candidate examined. A 10 A limit rules out two designs that each require 12 A, even though their current demands are equal. Conversely, a fixed interface can leave only one permitted design to check. Identify the bearer, conditions and architectural dependence. A proposed change in a characteristic needs grounds in that dependence before it can count as a gain.
A C.25 Q-Bundle is the structured form for a composite quality family. A scale row binds one characteristic or Q-Bundle slot to a bearer, scale form, use class, and receiving use. A row whose scale form exposes exception growth, interface variation, or another scale-sensitive characteristic remains a criterion row; a preference between architecture alternatives over a declared scale window is a separate C.31.ASAP claim. An architecture-characteristic eval program belongs to C.32.ACE; it frames evaluation of one declared row, coupled rows, Q-Bundle slots, or C.32 candidate palettes while each typed result remains under the pattern that defines and tests that result.
Criteria-set construction
Work in this order:
- Name the described holon, architecture use, and improvement cycle or one-pass eval use. Connect the receiving role and situation to a needed result, the quality that matters, the structures or binding constraints that determine it, and an observation, justified bound or existing result that can answer the criterion. Keep a Method, its description, a performed Work occurrence and the performing System distinct when choosing that bearer. For every admitted row, bind the claim scope and relevant context slices, effective reference scheme and applicable plane, with any qualification or evaluation window required by the criterion or receiving use. Preserve every declared slice selector. A proposed row can retain a missing binding as an explicit question until admission. Designate a selected A.1.1
BoundedModelUseStructureonly when it independently changes that row's interpretation. - Use C.32.HCS when a family starting point is needed. Directly recoverable project pressure or adequate existing criteria can supply the starting point. Source catalogues supply possible heads whose project bindings need examination.
- Build draft project criteria rows. There may be dozens of draft rows when broad scanning is needed, but each row must have a possible bearer, use reason, and pattern for the next question.
- For each source or starter head, decide whether it is one architecture characteristic, one C.25 Q-Bundle, one Q-Bundle slot, or only source vocabulary.
- Narrow the optimization-indicator core. Three to five rows is an attention heuristic. More rows require a use reason, such as a regulated trade-off study or a multi-team decision. Preserve every applicable non-compensable constraint and necessary guardrail; the small core does not limit their number.
- Classify remaining admitted rows as
monitoredGuardrailorcontextOnly. A guardrail protects a binding condition or guards against a loss caused by optimizing another row; a context-only row helps interpretation but does not drive optimization now. - Bind each admitted row to its bearer or selected structure, scale form, applicable polarity, current reading or no-reading reason and receiving use. Examine material proxy risk and protect the characteristics or constraints it could compromise. Add a source-return condition when returning to omitted source detail can change that use.
- Reference
C.32.ACEonly after the row exists and an eval program is needed for current characterization, candidate comparison, monitoring, or preparing inputs forA.19.SelectorMechanism. - Reopen the criteria set when the holon family changes, a B.2 whole reidentification changes the bearer, a guardrail degrades, an eval program no longer fits its use conditions, including a parity frame when the inference relies on one, or the source-currentness relation changes the acceptable trade-off.
Row use classes
Use optimizationIndicator only when the row can responsibly guide architecture changes now. A project normally carries only three to five such rows.
Use monitoredGuardrail when the row protects a binding condition or guards against a loss caused by optimizing another row. Guardrails can have readings and eval results, but they do not define the cycle's optimization direction.
One quantity can serve both decisions. If the service must respond within 100 ms and faster service is preferable among qualifying designs, use one row for admission and a second for the preference, with the same applicable latency measure. A lower price cannot compensate for breaching the hard limit. Identify whether that limit applies to every request, to a specified proportion or to the declared aggregate.
Use contextOnly when the row helps interpretation but should not drive improvement, comparison, or selection in the current cycle.
Stop condition. Stop C.32.ACS when the criteria needed for the receiving question have their use class, bearer or selected structure, scale form, relevant scope and conditions, material proxy risk and protected losses, plus any ACE, Q-Bundle or source-return contribution that the use needs. A precise unresolved draft can finish a question about what remains missing; it cannot stand in for a usable criterion.
Lowering condition. Remove optimization use when a row no longer guides the next change or its material proxy risk is uncontrolled. Classify it as a guardrail or context only when it can support that use; otherwise retain the unresolved question and stop reliance on it. A missing scale or other binding can remain in a draft for repair. Keep a term as source vocabulary only when its project concern or use cannot be recovered. Use C.32.HCS for a missing family starting point and C.25 for a composite quality; measurement, evaluation, comparison and decision uses retain their own conditions.
Improvement-cycle use
When a row is used inside an improvement cycle, add:
The row prepares improvement work. It does not carry a claim outside its declared scale and use. A reading or other typed eval result over a declared row retains its result kind and conditions; another pattern may use it as source material for bounded claim reliance under A.10, improvement feedback, comparison input, selection input, or decision input only when that pattern for the next question is named by value. It does not become the characteristic, the declared architecture concern, the architecture choice, or the optimization direction.
Worked slices
Manufacturing cell. If the operator must keep the cell available during setup changes and wants to reduce work per change, evaluate continuity under the permitted change and compare setup work among the designs that qualify. A further criterion belongs in the set when its outcome could change the commissioned finding or choice. C.32 can use the resulting criteria to develop fixture candidates.
Method-family architecture. A review lead needs a method that detects affected dependent claims without repeating unrelated reading. HCS supplies evidence reuse, change reach and exception growth as possible heads. ACS prepares criteria for comparing a fixed local search with a method that follows the changed claim into its direct uses. The bearer is the Method; the relevant structure is the connection between source recovery and dependent-use comparison. Candidate rows concern missed dependent claims under the declared task family and effort to obtain a useful review, with preservation of correct content as a guardrail. Known defective and difficult admissible cases can support a bounded comparison; a claim about actual release effort requires observations from performed Work. Teachability remains a C.25 question when learner preparation and support matter.
AI-agent architecture. Suppose a service helps archivists answer from the applicable edition of a rule before their decision deadline. Freshness and timely retrieval can then affect the same answer. If the commissioner also wants local policy owners to correct an obsolete routing rule, evaluate that ability in the configurations proposed for use. Keep observability or rollback as design options until a required use or constraint gives them a role in comparison. For a reading from a single benchmark run, state what was evaluated, the scale, the workload and what the reading can establish. A comparison across designs also depends on the operating conditions that must correspond for that comparison.
Team and assignment architecture. A hospital escalation team starts from coordination load, accountability clarity, decision latency, evidence custody, substitutability among local system-role kinds, and continuity of assignment occurrences and their holder Systems. ACS creates separate criteria because A.2.7 can compare or relate local kinds but does not substitute holders. The kind-substitutability row binds the exact local kind-relation structure and predicate; the assignment-continuity row binds the exact assignment occurrences, holder Systems, and continuity predicate. ACS marks decision latency, accountability clarity, and evidence custody as optimization indicators, keeps patient-safety loss and assignment-continuity loss as guardrails, and leaves staffing choice to the receiving decision pattern.
Kind and Receiving-Claim Boundary
C.32.ACS governs project criteria-set construction for architecture improvement. It does not govern:
- holon-family starter packs, governed by
C.32.HCS; - architecture-characteristic eval programs, governed by
C.32.ACE; - C.25 Q-Bundle normal form, governed by
C.25; - C.16 measurement templates or readings, governed by
C.16; - C.31 modularity and reusable-structure characteristic repair, governed by
C.31; - C.31.ASAP scale-preference claims, governed by
C.31.ASAP; - E.22 question framing and E.23 repeated improvement method, governed by
E.22andE.23; - C.32 candidate synthesis, governed by
C.32; - A.19.CPM comparison, A.19.SelectorMechanism selection, C.11 local choice, G.5 selected-set result declaration, E.17 source-backed publication-face and source-return work, E.24.PUB publication-occurrence and audience-availability work, or architecture-decision work for
C.32.PAD.
Conformance requirements
Common failures and repairs
Consequences
Rationale
Architecture optimization is meaningful only after the criteria are named. ACS supplies a project criteria set that can guide synthesis, residual reduction and repeated improvement.
The pattern stays holonic by allowing starter heads to recur across holon families while requiring bearer and scale rebinding. It stays action-facing by limiting optimization indicators and keeping non-optimized criteria rows as guardrails.
SoTA-Echoing
These rows document how source practice contributes to criteria-row fields, use-class rules, and receiving-pattern boundaries in C.32.ACS.
For the role-to-result construction in §4.2, Crawley, Cameron and Selva, System Architecture, Global Edition (2016), chapter 11, supplies a developed systems treatment of beneficiary, operator and supplier interests, benefits, goals and constraints. Adapt: recover these distinct interests before selecting structure and observation. Richards and Ford, Fundamentals of Software Architecture, second edition (2025), chapters 4–6, supplies the software branch of structural relevance; its non-domain criterion is not generalized to all bearers. Ciceri et al., Software Architecture Metrics, finished May 2022 edition, chapters 7 and 10, supplies the goal/question/measure and collection-cost connection. Together these contributions support §4.1’s use-relative architecture condition; they do not establish a universal catalogue or arithmetic over ordinal preferences. Reopen the transfer when the target bearer, domain or decision cannot be represented by that connection.
Source-currentness boundary. Use each source row only for the ACS field, use-class rule, or receiving-pattern boundary named in that row. Recheck the row when a named standard, book edition, source presentation, FPF pattern for the next question, or current architecture-evaluation line changes the transferred move. If the project wants measurement, eval-program design, comparison, selection, selected-set result declaration, actual publication, local choice, evidence, assurance, or decision use, leave ACS and open the pattern for the next question.
Relations
- Builds on:
C.32.HCS,A.17,A.18,A.2.6,A.19,C.16,C.16.P,C.25,C.30,C.30.P,C.31,C.31.ASAP,E.13,E.22, andE.23; uses A.1.1 only when a selectedBoundedModelUseStructurechanges one row's interpretation. - Receiving uses:
C.32.P2Sproblem-to-structure architecturing flow,C.32candidate synthesis,C.32.MLAOmultilevel residual work,C.32.CONWAYcorrespondence frames,C.32.FAILrepair cues,C.32.ACEeval programs,A.19.CPMcomparison inputs,A.19.SelectorMechanismselection inputs,C.11local choice inputs, inputs for selected-set result declaration underG.5, source-backed publication-face and source-return inputs underE.17, publication-occurrence and audience-availability inputs underE.24.PUB, and architecture-decision inputs forC.32.PAD. - Starter-pack boundary: Use
C.32.HCSwhen the project needs a holon-family starting set before criteria rows exist. - Q-Bundle boundary: Use
C.25when the architecture characteristic is a composite quality family with several measures, scope slots, mechanisms, statuses, qualification windows, or evidence. - Scale-preference boundary: Use
C.31.ASAPwhen a project claims that one architecture alternative is preferable over another under a declared scale window; the ACS row supplies a criterion, not that preference. - Eval boundary: Use
C.32.ACEwhen a project wants eval-program framing over declared rows, Q-Bundle slots, candidates, or selected-structure changes; state each typed result separately under the pattern that defines and tests its kind and exact predicate or constraint. - Measurement boundary: Use
C.16when a reading, coordinate, unit, threshold, score, or cross-case comparability claim is made. - Structural-information boundary: Use
C.33orC.34when the issue is captured structure, lost structure, or preservation adequacy before a criterion row exists. Use C.32.ACS only when that structural-information or preservation concern becomes a declared architecture-characteristic criterion row. UseC.35only for kind restoration and architecture-use adequacy of a generated or discovered result before C.32 or ACS uses a criteria-bearing claim based on that result. - Proxy boundary: Use
E.13when an optimization indicator, score, eval result, or dashboard state begins to replace the declared architecture concern. - Synthesis boundary: Use C.32 when characteristic pressure is recoverable and the next useful work is to synthesize candidate structures. Retain missing criterion bindings as unresolved; do not treat an exploratory candidate as an established improvement.
- Decision and publication boundary: Use
A.19.CPMfor comparison,A.19.SelectorMechanismfor selection,C.11for choice,G.5for selected-set result declaration, andC.32.PADfor an architecture decision. For publication, useE.24.PUBfor the publication occurrence and audience availability, and E.17 when a source-backed face and source return are needed.
C.32.ACS:End
Architecture Characteristic Eval Programs
Type: Architecture eval-support subpattern under C.32 Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when an architecture team has a criterion for an architecture characteristic, either in a project criteria set or in a declared C.25 Q-Bundle slot, and must evaluate the current architecture, compare candidate architectures, monitor evolution, or prepare a selection input.
Primary working reader: an architect or evaluator preparing readings over declared architecture-characteristic criteria without turning those readings into the criteria or the decision.
Typical entry phrases:
First-minute use slice. A product-family team must compare two architectures for substitutability, evidence reuse and latency, while preserving a safety constraint. Recover the workload, latency quantity and operating conditions of the criterion. A warm-start observation covers the warm start it measures. If the design also has an applicable bound of at most 10 ms extra for a cold start and at most 70 ms for a warm start throughout the required workload, the team can use the resulting 80 ms upper bound for a 100 ms limit. If starts may exceed that extra time, obtain evidence for the uncovered conditions. Check the test package against each changed interface for the separate evidence-reuse question. Each result keeps its subject, conditions and grounds. The bound settles the latency limit for the covered design. To compare both designs, obtain the grounds needed for each remaining criterion. For a configuration the team can commit to, establish that the results for the required criteria hold together there. For performance promised across operating modes or choices left to others, establish the requirement throughout those conditions.
This pattern frames architecture evaluation over declared criteria rows, Q-Bundle slots, candidates, bearers or selected structures. Use a parity frame when the receiving inference depends on comparable conditions. The optional record below uses a C.32.ACE-local form and introduces no new U-kind. Its program label establishes neither Method or MethodDescription membership, a WorkPlan, dated Work, nor an evaluation result. When measurement validity, comparison policy, a selection result, G.5 result declaration, publication, or architecture decision is current, use the definition and test for that claim.
Ordinary working move: select the declared criteria rows or Q-Bundle slots and bind the claim scope, relevant context slices and reference scheme, adding a plane, evaluation window or input projection when the evaluation or receiving inference depends on it, and establish parity for the variants when comparison depends on it. When a particular evaluation Work occurrence is claimed, recover each evaluator through A.13 and let A.15.1 independently admit the U.Work occurrence. Add evaluationWorkAttributionRefs only when the program or receiving use represents precise assignment-bound attribution; missing or failed F.6 attribution alone leaves the Work intact while its independent A.13/A.15.1 grounds hold. Any A.6.1 operation application is an additional relation, not a substitute for the Method enacted by the Work. Return the typed results established through their result patterns as feedback for comparison or the next synthesis pass.
The first useful result explains how a bounded architecture evaluation will obtain the result needed by the receiving use. An ordinary account can suffice; use ArchitectureCharacteristicEvalProgram@Project when a record helps preserve that framing. When a reusable way of evaluating is current, identify that separate U.Method under A.3.1; when a claim-bearing episteme describes that exact Method, test the same episteme for U.MethodDescription under A.3.2. A planned evaluation belongs to A.15.2, an actual dated evaluation to A.15.1, and each result to its measurement, comparison, evaluation, or assertion pattern. Use the record to frame readings of characteristics through rows, slots, candidates, or structures; it is not any of those neighboring objects.
For a first pass, make the evaluated rows or Q-Bundle slots, candidates or structures, purpose, operation, result form, receiving use and necessary conditions recoverable. Bind the claim scope, relevant context slices and reference scheme, including every declared selector. Include a plane, temporal window, input projection, parity frame or refresh condition when the evaluation or its use depends on it. A performed measurement retains its C.16 time stance; a dated Work occurrence retains its own extent. Add project-use, Method, MethodDescription, Work, operation-application and typed-result references only for those current claims.
Here @Project is a compatibility and retrieval cue only. It supplies no project entity, composite-work identity, context, authority, viewpoint, or parthood. A program local to one project names both the composite U.Work in projectWorkOccurrenceRef and the obtaining program-use relation in architectureEvalProgramProjectUseRelationRef; either field alone is insufficient. evalOperation states the intended operation family in the program record, not a run or application. Each evaluationWorkRef names one independently identified U.Work occurrence whose performers have A.13 cores and which A.15.1 admits independently. evaluationWorkAttributionRefs are optional and appear only when the record or receiving use represents precise assignment-bound attribution; any present ref resolves through F.6 to the same obtaining A.13 assignment, and its absence or failure alone leaves independently admitted Work intact. If that same finding questions a premise consumed by Work admission, recheck the premise and dependent conclusions. Any A.6.1 application binding and each typed result remain under the applicable patterns. A program, Method, MethodDescription, Work occurrence, operation application, assignment, attribution, and result never substitute for one another.
What goes wrong if C.32.ACE is missed: a project has architecture-characteristic rows but treats a test, monitor, dashboard, or source-side "fitness function" as the criterion or as the decision. The team may then reject useful losing variants as errors, optimize one indicator, or choose a candidate without fair comparison.
What C.32.ACE buys in practice: eval work is framed as typed evaluation over declared architecture criteria. A losing candidate can add knowledge about the solution space, while an error remains a failure against an expectation that causes unplanned rework.
The result identifies the candidates, bearers or structures, the conditions of evaluation, the intended result form and its receiving use. A comparative inference retains its parity conditions. An actual result is established separately under the rule for its result kind.
If neither an applicable C.32.ACS criterion nor a declared C.25 slot supplies what is to be evaluated, establish that criterion first. For measurement validity, composite-quality modeling, explicit comparison, selection, publication, evidence, assurance or an architecture decision, use the pattern governing that result.
Common exits by claim kind:
C.32.HCSandC.32.ACSwhen the evaluation still needs a criterion. A declaredC.25Q-Bundle slot can already supply it.C.16for measurement validity, readings, units, uncertainty, or comparability claims.C.25for Q-Bundles and composite quality families.E.13when an eval result or dashboard starts replacing the declared architecture concern.C.32for candidate synthesis,C.32.MLAOfor residual input, andE.23for repeated improvement feedback.A.19.CPMfor explicit comparison,A.19.SelectorMechanismfor set-returning selection,C.11for local choice, andG.5for selected-set result declaration. For publication, use E.24.PUB for the occurrence and audience availability, and E.17 when a source-backed face and source return are needed.- The subject rule for an evidence claim's required support;
A.10when its source-to-use account needs recovery or replay, andB.3for a named assurance claim. C.32.PADfor project decision.
Problem
Architecture synthesis is an optimization and learning activity under competing characteristics. It needs evals: deliberate measurement, simulation, benchmark, scenario, review, or monitoring runs that say how a current architecture or candidate architecture reads against declared criteria.
Testing for errors is a neighboring use. A test asks whether an expectation is violated. An eval asks how variants compare, how a candidate changes the trade-off front, which constraint is hit, or whether the next synthesis pass should open. A candidate that loses the eval is not automatically an error; it may be a deliberate probe that improves the architecture team's knowledge of the solution space.
Evolutionary-architecture sources often say "fitness function". Recover the operation, rule or expression meant by that source term before identifying its FPF use. When an architecture evaluation needs framing over declared criteria or structures, use this pattern; an individual test, monitor or expression keeps its own meaning within that framing. Automation does not change the kind of the program record.
Forces
Solution
Identify at least one declared criterion for the evaluation. Use a C.32.ACS row, a C.25 Q-Bundle slot, or both when the evaluation needs both.
Work in this order:
- Cite each evaluated criterion where it is defined. For an ACS row, include the set it belongs to. For a Q-Bundle slot, identify the bundle and slot. Recover the result it asks for, such as a bound for every request or the share of requests within a limit.
- State the eval purpose: current characterization, candidate comparison, portfolio-frontier work, post-change impact measurement, monitoring, or trigger for the next synthesis pass.
- Name the candidates, bearers, and selected structures being evaluated.
- Establish the exact
U.ClaimScope, relevant A.2.6U.ContextSlicemembership and effective reference scheme. Retain every declared selector and the plane, time stance or window, input projection, resource budget, applicable units, admissible inputs and missing-data treatment required by the chosen operation and receiving inference. Establish parity when that inference depends on comparable conditions;parityFrameRefcan retain it but does not replace the underlying facts. An atemporal structural judgement needs no invented temporal window; actual measurement and dated Work retain their own conditions. - Choose the evaluation scope that can answer the question. If the same separate readings could occur in a configuration that fails the joint requirement, identify and evaluate the relation that distinguishes it. An applicable result for the whole can already cover several criteria.
- Choose eval operations. An evaluation can derive a consequence from qualified premises, measure performance or examine an existing result for applicability. Choose measurement, simulation, benchmark, walkthrough, monitoring, review or evidence audit according to what remains open. Name the derivation and its premises when relying on a calculated bound. An existing result can settle the criterion directly. Use
testonly when the intended operation checks an expectation or hard constraint. When a particular evaluation Work occurrence is claimed, recover each evaluator's A.13 core and let A.15.1 independently admit the occurrence from its history, Method followed, extent and containing-System relation. AddevaluationWorkAttributionRefsonly when the record or receiving use represents precise assignment-bound attribution. Add an operation-application relation only when that claim is needed and its predicate obtains. Establish the typed result separately; the program record does not run. - Declare the result form. Examples include a reading, band, rank, dominance relation, trade-off front, qualitative state, or evidence finding; use the definition and test for the result kind.
- Examine material proxy risk before the eval result drives work, and protect the characteristics or constraints it could compromise. If no material risk is found, retain that basis. Optimize only the chosen indicators; keep binding constraints and other necessary safeguards visible.
- State the receiving use:
C.32synthesis input,C.32.MLAOresidual input,E.23improvement feedback,A.19.CPMcomparison input,A.19.SelectorMechanismselection input,C.11choice input, input for a selected-set result declared underG.5, or architecture-decision input forC.32.PAD. For publication input, distinguishE.17source-backed face and source return from theE.24.PUBpublication occurrence and audience availability. - When a criterion, candidate, bearer, structure, environment, parity condition or source premise changes, identify which evaluation dependence is affected. Revise or retire the program if its operation or conditions no longer support the receiving use; retain it when its grounds apply.
Stop condition. Stop C.32.ACE when the account makes the evaluated rows or slots, subject, purpose, operation, result form, receiving use and necessary conditions recoverable, including parity for a dependent comparison and material proxy risks or protected losses. Retain an adequate existing evaluation program; a changed premise identifies what to revisit.
Lowering condition. Use each result only for claims supported by its evaluated subject, criteria, operation and conditions. Missing data, uncontrolled proxy risk or failed parity blocks the inference that depends on it; retain independently usable readings and qualified reports. Publishing such a report does not itself claim the unavailable comparison or decision. Reuse a program for new candidates when its conditions apply, or revise the affected operation when they do not. Use ACS for a missing or wrong criterion, C.16 for measurement validity and C.25 for a composite quality; comparison, evidence, assurance and decisions retain their own rules.
Worked slices
Latency candidate. For a promise that every applicable request finishes in under 100 ms, one verified 240 ms request in the required operating conditions violates the promise. A computed 240 ms case in the design model defeats that model’s guarantee if the case is admitted. A claim about the installation needs grounds connecting the case and the response to that installation. For a requirement that at least 95 percent of this period’s requests finish in under 100 ms, a complete period of 99 requests at 50 ms and one at 240 ms passes. The slow request by itself leaves the share open. Keep the operating requirement when exploring alternatives. A candidate that fails it can still help explain the resource or interface trade-off. A different requirement needs a decision by whoever can change the commitment.
BIM digital twin. A built-asset team compares architecture candidates that combine placement, schedule, use-phase, maintenance, and cost structures. ACE does not treat the number of dimensions as the evaluation. The practitioner defines a parity frame and evals the ACS rows declared for the project, such as access, source-return cost, observability, and maintenance reach, then records results with the parity-frame and result-form fields needed by A.19.CPM.
Method-family architecture. A review-Method family has ACS rows for evidence reuse and change reach. If source wording also says “role substitutability,” use E.10.ROLE and C.32.ACS to bind the exact recovered subject and predicate—such as substitutability among local system-role kinds under A.2.7, not among holders or assignments—before ACE evaluates it. A separate C.25 bundle covers teachability. ACE defines a batch evaluation over three Method variants. One variant loses on teachability but reveals a reusable evidence relation; C.32 may use it as a stepping stone.
If the comparison asks whether a learner can design a review, determine which criteria, alternatives and checks the teaching material already supplied. A correct choice among those alternatives can demonstrate their use. A design the learner developed and justified, with its available help recorded, provides evidence for that part of the design claim. The three Method variants in the evaluation alone establish no learning effect.
AI-agent workflow. A model-supported workflow has candidates with different function graphs and tool boundaries. ACE evaluates latency, evidence refresh, policy controllability, and rollback under the same task set and evidence window. A benchmark score is not the architecture decision; it supplies one eval reading inside the parity frame.
Hospital escalation. A hospital escalation team has ACS rows for decision latency, accountability clarity, and evidence custody. Any source “role continuity” or “role-boundary” criterion first goes through E.10.ROLE and C.32.ACS, which binds the exact subject and predicate—such as continuity of assignment occurrences and their holder Systems, or a boundary among exact participant relations. ACE evaluates two recovered architecture variants under the same incident scenarios and handoff evidence window. The result can feed comparison or the next synthesis pass; staffing choice remains with the receiving decision pattern.
Kind and Receiving-Claim Boundary
Use C.32.ACE to frame the architecture evaluation, using its local record when useful, and to keep criterion, reusable evaluation Method, MethodDescription episteme, intended operation family, planned evaluation, dated evaluation Work, operation application, typed result, comparison input, selection input, and decision input distinct. The local record admits no new U-kind. Use A.3.1, A.3.2, A.15.2, A.15.1, and A.6.1 for their respective objects and relations; use the pattern that defines and tests each result. C.32.ACE does not define starter characteristic selection, ACS scale-row construction, measurement validity, Q-Bundle normal form, candidate synthesis, comparison policy, final selection, local choice, G.5 selected-set result declaration, publication, or an architecture decision. When those claims are current, use C.32.HCS, C.32.ACS, C.16, C.25, C.32, A.19.CPM, A.19.SelectorMechanism, C.11, G.5, or C.32.PAD as applicable. For publication, use E.24.PUB for the occurrence, form, carrier, audience, bounded use and availability, and E.17 when a source-backed face and source return are needed.
Conformance requirements
Common failures and repairs
Consequences
Rationale
An architecture-characteristic eval program connects declared criteria to the evaluation result needed by receiving work. It explains what will be observed or derived, under which conditions, and how that answer can be used. A comparison adds the parity conditions on which its inference depends; every answer comes from its independently established result kind. It does not answer "what is the criterion?", "is the measurement valid outside this use?", or "which architecture must be chosen?"
The pattern is architecture-specific because it evaluates selected structures and architecture characteristics. The same eval form can be used in system, organization-as-system, built-asset, episteme, work-occurrence, discipline, AI-agent-setup, and C.36-recovered cultural-evolution cases once the admitted holon being described has been identified in each case. It can also evaluate Method-side, Work-side, evidence-side, local-kind, classification, or assignment structures after the described holon, the defining or constraining ClaimGraph located through the subject pattern, the bearer and predicate, and the scale rows are rebound. Unresolved “role” wording goes through E.10.ROLE; the pattern does not admit Methods, roles, practices, or cultures as holon kinds by label.
SoTA-Echoing
These rows document how source practice contributes to eval-program fields, result-use boundaries, and refresh conditions in C.32.ACE.
Source-currentness boundary. Use each source row only for the ACE eval-program field, result-use boundary, or refresh condition named in that row. Recheck the row when a named book edition, source presentation, FPF pattern for the next question, metric practice, or evolutionary-architecture practice changes the transferred move. If the project wants criteria-row admission, measurement validity, Q-Bundle structure, explicit comparison, selection, selected-set result declaration, actual publication, local choice, evidence, assurance, or decision use, leave ACE and open the pattern for the next question.
Relations
- Builds on:
C.32.HCS,C.32.ACS,C.16,C.16.P,C.25,A.2.6,A.19,E.13,E.22,E.23, andA.19.CPM; coordinates with A.3.1, A.3.2, A.15.2, A.15.1, and A.6.1 only for separately current Method, MethodDescription, planned Work, dated Work, or operation-application claims. - Receiving uses:
C.32.P2Sactual-structure feedback and next-synthesis repair,C.32candidate synthesis,C.32.MLAOresidual optimization,C.32.CONWAYcorrespondence frames,C.32.FAILrepair,A.19.CPMcomparison,A.19.SelectorMechanismselection,C.11local choice, selected-set result declaration underG.5, source-backed publication-face and source-return work underE.17, publication-occurrence and audience-availability work underE.24.PUB, and architecture-decision work forC.32.PAD. - Measurement boundary: Use
C.16when a reading, coordinate, unit, threshold, score, uncertainty, or cross-case comparability claim is made. - Structural-information boundary: For an eval, these contributions are available only after
C.32.ACS,C.16, orC.25has declared what is being evaluated:C.33andC.34for captured structure, lost structure, and preservation adequacy, andC.35for recovering the kind of a generated or discovered result and assessing its adequacy for that architecture use. Use C.32.ACE for the eval-program frame, and accept each typed result under its own measurement, comparison, evaluation, or assertion definition and test.C.33,C.34, andC.35do not define eval programs. - Q-Bundle boundary: Use
C.25when the evaluated item is a composite quality family. - Test boundary: Use
testonly as an eval operation for a declared expectation or hard constraint. Error recognition and architecture-synthesis repair useC.32.FAIL; non-architecture defects use the local defect-subject pattern. - Decision boundary: An evaluation framed by an ACE record may reference separately admitted readings, ranks, dominance relations, trade-off-front descriptions, and other typed results. Establish the result under its measurement, comparison, evaluation or assertion rule. Add precise evaluation Work or operation-application identity when that claim or the receiving rule requires it. Such a result can support an evidence claim under its conditions; use A.10 when the source-to-use account needs recovery or replay. Use
A.19.CPMfor explicit comparison,A.19.SelectorMechanismfor set-returning selection,C.11for local choice,G.5for selected-set result declaration, andC.32.PADfor a project architecture decision. When audience availability is current, use E.24.PUB for the occurrence, form, carrier, audience, bounded use and availability, and E.17 when a source-backed face and source return are needed.
C.32.ACE:End
Architecture-Influence and Transformed-Architecture Correspondence
Type: Architectural subpattern under C.32 Status: Stable Normativity: Normative unless marked informative Tech-name:
Architecture-Influence and Transformed-Architecture CorrespondencePlain cue: compare an architecture that influences the change with the architecture being changed Lineage and search cue only: Transformer and Transformed Architecture Correspondence
Problem frame
Use this pattern when one architecture-side source — recovered as an exact described holon, selected U.Structure, and either an obtaining C.30 ArchitectureRelation or modal ArchitectureClaim — or another independently typed Work arrangement, communication structure, constraint, or candidate-synthesis result influences the candidate architecture of a changed referent, and the practitioner must decide what to change on either side without mistaking influence for action.
Begin with what constrains the proposed change and which arrangement could meet that constraint. To establish correspondence between two architectures, identify both structures and the relation that makes one relevant to the other.
Primary working reader: an architect or architecture-responsible practitioner who must compare one independently typed influence source with the current or modal architecture content of the changed referent and prepare candidate changes without turning an ArchitectureRelation, selected structure, claim, or architecture-bearing holon into an actor.
Typical entry situations include:
- a desired product architecture cannot be produced and verified by the current manufacturing and certification arrangements;
- chosen service boundaries force every delivery team to coordinate every release;
- a method family is proposed for changing documents, but the assigned review roles and evidence structure do not fit what the project must produce;
- an AI-agent toolchain is intended for Work on project products, but its control and evidence boundaries do not fit the changed product architecture; or
- the project needs a source-side, transformed-side, joint, or bounded-mismatch inverse-Conway candidate rather than another diagram of the desired target.
A clean-looking target architecture can be unbuildable or unproducible, untestable, hard to maintain or evolve, or hard to certify. Existing production, communication, approval, control, evidence, and operating arrangements can constrain the candidate and shift coordination into shared releases, approvals, evidence reconciliation, or exception handling. Treat each such arrangement as an independently typed influence source and recover its direct influence relation when that relation is asserted; the source architecture does not act, and mirroring alone does not establish architecture adequacy.
Start with the domain action: a manufacturing system builds a product, a compiler compiles a program, a service team changes a service, a clinical team treats a patient, or an instructional system teaches a learner. Identify the changed referent first. Only then name an acting system, exact system-role assignment, and dated Work when those facts are current. Separately name the architecture or other source that influences the candidate and the exact relation by which it does so.
First-minute use slice. A product-family team wants independently replaceable field modules. It identifies the changed referent as ProductFamilyFieldModuleBoundary@2026Q3. The influence side is the obtaining C.30 ArchitectureRelation(ManufacturingCertificationSystem@Plant-A, BatchLineSharedEvidenceStructure@Current); the transformed side is the obtaining C.30 ArchitectureRelation(ProductFamily@Current, FieldModuleBoundaryStructure@Current). The exact holons and selected U.Structure participants remain visible, and any desired replacement structure stays only in a separate ArchitectureClaim. No direct architecture-influence kind or predicate has yet been recovered, so the team keeps the pairing as a provisional independent-change pressure with missing-governor. To decide what could make modules independently replaceable, the team follows one permitted replacement through the certification and acceptance conditions. If separate qualification suffices, changing the shared test setup may free the boundary. If joint qualification is compulsory, a replacement must preserve it or obtain an authorized change. Evidence of a later implementation is needed to report which arrangement was used.
The primary working object is a local candidate-synthesis frame. It can pair actual architecture sides through exact obtaining C.30 ArchitectureRelation refs or carry candidate, required, desired, or expected structure only through separately identified ArchitectureClaim refs. When one exact architecture-influence or correspondence relation already obtains between two actual architecture sides, C.32.CONWAY is also the pattern for one reusable ArchitectureInfluenceTransformedArchitectureCorrespondenceRow@Context episteme about that exact occurrence. The frame, row, architecture relations, claims, selected structures, changing system, Work, actual transformation, changed referent, candidate palette, and any network that later cites the row remain different objects.
An error occurs when an architecture, organization chart, method family, toolchain, communication structure, or network record is called the transformer and receives agency, a system-role kind or assignment, Work, or participation in the change. Or the reverse happens: real performer and Work facts disappear behind a vague claim that one architecture shaped another.
The practitioner can prepare architecture candidates while preserving four independent questions—what changed, who acted or performed Work, which sources influenced the candidate, and which exact architecture pair the current correspondence row concerns.
Ordinary working move:
- name the changed referent and, only when actual change is claimed, the independently admitted
U.Transformation; keep every actor-side or Work-to-change relation separate; - name exact acting and performance facts only when current;
- name each influence source with its kind and direct influence relation;
- for an exact reusable row, select one pair of obtaining C.30
ArchitectureRelationoccurrences and keep each holon and selected-structure participant visible; when either side is only candidate, required, desired, or expected, keep the pair in the frame with its exactArchitectureClaiminstead; - prepare source-side change, transformed-side change, joint change, or bounded mismatch candidates.
Adoption test: a reader can tell which case passes, which does not, what the practitioner changes next, and whether the result is only local synthesis material or a reusable exact pair row.
Not this pattern when the current work is only bounded-change identification, system-role assignment or Work attribution, module-interface repair, mathematical structural similarity, local choice, or an architecture decision. Use the governing subject pattern for that work. Return here when a condition on producing, using or changing the target can change its candidate architecture. Trace the condition under :4.1.1. To claim correspondence between two existing architectures, establish the architectural relations on both sides.
Common exits by claim kind:
A.3.4orA.3.4.Pfor the bounded change and changed referent.A.12for acting-side externalization,A.13for exact performer recovery,A.15.1for independent dated-Work admission and distributed-performer forms,A.2.1for an exact assignment occurrence when separately claimed,F.6for a laterperformedUnderAssignment(W, RA)relation only when precise assignment-bound attribution is consumed, and the pattern that defines any direct actor-side or Work-to-change relation needed by the current use. F.6's holder projection only supports equality comparison with the already recovered performer; it identifies neither assignment nor performer.A.6.Mfor module-interface repair.C.32.ACSfor current architecture-characteristic criteria rows andC.25for any composite Q-Bundle and exact slot used by the trade-off.- The subject's mathematical treatment for structural similarity;
C.29when representation choice, transfer or reliance remains unresolved. A.19.CPMfor explicit comparison andA.19.SelectorMechanismfor set-returning selection.G.5for selected-set result declaration;E.17for a source-backed publication face and source return;E.24.PUBfor the publication occurrence and audience availability;C.18andC.19for archive, front, or pool-treatment policy.C.11for fixed local choice andC.32.PADfor a project architecture decision.
The first useful result explains which source-side condition constrains the candidate and which change could address it. An ordinary account can suffice. Use ArchitectureInfluenceTransformedArchitectureCorrespondenceFrame@Project when a record helps preserve that reasoning; the record does not establish an acting entity, relation occurrence, architecture decision or structural equivalence. When the constraining condition is not yet known, :4.1.1 recovers it from a needed change and its production, verification or approval dependencies.
For a first pass, recover the synthesis question, intended use, independently identified changed referent and typed influence source. State the source-side and transformed-side structures available for the question, and keep a missing structure explicit rather than inventing it. For an architecture side, distinguish an obtaining C.30 ArchitectureRelation from modal ArchitectureClaim content. Add current criteria or provisional characteristic heads, applicable candidate forms and the next needed result; include evidence and a window when the claim uses them. Assert an exact pair row only when its direct relation and both obtaining architecture sides satisfy :4.2. Otherwise retain the provisional pressure and its precise unresolved condition. The first-minute case above can be recorded as follows:
This sparse frame asserts no influence occurrence and no exact pair row. The four candidate refs are first-pass heads, not comparison-ready configurations. Add acting-system, exact system-role-assignment, dated-Work, exact-pair-row, C.29, network, publication, comparison-ready gain, loss, and preservation, and any additional source-return fields only when the corresponding claim becomes current; compatible additions refine the frame. Reconsider the affected claims if the referent, architecture pair or pressure changes. The complete extension schema is:
Project-local use keeps two separate fields. @Project remains a compatibility and retrieval cue only. If the frame is used in one project, projectWorkOccurrenceRef names the exact composite U.Work and architectureCorrespondenceFrameProjectUseRelationRef names the direct relation by which that Work uses the frame. The frame, synthesis Work, candidates, architecture relations, claims, selected structures, and project Work remain distinct. An ArchitectureRelation ref is affirmative only for an independently obtaining C.30 occurrence; candidate, required, desired, or expected architecture content stays in an ArchitectureClaim and cannot enter an exact pair row as though it already obtained.
TransformerTransformedArchitectureCorrespondenceFrame@Project and the former title “Transformer and Transformed Architecture Correspondence” are lineage and search cues only. They do not name the current Tech object, make any named value an actor, or establish an acting-system, system-role-kind, assignment, Work, or participation fact.
Problem
Architecture influence and action often occur in the same story but are not the same fact. A manufacturing architecture can constrain a product candidate while a manufacturing system performs production Work. A communication structure can influence service boundaries while people or teams perform change Work only through their admitted exact U.System identities. A method description can influence a work-product architecture without being a worker. A toolchain architecture can constrain project-task candidates while an admitted execution system acts.
Transformer and transformed wording can hide these differences. It can leave the changed referent implicit, treat an architecture bearer as the performer, omit the exact system-role assignment and dated Work, or call a source influential without a direct relation. It can also stretch one local architecture pair into a whole recursive transformation-flow network.
C.32.CONWAY repairs the problem by keeping the changed referent, actor and performance facts, influence-source facts, and one exact architecture pair separately recoverable. Conway and inverse-Conway practice then supplies candidate pressure, not a universal relation and not evidence that any source acted.
Forces
Solution
Begin with the source-side pressure and candidate changes, using the local frame when a record helps. When a new reusable relation kind is needed, admit it through the applicable E.24-family pattern and direct settlement; use an adequate existing relation directly. Use the relation's predicate and applicability to test the current pair. If current facts or constituting history satisfy the predicate, one world-side occurrence obtains and the row may cite its exact identity. If the predicate is false, no exact pair row is asserted. If the current facts do not decide it, keep the correspondence synthesis-local and name the missing grounding or information-sufficiency boundary. Use missing-governor only when no direct relation kind and predicate govern the intended pair and use.
Keep acting, influence, and correspondence facts separate
- Name the domain action and changed referent. Identify
changedReferentRefindependently. AddactualTransformationRefonly when A.3.4 independently admits one bounded change of that same continuing referent; keep actor-side and Work-to-change relations under their subject patterns. Architecture influence identifies none of those facts. - Add acting and performance facts only when claimed. Every precise performer is one exact
U.Systemrecovered through A.13. Claimed performance requires one exact datedU.Workindependently admitted through A.15.1 and the exact actor-side or Work-to-change relation needed by the claim. When this frame or its receiving use consumes precise assignment-bound attribution, separately represent the same obtaining A.13 assignment and its declared A.2.1 species, and test F.6performedUnderAssignment(W, RA); then compareSwithRA.HolderSystemSlot. F.6 identifies neither assignment nor performer, and missing or failed F.6 attribution alone leaves the Work intact while its independent A.13/A.15.1 grounds hold. Use A.15.1CC-A15.1-17when several systems jointly perform the top-level Work or when the use instead needs a parent Work with separately performed child occurrences. - Name every influence source by kind. Architecture, selected structure, Work, communication, constraint, and candidate-synthesis results retain their kinds and direct influence relations. Influence alone supplies no system identity, system-role kind or assignment, Work, performer status, changed-referent identity, or transformation participation.
- Select one architecture pair. For an exact row, name one obtaining influence-source C.30
ArchitectureRelationand one obtaining transformed-side C.30ArchitectureRelation, with each exact holon and selected-U.Structureparticipant. Their architecture-bearing holons may differ from every acting system. Record equality only when independent actor and architecture-bearer facts establish it. If either side is only candidate, required, desired, or expected, keep its exactArchitectureClaimand the pair in the synthesis frame; do not assert an exact pair row. - Map only structures and characteristics that change the candidate. Use the available source-side and transformed-side structures, intended gain and known loss. For an unexplained dependency, use :4.1.1 to recover its operative condition or a discriminating inquiry. Include a window or source return when the use depends on it. A comparison relies on its relevant bound C.32.ACS criteria and any declared C.25 slots; earlier candidate construction can retain provisional heads.
- Consider four candidate forms. Change the influence-source side, change the transformed architecture, change both, or keep a bounded mismatch with an explicit cost and reopen trigger. Develop the applicable alternatives and explain a consequential exclusion; an adequate existing candidate need not be rebuilt.
- Establish any mathematical correspondence under its definitions and assumptions. Use C.29 when choosing or transferring the representation, or relying on its result, leaves an unresolved question. A correspondence row alone establishes no homomorphism, equivalence or architecture adequacy.
- Stop at the next governed claim. Send comparison, selection, publication, choice, decision, evidence, assurance, gate, Work, or organization-governance claims to their patterns.
From a needed change to a correspondence candidate
Use an already warranted explanation and candidate directly. Otherwise start with one needed product change or receiving result and the available instructions, change records, test material and participants. For work and provision dependencies, use C.32.MWA.PP:4.1 and :4.4–:4.7 to trace the result backward to its supplying operations and follow their outputs forward to the receiving use.
- Locate the dependency that limits the intended change. Follow a representative actual change or explicit proposed change through production, integration, testing, evidence and approval as relevant. At a shared or unavailable contribution, inspect the rule and the input it needs. Ask its provider what would fail if this change proceeded separately. A wait may come from a shared resource, a required joint check, incompatible information or a decision right; those conditions lead to different candidates.
- Establish which condition must be preserved. Compare that explanation with the requirement, interface contract, test procedure or actual failed change. Distinguish a necessary domain or contractual dependency from the present arrangement used to satisfy it. If the choice turns on the reason, use the applicable rule, an observation or the provider’s explanation to distinguish the possible dependencies. A bounded trial earns its cost when its possible outcomes change what can be done and the trial is permitted. A candidate that satisfies every dependency still in question can be carried forward while the reason for the old arrangement remains open.
- Construct a change at that relation. A source-side candidate can provide independent access, qualified verification, information or decision authority where shared provision is the limitation. A transformed-side candidate can change a module boundary or interface so the required joint condition no longer crosses it. A joint candidate changes the interface and the way its production or verification is supplied. A bounded mismatch retains the dependency and states its actual cost and allowable limit; it is usable only if the receiving requirement permits that compromise. Develop the forms applicable to the recovered condition, including retaining an adequate arrangement.
- Follow each changed connection through to the result. Check that its input is available, its provider can supply it under the relevant capacity, timing and authority conditions, and its output establishes what the receiver needs. Preserve required joint checks or supply justified replacements. Explain the expected gain and any identified new burden or loss. A changed diagram alone does not establish these performance or causal claims.
The first result can be a candidate, a supported reason to retain the arrangement, or a precise open question with its obtainable source. For example, an unknown approval dependency can return “recover whether the approval rule requires a joint test of these two modules” to the procedure and its provider. An exact obtaining influence occurrence additionally retains the predicate and case grounds in :4.2; candidate construction does not require asserting it.
Exact reusable architecture-pair row
The row is a U.Episteme about one already obtaining direct influence or correspondence relation whose exact participants include the two obtaining C.30 architecture-relation occurrences required by this use. Because those occurrences fill participant positions of another relation, each is individuated under A.6.REL for this receiving use. The row neither creates the influence occurrence nor mints a universal Conway relation. Each C.30 occurrence keeps its exact holon and selected-U.Structure participants; the influence occurrence keeps its identity under its direct relation pattern and A.6.REL. This row only describes them for the current correspondence use. entityOfConcernRef, its kind, its governor, both C.30 occurrences and their participant pairs, and the changed referent are required. If the practitioner has only a useful local compound correspondence claim, or either architecture side is modal rather than obtaining, keep it in the frame for candidate synthesis. Assert the row only after the admitted influence relation kind's direct predicate is applicable and current facts satisfy it. If the predicate is false, assert no row; if facts are unresolved, keep the frame and name that boundary; if the kind or predicate is absent, return missing-governor. None of these branches permits inferring a relation from two architecture claims, structures, diagrams, or names.
Qualified network reading
The same exact pair row may appear in architectureCorrespondenceRowRefs[] of several TransformationFlowStructureNetworkRecord@Context values while its pair, relation occurrence, evolution window, correspondence use, and claim scope remain current. Each citation contributes only one qualified architecture reading. The row's optional singular networkCrossFlowRelationRowRef, when present, qualifies only the exact current record edition named by that locator; it does not qualify the row's citations from other records. No citation makes the pair row the network, adds a member, or satisfies the network's exact cross-flow-relation discriminator.
Set networkCrossFlowRelationRowRef only when the pair row's exact influence occurrence and architecture-relation participants are independently grounded in member-flow positions and the locator's transformationFlowStructureNetworkRecordRef names the same exact current record whose architectureCorrespondenceRowRefs[] citation this mapping is intended to qualify. Resolve that record first, then require exactly one crossFlowRelationRow to match the occurrence and complete ordered endpoint-binding identity. That row must preserve the same kind, governor, participant order, endpoints, and bindings as this correspondence row. Zero or several matches, a different record, or a stale record edition leaves the locator unresolved. Do not reuse one locator to qualify another record's citation. A record citation alone infers none of those facts.
Acting-system identity, system-role assignment, F.6 attribution, Work, actual transformation, actor-side or Work-to-change relation, influence relation, and network cross-flow relation remain separately governed even when one case cites all of them.
Candidate moves and repair rows
Plain text begins with the domain action—builds, assembles, repairs, configures, treats, teaches, compiles, or evaluates. It then names the acting system and Work only when those facts are current. In a separate sentence it says which architecture or other source influences which candidate through which exact relation. Creator, creation, producer, transformer architecture, and uses remain ordinary cues, not universal technical labels.
Choose only pressures that change the candidate or protect against a concrete loss. Every affectedArchitectureCharacteristicRefs[] value in an exact pair row or comparison-ready candidate must resolve to a current C.32.ACS criteria row; when the pressure is one slot of a composite quality family, also resolve the declared C.25 Q-Bundle and that exact slot. If those governed objects do not yet exist, put plain heads such as independent change, substitutability, evidence reuse, latency, coupling or cohesion, coordination load, and source-return cost only in the local frame's provisionalArchitectureCharacteristicHeads[]; never place them in affectedArchitectureCharacteristicRefs[]. These heads are discovery cues, not a universal catalogue and not criteria refs. Use C.32.ACS or C.25 before making a stronger comparison, selection, or decision claim.
Stop condition. A first-pass account may finish with the changed referent, typed source, relevant known or unknown structures, provisional pressure or grounded characteristics, applicable candidate forms and next useful action. Every asserted acting, performance or influence fact needs its basis. Before relying on a candidate comparison, explain which sides change or remain fixed, the expected gain and its grounds, known loss, relevant window and conditions of use. Add a source return where omitted detail can change that use. An exact pair row additionally requires the obtaining architecture sides and satisfied influence predicate in :4.2. Keep missing governor, unresolved grounding or information sufficiency, and a false predicate distinct in provisional correspondence material.
Lowering condition. Withhold an exact pair claim for a current use when its influence occurrence, either architecture occurrence or participant pair, changed referent or applicable window is missing or no longer supports that use. Keep supported historical claims within their conditions and retain useful provisional pressure. Reconsider a candidate when a proposed change, mismatch or known loss no longer fits the intended window; revise its conditions or retire it according to that result. Use A.3.4 or E.18 for a missing actual change or flow relation, A.12/A.2.1/A.15.1/F.6 for the particular acting, assignment, Work or attribution question, and C.29 for an unresolved mathematical representation choice, transfer or reliance question.
Worked Correspondence Cases
Recovering the Plant-A constraint. A request asks for independent field-module replacement. Begin with the present replacement instruction, the test package and a representative change record. Trace what happens after one module changes: which other modules must be present, which evidence is reused, and who can accept it. Suppose the test procedure evaluates an interaction across the current module boundary and the applicable certification rule consumes that joint result. The shared test is then a premise to preserve, not merely a scheduling habit.
A manufacturing-cell candidate must supply a qualified way to obtain the needed joint evidence with less shared work. A product-boundary candidate can place the interacting parts within one independently replaceable unit and check what that larger unit costs the field user. A joint candidate changes the interface and qualifies the corresponding separate test. Keeping the batch dependency is a bounded alternative only where the replacement requirement permits it; otherwise it does not satisfy that requirement. Follow each proposed change through production, test, acceptance and replacement before treating its benefit as established.
If the shared test exists for the convenience of the present fixture, independent provision may be possible without changing the product boundary. Suppose a qualified service can supply the required joint result within the field deadline under either account of the old fixture. The team can arrange that service on its terms. Replacing the joint result with two separate tests would require establishing what the certification rule accepts. The fact that past tests were shared settles neither that rule nor the reason for the old fixture. The exact positive case below uses independently established relation and performance facts.
Exact positive, distributed-performer, and network-local slice. In one Plant-A domain framework, PlantArchitectureInfluenceRelations-v3 defines BatchEvidenceArchitectureConstrainsModuleArchitecture(sourceArchitectureRelation, transformedArchitectureRelation, evolutionWindow). Its first participant is the obtaining C.30 ArchitectureRelation(ManufacturingCertificationSystem@Plant-A, BatchLineSharedEvidenceStructure@Current); its second is the obtaining C.30 ArchitectureRelation(ProductFamily@Current, FieldModuleBoundaryStructure@Current). Both occurrences are individuated under A.6.REL because this domain predicate uses them as participants. Plant-A facts satisfy that predicate over ProductFamilyModuleChange@2026Q3, so BatchEvidenceConstrainsFieldModules-17 is the exact obtaining influence occurrence and the EntityOfConcern of BatchEvidence-to-FieldModules-Row-17. This case-local predicate and occurrence do not mint a universal Conway relation.
The changed referent is independently identified as ProductFamilyFieldModuleBoundary@2026Q3. If the case also claims actual change, A.3.4 independently identifies FieldModuleBoundaryTransformation-17 : U.Transformation.
Before the candidate occurrence is called Work, A.13 recovers both collective performers for one shared scope, working situation, and window. The scope is ProductFamilyModuleChange@2026Q3; the situation is PlantAModuleTransitionArchitecturingSituation-17; and the window is 2026-07-14T09:00:00+03:00 through 2026-07-16T18:00:00+03:00. ManufacturingArchitectureTeam-A : U.System has boundary ManufacturingArchitectureTeamBoundary-A, comprising the rostered team members, their manufacturing-constraint decision channel, and the coordination interactions governed by ManufacturingArchitectureCoordinationMethod-v2. Its exact case-local action is ManufacturingBoundaryConstraintAction-17: select, revise, pause, or escalate manufacturing-feasibility constraints on the field-module boundary. It acts toward BatchFeasibleModuleBoundaryObjective-17 under the current batch-line capability, shared-evidence, and batch-exception conditions. CertificationArchitectureTeam-A : U.System has the distinct boundary CertificationArchitectureTeamBoundary-A, comprising its rostered members, certification-evidence decision channel, and the coordination interactions governed by CertificationArchitectureCoordinationMethod-v2. Its exact action is CertificationEvidenceConstraintAction-17: select, revise, pause, or escalate the certification-evidence obligations attached to the same boundary. It acts under CertificationEvidenceContinuityNorm-17, which requires every recommended boundary continuation to retain traceable applicable evidence and requires a hold when a material evidence gap remains. Its relevant conditions are the applicable certification obligations, current evidence links, and open evidence gaps. Neither System boundary includes the product family, either C.30 architecture-relation occurrence, the role-kind descriptions, or the assignment occurrences; neither action is yet called Work.
PlantArchitecturingSystemRoleKindDomain contains two exact local agential system-role kinds with independent membership criteria. ManufacturingArchitectureSystemRole requires the stable work-facing contribution of manufacturing-feasibility constraint judgment and goal-directed, condition-sensitive regulation toward BatchFeasibleModuleBoundaryObjective-17: the holder must select among admissible boundary continuations and revise, pause, or escalate its judgment when batch-line capability or exception evidence changes. CertificationArchitectureSystemRole requires the stable work-facing contribution of certification-evidence constraint judgment and corresponding regulation under CertificationEvidenceContinuityNorm-17: the holder must select among admissible evidence continuations and revise, pause, or escalate when an obligation or evidence gap changes. ManufacturingConstraintDecisionTrace-17 shows the manufacturing team comparing two admissible continuations, revising one constraint after new batch-exception evidence, and pausing the unsupported continuation. CertificationEvidenceDecisionTrace-17 shows the certification team revising the evidence obligation after an applicability change and holding the unsupported continuation pending a missing trace. A.10 evidence-use claims connect those traces and the respective boundary and coordination records to the two criteria. On that basis Plant-A separately classifies ManufacturingArchitectureTeam-A under ManufacturingArchitectureSystemRole and CertificationArchitectureTeam-A under CertificationArchitectureSystemRole. The evidence and classifications are independent of the candidate Work and of either assignment. No Grade, autonomy result, characteristic profile, or stronger assurance claim is consumed here.
The two-participant assignment kind PlantArchitecturingWorkAssignment is declared under U.SystemRoleAssignment. Its HolderSystemSlot admits a U.System; its local AssignedSystemRoleKindSlot uses PlantArchitecturingSystemRoleKindDomain; and its predicate says that, for the declared scope, situation, and interval, the holder is selected to supply the contribution defined by the assigned kind to the Plant-A transition-architecturing action. ManufacturingArchitectureAssignment-17 assigns ManufacturingArchitectureSystemRole to ManufacturingArchitectureTeam-A; CertificationArchitectureAssignment-17 assigns CertificationArchitectureSystemRole to CertificationArchitectureTeam-A. Both occurrences obtain with their declared holder and assigned-kind values, all species-specific predicates satisfied, and maximal uninterrupted predicate-true intervals covering the full stated window. These are the same obtaining assignments later consumed by F.6.
Only after both A.13 cores are established does A.15.1 independently admit ModuleTransitionArchitecturingWork-17 : U.Work. The independently grounded joint performance history records the two actions above as one top-level occurrence; the Work runs from 2026-07-14T09:00:00+03:00 to 2026-07-16T18:00:00+03:00 and enacts ModuleTransitionArchitecturingMethod-v3. A locally declared relation places that Work within ProductFamilyEngineeringSystem-A for the stated engineering-system boundary and Work interval. The independently admitted ProductFamilyEngineeringSystem-A : U.System is the second participant of that relation. The assignments belong to the required A.13 performer bases; neither those assignments alone nor an F.6 conclusion establishes the Work occurrence.
F.6 then records that ModuleTransitionArchitecturingWork-17 was performed under each of the same obtaining assignments established in the A.13 cores. The case facts link the exact Work to both assignment occurrences; holder equality holds for ManufacturingArchitectureTeam-A and CertificationArchitectureTeam-A, and both assignment intervals cover the Work. The Plant-A relation ArchitecturingWorkChangesModuleBoundary separately states that this Work brings about FieldModuleBoundaryTransformation-17.
This is the A.15.1 CC-A15.1-17 joint-performer form. Neither a lead assignment nor the architecture pair substitutes for either performer. Taxonomy and scheme epistemes may help interpret an assertion, but they are not assignment participants.
ProductDevelopmentNetworkRecord-2026Q3 may cite BatchEvidence-to-FieldModules-Row-17 once in architectureCorrespondenceRowRefs[]. That citation contributes one reading of the exact two C.30 architecture-relation occurrences. The pair row remains an episteme, not the TransformationFlowStructureNetwork, not one of its members, and not a cross-flow relation. Its optional networkCrossFlowRelationRowRef stays absent unless this same influence occurrence is independently grounded at exact member-flow positions and the composite E.18.NET locator resolves exactly one matching row in that same current record edition.
Network-qualified reading.
A product-development TFS and a production-system-change TFS participate in one selected E.18.NET-conforming network. A current architecture pair row about manufacturing-architecture influence may be cited by the network record alongside a separately grounded obtaining production or project occurrence. If the pair row also carries networkCrossFlowRelationRowRef, that locator names this same exact current record edition and resolves exactly one matching row; it qualifies no citation from another record. The pair row remains one reading of one exact architecture pair. It is neither the network nor proof that the architecture-influence occurrence is the cross-flow occurrence.
Near miss. A diagram places a factory architecture beside a product architecture and labels the arrow “shapes”. No direct relation kind and predicate govern that pair and use. The frame may retain the pair as synthesis-local pressure, but the exact row and network cross-flow mapping remain absent with missing-governor; the diagram does not create an occurrence.
Correspondence Failure Modes
Conformance Checklist
Common Repair Cues
Consequences
Rationale
Architectures do not act. Actual Work requires the A.13 performer basis and independent A.15.1 occurrence admission; precise assignment-bound attribution additionally uses F.6. Architectures, selected structures, Work arrangements, communication structures, constraints, and candidate-synthesis results can nevertheless influence which transformed architecture is feasible. C.32.CONWAY is useful because it relates those facts without merging them.
The exact pair row gives one obtaining architecture-influence or correspondence occurrence a reusable episteme. The larger frame remains useful when a project has enough information to prepare candidates but not enough to assert that exact row. This preserves practical forward motion while keeping the exact relation status visible: missing governor, unresolved grounding, false predicate, or satisfied affirmative case.
The four candidate forms remain: change the influence-source side, change the transformed architecture, change both, or keep a bounded mismatch. The split between actor facts and influence facts changes their grounding, not their constructive purpose.
SoTA-Echoing
These source contributions guide the correspondence question and its candidate changes. The source family supports architecture practice; it does not decide actor identity or make a relation obtain.
Source-currentness boundary. When the referent, performer basis, influence relation, architecture pair, selected structure, window or source premise changes, reconsider the row uses that depend on it. If the source no longer supports the selected local pressure, lower it to background lineage; do not preserve a technical claim by name alone.
Relations
- Builds on:
C.32for candidate architecture synthesis;C.30for exact described holons, obtainingArchitectureRelationoccurrences, selectedU.Structureparticipants, and modalArchitectureClaimcontent;A.3.4andA.3.4.Pfor the continuing changed referent and actual bounded transformation;A.12for acting-side externalization;A.13for exact performer recovery;A.15.1for independent dated-Work admission and distributed-performer forms;A.2.1for assignment occurrence identity when separately represented;F.6for a laterperformedUnderAssignment(W, RA)relation only when precise assignment-bound attribution is consumed; direct subject relation patterns for actor-side, Work-to-change, and influence occurrences;A.6.RELwhen this episteme consumes occurrence identity;E.18for one TFS; andE.18.NETfor network identity and exact cross-member relations. F.6's holder projection only compares equality with the already recovered performer. - Uses:
C.32.ACSfor current architecture-characteristic criteria rows;C.25for composite Q-Bundles and their declared slots;C.32.MWA.PPfor tracing work and provision dependencies and constructing changed arrangements;C.32.MLAOfor a cross-scope residual;C.32.FAILfor a correspondence repair failure;C.29when mathematical representation choice, transfer or reliance is unresolved; andA.6.P.WMRandA.6.RCDwhen a required direct relation cannot be recovered. - Patterns for the next questions:
A.19.CPMfor explicit comparison,A.19.SelectorMechanismfor set-returning selection,G.5for selected-set result declaration,E.17for a source-backed publication face and source return,E.24.PUBfor the publication occurrence and audience availability,C.18andC.19for archive, front, or pool treatment,C.11for fixed local choice,C.32.PADfor architecture decisions,A.10for evidence,B.3for assurance,A.21for gate decisions, the applicable authority or permission relation for release, and the method, Work, or organization-governance patterns when those claims are current. - Network boundary: an
ArchitectureInfluenceTransformedArchitectureCorrespondenceRow@Contextmay be cited as one qualified reading inarchitectureCorrespondenceRowRefs[]; it is not the network and does not satisfy an E.18.NET cross-flow relation without the separately grounded obtaining occurrence and endpoint bindings. Its optional singular row locator qualifies only the exact current citing record it names. - P2S docking:
C.32.P2Smay use C.32.CONWAY when a problem-to-structure flow needs one exact architecture-influence/transformed-architecture pair. It does not infer performer or influence facts from the flow card. - Boundary: C.32.CONWAY governs correspondence framing and one exact reusable architecture-pair episteme inside candidate synthesis. It does not govern actor identity, Work occurrence, organization redesign, authority, evidence sufficiency, assurance, gate passage, release, structural-equivalence theory, final architecture decision, or transformation-flow-network identity.
C.32.CONWAY:End
Architecture Candidates to Reduce Cross-Scope Residuals
Type: Architectural subpattern under C.32 Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when a practitioner has a recoverable cross-scope or interlevel architecture residual and needs candidate architecture changes that reduce that residual under a declared evolution window.
Primary working reader: an architect or architecture-responsible practitioner who has already recovered a residual and must prepare candidate changes without calling a local improvement a whole-holon optimum.
Typical entry phrases:
First-minute use slice. A product-family team finds that local variants ship earlier while the family spends more time on certification exceptions. A narrower interface might eliminate a class of exceptions. Qualifying the common evidence once might remove repeated checks while retaining the local checks that depend on installation. Keeping an uncommon variant can also be adequate. For each candidate, follow one affected delivery through certification and identify the change that would remove a check or add work. Each trace can establish where work is removed, added or transferred in that delivery. For a conclusion about the family, establish which deliveries the claim covers. Recover their mix or frequency when it could change the comparison. A common requirement can establish the same conclusion across all permitted mixes. Use applicable rules, bounds or records of the mix to determine whether the choice holds across the deliveries in question. The team can make this comparison before installing either design.
The primary EntityOfConcern is a residual-reducing candidate frame for one grounded architecture question. The frame identifies conditions that conflict when the required contributions are combined. Each pair may be compatible while the whole combination exceeds a shared limit. Identify the combination the original question actually requires. A candidate changes their connection, provision or requirements. When previous work transferred a burden, follow it to the participant who now bears it. The described holon can be a system, organization-as-system, discipline, AI-agent setup, built asset, episteme, work occurrence, or an instance of another admitted holon kind. Recover source labels such as practice, culture, tradition, style, or method before architecture use. Use E.10.ROLE for unresolved role wording: it may resolve to a local system-role kind, an assignment occurrence, a participant position in a direct relation, a function claim, an organization or representation position, or ordinary wording. Carry only the recovered object or relation, never a generic role-side structure. Candidate Systems, assignments, Methods, plans, and structures remain modal content until their own facts obtain; a publication family appears only under its applicable pattern. Use E.24.PUB for a publication occurrence and audience availability; add E.17 when a source-backed face and source return are needed. C.32.MLAO is not a universal optimizer, adequacy claim, selector, decision, assurance argument, publication pattern, or software-system-only pattern.
What goes wrong if C.32.MLAO is missed: local success is called whole-holon architecture success, or an optimization phrase hides the residual that shifted to another declared holon-level ref or declared scope ref.
What C.32.MLAO buys in practice: the practitioner can prepare residual-reducing architecture candidates for later comparison by explaining expected residual reduction, possible new burden, affected scopes and relevant preservation or loss. Include source-return conditions when later use needs omitted distinctions.
Begin with the scoped conditions the affected contributions must meet together. Trace their connection to the recipient and propose a change that could make the conditions jointly achievable. Include any work, delay or loss a proposal moves to another participant.
Adoption test: after using C.32.MLAO, a reader can see the expected residual reduction, possible new burden, affected scopes, relevant preservation or loss, and evolution-window stop condition for each candidate, including what remains unresolved.
Use C.32.MLAO only after residual triage. Do not use it to recover the residual, justify a mathematical lens, compare or select candidates, choose locally, declare a selected-set result, publish it to an audience, or decide the project architecture.
Common exits by claim kind:
C.30.ILCwhen the residual is not recoverable yet.C.32.ACSwhen architecture-characteristic criteria rows are missing.C.32.ACEwhen eval programs or eval results are the current claim.C.29when a question remains about choosing, transferring, or relying on a mathematical representation; a sufficient subject-level mathematical result can finish without that separate account.A.19.CPMfor explicit comparison,A.19.SelectorMechanismfor set-returning selection,C.11for local choice, andG.5for selected-set result declaration.C.18andC.19for archive, front, pool treatment, or stepping-stone retention.C.30.ADfor architecture-description work,E.17for a source-backed publication face and source return, andE.24.PUBfor the publication occurrence and audience availability.C.32.PADfor project decision.
The first useful output is an account of candidate changes, the residual each addresses and its possible new burden. An ordinary explanation can suffice. Use MultilevelArchitectureResidualOptimizationFrame@Project when a structured record helps retain or compare these contributions:
For a first pass, name the residual, affected holon, scopes and structures, candidate changes, expected burden and relevant window. Keep unknown structure, criteria or feasibility as questions to resolve. Before a candidate supports a comparison or stronger claim, supply the criteria, ClaimScope and other bindings that use requires. Add front, archive, NQD, OEE, C.29 lens, ideality, scale-amenability, function-bearer, and architecture-influence-correspondence refs only when that support is current for the candidate being framed.
Here @Project is a compatibility and retrieval cue only. It establishes no project entity, composite-work identity, context, authority, viewpoint, or parthood. When the frame is used in one project, projectWorkOccurrenceRef identifies the exact composite U.Work and residualOptimizationFrameProjectUseRelationRef identifies the direct relation by which that work uses the frame. The frame, the residual-reducing synthesis work, the candidate architectures, and the project work remain distinct.
Problem
Many architecture problems are residual problems. A local module boundary improves one team and creates integration exceptions elsewhere. A control relation improves response and creates audit or timing burden. A platform improves reuse and creates evidence decay. A method family speeds authoring and harms transfer.
A local characteristic can improve while the joint requirement still fails. A conflict can also be present in the original design. Identify which conditions meet at the relevant boundary and follow each candidate change to the required result, its costs and its remaining limits.
Forces
Solution
Develop candidate changes around one recoverable residual; record them in the form needed for their use.
Work in eight steps:
- Start from a
C.30.ILC-compatible residual triage. - Name the affected levels or scopes and the selected structures that carry the residual. Supply declared refs when the receiving use requires them.
- Explain why reducing the residual matters. Reuse adequate architecture-characteristic criteria and Q-Bundle slots; develop missing criteria through
C.32.ACSwhen needed for the proposed use. - Develop or reuse a C.32 candidate palette. Investigation of a candidate may expose structure or criteria not yet understood.
- For each candidate, state the residual it is expected to reduce, the selected structure changed, and the characteristics or criteria affected.
- Assess the new burden, loss, exception, or source-return load that the candidate could create within the declared evolution window. State what is expected and unresolved; if none is identified, say which affected structures, scopes, and criteria were inspected.
- Record the evolution window and any support that only keeps candidate plurality or directionality alive, such as NQD, OEE, an archive or front, stepping-stone retention, ideality, or BLP.
- Stop at the frame, or name the pattern for the next question when a later claim is current: use
A.19.CPMfor explicit comparison,A.19.SelectorMechanismfor set-returning selection,G.5for selected-set result declaration,C.11for local choice,C.32.PADfor an architecture decision,C.30.ADfor architecture-description work, andC.29for a remaining question about choosing, transferring, or relying on a mathematical representation. For publication, useE.24.PUBfor the occurrence, form, carrier, audience, bounded use and availability; addE.17when a source-backed face and return to source are needed.
Before relying on a candidate for residual reduction, answer the working questions: which scope and selected structure change, which architecture characteristic or Q-Bundle slot is at stake, what residual is expected to decrease, what structure is preserved or lost, and what new burden is expected or why none is currently identified. A provisional candidate can retain unanswered questions; it cannot supply the conclusions those answers would support.
For a claim of improvement across scopes, identify what is being compared and for whom. Reviewer effort, author effort and elapsed time to acceptance can give different orderings even for one document. Use the work and availability in each candidate to determine the ordering the recipient needs (C.32:5.1). A count of recent exceptions can locate a burden. To attribute the change to a shared design, compare the admitted mix of products and check which obligations the design actually changes. A new mandatory check on every product can raise the count under either design.
Comparison-input boundary. C.32.MLAO prepares comparison inputs; it does not run the comparison or choose a candidate. Its comparison inputs state expected residual reduction, possible new burden, selected structures and relevant preservation or loss. Include source-return conditions when a later use needs omitted distinctions, and C.29 lens-output references when used.
Those references are diagnostic inputs only.
Admitted profiles and a ComparatorSpec belong to the receiving explicit-comparison pattern.
If the current claim is explicit comparison, use A.19.CPM with admitted profiles and a declared ComparatorSpec. If the claim is local choice over an existing option set, use C.11. If the claim is set-returning selection, use A.19.SelectorMechanism. If the claim is selected-set result declaration, use G.5. For publication, use E.24.PUB for the occurrence and audience availability, with E.17 when a source-backed face and source return are needed.
Lens-output discipline. Graphs, fronts, residual vectors, DSMs, RG-like descriptions, and frustration language are C.29 lens outputs, structural descriptions, or diagnostic signals after their architecture use is typed. The failure is proxy preference: a candidate is preferred because the output looks better while selected structures, lost structure, architecture characteristics, and pattern for the next question remain unnamed. The repair is to interpret the output over selected structures and state what residual or loss it exposes; any comparison, selection, or choice claim then belongs to its pattern for the next question.
Method, culture, and episteme discipline. For a Method-family, culture, practice, or episteme case, first name the described holon, optimization question and use, selected structures, and the pattern that defines or constrains each load-bearing claim. If a publication family or publication face is in view, decide whether it is the described holon, a selected structure, an architecture description, or an MVPK face before using it. C.32.MLAO defines only the residual-reducing architecture candidate frame; use the applicable patterns for Method, Work, publication, evidence, ethical, and decision claims.
Dynamic candidate discipline. A preferred or retained candidate is bounded by an evolution window, source conditions, and the pattern for the next question that admitted the preference or retention. NQD, OEE, C.18, and C.19 can keep a front, archive, pool, or stepping stone visible; they do not select the architecture and they do not turn a front member into a durable optimum.
Ideality and BLP discipline. TRIZ ideality can suggest residual-reducing candidate changes: remove a support bearer, transfer a useful function onto an existing resource, or generalize a bearer so fewer selected structures carry more useful functions. BLP can prefer a more general scale-amenable bearer only inside its declared scale window and audit boundary. Both lines guide candidate generation. When removing a bearer, inspect transferred functions and possible coupling, evidence, control and source-return burdens; retain the applicable scale boundary and receiving comparison or choice rules.
Function-bearing feasibility. Identify what would perform the required function under the applicable interface, placement, resource, control, information and evidence conditions. A compatible qualified bearer can support a design-time feasibility claim. An assignment alone leaves its capability open. Failure to find a bearer leaves a search question. When a named bearer fails, reject that configuration for the stated use. To rule out the design across all permitted bearers, establish that the constraint excludes them all. A changed bearer, split function or different placement can be a further candidate. A reduction in the required function changes the promise and needs the authority to revise it.
Architecture-influence and transformed-side discipline. When a residual is carried by one independently typed architecture-side source constraining architecture content for a changed referent, use C.32.CONWAY. For each actual side, keep the exact C.30 described holon, obtaining ArchitectureRelation, and selected U.Structure together; keep candidate, required, desired, or expected content in an exact C.30 ArchitectureClaim. Keep the changed referent and any actual A.3.4 U.Transformation separate. Then prepare residual-reducing candidates that change the influence-source side, the transformed side, both sides, or a bounded mismatch as comparison inputs or downstream candidate alternatives. Influence, transformation, flow, Work, and module-interface claims belong to their exact relation pattern, A.3.4, E.18, A.15, or A.6.M when current. Use C.29 when a separate question remains about choosing, transferring, or relying on a mathematical representation of structural similarity.
Level, stratification-term, and whole-reidentification discipline. When stratification wording leaves the relevant object or relation unclear, use E.10.ARCH and C.30.STRAT. When BOSC, MHT, MET, MFT, emergence, boundary-crossing or promotion wording leaves the claim kind unclear, use E.10 and B.2.P. Use the recovered subject pattern directly when the meaning is already clear. Use B.2 only when a whole-reidentification question remains after the existing-whole explanation check; otherwise use the subject pattern for architecture, boundary, capability, function, measurement, publication, work, or lens claims.
Stop condition. Stop when the needed candidate account explains the residual, affected scopes and structures, proposed changes and possible new burdens. Preserve the distinctions, loss and source-return conditions needed by its receiving use. An exploratory account may finish with unresolved questions; comparison or decision inputs must meet their receiving rules.
Reconsideration condition. A changed residual, structure, criterion or evolution window reopens the affected candidate claims. Preserve any supported account and useful provisional questions. A missing binding blocks the use that needs it; it does not erase independent information. Retain a candidate for a new window only when its premises hold, and revise or retire the affected claim when they do not. Use C.30.ILC when the residual is missing, C.32.ACS when criteria rows are missing, C.32.ACE when eval results are needed but not current, C.29 for a remaining mathematical-representation choice, transfer, or reliance question, and A.19.CPM, A.19.SelectorMechanism, C.11, G.5, or C.32.PAD when the downstream claim is current.
Worked Residual Cases
Residual And Trade-Off Failure Modes
Conformance Checklist
Consequences
Rationale
C.30.ILC names cross-scope residuals and first architecture repair directions. C.32 creates candidate palettes. C.32.MLAO is needed when the constructive candidate work is specifically about reducing a residual across declared holon-level refs or declared scope refs.
The work is to prepare candidate architecture changes for later comparison by naming residual reduced and burden created, not by using an optimizer phrase, scalar output, or locally improved structure as the candidate frame.
Multilevel learning, frustration, RG-like, DSM and Pareto material can suggest candidate changes. Recover the source contribution and its applicable meaning before use; a source cue does not itself establish an architecture claim. A comparison or stronger conclusion needs the relevant structures, scopes, criteria and loss conditions under its own rules.
SoTA-Echoing
These rows document transfers from source practice into C.32.MLAO. Each row names the part of the residual-reducing frame that uses the source's contribution; none imports its source-domain ontology into FPF.
Source-use boundary. Each row states a bounded contribution. Recheck a dependent claim when a source or application premise changes materially. An unsupported transfer may remain historical background; retain independently supported candidate reasoning and identify the question that needs further work.
Relations
- Builds on:
C.30.ILCfor residual triage,C.32for palettes,C.32.ACSfor architecture-characteristic criteria rows,C.32.ACEfor eval programs and eval results,C.29for a remaining question about choosing, transferring, or relying on a mathematical representation,A.19.CPMfor explicit comparison,A.19.SelectorMechanismfor set-returning selection,C.11for local choice over an existing option set,G.5for selected-set result declaration,C.19.1for scale-amenability preference claims, andC.31orC.31.ASAPfor characteristic or scale-preference claims. - Uses:
E.10.ARCHandC.30.STRATwhen stratification terms hide the recovered neighborhood;E.10andB.2.Pwhen BOSC, emergence-family, MHT, MET, MFT, boundary-crossing, or promotion-like wording hides the claim kind;B.2when the recovered facts leave a question about retaining or reidentifying the relevant whole;C.32.CONWAYwhen residual reduction requires co-synthesis of exact influence-source and transformed-side architecture content;A.6.M,C.30.LCA,C.30.TFS-REL, and method or work patterns when their structures are the affected selected structures. - Patterns for the next questions:
A.19.CPMfor explicit comparison,A.19.SelectorMechanismfor set-returning selection,G.5for selected-set result declaration,C.11for fixed local choice,C.30.ADfor architecture-description work,E.17for a source-backed publication face and source return,E.24.PUBfor the publication occurrence and audience availability, andC.32.PADfor project architecture decisions. - P2S docking:
C.32.P2Suses MLAO when cross-scope, interlevel, interlayer, or meta-holon residual pressure must become candidate-synthesis and repair content inside the wider architecturing flow. - Boundary: Use C.32.MLAO only for residual-reducing architecture candidate frames after residual triage. Use the applicable patterns for mathematical-lens adequacy, evidence, assurance, gate passage, ethical mediation, causal claim adequacy, work authorization, and final selection.
C.32.MLAO:End
Synthesize an Architecture Account of Methods and Their Use
Tech-name:
MethodArchitectureSynthesisFromSeveralStructuresPlain-name: connect methods with their performance, support and development in one usable architecture account Type: Method-description pattern underC.32Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when a decision about related methods and their use depends on several structures that do not line up one-for-one: how methods are composed, how work is performed, what supports it, and how methods are developed or retained. The question may involve different wholes and relations even when one source shows them in aligned rows.
The subject is one or several related Methods considered together with their performance, support and development. Identify the relevant Methods, performed or planned Work, participating or supporting Systems, descriptions and models, capabilities and providers, and cultural relations separately. Select only what can change the decision. The account can include relations beyond a Method's constituents: using a platform to perform a Method does not make that platform part of the Method. Each precise C.30 architecture claim names its own holon and selected structure.
Here method is the main name for a reusable way of doing. Practice can name the same way, often emphasizing its established use, as E.10:0.2c.21a explains. A claim about performed Work, a community or a description requires those separately identified objects; the wording itself introduces no broader Practice whole.
The primary working reader is an architect, methodologist, practice designer, or domain-framework author who has already found useful source material but cannot safely copy its table, stack, lifecycle, or diagram as the architecture account.
First useful move. Decide what you need to establish about the arrangement. A service agreement in force can already reserve a reviewer, a tool or a time window for the first job. To describe how a job was actually done, recover that performance. For a proposed arrangement, establish what would make its contributions available together. Different parts of one account can have different status. Select the structures that can change the decision.
What goes wrong if missed. A source layout is mistaken for the arrangement it describes. A list becomes levels, a procedure becomes a hierarchy, overlapping activities become a sequence, a model becomes the thing modeled, or a future design acquires fictitious performed Work. A local repair may also look successful only because its unresolved burden moved to another structure or scope.
What this buys in practice. The reader can explain how the selected Methods and their use produce a result that the recipient needs. An account of a working practice can finish by showing which connections sustain it. When a design choice concerns a conflict or a transferred burden, trace that condition through the alternatives and show how it changes the conclusion. Use the form in which the recipient can recover those relations.
Not this pattern when.
- Use the Method, Work, subject, description, capability, provider, or cultural-change pattern when only one such claim is unclear.
- Use base
C.32when the need is a palette of architecture candidates for one already grounded architecture question, rather than one coherent answer about Methods and their use from several non-isomorphic structures. - Use
C.30.ILCandC.32.MLAOwhen an already recovered cross-scope conflict or residual is the main subject. - Do not use this pattern for one clear Method decomposition, one procedure order, a carrier index, a universal level stack, a mandatory record schema, domain filling, product-roster generation, or lifecycle design.
- A practitioner can use this Method to prepare evidence for a framework or project decision. Making that decision, establishing a product, publishing a description, or putting a proposed organization of Method use into effect remains separate work.
Use an adequate existing account directly when it answers the question under the present conditions. For the smaller question of what encompassing work is being done through one current action, start with B.1.5.EW. It can expose a missing constituent or a changed condition without requiring this full synthesis. Use B.1.5.RS when the difficulty is preserving encompassing uses while replacing one constituent. Return here when the answer depends on several structures that do not correspond one-for-one.
The result is an architecture synthesis for Methods and their use: a readable account connecting the selected Methods, Work and other identified objects through the relations relevant to one architecture question, with their consequences for the decision. The account describes how these objects fit together; whether any Methods form a composite Method remains a separate claim under B.1.5.
Problem
Professional-practice sources commonly mix several useful accounts. One may show Method order, another Work participation, another organizational or technical subjects, another capabilities and providers, another descriptions or models, and another cultural generation or retention. Their rows can look aligned while referring to different wholes and relations.
Three failures follow.
- Layout substitution. The source's visible order or nesting is copied before the practitioner knows whether it means Method composition, Work order or overlap, subject composition, model grouping, provider dependency, cultural scale, or only chapter order.
- Present/future substitution. A proposed organization of Method use is described as though a representative Work occurrence had already happened, so the architecture gains evidence it does not have.
- Local-success substitution. A candidate resolves one visible conflict but hides the residual or burden it moves elsewhere. The account reports the gain and loses the trade-off.
A fourth failure appears only in some uses: a live product-boundary question is reduced to one favored publication arrangement, or a product comparison is forced into an architecture question that does not concern product identity. Both errors confuse a conditional comparison with a universal step.
Forces
Solution
To explain or design a practice architecture, begin with the work the whole practice must sustain. Identify which contributions must be available together, what can be supplied independently and how a change to one affects the others. For each relevant kind, follow its own constituent–whole relation to the extent the question requires. A model of one performance can explain a relation across several scales of that work. A reusable account of the Method has to identify the relation that carries across its admissible performances. Choose the ranges under C.39.SA when they are still open. Use C.37.MM:4.1.1 when the conclusion depends on several models together.
A mantra for the architecture of a practice. What work must this arrangement sustain? Which contributions belong to one reusable method, and which can be supplied independently? What connects them in a particular performance? What must be available together? Who can perform, learn or obtain each contribution? What changes for the whole if one method, provider or means changes? Which distinctions must its description preserve for this use? What was proposed, what actually happened, and what reached the recipient?
For the narrower question of bringing one known or missing way of working into use, follow the contribution through provision and performance.
A mantra for following a method from a need to its use. What work is needed, and what is missing? Is there a way to do it? If several ways are available, how do we choose? Where can we find or develop a missing way? How will the way reach whoever needs it? What must the performer learn, what can the performer discover through inquiry or practice, and what help will be available? Which method fits this case, and what must be set or changed? Are the means and authority in place? Who did what? What did the work produce, what reached the recipient, how was it used, and what effect was observed?
Use as much of the mantra as the job requires. If a cook already knows the recipe and has been denied access to the kitchen, the immediate task is to obtain entry from the agent authorized to grant it or find another available kitchen. Check readiness when the decision is whether the whole job can start. Develop an architecture account when the arrangement itself needs examination, for instance when changing who provides a step also changes who can learn or perform it. A methodologist can reconstruct a way of working from observations and examine the grounds for its applicability. A practitioner can find the same way in a discipline's literature, learn it from a demonstration or commission the work. Discovery by this practitioner establishes novelty for that practitioner. A claim of novelty for a discipline needs comparison with what its practice and sources already supply.
A discipline can describe one reusable method. If the decision calls for comparing several, recover what makes each one applicable. Call them a family when the grouping matters and its basis can be stated. A family catalogue may compare many methods. A.3.2 identifies a MethodDescription by the one Method it describes and what it says about doing it. To put a way into use, identify the method that fits the present requirements and the settings its applicability permits (A.3.1:4.7–4.8). Develop the missing way through C.39 if none fits. A description or demonstration may give the performer what is needed to try it. E.23.CDI:4.2.1 addresses acquiring the capability or obtaining continuing support. A.15.5 tests readiness for the particular job, and A.15.1 identifies its actual performance. Establish what the work produced, whether it reached the recipient, what the recipient did with it and what changed. A catalogue may exist before anyone receives it. A.6.P.WMR helps recover the particular relation when “result” leaves it open. If an accepted problem formulation is the starting point, E.18.1 carries its relevant claim to the particular decision that needs it. In teaching, follow the dependencies aloud. In practice, begin with the gap that matters. An observed failure can take the practitioner back to the choice of a method, the help required or an unmet condition of the job.
Eight actions
- Choose an entry. For a claim about performed Work, start with an identified occurrence and recover its A.13 and A.15.1 grounds. For a claim about what can already be obtained, check the applicable agreement, qualification, access and availability. For a proposed organization of Method use, start with the intended use or WorkPlan and identify what must become effective. An approval can settle a design obligation before the first use. To establish performance or its effect, recover evidence for that claim.
- Name the subject and participants. State the subject under discussion (
EntityOfConcern) and the kinds of the participants before classifying rows or positions. A source word such as practice, layer, role, model, or provider does not settle those kinds. - Build alternatives for the question. State whether the alternatives are competing architecture accounts of the same arrangement, proposed changes to that arrangement, or both. Competing accounts can reveal a relation that a sequence-only description misses; proposed changes alter how the Methods would be used or supported. Keep those comparisons distinct. Develop or reuse alternatives that can change the decision or fail on different cases. When NQD is the current generation test, retain NQD-distinct candidates: their novelty is not a relabeling, their use-value preserves the declared constraints, and their diversity exposes different failure modes. A first layout needs examination against the question; retain it when it supplies an adequate answer. Do not turn NQD into one scalar score.
- Recover each relation used. In ordinary words first, state whether the source claim concerns Method composition or order, Work parthood or overlap, a subject relation, description or model use, a capability or provider dependency, cultural dynamics, or another admitted relation. Use the pattern that defines or constrains that relation when the claim matters to the answer. For structures of a Method and its uses, C.30.ASV:4.5a helps select the question before drawing its representation.
- Separate the described objects from their descriptions. Keep identified Methods, Work, Systems and their relations distinct from the descriptions, models, views and coarse-grained accounts about them. A description can itself be used in Work; name that use separately from what it describes. State the correspondence and the distinctions each description preserves or loses.
- Resolve a whole-identity question when needed. Use
B.2when the current claim raises a consequential question about retaining the identified Method, Work, System or Discipline whole or identifying a new one. Keep the existing whole when a supported explanation is sufficient, and retain uncertainty when the needed basis is missing. Reidentify a selected structure underA.22's four identity discriminators when the claim requires a different selected structure. - Compare conflict and moved burden. Follow the change to the people, agents or equipment that receive the work. Establish what their extra effort or waiting changes for the required result. A shorter drafting step can yield a later accepted release when the reviewer becomes available late (C.32:5.1). Attribute the elapsed time of a completed release from its actual start and acceptance. If the claim concerns who designed the change, recover the choices already supplied and the ones that the performer made.
- Return to the work the arrangement must sustain. Explain what the proposed connections would supply or which connections held in actual use. Follow a changed contribution through the obligations that depend on it. When the question extends to cultural change, distinguish the project’s deliberate intervention from the subsequent transmission, use and retention of the practice.
Short working result
Write the first result as a short account that another practitioner can use without decoding a private ontology. This outline is a writing aid, not a mandatory schema:
A compact table can support this account. Symbolic predicates are optional assurance and belong after the readable explanation. The result stops at architecture evidence; comparison, selection, decision, publication, evidence, assurance, or gate claims use their own patterns when those claims become current.
Conditional product-boundary comparison
When deciding where practitioners should find and maintain domain guidance, compare the arrangements that could actually meet their need. Four possibilities have arisen in the FPF ecosystem:
- one domain DPF;
- independently maintained sibling DPFs;
- direct use of FPF plus exact domain sources; and
- a proposed common applied DPF.
For each serious possibility, state the recurring problem, the first useful result, who could maintain it, and what would be hidden or duplicated. A binding access condition can rule out an arrangement before it is developed further. Other products can have different useful boundaries. Identify alternatives from the uses and maintenance work of those products.
Stop, lower, and reopen
Stop when the short synthesis identifies the entry, selected structures and relations, consequential correspondences and losses, any conflict or moved burden, and the next useful action. Include different alternatives when the receiving question needs a comparison; an adequate account or retained arrangement can finish without inventing another option. Add a view only when it can change the answer.
Limit the conclusion that lacks its grounds. For performed work, identify the occurrence. For a right, booking or other effective provision, check its terms and availability. For a proposed arrangement, recover the intended use and what is still needed for it to work. Keep the other supported conclusions and obtain the particular missing source when the decision depends on it.
Reopen when representative Work contradicts the synthesis, a proposed organization of Method use is put into practice, a new structure changes the decision, a moved burden becomes material, a direct source changes a relied-on premise, or the product-boundary question enters or leaves scope.
Archetypal Grounding
Preparing borscht for eighteen guests. The cooking repertoire contains several established ways to prepare borscht. The host asks which regional style the guests prefer. Their stated preference helps the cook select a recipe. Eighteen portions determine a quantity. If the selected way admits that quantity and the pot has the required capacity, scaling ingredients is a setting of the same method. If the pot is too small, an allowed split into batches, another established method, or a trial of a changed order of heating are different responses. A trial may show whether the change works. On the day, access to the kitchen, ingredients, equipment and an available cook determine whether preparation can start. Serving eighteen bowls supplies evidence of delivery. To establish that the guests liked the dish requires evidence of their response.
Checking a terminology collection. A methodologist observing several editorial teams reconstructs how they test whether two terms denote the same kind. Another team can use the account to select a check for its language pair. A recording shows how an editor compared the criteria. The account and recording are available to the receiving team even if a contracted service performs the entire first check. Its returned list of incompatible definitions can meet the immediate editorial need. To establish the team's ability to repeat the check with its normal tools, examine the team's contribution on a representative assignment. The criterion for equivalent meanings may stay the same. If the current method already permits the extra test, using it is an allowed choice. Requiring a new independent test before accepting a match changes the operation or its stopping rule. Compare the revised way with the original under A.3.1:4.7.
Methods and their social-dance use. A five-minute social-dance occurrence sits inside a longer party. The architecture question connects dance methods, their performance, participant capabilities and the transmission of the style. Preparation has a first-then order, while movement regulation, figure performance, partner coordination, style-specific performance and event participation can overlap in time. A teaching diagram is a description; bodily and pair capabilities concern participating Systems; scene transmission and later style selection are cultural relations. The synthesis keeps these structures linked but not identical. A sequence-only stack loses simultaneity, while one culture-level stack loses the Work and subject relations.
Show — engineering Methods in performed Work. In a 90-minute power-converter integration run, model correction overlaps simulation, rig reconfiguration precedes closed-loop test, assurance overlaps later testing, and restoration of a simulation service overlaps the run without becoming part of the converter-integration Work. Reconfiguration changes the rig; the test uses that changed rig only after reconfiguration is complete and the rig is authorized for testing. Plant and controller models are descriptions used in the Work, not the converter or the Work. The team compares a sequence-only account with a linked account of Method, Work, subjects, models, platform service, capability, and evidence. The linked account wins for this question because it preserves the order and guarded enabling dependency between rig change and test, together with the overlapping activities. Those relations alone do not establish the internal organization of one Method or a transformation-flow structure. If evidence assembly is moved to the platform, the synthesis records platform availability and traceability as a moved burden rather than declaring an unqualified gain.
Show again — possible-future use of a Method. A five-day Method-adaptation sprint is obtaining Work. The candidate Method is described, but the proposed organization of its use has not yet been performed. The team anchors the architecture in the intended project use, the incumbent Method used during the sprint, the candidate MethodDescription, and a planned representative trial. It compares candidate organizations of Method parts, tool and description support, provider capability, trial Work, and later repertoire maintenance. It states which support and representative Work would put the proposed organization into use and what the trial must observe. It does not invent past Work enacting the candidate Method. If the question is only how the project will use the candidate, the four product arrangements are omitted; if the question becomes where an independently maintained Method family belongs, that conditional comparison is opened.
The dance example explains how several contributions can be active during one performance. The integration run distinguishes a constituent performance from an external service that happened at the same time. The adaptation sprint supports a proposal for future use. Each account identifies the relations needed for its particular whole. A claim that a practice has already changed calls for the corresponding evidence of actual use.
Bias-Annotation
Scope: Limited to synthesizing one usable architecture account of Methods and their use when several selected structures do not line up one-for-one. This is not a universal ontology, a required view set, a product-selection Method, or a complete architecting lifecycle.
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Consequences
The Method preserves useful differences among Method, Work, subject, description, capability/provider, and cultural structures while producing one decision-ready account. It prevents source tables, lifecycle diagrams, and models from silently defining the objects and relations under discussion. It also makes prospective claims and moved burdens inspectable.
The cost is a small amount of relation recovery and, when needed, alternative construction before a clean diagram can be accepted. Some attractive source organizations will remain unresolved or will be retained only as descriptions. That cost is proportional: the first result is short, structures that cannot change the decision are omitted, and formal assurance is optional.
Receiving domains supply their own Methods, institutions, evidence, constraints, providers, cases, and maintenance conditions. Reuse of this Method does not move those fillings into FPF.
Rationale
The eight actions follow the dependencies of a synthesis. The entry must distinguish obtaining from prospective before evidence is interpreted. Subjects and participant kinds must be known before a source layout is classified. Test the first layout against the receiving question and its required relations. Construct alternatives when that question needs them; the recovered relations and description correspondences then make those alternatives comparable. A whole-identity question is resolved on its own grounds. Conflict and moved-burden comparison expose the trade-off, and the final deliberate/distributed distinction prevents one project from being mistaken for cultural evolution.
The Method does not require all structures to fit one master diagram. Its common result is coherence through stated relations and correspondences, not visual isomorphism. This is why a short prose account can be the first useful result and why a matrix, graph, or formal model remains optional.
SoTA-Echoing
The sources below make complementary contributions; none is used as authority for the whole Method. The pattern composes only the bounded moves stated in the table, and the final row keeps current FPF distinctions authoritative.
Recheck these source uses when a newer applicable architecture-description standard, practice-architecture synthesis, work-system method, or multi-domain technique changes the practical move, or when unlike cases show that the eight actions no longer preserve a decision-relevant structure at proportional effort.
Relations
C.32.MWA.PP uses this multi-structure account to construct and compare feasible arrangements of Method performance and provision, including direct performance without an office, shared or specialized provision, federation and external supply. It adds the design of the result joins, combined demand, capability, means and authority needed to realize the selected network.
- Builds on:
A.3.1for Method identity,B.1.5for Method order/composition and Work enactment distinctions,A.13andA.15.1for the bases and independent admission of performed Work, withF.6afterward when precise assignment-bound attribution is claimed,A.22for selected-structure identity,B.2for whole reidentification,C.30andC.30.ADfor architecture and architecture-description distinctions, andC.36for cultural-change relations. - Coordinates with: base
C.32for general architecture candidate synthesis;C.17andC.18only when NQD generation, novelty/diversity, archive, or front claims are current;C.30.STRATwhen level, layer, tier, stack or similar source wording leaves the relevant distinction unresolved; andC.30.ILCplusC.32.MLAOwhen a conflict, residual, or moved burden needs its own treatment. - Provides a result to:
E.4.PFADwhen several practice structures change a framework-architecture answer;E.4.DPFwhen those structures change DPF identity, scope, pattern organization, first use, or use of named results;E.4.DPF.DAwhen package adequacy needs the resulting two-way sequence-versus-simultaneity evidence; andE.23.CDIonly when target-practice architecture must be recovered or compared before capability development. - Does not replace: the direct domain sources and Methods,
E.4product-family decisions, explicit comparison or selection, project architecture decision,E.24.PUBpublication occurrence and availability, evidence or assurance, or the currentness and refresh patterns.
C.32.MWA:End
Construct and Compare Method Performance and Provision Arrangements
Type: Method pattern Status: Stable Primary working result: a Method-architecture decision episteme explaining how related Methods would be performed and supported together: the decision-changing structures and direct relations, their truth statuses, serious alternatives, joint demands and moved burdens, the selected or rejected arrangement, and the conditions for its realization or reconsideration. Co-used Methods can remain plural; this result need not establish a composite Method.
Use This When
Use this pattern when individually plausible Methods or candidate accounts could be combined or co-used in materially different ways, and the receiving result depends on their composition, Work overlap or order, allocation, relations among the affected Systems and means, descriptions, provider access or cultural relations. A typical practice or development network can supply the intended transformations while leaving their local performance and support unresolved.
The practical question is how the related contributions can become obtainable together. A Method may need a qualified judgement before work begins, while the proposed expert is available only later. A common service may save preparation but delay local decisions. Learning may consume the only expert time needed for current delivery. Assigning names to boxes does not settle these interactions.
Begin with the obtaining or possible-future practice and one result at risk. Follow that result from its subject through the Method that could obtain it to the person, AI or other System that would perform the work. Recover one condition that can defeat the connection: missing competence, time, equipment, evidence, permission, a usable incoming result or an available receiving action. Construct alternatives by changing that connection and completing its consequences through the receiving result.
A comparison uses at least two serious arrangements, including the feasible incumbent where applicable, and exposes local gains and burdens moved across participants, scopes, Systems or times. It can retain an incumbent, select a conditional future arrangement, warrant a bounded attempt, return one missing contribution or reject an unavailable promise. It does not itself make proposed relations actual or provide every subject Method needed to realize them.
The intended user can interpret the supplied Methods or obtain qualified interpretation, discuss work with participants, and compare commitments. The user need not possess every supported mastery. The setting can be one person's work with AI, an enterprise, an external provider with clients, independent practitioners or a combination. These examples do not limit the Method to development, offices or one organization.
Use an adequate existing arrangement directly. An isolated access repair, routine assignment under sufficient capacity or qualified supplier comparison need not trigger another architecture exercise. Do not repeat A.3.1.QM's individual connection work. If the only question is whether a named relation or Method composite is supported, use C.30 or B.1.5 and stop with that clarification; there is no C.32.MWA.PP comparison to complete until a materially different arrangement could change the decision. Do not invent a rival merely to enter the pattern.
Use the strategic or subject practice first when the unresolved question is only whether the aim is worthwhile or what the Method should do. Return here when performance and support can change that answer. Project, process, case, lifecycle, table and diagram views are not architecture alternatives unless their underlying proposed relations differ. A mathematical graph is one possible representation when its nodes, edges, semantics and use have been selected; otherwise name the structures and relations.
Working Distinctions
Mastery in the development cases means a person's qualified ability to contribute to a family of tasks under stated conditions. Organization and AI capability claims keep their own holders, mechanisms and evidence. Support describes how a specified contribution makes other work possible or changes its conditions. State who performs the supplying work, what it returns, and how the receiving performer uses that result. Identify the means, access and authority separately when they affect the connection. Teaching, an expert's execution, a tool and permission to use it can all matter, but they occupy different positions in that explanation. A change to the supplying performer's capability or Method is another claim, requiring its own result. Cultural continuation concerns how practices are carried, understood, reproduced and changed; installing a procedure does not establish it.
Problem Frame
The practitioner receives identified Methods or candidate accounts, their intended effects and any qualified connections. These may be explained in an ordinary practice account or a typical transformation-flow network. The task is to construct and compare their performance and support arrangement while preserving the supplied contributions, conditions and unresolved gaps.
The pattern considers Methods, Work, descriptions, capability, tools, variants and simultaneous contributions at several scales. Recover the subject of each project, process or case account under A.15.6 before comparing structures. Several useful structures and descriptions can be connected without sharing a subject or aligning one-for-one.
A project needs decisions. It must choose which provisional results may be used, which Work may overlap, where authority stays, which support arrangement is acceptable, and where burden moves. The comparison must preserve the several structures while remaining small enough to act on.
When continuing development is the supplied practice, preserve its full contribution: noticing an emerging difficulty or possibility, forming a problem, constructing a worthwhile reachable change, arranging and enacting it, then making a further development choice from the changed position, including after success. Consolidation, ordinary work or ending a particular line can be justified continuations. C.32.MWA.PP receives that practice and arranges its performance; it does not construct the whole development network by default or reduce it to acquiring or introducing a known Method.
Problem
The same drawing is often expected to describe what changes, how it changes, what work occurs, who is competent, which unit employs whom, and who may decide. These are different relations. Turning them into one hierarchy makes missing contributions disappear.
For example, a box for “development” can stand for a Method, a team, a proposed programme, a person's capability or work not yet performed. Naming an expert beside it can conceal unavailable time. Placing the box centrally can conceal missing local knowledge. Distributing it to every workplace can conceal preparation and coordination that nobody supplies. A common portal can appear to own every subject it merely routes to another provider.
Separately plausible arrangements can conflict when combined. The same leader may be needed to run current work, help others and develop their own capability in the same hour. A local incentive to finish quickly can prevent the examination and revision the culture requires. A provider can complete its contracted activity without leaving a client able to continue. More capable AI can invalidate a human bottleneck assumption while leaving permissions, physical access or the learner's own practice unchanged.
The recurring problem is to construct and compare complete arrangements across these relations and scales, then supply a decision and requirements for a feasible local enactment. The comparison must be allowed to change the allocation, grouping, means, Method connection or intended remit. Otherwise architecture becomes justification of the first office proposal.
A one-stack account also mistakes Method composition for Work order and description layers. Different views of the same relations can be mistaken for architecture alternatives. The resulting decision can assert obtaining architecture before its relations exist or miss a first–then guard because everything is described as concurrent.
Forces
Method-architecture comparison also preserves simultaneous contribution without losing order, and useful co-use without inventing a Method whole. Several structures can expose different conflicts, while views that cannot change the choice add only cost. Proposed relations are useful design inputs but cannot be reported as obtaining practice.
Solution
C.32.MWA recovers the several structures and their actual or prospective status. The construction here uses that account to design and compare realizable performance and provision: connect each needed result to a Method, make the receiving joins work, and obtain a feasible allocation of capability, time, means and authority. An already adequate architecture account is reused; the detailed comparison below is needed only for the arrangement question it leaves open.
Connect the supplied practice to possible and actual performance, construct materially different arrangements, compare their interactions on one bounded basis, and return an explained decision with realization or reconsideration conditions. C.32.MWA supplies the several-structure synthesis, C.30.ASV:4.5a the choice of questions and relations, C.11.CRC the finite comparison, and C.11 or a sufficient domain decision Method the choice.
The sections below explain connected questions, not a fixed production pipeline. An adequate existing result can settle a question. A changed premise can return the work to an earlier question, and independent contributions can proceed separately. The supplied practice retains its dependencies; paragraph order adds none. In a familiar case, follow the result, locate the decisive interaction, compare complete alternatives and retain the necessary continuation. Open the fuller explanation where a connection is unfamiliar or its premise changes.
Recover the supplied network and the result this local arrangement must obtain
Name the obtaining or possible-future practice, situation, configuration, window, decision authority and receiving use. Name the receiving subjects and the change or preserved condition that matters. A customer result, a performer's acquired capability, a reusable Method, an organization's continuing capability and cultural transmission among participants can all matter. Keep them separate when one could be obtained without another.
Recover enough of the supplied practice account to identify:
- what is transformed or kept usable, and which subject continues through the change;
- the Method or professional operation proposed for that change;
- the input and receiving conditions that make the result useful;
- independent branches, repeated or conditional work, and returns from later observations;
- the continuing work and other contributions the arrangement must protect.
For an E.18 or E.18.NET account, use its qualified positions, members and relation meanings as supplied. A typical diagram is a reusable description, not evidence that those relations already obtain in the local organization. Keep proposed local bindings and future Work prospective until their conditions are met. C.30.TFS-REL governs a stronger architecture use of an exact selected transformation-flow structure or network.
When the question spans several organizations, independent communities or people, do not invent a containing organization to make the picture tidy. Name the participants and the agreements or other relations on which the proposal depends. A network is not the Agent that performs the work or holds all its obligations.
Walk one supplied path forward and one necessary result backward. Forward, ask what the receiver can do with each result. Backward, ask which operation supplies the needed condition. Stop at an understood operation or qualified existing result. Where “develop competence”, “create culture” or “provide support” hides the operation, recover it through section 4.2 before allocating its performance.
Keep the remit revisable. A provider may arrange access to every required specialty without claiming to perform all of them. A person may acquire one capability and use another through an AI or expert service. If the proposed local arrangement cannot jointly supply the required results, compare a different way or return the affected promise to its strategic owner; do not quietly remove a difficult result while keeping the same offer.
Establish the Methods and the connections that performance must preserve
Bring in the identified Methods, candidate accounts and individually supported local connections, preserving their prior status and unresolved conditions. A candidate account does not become an admitted Method through allocation. If actual Work is later described as enacting a Method, establish that claim for the Method and occurrence in question.
For each decision-changing contribution, inspect the explanation or demonstration, not its name. Identify what material or situation it works on, what a competent performer does, why the action could obtain the result, which conditions are preserved, and where the Method stops.
Try the operation on one representative input or a sufficiently explicit prospective case. The purpose is to locate the first unsupported connection. Different findings require different work:
C.39 supplies the general missing-way construction: search through the required change, recover a relevant operation, try its receiving use, and construct only the still-missing connection. Keep subject-method development distinct from pedagogical construction. A learning designer may choose a contrast or practice task without inventing a missing scientific criterion. Conversely, a subject expert's correct answer does not supply a teachable route by itself.
If the original difficulty is not yet a usable problem, B.5.PI supplies inquiry from a weak cue or an apparently completed result. C.40.CD can develop a new problem from an achieved result. For continuing development, return their question to the selected development practice, rather than treating question formation as the entire development transition.
Retain distinct sources of change. Revising a Method can change what future performances require. Training can develop a person's capability; a claim of capability change needs suitable performance evidence. Assigning the work to a qualified performer identifies an expected contribution. Establish the access and authority required for that work. Reuse adequate existing evidence and authorization. Improving a tool can change available means. These repairs can be combined, but none supplies the others merely by being in the same project.
Explain the important joins in ordinary language. For example: “The domain review must return the condition under which the proposed experiment is interpretable before the participant runs it. A general coach can help formulate the question, but cannot substitute for that domain judgement.” This tells the architect what kind of performer and timing are required. “Coach, expert, learner” alone does not.
Do not force every result-use or support relation into Method composition. Use B.1.5 for a proposed whole; B.1.5.CM can resolve its identity. When that is the only unresolved question, finish with the relation or composition result. A materially different way to obtain the receiving result, rather than a larger Method name, creates the architecture comparison here.
Keep the necessary structures distinct and connect their claims
Select only the structures that can change this design. A compact cross-reference is often enough. C.30.ASV:4.5a supplies the Method questions; C.32.MWA:4.1 relates them to distinct accounts of Work, Systems and means, and cultural relations.
An individual can hold several compatible assignments. Several Systems can jointly obtain a result. A specialist can serve several organizational units. A Method can be used by several independent performers. Keep these relations visible rather than forcing one exclusive-parent tree.
For each provision connection, make the local claim explicit: this available or proposed holder can supply this contribution, using these means, in this window, under this permission, and the named receiver can use it under these conditions. Keep obtaining, proposed, contradicted and unresolved relation claims distinct. State the evidence, including whether it comes from observation or a qualified provider, and the receiving use it supports. A table is useful when it makes this comparison easier; it is not a substitute for the explanation.
An actual transformation claim needs the subject, boundaries, before/during/after basis and continuity conditions required by A.3.4. For an actual Work claim, establish the performer's A.13 bases and independently admit the occurrence under A.15.1; add F.6 only when precise assignment-bound attribution is claimed. The current synthesis can remain a useful possible-future arrangement without pretending to supply either.
Retain description relations as well. A WorkPlan's intended dates, a reusable MethodDescription and a case account can concern connected but different subjects. Record the needed schema correspondences, version relations and losses rather than treating their similar layout as identity.
At a revealing moment, explain what larger work, if any, is being performed through the selected action and how that whole constrains it. The same action can perform constituent Work while evaluation or adaptation is under way; separate product or support Work can merely overlap it. Use B.1.5.EW when this relation is unclear. Keep first–then guards alongside simultaneous contribution. Section :5.2 shows why an answer-bearing prompt can be appropriate for supported execution but defeat a test of independent recovery of the rule.
Recover the incumbent and the work hidden by the proposal
Describe the current way at the grain needed for this decision. Use an actual representative occurrence where available; for a new use, begin with a prospective case and recover the existing means on which it would rely. Include competent informal help when it is available, and its burden. An empty baseline makes any new service look valuable; an idealized baseline makes needed support look unnecessary.
Follow one consequence into each receiving whole that could reverse the choice. An expert's review can support a learner's progress, product assurance and the expert's own professional practice at once. A proposed use of that hour can therefore change several contributions. It does not create three independently available hours.
Recover demands that the proposal might hide:
- preparation, subject interpretation and development of the means;
- the performer's own work, learning, recovery and contribution to others;
- access, intake, coordination, exception response and maintenance;
- local participation, explanation, disagreement and authorization;
- continuing production or service during the change.
Ask the contributors and recipients which condition could make the proposed arrangement fail. Their knowledge can change a candidate without giving every participant identical decision rights. Keep unknown voices and evidence material to the choice visible; do not treat a senior sponsor's description as the whole work.
Locate the difficulty. If a result is late, determine whether it waits for a scarce judgement, missing input, permission, incompatible priority or an unnecessary handover. These explanations imply different architecture changes. Attribute a capability gap to a performer only when the evidence identifies the missing operation.
For human capability, qualify the actual task and permitted help; obtain a separately qualified acquisition or transfer result only when the arrangement needs that capability change or new use. For AI, qualify the relevant configuration, tools, data, permissions and evaluation conditions. For an organizational capability, identify how the contributions become obtainable together and remain available beyond one exceptional performer. Individual competence alone supplies none of the missing interfaces or commitments.
Generate whole alternatives by changing an operative relation
Construct alternatives from the recovered difficulty by developing grouping, allocation, information, decisions and support together. For a stable result, C.38 constructs comparable complete human, AI, internal, external and combined obtaining ways. When different result families or a revised remit remain in question, retain those differences and return the proposed change to the purpose and choice owner; the arrangement cannot authorize a different promise by itself.
To generate a materially different candidate, change a premise of the obtaining mechanism:
- When frequent mutual adjustment is limiting, bring the necessary contributors into closer work, or change the result/interface so that less adjustment is needed.
- When scarce subject expertise is limiting, retain a specialist contribution and improve its access, timing or reuse; alternatively develop another holder or obtain a qualified provider result.
- When distant decisions lack local information, move information to the decider, delegate a bounded decision to where the information exists, or change the work so the joint decision is no longer needed.
- When repeated support work is limiting, compare a common means or service with local provision. Retain the subject-specific contribution that the common means cannot supply.
- When a learning demand consumes essential current capacity, compare a staged acquisition, different provider, changed participation or narrower current commitment.
- When no holder can perform the Method adequately, construct a capability-development or research contribution; do not merely allocate the same unsupported work more widely.
Complete the changed candidate through its receiving result. Moving judgement locally may require training, data access, an exception boundary and agreement on consequences. A shared tool may require maintenance and a qualified operator. An external service may require a different interface, lawful information access, available response and a usable exit. These are parts of the candidate's burden, not free remedies added after selection.
Compare the whole without a new office whenever existing workplaces and support can plausibly provide the result. Describe how self-development, help to others, specialist work, maintaining common means and resolving conflicts would be performed within that arrangement. “Everyone is responsible” leaves the obtaining connections unresolved. A named department is not necessary when those connections are supplied elsewhere.
Similarly construct AI-based arrangements as whole alternatives. Qualified agents may notice, formulate, choose within delegated authority, teach, coordinate and execute. An arrangement can combine personal agents with common tools and externally maintained subject services without recreating a human office in software. Give it access, authority, support, evaluation and recovery conditions. Preserve a human's own performance only where that is part of the result or a justified constraint.
A common carrier can be responsible for making help obtainable while independent providers perform it. Complete that candidate with how a question reaches a qualified provider, how the receiver recognizes a usable response, who handles refusal or failure, and what keeps access available. A portal with an unfilled request queue supplies only a portal.
When an architecture comparison is warranted, compare at least two serious syntheses whose proposed direct relations differ.
For a decision on an assignment, distinguish the options the brief provided from the construction needed to make them workable. A practitioner who chooses a supplied service on the stated criteria may have completed the selection. To credit the design of a different arrangement, identify a changed handover, allocation or dependency that the practitioner established and explain how it meets the receiving need. Use the same distinction when the purchaser commissions a design from an agent or a supplier.
Treat central, local, specialized, federated and external descriptions as properties of the resulting candidates. They are not a catalogue from which the Method starts. One candidate can centralize common tools, distribute performance and retain an external specialist at the same time.
Keep the adequate incumbent and its smallest useful repair in the comparison. Generate further alternatives only while a new mechanism or failure mode can change the choice. Do not manufacture token rivals to a sufficient answer.
Expose contradictions across structures, participants and scales
Use contradiction here for demands that cannot jointly be met under the proposed conditions, or for claims about the same contribution that cannot jointly hold. A difference in preferences is not automatically such a contradiction. It becomes decision-relevant when a required contribution, legitimate interest, protected condition or resource is affected.
For each claimed gain, perform three connected inquiries.
Follow the changed contribution outward. What receives it, what else uses its performer or means, and what later action depends on it? Preserve the relations instead of assuming every receiver belongs to one management hierarchy. A change can cross a personal task, a team, an enterprise, a professional community and a supplier without making them nested parts.
Ask what was moved. Did a shorter expert response move interpretation to an unqualified learner? Did local autonomy move risk or integration work to another unit? Did a cheaper office consume unpaid community work? Did a more demanding method improve an individual result while preventing sufficient throughput for the whole service? Keep the affected result and bearer identifiable. A gain reported at one scale does not cancel a loss at another.
Challenge the combined mechanism. Hold the stated conditions fixed and ask whether the candidate can obtain all its promised results. Follow a consequential case, including a plausible exception. If a needed contribution is removed, what fails? If local results are obtained, what makes their combination sufficient for the receiving result? Trace the required intermediate results, their receiving operations, interactions and shared conditions. Construct or obtain the missing connection through C.39, and use C.40.JF for unresolved joint reachability. A list of necessary parts does not establish that their combination is sufficient. Keep the whole claim unresolved while its obtaining connection is missing, preserving the supported contributions.
Then construct a repair that changes the proposed obtaining relation, not just the wording. Keep a causal hypothesis distinct from a qualified causal conclusion; C.28 governs any causal reliance used in the decision. Possible repairs include altering timing, distributing a decision, adding or replacing a qualified contribution, changing an interface, reserving shared capacity, separating conflicting roles, reducing simultaneous commitments, or revising the remit with its owner. Follow the repair through the same receiving wholes. An extra approval can reduce one error while adding a delay that defeats the useful result.
For example, a local manager is asked to meet a production deadline and release the only qualified worker for the same two-hour development episode. Both commitments may be individually reasonable; together they are unavailable. Naming a development office does not create another qualified worker. The comparison must change a deadline, supply another qualified performer, alter the episode or defer a commitment. Who may authorize each change matters.
Also compare time horizons. Training a second mentor may reduce present help while removing a recurring future bottleneck. Include both periods under the same scenario. Do not count the new mentor's future capacity before qualification or count the training hours as available client support. Conversely, do not reject the investment merely because a one-week comparison excludes its intended return.
C.32.MWA:4.1(7) supplies the moved-burden question. C.11.CRC:4 supplies the finite configuration, horizons, result and resource coordinates, interactions and future-option effects. Use qualitative explanation where it settles the choice. A multi-scale picture is useful only when it makes a decision-changing relation clearer.
Construct the jointly feasible workload and commitments
Turn the serious candidates into possible performance at the scale of the present decision. Begin with the promised result count or range, quality and response conditions. Recover the input demand each Method places on each qualified holder and means.
For a small case, place the required work in actual windows and compare it with available capacity. Ask:
- Which contribution can be shared once, and which must recur for each participant or case?
- Which demands can be combined in one episode, and which compete for the same performer, device or information access?
- Which result must exist before another action begins, and when can the receiver use it?
- Which qualification or authority prevents apparently spare capacity from substituting?
- What current contribution or later option is displaced by the proposed schedule?
Count common preparation once at the declared scale. Count recurring performance and support wherever they occur. If one group episode obtains several results, credit them only where each required participant contribution and receiving condition is preserved. Neither multiplying shared preparation by every user nor assuming all group learning is free counts the demand correctly.
A necessary capacity inequality can expose failure cheaply: total demand for one qualified resource in a window cannot exceed its available capacity. It does not establish an executable schedule. Precedence, simultaneous work, locality, recovery and incompatible access can defeat the candidate. For a fixed booking, find a feasible allocation. For a promise about future requests, establish which requests may coincide, who can allocate the resource and what they can know before committing it. The allocation rule must serve the permitted cases the promise covers. For a rate or forecast, evaluate the population and conditions specified in that commission.
For an overload, construct the alternatives: reduce starts, stagger work within a still-useful window, obtain another qualified supplier, combine compatible performance, change the Method or change the promised result. Recalculate the affected recipients and periods. A transfer to another team is unavailable if that team already spends the same capacity elsewhere.
Make the required commitment precise enough for its owner to judge. Name what is to be supplied, when and under what acceptance and failure conditions, which resources would be reserved, and which decisions would be delegated or retained. For an already available arrangement, recover the actual commitments rather than requesting them again. A budget owner, subject assessor, line manager, community participant and service provider can hold different rights. An attractive design is not their agreement.
If the candidate needs changed authority or a new agreement, include the obtainable route and delay in its transition. A future arrangement can be compared on explicit conditional premises before those relations exist. Its dependent enactment cannot begin until the conditions are effective. If a participant may refuse and no contribution is secured, keep that dependency conditional or select a surviving arrangement. Willingness to discuss the proposal is not provision.
Compare complete changes at proportionate effort
Use one current baseline, declared horizons and decision-changing scenarios. Compare each candidate's result and resource consequences, including the transition from the incumbent. C.11.CRC supplies the comparison; C.11 supplies the reasoned local choice. The rights and agreements needed to carry it out remain conditions established through §§4.7 and 4.9.
First distinguish an unavailable arrangement from a costly but possible one. Then retain the coordinates that could reverse the choice. These can include useful results, capability retention, delay, participant effort, qualified scarce time, money, continuity, autonomy, dependence, error exposure and future flexibility. Compare contributions to their actual receivers. A lower supplier cost that makes client performance impossible is not a cheaper equivalent service.
Do not add unlike ordinal scores to manufacture a winner. A candidate can be worse for one participant or horizon and still be selected through an explicit legitimate trade-off. Protected conditions retain their force. If the options promise materially different results, return to the strategic choice rather than silently treating them as equivalent staffing alternatives.
Challenge the sensitive premise. If common provision appears cheaper only when demand exceeds a certain level, compare lower demand. If local provision works only with a qualified leader, compare their absence. If AI eliminates a bottleneck only with unavailable data or authority, preserve that conditionality. Use the changed condition to revise or reject the candidate, not merely to add a warning.
Separate an ordinary operational check, specialist advice or monitoring, and an independent conclusion required by the receiving decision. Do not add the last contribution merely because another view might be interesting. When it is required, its obtaining conditions belong in the alternatives and their whole cost. Competence and spare time do not establish independence, access or a usable reporting route.
For example, a common service is qualified to support the practice, but its designer also promises an independent conclusion about that service while controlling the evidence released and permission to report a failure. The supported work may remain feasible while that promised conclusion is unavailable. Compare an existing sufficient engagement, an obtainable qualified arrangement with suitable safeguards, or a narrower reliance permitted by the decision's obligations. Outsourcing the report or naming another colleague does not by itself supply those conditions. Section :4.9 gives the corporate case its direct supplier.
Apply C.11.DUA before commissioning more inquiry. Name a plausible answer that would change the arrangement or permitted reliance; compare obtaining it with a smaller reversible start, a narrower promise, existing evidence, delay or stopping. Include the recipient's interruption and transition costs as well as the architect's analysis. A binding assurance obligation remains binding until legitimately changed, but an optional attractive proposal does not entitle its author to an unlimited validation programme.
The decision can select a complete available arrangement, authorize a discriminating bounded attempt, retain alternatives pending one supplier result, revise the remit or stop. A comparison is not a decision by format. Preserve the chosen trade-off and the material return condition in the ordinary result needed by the participants.
Keep the decision within its evidence and authority. A non-causal capacity comparison can warrant a bounded trial while an effectiveness claim remains unsupported. Consume a causal-use result only within its verdict boundary. A selected proposal establishes no obtaining ArchitectureRelation, and a possible-future assignment supplies no actual performer occurrence.
Implementation and trial are separate Work choices. Retaining an adequate incumbent, or keeping a proposal available under stated conditions, can finish the comparison. Select additional Work through C.11.DUA only when its obtainable contribution warrants its full burden and the needed access, allocation, authority and window can be supplied. Carry observations that could support or defeat the proposed relations, protected conditions, stop and recovery into that choice. After enactment has begun, recovery preserves performed history; it is not a return to an earlier alternative as though the intervening Work never occurred. Section :5.3 makes that difference explicit.
Give the selected realization its contribution and conditions
The architecture result explains what would need to hold and why the arrangement was selected. It can finish the current work without a new implementation or trial. When further realization is selected, give its practitioner a complete receiving contribution and the conditions that remain to obtain. An organization chart or a request to "implement the architecture" does not carry that requirement.
Begin with the useful result and its receiver. Explain which Methods and local actions must contribute, what each receives and returns, the important exception path, permitted variations, ordinary support, protected continuing work, and the decision and authority for the attempt. Include the decisive rejected alternative, accepted trade-off and condition for reconsideration. Keep current evidence beside unestablished relations so the receiving practitioner can reuse what is sufficient.
Before assuming a contribution, establish that it can be provided under the proposed conditions. Give those who must assess or agree to the arrangement enough information to understand and challenge the relevant proposal. A provider can reveal an unavailable window; a recipient can expose an unusable intermediate result; a practitioner can show that a suggested action changes the Method's criterion. Return the challenge to the affected comparison. Explanation, discussion, negotiation, facilitation and leadership use their own sufficient Methods where needed. C.32.MWA.PP supplies the arrangement and unresolved question; it does not make all those practices architecture steps. Voluntary participation needs agreement. An employer's allocation authority has a scope and does not create missing competence or make a subject claim true.
Select the receiving Method from the result still missing.
For a separately selected organizational realization, use the applicable organization-change Method. Give its practitioner the selected design or duly authorized bounded attempt, receiving use, participants and configuration, acceptance basis, ordinary and exception paths, support and service consequences. A first comparison can find that an existing implementation or introduction already supplies the required capability; then use that result without another exercise. Otherwise have the practitioner trace crossings with participants, obtain missing conditions through their owners, exercise the complete contribution, repair the defeated crossing and return a bounded capability result. C.32.MWA.PP explains which joint capability is needed; the organization-change Method supplies realization, including an unresolved return. An isolated tool or learning question without an organization-change difficulty uses its professional Method.
For an account or tool that performers cannot use, use A.15.8.MS. Give it the intended user, task, input, point of failure, required operation and support condition. Request material or tool use that lets the performer recognize the situation, choose a branch, obtain the input, perform the operation, interpret the result and return a failure. A short local instruction can suffice when it preserves the condition that changes the action and gives access to fuller explanation when needed. A.15.8.MS's result concerns use of the Method account and support; it does not establish mastery or continuing organization capability. A changed criterion or missing subject operation returns to its Method owner rather than being hidden inside instructional editing.
For help with introducing, using and revising a practice, use A.3.1.MI in FPF Core. Supply the target Method or candidate account and its status, actual receiving practice, observed difficulty or explicit proposed condition, current introduction/support results, those who can supply and receive help, available means, authority, protected work and intended later use. Keep the target practice distinct from the strategy used to help it become usable. The support practitioner chooses and adapts that strategy to the determinant; training is one possible contribution, not the default remedy. Use A.3.1.MI:4.1 to perform the selected introduction, examine later actual use and non-use, and return keep, revise, branch, replace or stop for the target or introduction strategy. Use A.3.1.MI:4.5 when a changed result or proposed adaptation leaves explanations that would change the next action.
For a separately selected corporate assurance engagement, use the applicable corporate governance Method to arrange and receive it. Supply the subject, question, criteria, period and intended reliance. Reuse an adequate existing result; otherwise obtain the appropriate mandate, qualification, access, capacity, safeguards for conflicts and reporting route. Preserve the provider's conclusion and its limits for the governing recipient's action, including a route protected from those whose work is examined. The corporate governance Method arranges and receives the contribution; the qualified professional Method governs the examination itself. A non-corporate or profession-specific engagement uses its institutional or professional supplier instead of inheriting corporate jurisdiction. This return adds no compulsory audit of every support or learning episode.
These receiving Methods can consume the same competent realization and its results. Compare the available result with the present claim, name the remaining gap and obtain only the additional complete contribution needed to settle it. If human acquisition is missing, give the qualified practitioner of human learning the later-work demand, subject criterion and permitted help; obtain the capability result and its supported use rather than crediting instruction or attendance. If work is missing to integrate Systems, supply an input, arrange access or schedule use, obtain that work through the corresponding engineering, administration, provider or operating Method.
The hand-back to C.32.MWA.PP must make local enactment and its limits interpretable. Before dependent work, the receiving practitioner obtains the required inputs and usable means, and secures effective assignments, permissions and commitments through their owners. A prospective walkthrough can show how a real or sufficiently explicit input would pass through local actions: which operation of the supplied Method each action performs, which condition it preserves and how its result becomes usable next. That walkthrough supports a construction claim, not an assertion that the occurrences happened.
When actual performance is examined, compare the actions with the operative conditions as well as with the obtained result. A good result reached by another way can be valuable without demonstrating enactment of the selected Method. An allowed variation can preserve it; a changed criterion or replacement operation can require another subject judgement, Method identity or replacement decision through A.3.1 and B.1.5.RS. Retain that difference in the returned result without demanding detailed fresh observation at every workstation.
Request the observation that answers the relied-on claim. For a delivered result, establish what the receiver could use. For human mastery, establish the person's contribution under the promised help through an appropriate assessment of performance, transfer or retention for that person and Work. For organization capability, use the joined contribution and relevant exception under ordinary provision. For AI capability, retain the qualifying configuration, tools, data, access and limits. For cultural continuation, distinguish availability from another carrier taking up, varying and continuing the Method. A successful episode supports the result and capability claim for which its conditions and evidence are sufficient; it does not by itself establish stable cultural transmission.
For continuing development, the first complete contribution can be an inquiry when the subject way is unknown; it returns what changed in the problem or available possibilities. After success, preserve the supplied connection to a further worthwhile question or justified continuation. Do not close that branch merely because the present implementation succeeded. Its performance and cultural continuation remain questions for the corresponding development and cultural Methods.
Require a usable defect return. If the action did not occur, inspect capability, access, priority or permission. If it occurred but failed locally, inspect its Method and conditions. Adequate local results with a failed receiving whole return to their connection and combination. A result sustained only by unpromised expert rescue requires obtaining that support as part of the arrangement or repairing the intended independence. Retain useful performed results and their limits while reopening the defeated claim. Compatible current evidence can settle these questions without new trials; stronger causal effectiveness claims need their research basis.
Change the arrangement while preserving its valid uses
Reopen the smallest affected connection when its basis changes, then follow the consequences far enough to protect receiving uses. A new Method edition can affect subject criteria, materials, tools, provider qualification and interpretation of earlier results. A provider change can leave the Method unchanged while changing access, support or authority. A shift in demand can invalidate capacity without invalidating competence.
Compare the same serious choices again when they become live. Increased local capability can permit less central execution. Recurring specialty failures can warrant separate expertise. Shared tooling can make previously duplicated work reusable. A qualified autonomous AI can make an office unnecessary or make a maintained common service more valuable. None of these changes imposes a universal direction of organizational evolution.
When a local repair repeatedly moves the difficulty elsewhere, reopen the encompassing arrangement or remit. Several individually useful support services may be jointly unaffordable. A provider's promise may be too broad, a recipient may need a different result, or a typical Method connection may need revision. Return the precise reason to the subject, strategy or network-construction owner and retain independently useful work.
Keep deliberate change separate from distributed cultural change. Participants may ignore, adapt or replace what was introduced. Examine that behavior for its effect on the intended practice and for useful new possibilities. Do not equate disagreement with a capability defect or declare cultural success from formal compliance.
Finish when the current recipient has an explained arrangement they can put into use, a warranted selected attempt, or the unresolved contribution that changes the decision. A compact result can state the supplied network and local use, proposed or obtaining performers and relations, the decisive interaction, the chosen commitment and the next observation. Maintain only records that are needed for performance, accountability, later change or justified reliance.
Return the architecture decision
Keep the description proportional to the receiving use. A compact decision can suffice when its reader can recover the contribution and the condition that changes reliance.
Practitioners choose which direct relations should hold for one result, which structures expose the conflict, and who carries the burden. Co-used Methods can remain plural; a proposed architecture can guide a trial without being reported as already real.
Archetypal Grounding
The first case shows a cheaper relation-only stop; the next preserves several work grains and a purpose-sensitive trial; the release comparison separates peak demand, total burden, evidence history and recovery. Later cases develop distributed provision, voluntary participation and qualified AI. They illustrate different uses of the same Method, not stages or its domain boundary.
Cheaper Clarification: Several Methods, No Composite
In a safety-relevant release, M-HW-Verify, M-SW-Integrate, M-Supplier-Approve, and M-Release-Authorize are identified Methods. Hardware verification and software integration can contribute during overlapping intervals from differently versioned evidence. Supplier approval produces a signed result later. Release authorization uses the safety result and signed-evidence condition.
The selected structures are:
ARS-EC417-RelationOnly-1 records the smaller clarification: retain the four Methods and their direct result-use, permission, and authority relations without naming a composite. The question here is only whether co-use supports a Method-whole claim. B.1.5 supplies the composition test; where an architecture claim is made, C.30 distinguishes the claimed structure from obtaining relations. No composite is established by the evidence shown, and no rival arrangement is at issue. Stop with that clarification, not a completed C.32.MWA.PP comparison. A later choice about provisional supplier evidence opens the genuine alternatives in :5.3; their comparison can also end without creating a Method whole.
Simultaneous Multigrain Contributions with Local First–Then Relations
In a separate five-day Method Engineering case, ME-W0 is the bounded project adaptation Work. ME-W1 repertoire-recovery Work, ME-W2 candidate-account formation Work, ME-W3 description/tool-proposal Work, and ME-W4 evaluation Work are parts at the next finer grain; trial Work ME-W7 is part of ME-W4. During the same five-day interval, continuing product-engineering Work ME-W5 and repository-maintenance Work ME-W6 overlap ME-W0 but are not its parts.
The repertoire-recovery Method contributes through ME-W1, candidate-account formation through ME-W2, description/tool proposal through ME-W3, and trial/evaluation Methods through ME-W7 and ME-W4. Product-engineering and repository-maintenance Methods contribute through the external performances ME-W5 and ME-W6. Each overlaps some of ME-W0. The composition scales follow from the chain through ME-W0, ME-W4 and ME-W7. To claim that the external performances also compete for a resource at a given moment, identify their relevant activities and intervals.
Two local precedence relations hold within the five days: ME-W1 precedes ME-W2, and ME-W3 precedes the relevant evaluation decision in ME-W4. The external activities can occur at different times within the five days. A teaching sequence can present recovery before evaluation. To assign a constituent to a Method, establish its part in the same reproducible way of working.
Consider a minute inside trial Work ME-W7. The participant follows the proposed unit-review description, converts a supplied measurement and compares it with the stated tolerance. Performing that comparison is part of carrying out the trial; the trial is part of evaluating the proposed description in ME-W4; that evaluation is part of the project adaptation ME-W0 currently under way. The encompassing evaluation and adaptation are ongoing through this performed action, although their conclusions have not yet been reached. This is a constituent connection, not an inference from the five-day overlap.
The evaluation purpose changes how the trial is performed. If the purpose is to test whether a reader can recover a conversion rule from the description, supplying the rule as a spoken prompt would defeat that use. If the purpose is instead to measure supported execution after instruction, the same prompt may be appropriate. The arithmetic can be correct in both cases while one trial fails its encompassing purpose. Keep the participant's result, the help actually supplied and the evaluation conclusion distinct. B.1.5.EW recovers this connection; B.1.5.RS applies when replacing a shared trial or checking Method across differently intended evaluations.
Possible-Future Alternatives: A, B, and B2
The EC-417 release scenario later asks how to handle provisional supplier evidence. Three alternatives differ by Work order and allocation:
For this constructed comparison, 2.07 h is safety-engineer time for traceability and safety-closure work: checking the current requirement–implementation–verification correspondences and preparing or recording the closure disposition. It excludes the separately counted signed-delta preparation and 0.33 h board, and any additional affected verification or integration rework.
The first 20-minute board is on D-21; the second is on D-8. They occur on different days. The peak-day comparison and per-release burden are separate measures. Each board remains below the 45-minute meeting limit. Under B or B2, signed evidence supersedes the provisional edition only for safety-closure reliance; the provisional edition, uncertainty, use in D-21 integration Work, and relation and delta to the signed evidence remain traceable.
CUR-EC417-CadenceEffect-1 returns unsupported for the interventional claim that earlier reconciliation reduces reopenings. AD-EC417-B2-Trial-1 therefore consumes no positive causal premise. It consumes separately named non-causal DC-EC417-CadenceMismatch-1, the capacity comparison, confidentiality, the covering trace/safety/release assignments, PERM-TraceAcceptReject-17, the two independently supported direct authority relations, and reversibility to choose only three bounded B2 trials.
At the D-21 pre-entry checkpoint, ReleaseDecider-17 performs the branch decision under ASG-ReleaseDecision-17 and AUTH-ReleaseDecision-17, after the required safety-evidence decision under ASG-SafetyReview-17 and AUTH-SafetyEvidence-17. B2 entry also requires TraceReviewer-17 to perform W-TraceAcceptReject-17 under ASG-TraceReview-17 and PERM-TraceAcceptReject-17 for every AI suggestion used by the branch.
Before entry, choose A only when signed evidence is available in time and the signed-first arrangement meets the required integration slot, capacity and authority conditions. A missing B2 premise—versioned provisional evidence, supplier preparation, confidentiality, trace-review permission or a required safety/release assignment or authority—blocks B2; it does not make A meet the D-21 slot while its evidence arrives at D-8. Recover the missing premise, construct another permitted arrangement, seek an authorized revision of the commitment, or hold the unsupported continuation. Once early B2 integration has occurred, A is no longer a possible history for that release.
Missing signed evidence at D-8 withholds release and starts recovery R. Preserve the performed integration record, exact provisional edition, uncertainty, and earlier use. When signed evidence arrives, record its relation and delta to the provisional edition, re-baseline, and repeat the comparison and affected verification. Retain or roll back/repeat early integration. For a full repeat of these recovery tasks, count 1.60 h supplier preparation plus 2.80 h safety preparation, board, and trace/closure work (0.40 + 0.33 + 2.07). Count affected verification or integration rework in addition; the case supplies no fixed duration for it. If recovery stops before those tasks finish, record the Work and burden that actually occurred rather than assigning the full repeat cost. Move only to the next authorized slot after repeated decision Work under the covering assignments and authority relations. Every repeated AI suggestion again requires the trace-review Work, assignment, and permission. Signed evidence supersedes provisional evidence for closure reliance without erasing the provisional history.
The decision is prospective. Implementation and the three trial releases are later Work. Signed evidence available at D-21 makes A eligible for comparison. The choice also depends on the capacity, authority and value conditions below. A change in future availability leaves the past release history intact.
In that pre-entry variant, suppose the existing signed-first arrangement has adequate current capability, capacity and release evidence, and the same comparison reveals no useful gain from provisional-first preparation. The team can finish by retaining A on that basis. Planning an extra B2 trial would consume the safety attention needed for the current release without changing this decision. Changed evidence timing or capacity can reopen the comparison. The baseline case has a thirteen-day mismatch and supports the three bounded B2 trials under their stated conditions.
The cases are constructed to show the synthesis and a changed condition. Their numerical demands and capability premises are stipulated task data, not empirical estimates for every development practice. The three-team case uses a supplied practice account in ordinary language; it does not claim that a diagram alone establishes an E.18.NET network.
Three teams: development in ordinary work without creating a new office
The supplied practice. Three engineering teams have adequate Methods for their current technical work. They want both reliable subject review and a culture of continuing development. A recurring pattern has become visible: a team's deliverable passes its own checks, but the receiver repeatedly performs extra interpretation before using it.
The supplied development practice connects several contributions:
- A participant examines a completed result and its receiving use, notices a consequential difference, and formulates a question. B.5.PI supports the move from the apparently finished deliverable to the receiver's remaining difficulty.
- The participant and a coach construct a bounded next attempt. The coach helps problem formation and choice; the relevant technical expert qualifies the subject condition that makes the attempt interpretable.
- The participant performs the agreed attempt in a permitted setting, examines what became possible and what failed, and updates the usable means.
- After a successful result, the participant formulates a further worthwhile question or gives a reason to consolidate. C.40.CD supports the new problem arising from the achieved result. The coach examines the participant's contribution without silently supplying the promised independent choice.
- Another participant can obtain the relevant Method and its limits through the ordinary shared means. Whether it is taken up and continued remains a separate cultural question.
These connections require different competence and authority. The coach cannot approve an unknown technical inference. The technical expert need not own the person's development choice. A team manager can reserve work time but cannot turn an unsupported experiment into valid subject evidence.
The receiving demand. The initial ambition is two bounded attempts per team each week: six total. Each normal attempt requires two hours of the participant's own work, one hour of qualified coaching split between initial formulation and examination of the result, and half an hour of the relevant technical expert's review before the attempt. Common access, scheduling and material maintenance require one hour per week. These are the qualified demands assumed for this comparison.
Each team has four approved participant-hours and one hour of its own technical expert's time. Each local leader has one approved hour and the competence to coach one such attempt. A practitioner who already helps several teams has four approved hours for this use. None of these hours may also be spent on protected current delivery.
Construct alternatives around the difficulty.
The distributed candidate does not require a separate organizational unit. The same people perform contributions through their ordinary workplaces, with explicit reservations and shared means. Putting these same assignments under an office label would not change the calculation. A new unit would earn its cost only through an additional needed relation, such as durable responsibility for access or an effective shared-priority decision. Calling this a federation would add no content unless actual independent services and their coordination require that description. If those relations did require a permanent service owner, that would be a further justified architectural choice.
Check time as well as totals. Participants prepare their initial questions on Monday. Each coaching-hour is split into half an hour before the technical review and half an hour after the attempt. Local leaders each supply those two half-hours for one case; the shared practitioner supplies three such pairs. Technical reviews occur before the permitted attempts. The participants' two hours cover initial preparation, the attempt and examination of its result. The common hour supplies access and scheduling before those uses.
The allocation preserves who performs the difficult operations. A coach who writes the participant's next question may obtain a better proposal but would change the mastery observation. A technical review arriving after the attempt cannot qualify its prior conditions retrospectively.
The arithmetic only covers normal bounded cases. At six starts, the shared practitioner and technical experts have no reserve for an overrun. It would be misleading to advertise unconditional six-case throughput from these sums.
Select a first commitment and begin. For the first week, the authorized managers choose four starts: one locally coached case in each team and one additional case coached by the shared practitioner. That uses three local coaching-hours, one shared coaching-hour and one common hour. The shared practitioner retains two hours for an exception or an observed defect. Technical review demand is one hour in the team with two cases and half an hour in each other team. Participant demand is four, two and two hours respectively. The six-case ambition remains conditional rather than being silently reported as fulfilled.
For this stipulated first-week decision, no deadline requires all six starts. Four complete attempts include every team and are sufficient for the managers' present receiving question. They judge that the two additional starts and released local coaching-hours offered by the external arrangement do not justify its 800-unit price and additional context work. They accept the three reserved leader-hours and slower initial coverage in order to retain the money for other work. This is the case's explicit value judgement, not a claim that distributed provision is generally cheaper or better. A time-critical need for six starts, a more valuable competing use of those leader-hours, an improved external offer or inadequate local coaching would reopen that comparison.
The two-hour reserve belongs to the shared coaching and common-support contribution. It supplies neither extra technical-review capacity in the team with two cases nor additional participant time there. If that technical review cannot be obtained before the attempt, stop the dependent attempt and obtain another qualified contribution or revise its timing and commitment; the coach cannot cover the gap merely by having spare hours.
This bounded start is not selected because four is a universal pilot size. It preserves a complete receiving result while leaving capacity to examine the uncertain normal-demand assumption. If reliable observations already established six-case demand without material exceptions, that additional restriction could be unnecessary.
In the first constructed use, a participant finds that the receiving team has to reconstruct why a parameter was chosen. The relevant technical Method can supply that reason, but the original handover omitted it. The participant changes the handover and tests it with the receiver. After successful use, the next question concerns whether the same explanation works when the receiving team uses a different model. That question is not answered by celebrating the first improvement. Its subject relation may require further construction before another attempt is authorized.
The observation can support the performed local result and the participant's demonstrated contribution. It does not by itself establish stable culture across all teams. The shared means make the qualified Method and limitation available to another participant; a later use can examine whether the practice continues without the original coach supplying its decisive choices.
Change a condition. Suppose the shared practitioner becomes unavailable for two weeks. The distributed six-case candidate loses a contribution. Local leaders can coach three attempts, but common maintenance and the remaining coaching are unprovided. The available external offer can preserve more work if its required receiving/context contribution is assigned and available. Alternatively, the managers can accept a smaller temporary result and supply the common work through another qualified holder. The earlier choice does not survive merely because its organization chart is unchanged.
If the local leaders instead acquire two available coaching-hours each and the common work is already supplied by an existing operations arrangement, six normal attempts may be obtained without the shared practitioner's coaching. That changed competence-and-capacity basis, not a universal preference for decentralization, justifies reducing shared execution.
A voluntary provider community: common access without ownership of every mastery
The supplied network and the boundary. A professional association helps independent engineers–managers develop their practice. Its account connects personal development, peer examination, creation of reusable methods, delivery of help and cultural continuation among members. Its provider work is distinct from the members' client work. Members are not employees of the association, and their clients have not delegated authority to it.
One member wants to improve decisions in a familiar project setting. Another encounters an unresolved technical relation in a new professional field. A third has successfully changed a practice and wants to explore the next attainable challenge. These are legitimate entry situations for the selected development practice; they are not three instances of the same course.
The association's coordinator can establish contact and obtain a qualified provider, but cannot certify every technical proposal. Examine these three proposals, retaining the missing contributions that can change the decision.
A universal internal expert team would need subject competence and maintained evidence across all requested fields. Those contributions are presently absent and have a substantial carrying cost. A common portal alone would be cheaper, but would leave questions unresolved after intake. A bounded common access function can instead obtain development facilitation from one provider, technical contributions from selected specialists, and research from another where a qualified Method is missing.
The third arrangement is not complete until a request can reach an available contributor and return in a usable form. For the familiar case, the subject Method is supplied, the member retains responsibility for their own project choices, and a coach helps them formulate and examine the next change. The client's permission for an intervention must be obtained where required. An association membership does not supply that permission.
For the unresolved technical relation, the available specialist explains what is not known and proposes an inquiry that could discriminate between two constructions. The association can offer that bounded research contribution if it is strategically included and resourced. It cannot advertise acquisition of a settled mastery before the subject Method exists. If it elects not to enter that field, it can provide a referral or decline the unsupported offer. The member's legitimate need and the association's strategic remit remain different questions.
Follow the moved burden. The proposed shared service initially assumes that members will donate unlimited specialist review. Looking from the common provider's scale, the arrangement is inexpensive. Looking from each member's client commitments, the same work displaces paid delivery. No enforceable or voluntary contribution has been obtained.
Repair the arrangement with bounded commitments: for example, a willing specialist accepts one named contribution within an agreed window, with the permitted source material and an explicit refusal or failure return. The coordinator can use that contribution under its limits, not promise the specialist's continuing unlimited availability. If the provider cannot sustain the receiving work and compensation or reciprocity, narrow the offer rather than treating participation as a moral debt.
For the successful third member, the next development question is constructed from what the successful change made possible. The member may use peers or AI, but the provider must not covertly decide every next challenge if independent self-development is promised. If the member chooses ongoing expert-led development instead, state that different assistance condition openly.
Change a condition. A member asks the association to require the same practice in all member firms. The current voluntary network provides no such authority. The provider can explain the proposal, seek agreements and arrange supported trials. A participating firm can then establish its own local assignments and resources. The association cannot skip those local architecture decisions by declaring a common culture.
This changed condition distinguishes cultural availability from corporate enforcement. It also shows why one shared Method and common access do not make all receiving organizations parts of one provider.
Capable AI changes the whole arrangement, not just the coach's speed
Return to the three-team case in :5.4, but change the available means. Stipulate that appropriately configured agents are qualified to help with problem formation, elicit and examine the participant's own judgement, retrieve the approved subject Methods and perform the relevant bounded technical review. They have authorized access to the needed data and can act within specified experiment limits. These are scenario premises to be established before reliance; cheap tokens alone do not supply them.
The new candidate does not need to route every case through a central human coach. Participants can work with their own authorized agents at the relevant workplace. A maintained common service supplies Method versions, access, evaluation and recovery. A provider may operate that service; it need not be a new internal office. Local managers retain or delegate the allocation decisions according to the current authority.
If the human mastery promise is unchanged, the human has to perform the contribution to be learned. The agent elicits that judgement before supplying answer-bearing help and examines later use under the promised support conditions. It can provide coaching and technical review without requiring a human to repeat each operation solely because the former arrangement used one. The participants' two hours per case remain in the capacity account; cheap AI has not made their practice free.
A materially different candidate gives the agents responsibility for recognizing opportunities, constructing and selecting bounded changes, conducting authorized work and returning results and exceptions. If the recipient needs an organizational development result rather than every employee's independent mastery, this can be a serious arrangement. It changes the human-learning promise and must be compared at that revised strategic result, not credited with human mastery the people did not acquire.
The central-office option can be better when subject evaluation, confidential information or expensive shared infrastructure must be maintained once for many receivers. Conversely, trustworthy local agents with available common means can eliminate the original coaching bottleneck and its associated coordination. Compare those complete arrangements; neither “keep people in control” nor “replace the office with agents” supplies the configuration.
Change a condition. The agents can generate plausible questions but cannot use the decisive field measurements, and those measurements require a scarce physical facility. The bottleneck moves into access and experiment performance. More autonomous reasoning does not increase facility capacity. Recalculate which attempts can be performed, obtain the required access or choose a different question. The AI-rich arrangement remains useful for some work without supporting the original throughput claim.
If an agent proposes to change the very Method or evaluation means on which its qualification depends, separate that development work from ordinary enactment. Obtain the relevant subject and assurance result before extending the permitted use. The common service's role may change from case-by-case help to maintaining qualified means, while the actual authority and receiving conditions remain explicit.
Obtain a weekly comparison and distinguish the learner's own contribution
The following constructed cases use stipulated requirements, task evidence and effort estimates. They illustrate choices under those premises, not measured savings or a general estimate of AI reliability. Use this Method and its full suppliers when adapting them.
Obtain a weekly comparison with limited specialist time
A team needs a comparison of changes in twenty supplier notices before its weekly planning decision. The result must preserve the applicable source version, changed conditions and any unresolved implication. The recipient needs the comparison, not personal mastery of every extraction operation. In this case, representative prior use supports two ways that meet the same acceptance conditions.
The assisted way takes another 1.5 specialist hours to set up. Only three are free each week after standing commitments. Setup and the first comparison together need four, so the owner reserves the setup time the week before. The supplier releases the twenty notices by 09:00 Thursday. The specialist can prepare the comparison from 09:00 to 09:30, check the draft from 10:00 to 11:00 once the agent has returned it, and correct and use it from 11:00 to noon. Those bookings and timely access to the draft make the noon decision feasible under the estimates. Three free hours on Friday would give the same weekly total and miss that decision. Two comparisons with no shared work would exceed even the weekly capacity.
The agent receives the twenty applicable notices and the comparison question. It may read the supplied material and prepare an internal draft; it has no authority in this case to amend supplier terms or publish externally. A source-checking return preserves a notice's condition: a changed lead time applies to new orders after a specified date. The receiver uses that condition to separate new orders from the existing backlog. A draft that applied the new time to every order would fail the receiving use even if generation completed without an error.
The worked arrangement ends with the used comparison and its unresolved items. If the supplier portal goes offline, a retained notice or prior check can still establish the edition and the order dates it covers. Check whether it meets the receiving terms. Reopen the affected conclusion if a consequential condition or the notice’s currentness remains unknown, or if the agreement requires a fresh check with the supplier. If correcting the drafts begins to exceed the estimate, reopen that candidate's effort and qualification. Faster generation alone does not establish the original capacity claim.
Change the purpose to the person's own capability
Keep the reporting task, but add a separately chosen purpose: an analyst must learn to identify when a changed source condition alters the comparison. The AI-assisted production arrangement can continue supplying this week's usable result while the learning work is arranged. Include the learning time in the same capacity account.
Use appropriate learning Methods to design representative practice and perform it with feedback and correction. Give the learner a relevant notice and its receiving question, with the assistance permitted for this learning purpose. Ask the learner to identify the condition, show which conclusion it changes and explain why. Feedback addresses the observed omission; a subsequent attempt tests the repaired operation. Giving the answer before an attempt may support production while defeating the particular evidence sought about the learner's own performance.
When the claim concerns unfamiliar work or retained ability, use an assessment Method for that transfer or retention claim under the required support and timing conditions. One successful attempt supports only what was exercised. If the later work permits a source or AI aid, test the person's required contribution with that support instead of silently changing the goal to unaided recall.
Suppose the learner now notices that two apparently comparable notices use different order-date conventions. Recover the conventions from their sources, or ask a domain practitioner for the missing interpretation. Use C.39.MC if the comparison Method needs to change, and preserve comparisons whose premises hold. Use each result at its supported scope: the report supports this week's decision, the observed learning response informs further practice, and a qualified Method change can be reused under its conditions.
Bias-Annotation
This Method starts from an intended contribution and can therefore underrepresent people who bear costs without being named as recipients. Recover affected participants when their workload, autonomy, exposure or legitimate interests can change the design. A sponsor's preferred culture is not the whole account of those consequences.
Existing organizational boundaries bias generation toward moving tasks inside the current chart. Existing framework or professional labels can create the opposite bias: every useful distinction becomes a separate permanent specialty. Test contribution and interaction before selecting either boundary.
Common-service advocates can count saved duplication while overlooking local interpretation and integration. Local-autonomy advocates can count responsiveness while overlooking common maintenance, incompatible decisions and loss of specialist depth. Apply the same whole-cost comparison to both.
An AI-rich candidate can appear superior by assuming perfect competence, free access or unchanged human learning. A human-centered candidate can appear necessary by retaining reviews whose contribution is no longer established. Qualify actual capability and permissions, preserve the intended holder's result, and compare the best applicable alternatives.
A successful example can hide a narrow assumption: ready-made Methods, an employer's authority, qualified local leaders or stable demand. The Solution allows those conditions to vary and returns the resulting missing contribution. Renaming the example's organization is not a transfer test.
Conformance Checklist
Use these questions on the arrangement being constructed; they do not require a separate record for every local action.
Recognition and construction
- The result names an obtaining or possible-future practice, receiving result and decision boundary; a supplied typical network is distinct from proposed local bindings and actual performance.
- Each Method and candidate account keeps its prior status. A missing subject Method or join is named in the return, not replaced with a role or office label.
- A consequential result can be followed through its subject, applicable Method, intended Work, capable performer, means, authority and receiving condition.
- Each selected structure can change the decision. Method composition or unfolding, Work parthood, overlap and first–then guards, allocation, support, description and cultural relations retain their meanings.
- Serious alternatives differ in direct relations, not view or format; they retain the feasible incumbent and applicable ordinary-workplace, common, specialized, federated, external or AI-based provision without requiring a new office.
- The comparison exposes the interaction or moved burden that can reverse the choice, including qualified shared resources, recurring work, windows, continuing obligations and the transition.
- Commitments distinguish who can assess, assign, authorize, supply and accept. Agreements or changed authority remain conditional until effective.
- A sufficient current arrangement or realization is preserved. Additional implementation or trial has a worthwhile contribution, feasible conditions, observations, stop and recovery.
- The realization receives the remaining contribution and return condition; it distinguishes a supplied result, Method enactment, supported capability and unpromised rescue.
- A change to a Method, means, provider, demand or authority reaches affected receiving uses without erasing valid independent work or performed history.
Assurance at the claimed scope
An architecture comparison needs a credible basis for its inputs and explicit proposed conditions. A choice needs the relevant decider and authority. Actual performance needs its occurrence and result. Capability needs evidence for the named holder under its support conditions. A causal-use result is consumed only within its verdict boundary. A possible-future decision establishes no obtaining ArchitectureRelation.
When continuing development is promised, keep the return from an achieved result to a further worthwhile question or justified continuation. When human mastery is promised, examine the human's contribution. When cultural continuation is claimed, examine transmission and further use at the stated scale. These assurance questions apply to their respective claims, not automatically to every technical support arrangement.
Common Anti-Patterns and How to Avoid Them
Performance and support arrangements
Consequences
Architecture choices become concrete enough to realize and reconsider without forcing one universal decomposition. Guards survive concurrency, co-use survives without a fictitious composite, and moved burden is visible before local optimization is accepted. Missing subject Methods, learning needs, provider gaps and authority questions become different actionable returns.
The comparison can expose an unavailable volume, independence level or remit. A smaller complete commitment can be preferable to a broad service that cannot respond. Office creation becomes one possible arrangement choice. A first use can begin while later contributions remain conditional, provided its own conditions and protected work hold.
The cost is attention, consultation, recovery of relevant structures and sometimes modelling or inquiry. Some attractive alternatives remain only proposed. A relation-only stop can be more useful than an unsupported Method whole, and an adequate incumbent can make another comparison or trial wasteful.
Explicit commitments can also make informal help rigid or costly. Preserve useful discretion and change only what the receiving result needs. An architecture description does not itself produce capability, authority or cultural continuation.
Rationale
Method organization, dated Work, subject arrangement, allocation, descriptions, provider support and cultural continuation answer different questions. Treating them as isomorphic loses either useful simultaneous contribution or dependency. Select the structures that can change the decision and compare their relations in the receiving case.
Project, process and case views provide a practical test: when several views concern the same Work, their labels do not create architecture alternatives. The alternatives begin where proposed direct relations differ.
Begin with the work's subject and Method because allocation must preserve an obtaining way. A staffing chart can name a performer for an operation nobody knows how to perform. A service catalogue can promise a result without explaining the transformation that obtains it. Recovering the supplied network first exposes this difference. It does not require a complete description of every Method before a simple available contribution is used.
Connect several structures because each can defeat the others. A valid Method can have unavailable performers; competent performers can lack a usable interface; a reasonable local schedule can exhaust a shared expert; an authorized decision can rest on an unsupported subject claim. C.32.MWA and C.30.ASV provide the governing distinction. The support question needs their connections made practical through receiving results, candidate assignments and joint demand.
Generate by changing mechanisms because named forms are underdetermined. A centralized office can perform cases, provide common tools or only maintain access. Each has different consequences. Conversely, the same operative allocation can exist in a formal office or ordinary roles. Specialization, proximity, decision placement, provider relations and common means are useful design variables only through the contributions they change.
Allow the remit to change because feasibility can expose the wrong promise. Strategic purpose constrains the architecture, but an unavailable or disproportionate obtaining arrangement can properly return a different result, audience or horizon to that purpose. Return the changed contribution, its feasibility or burden and the affected promise to that purpose and choice owner; carry the resulting constraint back into the arrangement. The architecture does not grant itself authority to change the promise; it returns the reason to the relevant choice.
Keep development and support distinct because their recipients and results differ. A supplier can improve its capability to help while a client learns nothing. A client can develop through an existing provider without the provider changing. A community can transmit a Method without becoming the employer of its carriers. The network can connect those contributions while their Work, rights and evidence remain independent.
Use finite comparison and selective detail because the architecture itself consumes resources. The fullest possible model is not the goal. A contradiction that survives a cheap calculation should be repaired before collecting more data. A consequential uncertainty can justify inquiry when its plausible answer changes the choice. C.11.CRC and C.11.DUA preserve that proportionality.
A direct use of existing organization, service-design or operating Methods is preferable when it already supplies this connected answer. C.32.MWA.PP is useful where the unresolved question is how supplied Methods can be combined or co-used, including their distributed performance and support, without merging their structures or moving a difficulty out of sight. A bigger office model or a longer document has no advantage by itself.
SoTA-Echoing
C.32.MWA, C.30, A.22 and B.1.5 supply several-structure synthesis, architecture truth, selected structures and Method composition. C.32.MWA.PP applies them to construction and bounded comparison of Method arrangements. Organization-design Methods can already supply organizational mechanisms and complete same-result human, AI, internal and external provision; use a sufficient result directly. Organizational realization Methods, A.15.8.MS and the introduction Method described in A.3.1.MI provide realization at their respective scopes. Section :4.9 gives them the remaining contribution instead of adding another implementation practice.
The practice question is how to make a family of Methods performable through a jointly feasible arrangement while preserving subject competence, recipient results and authority. A useful constructive line combines service/component construction, organization design and context-sensitive support. Its alternatives include selecting a named office model, assuming universal decentralization or reducing every gap to training. The synthesis adopts particular operations from that line; it does not claim a comparative experiment proving one general organizational form.
AI performance and human acquisition need different evidence. The Cybernetic Teammate primary abstract supplies a bounded task-performance result, not whole-organization replacement or durable mastery. Anthropic's skill-formation study supplies immediate learning observations in one programming setting, not a universal prediction of future human–AI work. These limits motivate the separate promises and conditional capability premises in §5.6.
The source contributions justify candidate operations and useful distinctions. Transfer beyond their original fields depends on the Method, holder, result and relations recovered here. Reopen the relevant choice when a developed subject practice supplies a better complete construction, when observed work defeats the claimed connection, or when changed means alter the total contribution.
Reopen when a representative use exposes another non-isomorphic structure that reverses the decision, when moved burden remains hidden, or when an arrangement cannot be realized without changing its selected relation set.
Relations
The Core contributions below support the general construction. The domain entries describe possible separately selected work: obtain the particular result and conditions from the practice that governs the receiving case. A reference does not supply an unavailable performer, permission or professional judgement.
An applicable organization-change Method can realize the bounded organizational case illustrated in :4.9. A.15.8.MS supplies usable Method materials and support, and A.3.1.MI the complete introduction/use/revision Method. Section :4.9 gives each its receiving requirement and reuses sufficient implementation.
C.36 and C.36.RP supply the cultural-practice questions alongside the applicable Community Management Methods for actual community participation. A shared Method can continue among independent carriers without making them one community or organization. Their full Methods and evidence boundaries remain with those suppliers.
A research Method for explaining how a Method architecture differentiated may supply that account and a separate causal-use result. C.28 governs any causal conclusion consumed. C.32.MWA.PP consumes only the reliance allowed by the verdict and keeps its own decision authority.
When a separate mathematical construction is needed, use a mathematical Method to develop the operations, retained information and shared-resource constraints that derive a property or counterexample. A Method for designing work through mathematical transformation can then construct a changed way of working when that is the selected design question. The arrangement consumes only the qualified mathematical result, with its assumptions and limits.
C.32.MWA.PP:End
Architecture Failure Recognition and Repair
Type: Architectural subpattern under C.32 Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when a recurring architecture-synthesis failure needs a first justified repair action over a named architecture object. The action may recover a missing premise before a structural change can be chosen.
Primary working reader: an architect or architecture-responsible practitioner who sees a warning sign during synthesis and needs the first architecture repair action, not a larger risk catalogue.
Typical entry cues:
First-minute use slice. A team calls an ML model a module in a safety-relevant product architecture. Using C.32.FAIL, the practitioner names the architecture object under stress: a candidate module-interface relation for the described product holon. The blocked overread is: model file equals stable module. The first repair action is to recover interface behavior, admissible-use conditions, change policy, and evidence-decay boundary before using the model as a module. If a safety assurance claim is current, use its subject pattern with the identified architecture question and available evidence.
The primary EntityOfConcern is one repair cue for one architecture object under stress. The cue is a working repair aid, not a risk register, assurance case, selection result, release argument, or decision object.
Failure language can become a warning bank: the team can say what looks suspicious, but it cannot say which architecture object must be repaired or which pattern defines or constrains the next claim.
A practitioner can convert a vague failure signal into one typed repair action, keep the repair near the selected structure, and stop before nearby decision, release, or governance claims expand the case.
Ordinary working move: explain the symptom, affected architecture object, first justified repair action and its stop or continuation condition. Name a mistaken inference when it explains the failure.
After using C.32.FAIL, the reader can identify the affected architecture object, the first repair action and its stop or continuation condition.
Use another pattern when the current work is only lexical cleanup, evidence sufficiency, release, architecture description, MVPK publication face, comparison, selection, archive, front, selected-set result declaration, actual publication, local choice, or final architecture decision. Use C.32.FAIL only when the failure cue changes the first architecture repair action.
Common exits by claim kind:
C.30.P,A.6.F,A.6.M,C.31,C.32,C.32.MLAO, andC.32.CONWAYfor architecture or selected-structure repair.A.19.CPMfor explicit comparison andA.19.SelectorMechanismfor set-returning selection.C.18andC.19for archive, front, pool-treatment, or retained-stepping-stone claims.- Use the applicable subject rule for the evidence needed by the repair,
A.10when a replayable account of that claim's support is needed,B.3for assurance,A.20for internal-constraint validity, andA.21for a named gate decision under an applicable profile, including a release gate when current. Release requirements remain with the patterns that define them. C.30.ADfor architecture description,E.17for a source-backed publication face and source return, andE.24.PUBfor the publication occurrence and audience availability.G.5for selected-set result declaration,C.11for local choice, andC.32.PADfor a project decision. For publication, keep the distinct E.17 and E.24.PUB uses just named.
The first useful output names the affected architecture object and a justified next action. An ordinary explanation can suffice; ArchitectureRepairCue@Project can organize it when a record is useful:
Here @Project is a compatibility and retrieval cue only. It establishes no project entity, composite-work identity, context, authority, viewpoint, or parthood. When the repair cue is used in one project, projectWorkOccurrenceRef identifies the exact composite U.Work and architectureRepairCueProjectUseRelationRef identifies the direct relation by which that exact project Work uses the cue. Any separately claimed repair Work and its own cue-use or work-to-change relation remain under their governors. The cue, the repair Work, the architecture object under stress, and the composite project Work remain distinct.
Problem
Architecture synthesis often fails before formal evidence or decision work starts. The defect is not only that a word is vague. The practical defect is that the architecture object under stress is missing or misread.
Most first-contact failures cluster into a few repair-entry families:
- a proposed bearer, module, platform, or universal substrate hides interface behavior, variation pressure, function bearing, evidence burden, or new coupling;
- a proxy result, generated artifact, architecture description, graph, dashboard, front member, or workshop favorite is used before the selected structures, losses, and pattern for the next question are named;
- one structure, function, system-role kind or assignment, responsibility, control relation, evidence relation, or Method step is improved while the synthesis frame loses the architecture characteristics and other structures that made the trade-off real;
- a current candidate is treated as a durable optimum, or ideality pressure deletes a bearer without naming the function still carried, the lost structure, and the new burden;
- an influence-source architecture and the transformed-side architecture content collapse into one claim instead of opening C.32.CONWAY and separating the changed referent, each obtaining C.30 architecture relation or modal
ArchitectureClaim, any asserted direct influence occurrence, and any separately current actor, assignment, Work, or actual transformation facts.
These cues are useful only when each one is converted into a repair shape: symptom, architecture object under stress, first repair action, and stop or pattern for the next question.
Use C.32.FAIL for that conversion. It does not mint a local ontology of failure kinds.
Forces
Solution
Develop the first repair action from the warning cue. The following six moves guide that work:
- Establish what the recipient needed to accomplish and where the observed work or proposed arrangement stops supporting it. Follow that use to the condition on which it depends.
- Identify the holon and affected architecture relation as far as the evidence warrants. If the required explanation or choice of repair depends on which account holds, seek an observation, requirement or comparison that distinguishes the accounts. An intervention that meets the recipient’s needs under every remaining account can proceed on those grounds. A separate request for the cause remains open. Carry forward the claim scope or qualification window when it affects the inference.
- If a mistaken inference explains the failure, state it and the distinction needed to correct it.
- Use the subject pattern for the architecture object or lens relation when its definition or test is needed; cite it when the next user needs that route.
- Propose the smallest action that can resolve the architecture difficulty. If a consequential premise is unknown, first recover the observation or subject result that can settle it.
- State where to stop, or which neighboring pattern defines or constrains the next claim if another claim is already current.
Core repair families for a first repair-cue draft:
Admit a new repair family only when its row tells the practitioner what to repair first. A suspicious name alone is not enough; the row must name the architecture object under stress, the first repair action, and the stop or pattern for the next question.
Stop condition. Stop when the affected object and justified next action are clear. Carry a source-return condition or another pattern when the next use needs it. A precise unresolved question can be the useful return; the cue itself supplies no evidence, assurance or decision verdict.
Reconsideration condition. Keep the symptom, affected object, supported observations and justified action current. If the object or repair remains unknown, retain the observation and the question needed to recover it. Missing documentation alone does not invalidate a known action. After a repair, examine whether it resolved the failure before retiring the active cue; preserve useful history. A changed object reopens the affected use. Use A.6.P or E.10 when the case is only source-expression recovery, C.32 when candidate repair is current, C.32.MLAO or C.32.CONWAY when their residual or correspondence repair is current, and the named pattern for the next question when a stronger downstream claim is current.
Worked Repair Cases
Tell. C.32.FAIL is a repair-entry pattern. It takes a recognizable warning cue and returns one typed repair action over a selected architecture object. It is useful only when the repair action changes architecture handling.
Show-A - Safety-relevant model-as-module. A model file is being treated as a module in a product architecture. The repair cue names the candidate module-interface relation, blocks the file-equals-module overread, and recovers interface behavior, admissible-use conditions, change policy, and evidence-decay boundary. Safety assurance follows only through its subject pattern.
Show-B - Product-family platform with exception growth. Local records show shorter delivery effort and more exceptions requiring evidence. Check which variants were delivered and which obligations created the exceptions. If a new contract requires an on-site check under either design, keep that check in both candidates. If variant-specific fields repeatedly force reviewers to reconstruct common evidence, a qualified shared rule and a local check may remove that work. Compare those changes through C.32.MLAO for the affected deliveries. The count still serves a request for the number of recorded exceptions even while their cause is open.
Show-C - Responsibility change shifts coordination cost. A stream-aligned team improves local delivery flow, but release testing and evidence responsibility remain shared. The repair cue names the team or organization System, the coordination relation, and the module-interface and evidence structures under stress. A proposed responsibility retargeting names its direct predicate, the current and proposed participants, and the occurrence to replace. Return missing-governor only when no applicable predicate governs the relation; leave an unestablished application premise unresolved. Ordinary work organization, Method or plan structure, local kind, separate System-classification judgment, assignment, enactor relation, and actual Work network remain separate. C.32.CONWAY supplies only the architecture-influence synthesis frame or qualified pair row; it supplies none of those other facts.
Show-D - Generated architecture candidate. An agent system produces a high-scoring blueprint. The repair cue treats the blueprint as a source cue, recovers the selected-structure changes encoded in it, names preserved and lost structure, and recovers an adequate candidate palette for any comparison or decision that needs it. A publication can retain the proposal and its unresolved conditions under its own rules.
Show-E - Built-asset maintenance dashboard. A facility maintenance dashboard shows a dependency graph and freshness scores. The repair cue keeps the graph's mathematical-lens use bounded, recovers the selected structures under stress in maintenance work and asset interfaces, and keeps timing or evidence claims with their subject patterns.
Show-F - Function with no feasible bearer. A searched AI workflow adds a verification function after model output, but the edge device has no resource margin and the cloud placement violates latency. The repair cue names the function-bearing gap, then opens C.32. Candidate repairs can, for example, add a local bearer, split verification into local and cloud steps, change deployment placement, or reject the configuration for the current evolution window. A narrower verification promise belongs to a changed remit. The responsible authority must accept it before the reduced demand can govern a deployment.
Repair-Entry Failure Modes
Conformance Checklist
Consequences
Rationale
C.32 needs a failure-recognition subpattern because candidate architecture work repeatedly breaks at the repair-entry point. The useful work is to recover the architecture object under stress and make the next repair action reviewable.
C.32.FAIL does not establish failure, make a score-based risk finding, select a candidate, or authorize a release. It gives practitioners a way to go from "something is wrong here" to "this architecture object needs this repair, and this neighboring pattern defines or constrains the next claim if it is current."
SoTA-Echoing
These rows show how source practice informs C.32.FAIL repair fields, repair families, boundaries, and receiving-pattern exits.
Source-use boundary. Each row supplies a bounded repair contribution. Recheck a dependent use when a source or application premise changes materially. Preserve independently supported observations and repairs; an unsupported source transfer can remain historical background.
Relations
- Builds on:
C.32for candidate palette repair;C.32.CONWAYfor a synthesis frame or qualified pair connecting architecture influence with transformed-side architecture while keeping the changed referent, C.30 architecture relations or modal claims, direct influence occurrence, organization relations, local kinds, separate System-classification judgments, assignments, enactor relations, ordinary work or procedure organization, actual Work, direct responsibility relations, actual transformation, and any E.18.NET network distinct;C.30andC.30.ADfor architecture relation, claim, and description boundaries;C.30.ASVfor architecture structural views;C.31for module and interface architecture;C.32.MLAOfor cross-scope residual repairs;C.29for mathematical-lens use;E.17andE.24.PUBfor publication-face boundaries; andA.6.P,E.10, andE.10.ROLEfor source-expression and role-word recovery. - Coordinates with:
A.6.Fwhen function and architecture-characteristic wording is mixed;A.6.Mfor module-interface repair;C.19.1for a general scale-amenable bearer or Method; A.2/C.3 for a local system-role kind and any separate System-classification judgment; A.2.1 for an assignment species and occurrence; A.13 for every precise performer's core, A.15.1 for independent actual-Work admission, and F.6 only for current precise assignment-bound attribution; and the exact enactor, coordination, or responsibility predicate, or A.6.RCDmissing-governor, when that direct route is absent. UseE.10.ROLEonly for unresolved claim-bearing role wording. Ordinary work or procedure organization may remain ordinary. For other current claims, use the applicable exits in §1;C.27additionally governs temporal adequacy,E.18transformation-flow structure, andC.32.P2Sreopened carry-through. - Patterns for the next questions: Use the applicable exits in §1 for the needed contribution, including evidence that can determine the repair action.
- Boundary: C.32.FAIL contains repair cues for architecture-synthesis failures. It does not decide final candidate selection, evidence sufficiency, assurance, gate passage, release, or an architecture decision.
C.32.FAIL:End
Project Architecture Decision After Candidate Synthesis
Type: Architecture decision pattern under C.32 Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when a project must commit to an architecture that will guide design, construction, operation or change. A commission to choose any conforming design can be satisfied with one qualified option. Choosing the best of several designs calls for a comparison. An agreement may already fix a design, leaving the project to establish feasibility and decide how to put the commitment into effect.
An architect or another practitioner responsible for the decision uses the pattern to settle what the project will pursue, which conditions bind it and what can be decided during implementation.
Typical entry phrases:
First-minute use slice. A product-family team has evaluated three arrangements for modules, placement and supporting evidence. Within the authority granted for this release, its architect adopts one configuration. The decision fixes certain interfaces, accepts a stated loss of evidence reuse and leaves product teams free to choose local implementations within those constraints. Record the governing act, scope, remaining choices and condition for reconsideration. If the architect was commissioned to recommend, return the preferred configuration and its grounds to the person authorized to commit the project. An ADR can describe either status.
The primary EntityOfConcern is ArchitectureDecisionRelation@Project: an architecture decision relation over one bounded architecture question with one exact composite project U.Work as a participant. It links that work, the decision subject, candidate basis or binding condition, selected architecture option, affected structures, architecture characteristics, rationale, accepted losses, consequences, method and work expectations, publication projection, evidence or eval exits, and reopen conditions.
ArchitectureDecisionRelation@Project is not a new U.* kind. @Project is a compatibility and retrieval cue, not the source of project identity or scope. The relation is project-local only when projectWorkOccurrenceRef identifies the exact composite U.Work that participates in it. When another slot becomes load-bearing as an FPF object, recover the subject pattern for that object.
The system-of-interest (SoI) is the product System that a particular extended enterprise is organized to create, change, sustain in use or retire. A.1.STM:4.1 explains how to recover that organization, the agents' objectives and agreements. The decision's described holon can be the creator organization itself. Choose its architecture in light of the capability needed for the product use, and state which local result the work must establish and how it will be accepted under A.1.STM:4.1.1. The organization remains a creator when its development has a separate budget, contract or external provider, or when the business is sold. A distinct usable machine or component supply can have its own product SoI under that pattern's receiving-use and fitness grounds. Keep every product demand and shared condition that can change this decision. The singular System reference below addresses one designation; preserve other relevant designations and relations separately rather than binding a plurality to it.
When this decision designates a project system-of-interest, projectSystemOfInterestRef? names only an independently admitted existing U.System. Before identity inception, keep the intended referent in intendedProjectSystemClaimRef? as U.WorkPlan, decision, system description, or other claim content. A local kind is named separately in systemOfInterestKindRef?, while systemOfInterestClassificationRef? cites an independently obtaining classification judgment; neither entails an assignment. A current assignment uses separate species and occurrence refs, and the occurrence's holder is that System. Route any source phrase such as SystemOfInterestRole through E.10.ROLE before filling these fields. A taxonomy or scheme is not assignment content. Designation, kind, classification, assignment, Work, change, use facts, and the decision remain distinct. When a compound project-selection claim is needed, use A.15.6:4.1a to construct or reuse its sufficient local semantics and test the required facts. Return missing-substrate[project-selection-conjunction] only when the needed operator semantics are unavailable; retain unknown or failed case facts as those distinct outcomes.
When the decision uses a transformation-flow network, transformationFlowStructureNetworkRef? names only an independently selected E.18.NET TransformationFlowStructureNetwork@Context <: U.Structure; projectNetworkSelectionResultRef? may cite the separate C.2.1 judgment about why that network answers the project question, and architectureTransformationFlowStructureRelationRef? cites C.30.TFS-REL when architecture use is current. A network record, a C.32.CONWAY frame or exact pair row, and this decision create no network member, cross-flow occurrence, architecture-influence occurrence, architecture relation, or other world-side fact.
A team writes an architecture record, diagram, shortlist, ranking, or local choice without a recoverable architecture decision relation to exact project work. Later workers cannot tell which architecture configuration is selected, which structures are affected, which method they must use, which losses were accepted, or when the decision must be reopened.
Practitioners performing the project work can turn a candidate palette into one governed decision relation that can guide work, be described in an ADR-like record, support review, and be reconsidered as the architecture evolves.
Recover the decision to be made. Compare the feasible options if their differences affect the choice. If a binding condition fixes the design, identify it and check whether the project can fulfill it. State what the project commits to, how that affects subsequent work and what could reopen the decision.
Another practitioner can recover why the project will pursue this design, the structures and commitments affected, any alternatives and losses considered, the consequences for methods and work, and the conditions for reopening it.
Not this pattern when the current work is candidate synthesis, architecture-description adequacy, ADR publication projection, adequacy evaluation, evidence, assurance, gate passage, local choice, or performed work. Use the pattern for the next question named in Relations for those claims.
The first useful output is a recoverable project architecture decision. The following form can describe ArchitectureDecisionRelation@Project; ordinary prose can carry the same required values, and optional or inapplicable branches need not be filled:
The decision result must make its selected option, affected structures, criteria and trade-offs, scope, window, consequences and status recoverable. The relation and a description of it remain distinct; no second generic result or context record is required.
The field names in this first-output form are publication-friendly filled-reference fields. Durable relation positions must be expressible through A.6.5 SlotSpecs: each position has a local SlotKind, an admitted ValueKind, and a by-value or concrete RefKind filling mode. A field name such as decisionSubjectRef is not a SlotKind, not a U-kind, and not an ADR heading; it is the filled-reference field by which this relation record points to the value governed by the slot-bearing relation.
When an instruction cites actual implementation Work, recover each performer through A.13 and let actualImplementationWorkRef name the U.Work occurrence independently admitted through A.15.1. Assignment species and occurrence fields remain optional neighboring claims. Add actualImplementationWorkAttributionRef only when the instruction or receiving use represents precise assignment-bound attribution through the same obtaining A.13 assignment; missing or failed F.6 attribution alone leaves the Work ref intact while the referenced Work's independent A.13/A.15.1 grounds hold. The decision record creates none of these facts.
Problem
Architecture synthesis produces candidates; the Systems performing project Work need a decision, while any local system-role kind, assignment species, authority, or responsibility claim remains a separate fact established through its own pattern. The decision is not the candidate palette, the declared selected-set result, its publication, the architecture description, or the ADR file. It is the architecture decision relation that identifies the composite project Work, says which architecture option is now pursued for it, and records what follows from that selection.
The problem is difficult because architecture decisions sit between structures and Methods. C.30 keeps an obtaining ArchitectureRelation with its holon and selected U.Structure separate from an ArchitectureClaim carrying candidate, required, desired, or expected content. A project architecture decision can tell intended developer Systems which Method description, architectural style, pattern use, or work boundary to follow so that later work aims to produce or preserve the intended structures. For example, "use the client-server style here" is a Method-use instruction whose intended result is a module and interaction structure of the described System. The decision relation must keep actual or modal structure content, intended Systems, local kinds, separate System-classification judgments, assignment requirements and current assignment occurrences, plans and commitments, permissions and authority, and actual Work as separate branches. Route unresolved role wording through E.10.ROLE. When C.32.CONWAY supplies an influence-source architecture or selected structure, that source remains non-agentive and does not become the performer.
The architecture decision can fix selected structures while leaving other design choices open, either for the same holon or for its constituents. In this pattern, “refinement left open” names those unresolved design choices. State the fixed structures, the open choices, the source detail that must remain recoverable and the result that can reopen the decision. A constituent is located by the composition relation; any additional claim of model refinement requires B.5.MR's comparison. If these boundaries are missing, the decision becomes either empty advice or uncontrolled micro-management.
Finally, architecture decisions are evolutionary. They are made under current candidate knowledge, current characteristic criteria, current eval readings, and current organization or tool constraints. They should be explicit enough for present work and cheap enough to supersede when a better candidate, changed characteristic pressure, or architecture-influence/transformed-side fit changes.
C.32.PAD makes the decision relation recoverable for later use. Drafting its description can help develop the decision; a published account must distinguish a proposal from a decision that has been made.
Forces
Solution
Establish the decision relation with its exact composite project U.Work, candidate basis or binding condition, selected option, affected structures, any trade-offs, rationale, consequences and reopen conditions. Include the Method expectations and work split needed to guide its intended use. A single written account can develop and record these values; its form does not establish the decision.
Work in this order:
- Name the composite project
U.Workparticipant and the decision subject: described holon, decision question, intended use, ClaimScope, decision window, and status. If the decision designates a project system-of-interest, recover the named extended enterprise and agreement grounds under A.1.STM:4.1; cite the existingU.Systemor the pre-inception intended-system claim. Add a local kind only when it is current, add a separate System-classification judgment only when that judgment independently obtains, and add an assignment only through its separately declared A.2.1 species and obtaining occurrence whose holder is that System. RouteSystemOfInterestRolesource wording throughE.10.ROLE; establish every Work, change, and use fact through its own predicate and pattern. For a needed compound project-selection claim, apply A.15.6:4.1a to its sufficient local semantics and case facts. Returnmissing-substrate[project-selection-conjunction]only for unavailable required semantics; missing or failed facts retain their own outcomes. - Give the basis for the chosen design. If the commission permits any conforming option, identify the option and how it meets the binding conditions. If the decision turns on a preference, cite the alternatives compared. Identify the condition that fixes the design when a prior commitment settles it. For a binding choice, show which implementation remains feasible and what the project can still decide. When the decision calls for comparing alternatives, use
C.32for the candidate palette,C.32.MLAOfor residual-reducing multilevel candidate frames,C.32.CONWAYwhen an influence-source architecture and transformed-side architecture content shaped the candidate, andC.32.FAILfor repaired candidate errors. Cite a C.32.CONWAY synthesis frame while either side is modal or the direct influence relation is unresolved; cite an exact pair row only for its already obtaining direct occurrence and two obtaining C.30 architecture-relation participants. - Cite comparison or selection input only when it exists. Use
A.19.CPMfor explicit comparison,A.19.SelectorMechanismfor set-returning selection,G.5for selected-set result declaration, andC.11for local choice. For publication, useE.24.PUBfor the occurrence and audience availability; addE.17when a source-backed face and source return are needed. - State the selected architecture option or bounded exception. Name the affected selected structures and the subject pattern for each structure claim.
- Record the architecture-characteristic trade-off, or why no live trade-off is needed. Reuse adequate criteria, readings and explanations. Use
C.32.ACS,C.32.ACE,C.16,C.25orC.31for the particular missing criterion, evaluation, measurement, quality bundle or modularity/scale claim. Recover description correspondence and relevant loss throughC.30.AD; useC.29when mathematical representation choice, transfer or reliance remains unresolved. None of these inputs decides the architecture by itself. - Record the rationale and consequences. Identify any rejected options and the losses the project accepts. For organizational architecture, explain how the selected structure can provide the needed capability for the named products, which provision and cooperation it relies on, and who is entitled to accept a consequential change to their commitments. Keep expenditure, financing and expected product benefit distinguishable. A rejected option can remain useful as a stepping stone or archive item; do not turn it into a failure unless the receiving failure pattern is triggered.
- Relate the decision to architecture descriptions it uses. Use
C.30.ADfor architecture-description adequacy andC.30.ASVfor selected-structure view adequacy. A diagram, model, file, or view can describe the decision basis; it does not become the decision relation. - Bind the decision to method-use instructions when the architect needs developers to use a method, pattern, style, toolchain step, or work practice so the described or transformed-side holon is intended to gain or preserve the named structure. Use
A.15,A.15.1,A.15.2,A.15.5,A.6.M,E.8,E.11.PUR, andC.24according to the live claim. - State the architecture-to-refinement boundary. Name selected structures fixed by the decision, refinement scopes left open, source-return conditions, readiness exits, and patterns for any later governance question. For creator development, carry the finite local result and the evidence needed to accept its promised capability into the work boundary. Remaining product-use uncertainty does not make that local work endless; local completion does not establish every later product effect. When level or whole-change wording leaves the boundary unclear, use
C.30.STRATorB.2.Pto recover its meaning. UseB.2for a remaining question about retaining or reidentifying the whole; keep an existing-whole answer when its supported explanation is sufficient, and preserve unresolved grounds; cite the resulting claim inholonTransitionOrBOSCTriggerRefs?when needed. - Use
C.32.ADRwhen an ADR-like description is needed, andE.24.PUBfor actual publication and audience availability. AddE.17when a source-backed face and source return are needed. The description must retain whether the decision is proposed, made or superseded. - Add the evidence, assurance, gate or governance grounds required by the claims being made. Use the subject rule for evidence support,
A.10for a needed replayable source-to-use account,B.3for assurance,A.21for a gate and the applicable governance rule for authority. - Write reopen and supersession conditions. Reopen when the candidate basis changes, a protected architecture characteristic crosses its guardrail, an independently typed influence-source structure or arrangement no longer fits the transformed-side actual or modal architecture content, a stronger source changes the accepted loss, or the decision's method-use instruction proves unusable.
If one project question uses an E.18.NET network, first preserve that network's independent A.22/E.18.NET selection. A persistent project-network judgment stays in its C.2.1 result episteme under A.15.6, and architecture use docks through C.30.TFS-REL. A C.32.CONWAY exact pair row may be cited in architectureCorrespondenceRowRefs[] of a network record, but that citation is only a qualified reading: it adds no network member or cross-flow occurrence, and PAD repeats none of the network's member, relation, constraint, endpoint, or use-frame fields.
Decision readiness
A C.32.PAD decision is ready to draft when the current decision relation identifies the composite project U.Work participant and can identify the candidate basis or the binding condition that settles the design, one affected selected structure, one architecture-characteristic trade-off or declared reason for no live trade-off, one expected work consequence, one reopen condition, and any triggered holonTransitionOrBOSCTriggerRefs? or structuralInformationLensUseRefs? needed to preserve source return. When system-of-interest local-kind, separate System-classification, assignment, architecture-influence, or network fields are present, the applicable A.15.6, A.2 and A.2.1, C.32.CONWAY, E.18.NET, and C.30.TFS-REL preconditions must already be satisfied or the reference remains absent.
If a contract already fixes the interface, establish which interface it requires, whether the design can satisfy it and which choices remain with the project. A unique admissible option needs grounds for its uniqueness. One option found so far can suffice to show feasibility, while a claim that it is best among further options remains open. Use C.32 to construct alternatives when the decision can still turn on their differences. If the decision lacks adequate architecture criteria, use C.32.ACS or C.25; a separate new row is unnecessary when the needed criterion is already recoverable. If the decision only says "the metric is best", recover the affected structure, criterion, loss and comparison basis through the applicable C.32.ACE, C.16 or A.19.CPM contribution. If a required work Method is unclear, use its subject pattern and A.15 for the relevant work claim. If an existing system, role assignment, project-network judgment, network selection, architecture use, or influence pair is unresolved, require its subject pattern and keep only the designation, modal claim, candidate frame, or explicit stop in PAD.
A creator-architecture decision uses the product-use, capability and local-result connection from A.1.STM:4.1.1. If the needed resources, cooperation or authority are unresolved, keep the dependent capability or implementation commitment conditional. The decision can select bounded research with its own epistemic acceptance or another supported local result. For shared creators, an accepted loss must identify the affected product commitments and the decision basis that permits changing them; one product's improved metric supplies neither.
Constructive architecture decision path
Some architecture decisions are constructive: they prescribe Methods that intended developer Systems are expected to use so that later work aims to produce or preserve intended structures. A decision may name intended Systems, local-kind requirements, separate classification requirements, assignment requirements, plans, commitments, permissions, or authority before any assignment or Work obtains. Admit that path only when the decision keeps those claims separate and names:
- the obtaining architecture relation and selected structure to be preserved, or the exact modal
ArchitectureClaimstating the intended production or preservation effect; - the method description, architectural style, pattern use, or work practice to be used;
- the intended System when known; an optional local kind and an independently optional System-classification judgment; any current assignment through separate species and obtaining-occurrence refs whose holder is that System; and any merely intended assignment as plan, policy, or decision content rather than an occurrence;
- any responsibility or authority claim under its direct predicate, with the required participants, applicability and identity. An obtaining claim needs satisfying case facts; a prospective allocation states its intended participants and conditions without asserting a current occurrence. Use A.6.RCD for a missing predicate, and distinguish that from unknown facts or a failed applicable test;
- the expected structure effect on the described or transformed-side holon, kept modal until its direct C.30 architecture predicate obtains;
- the work-planning boundary and readiness or gate exit;
- the source-return condition and reopen trigger.
These requirements connect the architecture decision to work while keeping its expected structure effect distinct from any independently established actual effect.
Minimum sufficient relation and slot-change impact
A small complete PAD instance can be this short:
For this decision a developer may add optional detail while preserving every order form the approved contract accepts. A newly required field can exclude a previously valid order even if the boundary, service name and version label stay the same. Check that ordinary and exception orders still have a feasible route through the new schema. When teams may choose their schemas independently, establish compatibility for the combinations permitted by those choices or provide an authorized way to coordinate them in time. A commission for one agreed pair can be met by establishing the compatibility of that pair. If the field changes a fixed commitment, return that change to its owner. Inspecting a deployment can establish the schema in use. To establish that it made service replacement easier, identify the replacement work and a supported comparison with the previous arrangement. The schema could already have been present before adoption.
When a filled field changes, repair the smallest declaration or claim record that carries the changed content:
Archetypal Grounding
Software service architecture. A platform team compares synchronous service calls, event-carried integration, and a bounded shared kernel. The selected option is event-carried integration for order events with a bounded exception for payment authorization. C.32.PAD records affected module and information structures, latency and substitutability trade-offs, the method-use instruction for service teams, the event-schema source-return condition, and the reopen trigger when payment volume crosses the declared eval band.
Manufacturing fixture architecture. A production architect compares a dedicated fixture per product, a universal fixture with adapters, and a mixed cell layout. The selected option uses a universal fixture only for products inside a scale window. C.32.PAD records module, placement, maintenance, and evidence-structure effects, the accepted setup-time loss, the method instruction for cell design, and the trigger for returning to candidate synthesis when adapter complexity exceeds the guardrail.
Project system-of-interest and network. A plant-modernization decision designates already admitted PumpUnit-3 : U.System as the project system-of-interest and cites exact composite PumpUpgradeWork-7. The source label SystemOfInterestRole is recovered through E.10.ROLE. In this constructed case, the operator, controller supplier and maintenance organization have agreed the alteration, provision and qualification responsibilities that organize their extended enterprise. The local SystemOfInterestSystemRole criterion selects the product System of that enterprise: the intended delivery use, designation, agents' decisions and agreed contributions are recoverable. Merely being under investigation would not satisfy that criterion. The kind is cited through systemOfInterestKindRef, and the separate judgment classifying PumpUnit-3 under it is cited through systemOfInterestClassificationRef. This optional technical account serves a receiver that needs the classification; ordinary project reasoning need not introduce it. For this constructed case, the complete A.2.1 declaration of PumpQualificationSystemOfInterestAssignment is independently established: its assigned-kind domain contains SystemOfInterestSystemRole, and its predicate, applicability, participant meanings and occurrence-identity rule are settled. Case facts establish that predicate throughout the cited occurrence’s interval for PumpUnit-3 as holder, the assigned kind and every other required participant. PAD cites the species and its obtaining occurrence PumpUnit-3-QualificationAssignment-7 separately. A taxonomy or scheme is not an assignment participant, and neither kind, classification, nor assignment establishes Work, responsibility, or authority. Each Work-to-pump, Work-to-change, evaluation, production, or use fact keeps its own subject pattern. An independently selected E.18.NET network may connect production-system change, pump change, and qualification flows for the named project question. PAD cites that exact network, the optional C.2.1 selection judgment, and the C.30.TFS-REL architecture-use trace; it does not recreate network identity or infer cross-flow relations from a record or C.32.CONWAY pair row. If the decision needs a combined project-selection claim, apply A.15.6:4.1a to recover its local constructor semantics and required case facts. A separately materialized substrate is needed only under that pattern’s stated conditions. Return missing-substrate[project-selection-conjunction] only when the needed operator semantics are unavailable; distinguish unresolved case facts from an established failure of the claim.
Method-family architecture. The team applies a review Method to compare specialized review contributions, peer rotation, and tool-supported triage. E.10.ROLE first recovers any local kind, assignment, direct-relation position, function claim, or ordinary label hidden by role wording. The selected option uses peer rotation plus a tool-supported evidence handoff. C.32.PAD records those recovered relations, Method, evidence, and information structures, the trade-off between teachability and evidence custody, and the refinement scope left open for local checklists.
Architecture influence and transformed-side fit. An automation project changes both a toolchain and the product it is used to change. C.32.CONWAY supplies a synthesis frame or, only after direct settlement, an exact architecture-influence pair row over the obtaining source-side and transformed-side C.30 architecture relations. C.32.PAD records which toolchain structure is decision-relevant, which product-side structure is selected or intended, which Systems are intended, which assignment requirements and Methods are planned, and which fit change reopens the decision. A current assignment or actual Work appears only when it independently obtains. Performance and actual transformation require their own obtaining relations.
Digital-twin structural information loss. A built-asset team publishes a 6D-style digital-twin decision view for construction planning. The view intentionally hides supplier-agreement and temporary-work structures. C.32.PAD records the selected building, placement, schedule, cost, operation, and evidence structures that the decision uses; C.30.AD keeps the view correspondence and recoverable source structures explicit, and PAD states which accepted loss reopens the decision. Use C.29 additionally when mathematical representation choice, transfer or reliance remains unresolved. The view count, file, and model do not become the decision authority.
Workshop development and local completion. The workshop and receiving organizations have agreed component-supply and acceptance obligations that organize their extended enterprise. During an admitted workshop-development project, the architect compares an additional machine, an integrated production-and-measurement arrangement and external machining for components with an agreed surface finish and order flow. The machine-only option lacks a qualified measurement Method and operator, so it cannot support the promised workshop capability. The selected option includes those contributions and material access, with capacity, cost and handover conditions checked for the affected products. The decision fixes the measurement and production interfaces, leaves detailed scheduling open within the capacity limit, and gives the responsible performers the corresponding work expectations under the independently established authority. Acceptance requires evidence that the combination can meet the agreed capability conditions; buying the machine alone closes only the acquisition result. The organizational project can finish before future components operate. A changed workload, loss of operator availability or failed product-use hypothesis can reopen the relevant decision. The components remain the product SoI throughout; the workshop's own architecture and acceptance remain explicit.
Bias-Annotation
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Consequences
Rationale
Using C.32, practitioners synthesize candidate architecture configurations. Using C.32.PAD, the architect selects an option or bounded exception for the project work and states the resulting Method-use instructions and work consequences.
The pattern distinguishes four objects: an obtaining C.30 ArchitectureRelation over one architecture-bearing holon and selected U.Structure; any ArchitectureClaim that states actual, negative, unresolved, candidate, required, desired, or expected content about the holon, relation, or structure; ArchitectureDecisionRelation@Project, which connects composite project Work to the selected option and declared work consequences; and ArchitectureDecisionDescription@Project, which can be published in ADR-like or other forms and whose project use requires the exact composite U.Work and an independently obtaining use relation defined by its own pattern. Optional system-of-interest, local-kind, System-classification, assignment-species, assignment-occurrence, architecture-influence, and network references retain their A.15.6, A.2 and A.2.1, C.32.CONWAY, E.18.NET, and C.30.TFS-REL subject patterns. This lets FPF reuse its existing architecture, description, Method, work, evidence, assurance, measurement, publication, project, and network patterns instead of creating a separate architecture-decision ontology for those facts.
The pattern is architecture-reusable across holon kinds, not because every decision target is itself a holon kind. The same decision relation can apply to admitted holons such as systems, organizations-as-systems, built assets, AI-agent setups, epistemes, work occurrences, or disciplines. It can also concern Method, evidence, or an exact object or relation recovered from role wording, provided those values stay under A.3.1, E.10.ROLE, A.2.7, A.10, and A.15 rather than being admitted as holons by label.
SoTA-Echoing
These sources inform the decision-relation fields, boundaries, and reopen conditions used below.
Source-currentness boundary. Recheck a source row when an ADR template, architecture-description standard, evolutionary-architecture practice, FPF pattern, or project governance practice changes the decision field, method-work boundary, or reopen condition that PAD uses. Reopen only the affected optional docks if A.15.6 changes the actual or intended System or project-selection stop, A.2 or A.2.1 changes the role or assignment boundary, C.32.CONWAY changes its frame or qualified-pair threshold, or E.18.NET or C.30.TFS-REL changes network identity or architecture-use requirements.
Relations
- Builds on:
A.1.STM,A.15.6,A.2,A.2.1,C.30,C.30.ASV,C.30.AD,C.30.TFS-REL,E.18.NET,C.32.P2S,C.32,C.32.MLAO,C.32.ACS,C.32.ACE,C.32.CONWAY,C.32.FAIL,C.25,C.16,C.29,C.31, andC.31.ASAP. - Comparison and selection boundary: Use
A.19.CPMfor comparison,A.19.SelectorMechanismfor set-returning selection,G.5for selected-set result declaration, andC.11for local choice. When audience availability is current, useE.24.PUBfor the publication occurrence, form, carrier, audience, bounded use and availability, withE.17when a source-backed face and return to source are needed. PAD records the architecture decision relation with its exact composite project-work participant after those inputs are sufficient. - Description boundary:
C.30.ADandC.30.ASVgovern architecture-description and selected-structure view adequacy. PAD may cite those descriptions but does not replace them. - Structural-information boundary: Use
C.33andC.34for captured structure, lost structure, and preservation adequacy used by the decision relation. UseC.35for the exact generated or discovered result, the organization concerned and the status of its actual or proposed claim, including any unresolved basis, its next-use condition, and the limit or return. Resolve representation, publication-form, or carrier questions separately when the decision's use depends on them. Use PAD for the decision relation, rationale, consequences, accepted losses, Method consequences, Work consequences, source return, repair, and supersession claims. - Publication boundary: Use
C.32.ADRfor an ADR-likeArchitectureDecisionDescription@Project,E.24.PUBfor the publication occurrence and audience availability, andE.17when a source-backed face and source return are needed. - Adequacy boundary:
C.32.ADAevaluates a PAD decision relation, method docking, and publication projection for a declared use. - P2S docking: P2S reaches PAD only when implementation commitment is live; PAD records the decision relation and returns reopen conditions to P2S when actual structures, eval results, or source-return change the architecture question.
- Project system-of-interest boundary: Use
A.1.STM:4.1–4.3for the common product-use purpose and the engineering and finite acceptance of its creator. UseA.15.6for composite project Work, actual-versus-intended System designation, independent Work, change and use facts, project-network judgement, and a needed compound project-selection claim; its missing-substrate return applies only when the required semantics are unavailable.E.10.ROLErecoversSystemOfInterestRolewording; useA.2andA.2.1separately for local kind, classification, and obtaining assignment. PAD cites those objects only when the architecture decision uses them and proves none of them. - Network and architecture-influence boundary: use
E.18.NETto identify the selected network, its members, obtaining cross-flow occurrences, constraints, endpoints, and use frame;C.30.TFS-RELdefines architecture use;C.32.CONWAYis the pattern for the synthesis frame and exact qualified pair row. PAD cites the smallest exact refs and never turns a record or row citation into network membership, a direct relation occurrence, actor identity, performance, or actual structure effect. - Method and work boundary:
A.15,A.15.1,A.15.2,A.15.5,E.8,E.11.PUR, andC.24govern method descriptions, work plans, readiness, pattern-use recommendations, and agentic tool-use work. - Evidence, assurance, and gate boundary: Evidence claims retain their direct support;
A.10supplies a replayable source-to-use account when needed,B.3assurance calculus andA.21gate profiles for those current claims.
C.32.PAD:End
Architecture Decision Record Projection
Type: Architecture publication pattern under C.32 Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern to publish a project architecture proposal for discussion or an effective project architecture decision for people who will act on it. A team can use an ADR, a decision memo, a trade-study record or a certification rationale. For a decision in force, identify the ArchitectureDecisionRelation@Project or the equivalent decision under the project’s governing practice.
Primary working reader: an architect or architecture-responsible practitioner preparing a decision record for developers, reviewers, maintainers, operators, certifiers, or future architects.
Typical entry phrases:
First-minute use slice. A platform architect has a C.32.PAD decision relation selecting event-carried integration with a bounded exception. Using C.32.ADR, the architect creates an ADR projection with section functions for problem frame, candidate options, decision outcome, rationale, consequences, method-use instruction, work split, confirmation eval, source-return links, and supersession. The file is short enough for developers to read, but it remains a publication projection of the decision description, not the decision relation and not the architecture itself.
The primary EntityOfConcern is ArchitectureDecisionRecordProjection@Project: a publication projection of ArchitectureDecisionDescription@Project into an ADR-like record or package. Select this pattern only when the work is to publish or package that decision description for a declared reader use; generic ADR advice that cannot be mapped to decision-section functions stays outside C.32.ADR.
ArchitectureDecisionRecordProjection@Project is not a new U.* kind and not a new root publication ontology. It publishes a project decision description for the declared reader use. Use E.24.PUB for actual publication and E.17 when a source-backed face and source return are needed; use C.32.PAD for the decision relation.
A project can publish a record-shaped text that hides the actual architecture decision, or copy architecture descriptions, diagrams, and method material into a record without telling the reader what decision was made and what work must change.
A decision record can be small, readable, updateable, and tied to candidate synthesis, selected structures, architecture characteristics, method-use instructions, work split, confirmation evals, and source-return.
Start with the proposal readers are to discuss or the decision they can act on. Establish its status and choose the content readers need to examine, use or reopen it. For an effective decision, make the PAD decision relation recoverable.
After using the record, a reader can identify the decision to be made, the options and what can be done at the stated status. For an adopted design the reader can recover the outcome, grounds, consequences, changes to methods or work, how to check implementation and when to reopen the choice.
Use C.32.PAD when the decision content needs development; drafting its record can support that work without a second artifact. Keep an unresolved proposal visibly proposed. Use the subject pattern when the current question is architecture-description adequacy, a general MVPK face, evidence, assurance, gate passage, local choice, performed work or pattern authoring. Use the pattern for the next question named in Relations.
The first useful output is the readable decision record. ArchitectureDecisionRecordProjection@Project can organize its content when a structured account is useful; include the fields needed by the declared use:
Here @Project is a compatibility and retrieval cue only. It does not make the projection a project, project work, or the architecture decision relation. When this record projection is local to one project, projectWorkOccurrenceRef identifies the exact composite U.Work, while architectureDecisionRecordProjectUseRelationRef identifies the direct relation by which the record is published for, relied on by, or otherwise used in that work under the corresponding subject pattern. The decision relation, its description, the publication projection, any publication occurrence, and the composite project work remain separately identifiable.
Problem
ADR practice is useful because it makes architectural decisions small enough to read and update. It is also easy to misuse. A record can become a substitute for the decision relation, a loose essay about architecture, a copied architecture description, or a method prescription with no recoverable target structure.
C.32.ADR treats ADR as a publication projection. The project decision relation belongs to C.32.PAD. The architecture description belongs to C.30.AD and related view patterns. The method description or pattern-use recommendation belongs to A.15, E.8, and E.11.PUR when those claims are live. The ADR-like record publishes a decision description for a declared reader and use.
For a principle framework, use C.32.ADR only in the exceptional case where the accepted framework-architecture answer is also a project architecture decision with the ArchitectureDecisionRelation@Project and ArchitectureDecisionDescription@Project required by this pattern. Its prior basis may then cite the accepted answer and the exact E.9 DRR that records it; acceptance remains a separate decision, and E.4.PFAD only profiles the framework-specific content. An ADR-like publication may project the question, selected answer, alternatives, rationale, consequences, status, links, and supersession conditions for declared readers. That projection remains separate from the answer, its acceptance, the DRR, framework realization, pattern quality, and publication adequacy. When the principle-framework answer is not a project architecture decision, publish the selected decision episteme or reader-specific projection through E.24.PUB, with E.17 when its face/source-return conditions apply; do not use C.32.ADR.
The section question is therefore not "which headings are allowed?" The section question is "which decision functions must a reader recover?" Headings can vary by organization or industry. Preserve the decision question, outcome, rationale, material consequences, status and update conditions; include options, Method/work instructions, confirmation and source return when they are needed for the declared use.
Engineering trade studies, safety-certification rationales, design review memos, BIM decision logs and method-governance records can each carry a proposal or an effective decision. Identify what the intended reader can review, authorize or act on and what gives the document that effect. Choose the sections that supply the grounds for that use.
Forces
Solution
Publish the content and status the intended readers need. A consultation draft can give them the proposal to examine. Where the governing rule makes an authorized signature on that document effective, the signed document can also establish the decision, its scope and effective time. Identify the authority, act and applicable rule for that use. Two copies with the same proposed design can therefore warrant different actions. Whether an installation already implements the decision is a further claim about that installation.
Work in this order:
- Name the publication carrier and intended readers. Use a file or other carrier for the Markdown ADR, decision memo, trade-study record, engineering change note, certification rationale, design-review record, or other decision-description form.
- For a proposal, identify what readers can discuss and what action could adopt it. For an effective decision, give the decision and its description by reference or by value, with its applicable status. Use C.32.PAD to develop missing decision content.
- Choose the smallest scope that lets intended readers use the decision. Include the explanatory slice they need, with access to the fuller architecture or Method description when needed; avoid duplicating material that adds no useful content.
- Choose headings or carrier positions that make the needed decision functions recoverable. A separate mapping table is useful only when the relationship would otherwise be unclear.
- Carry the candidate basis. Record candidate options from
C.32or the reason no candidate-set question is live. Do not invent options in the ADR after the decision. - Carry the outcome that has been reached. A consultation can invite a choice among open options. For an adopted decision, identify the selected architecture option, bounded exception or supersession from PAD.
- Carry rationale, accepted losses, and consequences. Include architecture-characteristic trade-offs and guardrails, not only benefits.
- When the decision guides implementation, state what it fixes and which choices remain with the developer. Where the developer can choose among implementations, try a different permitted one. For a fully specified implementation, trace that one. Can its developer recover the obligation and any applicable exception in time to act? Include the required relation or provide access to its grounds.
- Carry the confirmation, evaluation or violation-detection conditions needed by the decision's use. Use
C.32.ACEfor evaluation framing,C.16for measurement,A.10for a needed replayable source-to-use account, andB.3,A.21or the governance rule for the respective assurance, gate or authority claim. Preserve the support each claim requires. - Carry the applicable publication and source-return boundaries. Use
E.24.PUBfor publication occurrences and audience availability,E.17when a source-backed face and source return are needed, andC.30.ADfor architecture-description claims. - Carry status, supersession, and update conditions. Old records remain useful as history when superseded; the active decision relation tells which one governs current work.
Required section functions
Select the following functions by the declared reader use, retaining what is needed to understand, apply, check or reconsider the decision. Explain a consequential omission when the reader could otherwise infer a missing option, instruction or confirmation. The table does not require a heading or an omission statement for every inapplicable function.
ADR package use
When readers need navigation among several records, provide a package map or an adequate existing index that identifies active, proposed, superseded and related records. A package map is a publication navigation aid. It does not merge decisions, replace PAD relations, or decide record priority by file order alone.
When one decision changes another, use explicit supersession or amendment links. Do not rewrite history by deleting the old record unless the project has a governed archival policy.
Archetypal Grounding
Software ADR. A team records a choice of event-carried integration. The developer must retain the event identity across consumers and may choose the local storage layout. To assess another permitted layout, the developer needs the identity requirement, applicable exceptions and access to their grounds. A separately commissioned notice of adoption can be complete once it identifies the effective decision and its scope.
Manufacturing trade-study record. A fixture decision is published as an engineering trade-study memo rather than a Markdown ADR. The memo carries candidate fixture variants, selected universal-fixture scope, accepted setup-time loss, cell-design method instruction, evidence links, and reopen threshold. C.32.ADR admits the memo because it projects the decision description for the intended reader.
Certification rationale. A regulated product records a safety-architecture decision in a certification rationale. The record carries the decision outcome, rationale, evidence refs, architecture-description refs, and confirmation path, while evidence validity retains its direct grounds and assurance uses B.3. Use A.10 when the source-to-use account needs recovery or replay.
Method-use record. A project decision requires reviewers to use an evidence handoff pattern before final review. The ADR-like record cites the Method description and expected evidence-structure effect; the instruction does not by itself establish performed review Work.
Bias-Annotation
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Consequences
Rationale
ADR practice is valuable when it makes architectural decisions communicable and revisitable. It becomes weak when a record is treated as the decision itself or when a template substitutes for decision work.
C.32.ADR therefore uses the record as a projection. C.32.PAD governs the decision relation; the record publishes its description for a declared reader. Drafting can help develop that decision without establishing it by itself. This preserves the ADR practice of keeping records small and updateable, while adding FPF kind control for architecture descriptions, method descriptions, evidence, assurance, gate, publication, and performed work.
The pattern also generalizes ADR practice beyond software by using section functions rather than software-specific carrier assumptions. A record can be a Markdown file, engineering memo, or certification rationale if it projects the decision description and keeps receiving claims with their subject patterns.
SoTA-Echoing
These sources inform the section functions, projection boundaries, and update conditions used below.
Source-currentness boundary. Recheck a source row when ADR template practice, decision-record tooling, violation-detection practice, architecture-description practice, FPF publication patterns, or project governance changes the section function or update rule used by C.32.ADR.
Relations
- Builds on:
C.32.PAD,C.32.P2S,C.30.AD,C.30.ASV,E.17,E.24.PUB,A.15,E.8,E.11.PUR, andC.32.ADA. - Decision boundary: Use
C.32.PADfor the project architecture decision relation. C.32.ADR publishes anArchitectureDecisionDescription@Project; it is not generic ADR guidance and not a second decision authority. - Structural-information boundary: ADR-like projections may cite
C.33orC.34to show captured structure, lost structure, or preservation adequacy behind the projected decision. CiteC.35for the exact generated or discovered result, the organization concerned and the status of its actual or proposed claim, including any unresolved basis, its next-use condition, and the limit or return. Resolve representation, publication-form, or carrier questions separately when the receiving use depends on them. The ADR projection publishes a decision description; useC.32.PADfor the decision relation andC.35for the cited result's admissible architecture use. - P2S docking: P2S may cite an ADR projection as one stage where decision, rationale, method expectation, and source-return are published for readers; ADR does not carry the whole architecturing flow.
- Architecture-description boundary: Use
C.30.ADandC.30.ASVfor architecture-description and view adequacy. ADR carries refs and reader-use slices, not full description authority. - Pattern and method boundary: Use
E.8when the published object is an FPF pattern,E.11.PURfor pattern-use recommendation, andA.15for method and work claims. - Publication boundary: Use
E.24.PUBfor publication occurrences, forms, carriers, bounded use and audience availability; addE.17when an MVPK face and source return are needed. - Evaluation boundary: Use
C.32.ADAfor decision adequacy,C.32.ACEfor evaluation framing andC.16for measurement. Evidence claims retain their direct support; useA.10for a needed replayable source-to-use account,B.3for assurance andA.21for a gate claim. - Package boundary: A record package map aids navigation among records. It does not decide active architecture by file order; PAD relations and status refs remain governing.
C.32.ADR:End
Architecture Decision Adequacy Scales
Type: Architecture evaluation pattern under C.32 Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when a project architecture decision, its method docking, or its ADR-like publication projection must be evaluated for adequacy before use, review, handoff, governance, or improvement.
Primary working reader: an architect, reviewer, or architecture-responsible practitioner checking whether a project architecture decision is good enough for a declared use and which repair should happen next.
Typical entry phrases:
First-minute use slice. An OrderFlow decision is to guide developer work, but its ADR says only “use events”. Read the decision and the available implementation instruction against the intended result: can developers recover the Method, inputs, readiness conditions and expected structural effect? If the instruction supplies part of that way but omits a needed readiness condition, assign the supported docking value and name the missing condition and its source. If the cited instruction is unavailable, leave that value unestablished and request the needed material; unavailable evidence is not established absence. Continue through the complete coordinate set, preserving supported results and the declared readiness floor. The evaluation yields bounded repairs or a use conclusion; approval and implementation commitment retain their own grounds.
The primary governed object is ArchitectureDecisionAdequacyEvaluation@Project: a C.32.ADA-local evaluation record over one ArchitectureDecisionRelation@Project, optional ArchitectureDecisionRecordProjection@Project, and declared use. It is not the evaluated decision, evaluation Work, or result episteme.
ArchitectureDecisionAdequacyEvaluation@Project is a local record form, not a new U.* kind, gate, evidence, assurance, pattern-quality evaluation, or replacement for C.32.PAD. Its coordinate table expresses the ADA result content; when that result must be a durable claim, one separately identified C.2.1 episteme states it. The dated evaluation remains separate U.Work, and any actual evaluation operation application remains with its subject pattern.
A decision can appear complete because it has a record, rationale, or diagram, while it is unusable for the declared work. Weak candidate basis, hidden trade-offs, missing method instructions, absent source-return, and vague supersession conditions remain invisible until implementation or review fails.
The project can evaluate architecture decisions by complete coordinate set, keep kinds distinct, and repair the weakest live coordinates without turning adequacy into a single score.
Ordinary working move: declare the evaluation use, assess every coordinate under :4.2, and give each supported value its rationale. Retain a grounded notTriggered disposition or an unestablished dependent value when appropriate. State the needed repair or missing basis at its smallest useful source locus.
Adoption test: after using C.32.ADA, another practitioner can recover the declared use, complete coordinate results with their rationales or missing bases, repair targets, and the supported use conclusion or reason it remains open.
Not this pattern when the current object is FPF pattern quality, measurement validity, evidence support, assurance, gate passage, candidate synthesis, comparison, selection, local choice, or ADR publication projection itself. Use the pattern for the next question named in Relations.
The first useful output is the complete coordinate evaluation and its use or repair conclusion. ArchitectureDecisionAdequacyEvaluation@Project can organize it when this structured record is useful:
Here @Project is a compatibility and retrieval cue only. A project-local ADA record names both the composite U.Work in projectWorkOccurrenceRef and the obtaining record-use relation in architectureDecisionEvaluationProjectUseRelationRef; the evaluated decision's own project relation, the suffix, or either field alone is insufficient. evaluatorSystemRoleKindRef and evaluatorSystemRoleClassificationJudgmentRef remain optional and separate. When a particular dated evaluation Work occurrence is claimed, the evaluator first has its A.13 core and evaluationWorkRef names Work independently admitted under A.15.1. An ordinary evaluation retains the operation, material conditions and support needed for its result. Assignment species, occurrence, and evaluationPerformedUnderAssignmentRef are optional and appear only when the record or receiving use represents precise assignment-bound attribution; any present F.6 ref uses the same obtaining A.13 assignment, and its absence or failure alone leaves independently admitted Work intact. If the finding questions a premise consumed by that admission, recheck it and dependent conclusions. Any operation application and result episteme remain separate. Kind, classification, assignment, Work, attribution, application, responsibility, and result do not substitute for one another.
Problem
Architecture-decision adequacy concerns several distinct objects. A decision relation can be strong while its ADR projection is weak. A record can be readable while the decision lacks candidate traceability. A candidate basis can be strong while method docking is absent. A method instruction can be clear while the architecture-characteristic trade-off is hidden.
A single overall grade or average score hides these differences. Adequacy must be evaluated by coordinates tied to the declared use. "Ready for internal architecture review", "ready for developer work", "ready for ADR publication", and "ready for governance enforcement" can require different stop conditions, but each use needs complete coordinate inspection.
C.32.ADA supplies an E.21-shaped ordinal evaluation pattern for architecture decisions. It uses the E.21 value domain and labels directly, then defines architecture-decision coordinates over PAD relation, method docking, publication projection, structural description, characteristic trade-off, and evolution. Weak coordinates point back to C.32.PAD, C.32.ADR, C.30.AD, A.15, C.32.ACS, C.32.ACE, C.16, C.25, or another subject pattern.
Forces
Solution
Evaluate the complete coordinate set for one declared use without averaging its values. Identify the blocking weaknesses and the repair that can make the intended use adequate. Retain a sufficient result; select further improvement when its useful contribution warrants the work.
Shared value meanings
Use the same ordinal value domain and labels as E.21. ADA specializes what counts as expression for architecture-decision adequacy; it does not create a second scale.
Values are ordinal content evaluations. They are not measures, averages, votes, maturity ladder names, evidence weights, assurance levels, gate statuses, or implementation approval.
The result-bearing coordinate row uses the E.21 label domain with an architecture-decision coordinate:
5 is not required for every use. Stop conditions are declared before evaluation. A lower diagnostic floor may be used for exploration or internal discussion, but it does not make the decision ready for developer work, implementation commitment, or governance enforcement.
Complete coordinate set
Evaluate every coordinate. If a coordinate is not live, mark it notTriggered only with a short reason grounded in the declared use. An established absence of required content can warrant a low value; an unperformed check or unavailable basis leaves the dependent value unestablished. Name that missing judgement, retain supported coordinate results and keep the complete evaluation as draft until its required basis is available. Do not use notTriggered to conceal that gap.
Assess each statement at the source needed for the decision. An omission in a memo can make the memo unusable for its audience. It leaves open whether the governing decision contains the missing provision. To judge transfer to another team, identify which information and help that team would have and which choices it must recover. Reuse adequate evidence for that coordinate and its stated use. Claim that another team actually succeeded only when its performance is evidenced.
Use-specific stop conditions
Declare the use before scoring. Common uses:
Use these as ordinary defaults. A project can declare stricter stop conditions. It must not weaken a triggered coordinate by hiding it under an average, and it must not call a diagnostic result ready for developer work or governance enforcement.
Small complete evaluation slice
This slice additionally records a dated evaluation and precise assignment-bound attribution. ArchitectureReviewService-4 has the A.13 performer basis for decision-adequacy evaluation. A.15.1 independently admits DecisionAdequacyEvaluationWork-12, performed from 10:00 to 10:20 on 2026-08-12, enacting DecisionAdequacyEvaluationMethod-2 within ProjectArchitectureReviewService-4 : U.System. The obtaining ArchitectureReviewerAssignment-6 is of the declared species ArchitectureReviewerAssignment, has that recovered performer as holder and covers the Work. The separate F.6 relation uses that same obtaining A.13 assignment. These are the grounds for the Work and attribution fields below; the coordinate result is a separate episteme.
In this example the implementers can recover the governing grounds, but have to reconcile parts of the record for the same release. The rows identify that work where it affects a rating. For a question confined to this release, excellent expression can be established within this release.
PAD adequate, ADR weak. A fixture architecture decision relation can reach 4 wellExpressedForDeclaredUse on every triggered PAD, Method, work-split, trade-off, and reopen coordinate while the trade-study memo omits status and supersession. ADA identifies only the missing publication-projection assertion and cites C.32.ADR; it does not rewrite the PAD relation.
ADR readable, PAD weak. A Markdown ADR can have clear headings, status, context, decision, and consequences while the project relation lacks candidate basis, affected selected structures, and Method/Work docking. ADA identifies those missing assertions and cites C.32.PAD, C.32, and A.15 as their subject-pattern locators; template completeness does not make the architecture decision adequate.
Archetypal Grounding
Developer-work readiness. A service decision explains the candidate and trade-off, but the instruction “teams should use events” leaves the event boundary, Method, required capabilities and start conditions open. Its MethodAndWorkDockingAdequacy for the intended implementers is 2 partiallyExpressedForDeclaredUse. The decision can prescribe how a future shift should implement a change before its members are assigned. For an instruction to start now, establish who can and may perform it. If the report goes on to claim actual Work or its performance under a particular assignment, establish those occurrences under A.13/A.15.1 and F.6 as applicable. Any claim of responsibility has its own domain predicate.
ADR-publication readiness. The manufacturing trade-study memo omits status and supersession, so its readers cannot establish which instruction is in force from this copy. For that use the projection merits 2 partiallyExpressedForDeclaredUse. Open the decision or another authoritative record if the task also asks how adequately the underlying decision sets out its reopen conditions. A complete decision with an incomplete memo and a decision that itself lacks those conditions require different repairs. Until the relevant decision content is available, leave its coordinate unresolved. When it is available, assess that content and repair the memo through C.32.ADR.
Architecture review. A method-family architecture decision has candidate options and Method instructions, but no declared architecture characteristics. ADA gives ArchitectureCharacteristicTradeoffAdequacy = 0 absent. The repair states the missing characteristic assertions using C.32.ACS and C.25 before review can judge the decision.
Governance enforcement. A toolchain-product correspondence decision depends on team and tool structures. ADA evaluates TransformerTransformedCorrespondenceAdequacy; if the correspondence refs are absent, the repair states the missing correspondence assertion using C.32.CONWAY before institutional governance can constrain Method use.
Bias-Annotation
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Consequences
Rationale
C.32.ADA exists because architecture-decision adequacy is not one property. It is a family of recoverability and use-readiness coordinates across decision relation, selected structures, architecture characteristics, method docking, work split, publication projection, evidence or eval exits, and evolution.
The pattern follows the same discipline as FPF quality evaluation: declared use first, complete coordinate set, ordinal values with rationale, no average, and targeted repair. It remains architecture-specific because the coordinates are tied to PAD, ADR projection, architecture descriptions, candidate synthesis, method-use instructions, and architecture-characteristic trade-offs.
SoTA-Echoing
These sources inform the coordinates, value meanings, stop conditions, and repair exits used below.
Source-currentness boundary. Recheck a source row when FPF evaluation discipline, architecture-decision practice, ADR violation checking, evolutionary-architecture eval practice, or project governance changes a coordinate, value meaning, or repair exit.
Relations
- Builds on:
C.32.PAD,C.32.ADR,C.32.P2S,C.32,C.32.MLAO,C.32.ACS,C.32.ACE,C.32.CONWAY,C.32.FAIL,C.30.AD,C.30.ASV,A.2.1,A.2.6,A.15,A.15.1,A.19,C.2.1,C.16,C.25,C.29,E.17, andE.21; uses A.1.1 only when a selectedBoundedModelUseStructurechanges evaluation interpretation. - Evaluation boundary: Use ADA only to evaluate architecture-decision adequacy for a declared use. It does not perform candidate synthesis, comparison, selection, selected-set result declaration, actual publication, local choice, evidence support, assurance, gate passage, governance enforcement, or pattern-quality evaluation.
- Decision and projection boundary: Use
C.32.PADto repair the decision relation andC.32.ADRto repair ADR-like publication projection. - Description and structure boundary: Use
C.30,C.30.AD, andC.30.ASVfor architecture claim, description, and view adequacy. - P2S docking: Use
C.32.P2Swhen a weak decision-adequacy row must reopen the connected architecturing flow rather than only repair the decision record. - Method and work boundary: Use
A.15,A.15.1,A.15.2,A.15.5,E.8,E.11.PUR, andC.24for method, work, readiness, pattern-use, and agentic tool-use claims. - Characteristic and eval boundary: Use
C.32.ACS,C.32.HCS,C.25,C.32.ACE,C.16,C.31, andC.31.ASAPfor characteristic rows, Q-Bundles, eval-program framing, typed measurement or evaluation results, modularity, and scale preference. ADA may inspect references to those objects as coordinate evidence; use their subject patterns for any definition change or actual operation. - Evidence, assurance, gate, and governance boundary: Use
A.10,B.3,A.21, and local governance patterns when those claims are live.
C.32.ADA:End
Assess Structural Information for Architecture Use
Type: Architectural pattern Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when an architect has a description, view, decision record, ADR-like projection, eval report, method handoff, generated relation graph, source model, or realized holon observation that carries or describes selected architecture-relevant structure and needs to know which selected structure is recoverable for the next architecture use.
Use the same pattern when the carrier is a narrative rendering or principle-framework publication for architecture work: the carrier may preserve a problem-to-structure ordering, problem-situation architecture, solution-move architecture, candidate trade-off, decision rationale, or missing-structure cue, but it may also hide selected structures that the next architecture use still needs.
In architecture-mediated rendering, inspect the chain from carrier to architecture description or view, then to the described holon and its architecture-relevant selected structures in context, then to wider selected source structures, because each relation may have captured and lost different structure.
Primary working reader: an architect, architecture reviewer, method steward, or AI-assisted architecture worker who must use one carrier or observation without letting it stand for the whole architecture, the project decision, evidence sufficiency, or realized structure.
Typical entry phrases:
The first useful result states which captured structure supports the intended architecture use and which missing structure would limit it. An ordinary explanation can suffice; StructuralInformationAdequacyNote@Context is an optional working form. A measurement, assurance conclusion, project decision or architecture description has its own requirements. If an existing account already answers the question for this use, use it directly.
When the structured form helps, include the values needed for the intended use:
After using C.33, another practitioner can recover the captured structure and the use it supports. State missing structure, excluded uses and the question or test needed to recover it when they affect that use or a plausible continuation. Record a stable reference only when that next use must travel independently.
What C.33 buys in practice: the practitioner can use a partial carrier within its limits. The pattern turns "this diagram, ADR, graph, report, or observation is useful" into a reviewable statement about captured structure, missing structure, missing-structure return, and the next claim or test that needs that structure.
Read the statement, diagram or observation for the decision at hand. Identify the relation it shows and how the source established it. If the decision turns on an omission, find out which relations the source would have shown.
For a reader-relative comparison of structural amount in an expressed account, use C.2.8 with the reader's preparation, access and budget. State a qualitative comparison directly in the note, or cite an existing measurement or evaluation through measurementOrEvalRefs? when that separate result is needed. Then judge whether the captured structure is sufficient for the declared architecture use. The note retains its capture-and-return purpose when it uses a characteristic result.
When the current question concerns architecture, record, lens, reading, decision, authorization, or publication admissibility beyond structure-capture adequacy, use the pattern that defines or tests that object and question. Use C.33 for the captured and missing structural content on which that use depends.
Problem
Architecture work depends on partial carriers. Diagrams, views, relation graphs, ADRs, model queries, code-agent probes, neural-network architecture reviews, eval reports, method descriptions, and operation observations can carry enough structure for one action while losing structure needed for another action.
The practical problem is not "is the carrier good?" The problem is: what selected structure can be recovered from it for this declared architecture use, and what missing-structure return is needed before relying on it further?
Common failures:
- a diagram, model, generated graph, ADR, or benchmark trace starts acting as architecture by presentation;
- structural information is confused with a score, entropy value, epiplexity estimate, dashboard reading, or eval result;
- hidden structure becomes invisible when a later candidate, decision, or work method depends on it;
- source labels such as layer, router, expert, cache, memory, block, gate, SSM, pruning, distillation, or architecture search are copied as FPF ontology instead of being recovered through current FPF rules that define the relevant structures and relations;
- partial-observation outputs from code agents or AI tools are treated as internal belief proof, safe-change authority, evidence sufficiency, or release confidence.
Forces
Solution
Start with the decision the reader must make. Which relation would settle it, and what does this view record about that relation? A confirmed entry can establish that one connection exists within the stated scope. To infer from an empty result that there is no such connection, establish that the source covered the relevant entities, relation and period and that its selection and display would have exposed a match. A filtered view can be complete for that particular inquiry.
Compare the result the recipient needs across the arrangements still consistent with the record. If one action is justified in every case, a request to choose that action can finish. To describe the actual arrangement, list all permitted alternatives or explain why an outcome occurs, preserve the distinctions that change that answer or its grounds. Look for the assignment, rule or observation that can resolve a consequential difference. It may already be in the record.
Work in this order:
- Name the architecture claim or pre-claim, described holon, architecture concern, intended use, and any ClaimScope or qualification window that changes the adequacy judgment.
- Identify the selected structures being relied on. If a needed structure is not recoverable, leave the dependent adequacy conclusion open and recover it through
C.30,C.30.ASV,A.22orC.32.P2Sas appropriate. Retain independently supported uses. - Name the carrier, selected source structure, description, view, narrative rendering, decision record, eval report, method handoff, generated relation graph, or realized observation being used.
- State the captured selected structure in relation terms: relations, constraints, invariants, allocations, compositions, variation classes, operations, dynamics refs, or preserved organization.
- State the expected but uncaptured structure when the next use needs it: hidden placement, data custody, runtime dependency, transformation-flow relation, unexplored region, or missing bearer. State any unresolved source-label semantics or missing confidence class separately when it changes that use.
- State lost or hidden structure when it limits the intended use or could mislead its reader. Explain the basis of the adequacy conclusion; no invented loss is required when the captured structure suffices.
- Add the observer and budget conditions needed to interpret a
C.2.8comparison or a source obtained from a bounded observer, learned representation, probe, relation graph, or epiplexity-style lens. Keep preparation, source access and actual help visible when they can change the recovery claim. - Add source label recovery when unresolved terms from a domain practice such as neural-network architectures, software modules, built assets, organizational roles, methods, or work change an architecture claim or its next use.
- Use the pattern that defines or tests each mathematical-lens, measurement, eval, decision, evidence, assurance, gate, release, method, work, or publication claim when one of those claims is current.
- Stop when the supported use is clear, with any material limitation and the missing structure needed to resolve it. Apply a further rule or test only when the receiving claim needs it.
CGUS-aware neighbor use: when a carrier, route card, narrative rendering, architecture description, framework publication, or generated relation graph is relied on because it preserves a constraint-governed unfolding structure, C.33 records only what that carrier captures and loses. The selected structure remains ConstraintGovernedUnfoldingStructure@Context or a local U.Structure block governed by A.22.CGUS, E.18.3, C.32.P2S, E.23, or another structure pattern.
When the carrier is a narrative, A.6.3.NAR governs only its selected-source carry-through, ordering and connective account, loss, reader use, and source return; it neither selects nor admits the structure.
A DemonstrativeUnfoldingSlice@Context may be the U.Episteme slice or presentation whose captured structure and lost structure C.33 records; it is not the selected U.Structure by itself. C.33 does not admit the CGUS; it tells the practitioner handling the next question what the carrier preserved and where missing selected structure must be inspected or repaired.
Archetypal Grounding
C.33 assesses what selected structure a partial carrier captures, what it leaves out, and which practical use it supports. The carrier may be a diagram, decision record, query result, eval report, code-agent map, neural-network architecture review, method handoff, or observation of the realized holon. Apply the rule or test needed by a further claim. Add C.2.1 identity only if that claim must travel independently.
An integration decision. An approved decision record says “use event-carried integration with bounded exceptions.” A team can cite the adopted integration style. To decide whether a particular interface qualifies for an exception, the team needs the condition defining the exceptions. If the record gives that condition, the team can apply it to the interface. If the record only says “bounded,” check whether an applicable contract or policy already decides the case. If the applicable rule gives the team discretion, the team can exercise it. If its authority or the applicable rule is still unknown and the answer depends on it, recover that basis or obtain a determination from the decision holder.
To answer which interfaces actually use event-carried integration, inspect evidence about the installed interfaces. A verified as-built inventory attached to the same record can already supply it. Use C.30.AD for the architecture description or C.32.ADR for the decision record. Reopen the decision under C.32.PAD if the rule itself needs to change.
Episteme case. A relation graph records observed dependencies and inferred module roles. Identify the basis for each relation used in the answer and which regions the inquiry covered. The same graph can contain confirmed relations and leave other modules unexamined. Record the consequential gap, including runtime or deployment relations the graph cannot reveal.
The graph can inform C.34 preservation checks or be assessed under C.35 for its next architecture use; it does not by itself prove an internal-belief claim, qualify as release evidence, or establish full architecture adequacy.
Neural architecture case. A neural-network architecture review says a model changed attention, SSM block, router, cache placement, pruning mask, and distillation path. C.33 recovers the selected structures and changes being described: dataflow relation, path-selection relation, memory placement, cache placement, and block substitution.
C.33 separately names affected characteristics such as latency, compute, memory, and robustness. Retain clear source-local meanings. Use C.30.STRAT, C.30.TFS-REL, C.31, C.32, C.16 or C.32.ACE when a structure, relation or characteristic needed for the current claim remains unresolved.
Architecture narrative case. A team narrative says "we moved from candidate A to candidate B because data custody forced placement P and made latency trade-off T acceptable." C.33 records the structural content and decision context the narrative captures: the move from candidate A to candidate B, data-custody constraint, placement constraint, and trade-off.
The statement gives the team's reason for changing candidates. To use B under a tighter latency limit, establish that its latency meets the new limit under the intended load. An existing validated bound can settle that question. If both A and B remain eligible and the commission calls for choosing between them, compare them under the criteria that decide this choice. To establish which placement was installed, check applicable as-built evidence. Use the evidence already in the account when it settles the question.
The account is sufficient when it establishes what the carrier supports for the intended use. A further claim may need an architecture description, view, evaluation, evidence, decision or authorization under its own rule; C.33 identifies the missing structural contribution and returns that question to the applicable pattern.
Bias-Annotation
Conformance checklist
Common Anti-Patterns and How to Avoid Them
Consequences
Positive consequences:
- A partial carrier becomes usable without becoming authoritative. The architect can take the structure that is recoverable and stop before overreading the carrier.
- Missing-structure return becomes local and reviewable: the note says which missing structure requires C.30, C.30.ASV, C.32.P2S, C.32, PAD, ADR, C.29, C.16, ACE, evidence, assurance, or work checks before the next use.
- AI-produced maps and maps derived from named source publications, source models, or source codebases become safer architecture inputs because observation class, confidence, unexplored regions, and budget boundary are visible.
- Neural-network and code-architecture source labels become usable without importing those labels as FPF ontology.
Costs and trade-offs:
- The assessment takes attention and may need a retained note. Its cost is justified when a structural uncertainty could change the next use; an already adequate account can be reused.
- The note can be too weak for decision, evidence, assurance, eval, release, or realized-structure claims. In those cases C.33 should stop early and use the pattern that defines or tests the current claim.
- A familiar diagram or ADR may leave structure needed for the intended use unestablished. Recover that structure before relying on the dependent conclusion.
Rationale
A partial carrier can preserve the relations needed for one architecture use while omitting those needed for another. Assessing its captured structure against the intended use makes that boundary explicit and identifies any missing structural contribution. A description, measurement or evaluation remains available under its own requirements.
This split keeps P2S as the whole architecturing spine and C.32 as the pattern that describes candidate synthesis. C.33 does not synthesize architecture and does not decide the project architecture. It explains what the carrier contributes and what the next use still needs.
The source choices explain the fields. Epiplexity supplies a computational model for observer-bounded structural information; C.2.8 defines the general characteristic consumed here when the source is an expressed episteme. Captured amount informs the architecture-use adequacy judgment without replacing it. Multi-relational structural entropy motivates relation-kind awareness but not adequacy by number. Sapunov and ToCS motivate partial observability, active-passive gap, invariant fields, confidence, and unexplored regions. GonzoML motivates richer neural architecture operation language without making those labels FPF ontology.
SoTA-Echoing
Relations
- Builds on:
A.22,C.30,C.30.AD,C.30.ASV,C.32.P2S, andC.32. - Uses:
C.2.8for extractable structural amount in an expressed episteme;C.29when choosing, transferring, or relying on a mathematical representation remains unresolved;C.16,C.25, andC.32.ACEwhen a claim about captured or lost structure is recorded as a measurement, Q-bundle slot, criterion, or eval reading. - Coordinates with:
C.30.STRAT,C.30.TFS-REL,A.6.M,A.6.3.NAR,C.31,C.31.ASAP,C.32.PAD,C.32.ADR,G.5,C.18,C.19,E.18,F.9, andF.15. - Boundary: C.33 governs structure-capture adequacy and missing-structure return for a declared architecture use. It does not ground architecture, select candidates, decide projects, publish records, measure values, supply evidence or assurance, authorize work, or claim realization.
C.33:End
Assess Structural Correspondence for Architecture Use (Equivalence and Morphisms)
Type: Architectural pattern Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when two descriptions, views, models, generated outputs, or realized observations that carry or describe selected structure are being treated as same enough for architecture work and the practitioner must say what selected structure is preserved, what is lost, and which use the correspondence licenses.
Primary working reader: an architect, reviewer, or model-assisted practitioner comparing views, descriptions, source models, generated graphs, candidate architectures, realized structures, abstraction levels, coarsened models, or transformed models.
Typical entry phrases:
The first useful result explains which correspondence supports the intended architecture use. An ordinary account can suffice; reuse an adequate existing account. StructuralPreservationAdequacyNote@Context is an optional working form:
After using C.34, another practitioner can recover the claimed correspondence, preserved structure, direction and scope, any loss or uncertainty that limits reliance, and the use it supports.
What C.34 buys in practice: the practitioner can say "same enough for this use" without implying stronger equivalence. The pattern makes sameness conditional on preserved relation, declared loss, and receiving use.
Name the conclusion you want to carry from one model to the other. Identify what makes it true in the first. Look for that relation, or a rule from which it follows, in the second.
For a current claim limited to mathematical-lens use, generic bridge translation, measurement, structural view adequacy, architecture-description correspondence, candidate synthesis, decision, evidence, assurance, gate, release, or work authorization, use the pattern that defines or tests it. Use C.34 only for the architecture-specific preservation claim.
Problem
Architecture work often needs "same enough" claims. A view should correspond to a description. A generated graph should preserve selected dependencies. A candidate should preserve required interfaces while changing placement. A realized structure should match an expected selected structure enough for an evaluation or decision repair. A neural-network substitution should preserve dataflow or routing while changing memory and compute trade-offs.
Visual similarity, label sameness, graph isomorphism, or formal vocabulary does not by itself establish architecture adequacy. An edge-isomorphic graph can lose relation semantics. A projection can preserve module names while dropping control authority. A category-theoretic morphism can be useful as a C.29 lens without proving architecture equivalence. A DSM cluster can preserve co-change pressure while losing functional bearer semantics.
C.34 makes the preservation claim explicit before the result is used.
Forces
Solution
State the decision that will use the correspondence. Identify the entities being matched and the relation on which the decision depends. A conclusion about the source carries over in the direction for which the mapping preserves the facts needed to reach it. To infer absence in the target from absence in the source, also show why any relevant target occurrence would have a source counterpart. Extra target connections can change the answer.
For a grouping, ask how membership is determined. If both models define membership by the same rule and the correspondence preserves everything that rule uses, derive the target grouping. If membership is assigned independently, inspect the assignments or an applicable allocation rule and the facts it uses. For an assignment whose preservation is in doubt, keep the matched entities and links fixed. Could two different assignments fit them? When this changes who may act, the correspondence alone leaves that decision open. Correspondence between two disciplinary models can matter even when both are at L2. A claim of model refinement has its own test in B.5.MR:4.3.
When reviewing a comparison someone produced, distinguish the correspondence supplied in the assignment from the choices the person or AI agent made. Applying a supplied match can establish that the comparison gave the right result. To credit someone with constructing a disciplinary model of correspondence, recover how they selected the entities, interpreted the links and justified the matching rule for further comparisons. An earlier decision record can show those choices. A new task can leave them open. Use C.39.QR:4.3 when the attribution changes what you entrust to the agent.
Work in this order:
- Name the selected source structures and selected target structures. Do not start from labels, diagrams, or tool objects alone.
- Name the intended architecture use: view correspondence, candidate comparison, structure recovery, generated-output admission, realization check, eval support, decision repair, or another receiving claim.
- Select a warranted mapping mode sufficient for the use. A weaker sufficient correspondence can finish the work; an already established stronger result can be reused within its scope. Use
exactEquivalenceonly when preservation of the entire selected structure is established within the declared scope. - State preserved relations or constraints in domain terms. Use FPF distinctions where the claim depends on them. Include relation-type semantics when edge or link meaning changes the use.
- State directionality and scope or scale window, with any material lost or hidden structure. Distinguish an observed loss from preservation left unestablished by incomplete observation. Neither supports a claim requiring the missing preservation, but each selects a different next inquiry or repair.
- Retain the mathematical definition, mapping and proof conditions needed by a formal preservation claim. Use
C.29when choosing, transferring, or relying on that mathematical representation remains unresolved. - Cite
C.30.ASV,C.30.AD, or their correspondence records when the relation is view or architecture-description correspondence. - Cite
A.6.3.NARwhen the target episteme or representation is a narrative rendering whose ordering rationale, preserved selected source structure, source-return condition, and any unresolved stronger assertion with the pattern that defines it must stay inspectable. - Use
F.9for cross-context or cross-tradition substitution or claim transport that requires a Bridge. UseF.15when the receiving use needs its static or regression conformance tests. - Stop when the supported use and its grounds are clear, with any material limitation and the condition for resolving it. Apply an additional mapping, Bridge or conformance rule when that use requires it.
CGUS-aware neighbor use: when a route-shaped publication card, narrative sequence, generated route card, framework publication, or demonstrative slice is claimed to preserve a constraint-governed unfolding structure, use C.34 only to check the sameness relation. The result names selected source and target structures, mapping mode, preserved constraints, preserved ordering or branching relations, lost alternatives, directionality, and admissible use.
A.22.CGUS, E.18.3, C.32.P2S, or another structure pattern continues to define or constrain the selected unfolding U.Structure; cite a ClaimGraph only if that structure claim must travel independently.
A demonstrative slice is a U.Episteme that presents one traversal of a qualified CGUS. Its correspondence to that structure may be checked here; DemonstrativeUnfoldingSlice@Context is a lineage cue, not a slice identity.
When that presentation is a narrative, A.6.3.NAR defines its source selection, ordering and connective account, preservation/loss, use, and return, not the selected structure.
The C.34 result says only whether the target is same enough for the declared architecture use.
Archetypal Grounding
C.34 is the pattern for a declared architecture preservation claim. It is used when a practitioner says that one description, view, model, generated output, or realized observation is same enough as another for a specific architecture use. The account identifies a sufficient warranted mapping, preserved structure, direction, scope and supported use, together with any material loss or uncertainty and the rule needed by further reliance.
What a pair of diagrams preserves. Two diagrams show all three stations A, B and C, with links A–B and B–C. In one, a link records data dependency. In the other, it records control authority. For the question “which stations are joined by some path when link direction and meaning are ignored?”, both give the same answer. Matching all the stations and links preserves connected components under that definition.
A planner now needs the crew responsible for each station. One crew could have A and B and another C, or one crew could have all three, with exactly the links shown. Connectivity leaves both crew assignments possible. The current allocation record can identify the responsible crew, as can a rule assigning stations by connected component together with the current assignment of each component to a crew. To dispatch a crew, use the duty rule for this shift. If it sends the responsible crew, the assignment settles whom to send. If it provides cover for an unavailable crew, also check who is on duty. If the decision concerns who may authorize a control action, inspect the control relation with its direction and meaning intact. Use C.30.ASV or C.30.LCA when that relation needs recovery.
Source model and generated graph case. A code-agent dependency graph matches module names in the model used as the source, but marks several edges inferred and several regions unexplored. C.34 records the observation class and directionality, the supported dependency correspondence, and the dynamic wiring whose preservation is still unestablished.
The graph may help inspect candidate dependencies; safe-change or release reliance needs the missing structure and the assurance grounds for that use.
Candidate and realized structure case. A candidate architecture promises that a service split preserves interface substitutability, but the realized structure adds shared storage and a hidden orchestration dependency. C.34 records preserved interface signatures, lost runtime independence, changed coupling, and a preservation-loss condition that reopens the dependent decision premise before its reuse.
Use A.6.M for substitutability, A.6.F for a function or bearer question, C.31 for changed coupling, C.30 for the changed structure and C.32.PAD for the affected decision as each question requires.
Neural substitution case. A candidate replaces an attention block with an SSM block. C.34 asks which selected structures are preserved: sequence dataflow, routing interface, memory access, latency envelope, training resource boundary, or inference resource boundary. Shape sameness or benchmark improvement does not by itself preserve the architecture relation needed by the next claim.
Selected source structure and narrative structure case. An architecture narrative uses the order “pressure, alternative, trade-off, decision, residual.” To recover the grounds for a decision, match the option, condition and reason in the narrative to the source. Suppose the source records that B was selected because it met the custody condition and also gives the evaluations of the alternatives. The brief narrative preserves B and its reason and omits those evaluations. The resulting account includes the following.
The narrative order is not exact equivalence and does not license implementation, evidence, or assurance use without the rule that defines or tests that downstream claim.
Bias-Annotation
Conformance checklist
Common Anti-Patterns and How to Avoid Them
Consequences
Positive consequences:
- Architects can use partial sameness without asserting identity. This keeps comparison, projection, generated-output admission, realization checks, and decision repair usable.
- Formal methods can support the mapping without becoming architecture ontology. Examples include graph matching, category-theoretic morphisms, entropy, and simulation relations.
- Cross-context and source-tradition risks are visible early because directionality, scope, bridge loss, and the conformance rules for the receiving use are named.
- Later decisions can be repaired locally: the preservation note says which relation failed, which structure was lost, and which claim or check must be reopened.
Costs and trade-offs:
- Assessing correspondence takes attention. Reuse an adequate account; investigate further when a plausible preservation difference could change the intended use.
- The pattern does not prove formal equivalence by itself. When proof, measurement, evidence, assurance, gate, release, or work authorization is current, apply the corresponding rule or test.
- Some comparisons will lower from equivalence to correspondence or near-sameness. That lowering is a success when it prevents a false downstream claim.
Rationale
Architecture preservation is use-relative. The same two structures can be equivalent for one use, merely corresponding for another, and unusable for a third. For this architecture-specific claim, start from source and target selected structures, then establish a warranted correspondence sufficient for the next architecture use.
This keeps C.34 separate from its neighbors. C.29 defines mathematical-lens-use accounts. C.30.AD and C.30.ASV govern architecture-description and structural-view adequacy, with records for uses that need them. F.9 defines cross-context Bridges. F.15 supplies static and regression conformance tests over a finite slice of governed objects and versions for one receiving use. C.32 describes candidate synthesis. C.34 contributes the preservation claim those uses may need, but it does not replace them.
The source families explain the safeguards. Structural-equivalence research shows that symmetry can compact search only under explicit conditions. Applied category theory explains the capabilities of preservation maps; they remain formal lenses until tied to the architecture use. MBSE view practice makes projection and omitted structure ordinary. Sapunov and ToCS, plus GonzoML, show why observed relation maps and neural substitution labels need typed relation, confidence, and source-label recovery before architecture use.
SoTA-Echoing
Relations
- Builds on:
A.22,C.30,C.30.ASV,C.30.AD,C.29, andF.9. - Uses:
C.16,C.25, andC.32.ACEwhen a preservation, similarity, distance, entropy, loss, or compression claim is recorded as a reading or eval result. - Coordinates with:
C.32,C.32.PAD,C.32.ADR,C.30.TFS-REL,C.30.STRAT,A.6.M,A.6.3.NAR,C.31,C.31.ASAP,E.18, andF.15. - Boundary: C.34 tests declared preservation adequacy for an architecture use. It does not make a formalism ontology, select a candidate, decide a project, establish evidence or assurance, or authorize substitution across contexts; that substitution requires the Bridge predicate and conformance rule.
C.34:End
Assess Generated or Discovered Results for Architecture Use
Type: Architectural pattern Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when a generated, searched, clustered, queried, learned, transformed, simulated, or discovered result may seed or inform architecturing, and the practitioner must decide what that result is and whether it can enter architecture work before or around C.32 candidate admission.
Primary working reader: an architect, architecture researcher, AI-assisted architecture worker, model-based engineer, or reviewer receiving a result from DSM and MDM modularization, MBSE query and view generation, graph grammar, model transformation, NAS, DSE, QD, OEE, and NQD search, LLM-assisted architecture design, code-agent mapping, simulation, benchmark trace, or source discovery.
Typical entry phrases:
Primary working object. The generated or discovered result on which the next architecture use would rely.
First useful move. Identify the result and the organization it concerns, then state its supported next use and any condition or limitation that affects reliance. Distinguish supported, refuted and unestablished claims about existing organization from proposals for a possible organization. Keep a proposal modal; add C.30 claim identity when the receiving use needs independent reidentification.
Keep a representation distinct from its represented object when correspondence or representation operations matter. Use C.30.AD or C.30.ASV for architecture descriptions or views, and C.29 for a remaining mathematical-representation choice, transfer, or reliance question. Publication detail enters only when availability or form changes the use.
Normal first result. A short sentence can suffice; reuse an adequate existing account. When separate lines help, use:
Stop there when another practitioner can make the intended next move safely. Only when result identity, branch evidence, or a receiving claim must be reidentified independently, extend those four facts into the optional StructuralSynthesisDiscoveryAdequacyNote@Project C.2.1 episteme:
The note retains the result, organization, next use and any material limit or return; its fields are included when the selected branch or receiving use needs them. Its EntityOfConcern is the result designated by resultRef, not the reference value, representation, publication occurrence, form, or carrier used to reach it.
When publication detail is relied on, publicationFormRef, publicationOccurrenceRef, and presentationCarrierRef refer to the form, occurrence, and carrier, respectively; omit every one that the receiving use does not need.
The note's ClaimGraph states the organization concerned, supported next use, and any material limitation or unresolved condition. A plausible mistaken inference is corrected when that changes reliance. Do not open or fill the dossier merely to prove completeness.
Here @Project is a compatibility and retrieval cue only. It establishes no project entity, composite-work identity, context, authority, viewpoint, or parthood.
When the note is used in one project, projectWorkOccurrenceRef identifies the composite U.Work and structuralSynthesisAdequacyNoteProjectUseRelationRef identifies the relation by which that project Work uses the note. The suffix or either reference alone establishes no project locality. The admission-note episteme, its result, and the composite project Work remain distinct.
When reliance depends on the result's generation or discovery history, retain the operation and the material source, input, performer and condition facts needed to establish that history and its bearing on the result. For an express claim about a U.Work, generationOrDiscoveryWorkRef names that independently admitted occurrence: its performers have A.13 cores and A.15.1 supplies Work admission.
Include the Method reference when the receiving claim depends on that Method. generationOrDiscoveryWorkAttributionRefs appear only for precise assignment-bound F.6 attribution. Missing or failed F.6 attribution alone leaves independently admitted Work intact; recheck an admission premise questioned by the same finding and any conclusion depending on it. The operation, Work, attribution, admission-note episteme, result, representation and publication objects retain their distinct identities.
actualTransformationRefs cite independently identified A.3.4 bounded changes. A positive link from Work to an actual transformation or produced entity uses the declared predicate governing that link and its supporting facts, including A.15.PROD when its branch is selected. Return missing-governor under A.6.RCD when no predicate governs the link; leave missing case facts unresolved and reject a link contradicted by them.
An entry in obtainingStructureRefs resolves to an A.22 U.Structure with independently identified constituents, obtaining relation occurrences, applied constraints and a named use frame. An incomplete basis for a claim about existing organization leaves that claim unestablished.
A proposal records a possible or intended organization. To say that this organization now obtains, identify the arrangement actually in use and the grounds for the claim. Supply modalArchitectureClaimRef when the proposed arrangement must be independently reidentified. A result, representation, publication item, graph, cluster or description supplies only the structural facts it establishes.
Adoption test. After using C.35, another practitioner can state the four-line minimum or an equivalent sentence: the result, the organization concerned and whether the claim about it is established, refuted or unestablished, or presents a proposal, the supported next use and any unresolved condition, material limitation or required return. That practitioner can also distinguish claim content, an obtaining A.22 structure, its representation, and a publication-side object when those distinctions affect use. Additional identity, branch, Work, bearer, publication, evaluation, or next-claim detail appears only when the receiving use relies on it.
What C.35 buys in practice. The practitioner can keep a useful generated or discovered result without treating it as authority for architecture claims. Architecture use attaches to the result; changing a rendering or file does not silently change the admitted claim, and admitting a carrier does not silently admit claim content.
Ordinary working move. Explain the organization concerned, its actual or proposed status, what remains unestablished, and the use the result supports. Stop when the next move and any material limit or return are clear. Claim an obtaining A.22 structure only when its four identity discriminators resolve. Preserve an unestablished claim about existing organization as unresolved and a proposal as modal. Add claim identity, representation, branch, Method, Work, bearer, publication or evaluation detail when the intended use depends on it; use C.29 for an unresolved mathematical-representation choice, transfer, or reliance question.
Not this pattern when. If the current question is how to search, choose, measure, decide, authorize, publish, govern a reusable generator, govern a cultural-evolution case, or run the work itself, use the pattern that defines or decides that question first, including C.36 for the cultural-evolution relation bundle. Use C.35 only when a generated or discovered result must be admitted or rejected before another architecture claim relies on it.
Problem
Modern architecture work receives claim-bearing outputs, representations, publication items, and other results from DSM clusters, MDM slices, MBSE queries, generated views, graph grammars, model transformations, LLM architecture proposals, AI-assisted ADD, code-agent relation graphs, NAS graphs, DSE traces, Pareto fronts, QD archives, benchmark traces, simulations, and source-corpus mining.
These results can expose candidate decompositions, relation gaps, hidden invariants, feasible search regions, trade-off points, source labels, or overlooked organization. They are not automatically obtaining A.22 structures, realized holon structures, eval results, evidence sufficiency, or decision authority.
C.35 handles the gap between result identity and architecture use. It first asks what the result is; which claim, represented object, or proposed organization it carries; whether representation or publication distinctions matter; and which architecture use is proposed. It then asks only the questions belonging to the result's branch:
- A transformation names its source and result objects, trace, preservation, and loss.
- A discovery names its observation or extraction basis, what is observed, inferred, or unknown, the covered and unexplored region, uncertainty, and validation.
- A generative proposal names its constraints, proposed organization or claim content, known omissions, and validation needs; identify the source of a constraint when its authority or edition matters to the recipient. A claim to preserve structure also needs an identified baseline.
Method, performed Work, attribution, bearer feasibility, and publication detail enter only when the receiving use relies on them.
Forces
Solution
Ask what decision the result can settle. One candidate checked against every applicable constraint can establish feasibility within the model. It may complete a commission to find any suitable design. A claim that no design qualifies requires grounds that cover every admissible alternative, such as an exhaustive search or a proof of incompatible constraints. A search that stops without a candidate can still tell the next designer which alternatives were examined. When one design uses less energy and another takes less time, both can be nondominated. To call a design nondominated, exclude alternatives that do at least as well on every stated criterion and better on one. State the set over which this was established. A claim that a design is best needs a ranking or trade-off rule for the commission and grounds that the design ranks at least as high as every admissible alternative. A verified bound may settle that comparison. If the leading alternatives remain incomparable, the requirement for a greatest design can have the answer that none exists. A commission to select one can leave the choice to the authorized recipient or require a further preference. A request for the nondominated set can be settled under the given criteria. A unique optimum also excludes ties for the best value. A unique feasible design excludes every other feasible design regardless of preference.
Judge the result and its production separately. A verified design can serve the commission whoever suggested it. To credit an agent with designing the organization, identify the decisions left open by the input, the method or generator, and the evidence that the agent made them. A fixed input that already specifies the organization can support an assessment of how the agent checked it. Claims about the installed organization depend on evidence about the installation. State any material limit or next check in the existing account when possible.
Use the smallest sufficient path:
-
Give or reuse the sufficient account. If it lets the receiver make the intended next move, stop.
-
When organization status affects use, distinguish an obtaining structure, a refuted or unestablished claim about existing organization, and a proposal. Apply A.22's four discriminators to a positive structure claim. Add C.30 claim identity when the receiving use needs independent reidentification. Distinguish the representation when correspondence or representation operations matter; use C.29 for a remaining mathematical-representation choice, transfer, or reliance question.
-
When the receiving use relies on how the result arose, select the branch for that claim. A complex result may need separate accounts for claims arising through different operations; do not infer one branch's basis from another.
- A transformation adds source and result objects, trace, preservation, and loss.
- A discovery adds observation or extraction basis, what is observed, what is inferred, what remains unknown, the covered and unexplored region, uncertainty, and validation.
- A generative proposal adds constraints, proposed organization content, known omissions, and validation needs; state the origin of any input on which the commissioning decision depends. For preservation, also identify the baseline and the relation to what is proposed.
-
When reliance depends on how the result arose, establish the operation and the history facts material to that use. Add the Method, independently admitted dated
U.Work, and production, discovery-use or Work-to-change relation only when those claims are being made; retain the premises and evidence each requires. -
Add bearer feasibility, realized structure, publication, archive or front policy, evaluation, measurement, decision, reusable-generator, or next-claim references only when the intended use relies on that claim. Use the pattern for that claim rather than copying its dossier into C.35.
-
Stop when the receiver knows the supported next move and any unresolved condition, material limit or return. Materialize the optional full note only when those facts or their relied-on branch details need an independently reidentifiable result.
Use these patterns when their questions arise:
Use concerning CGUS. When a result is useful because it describes, compresses, or demonstrates a constraint-governed unfolding organization, C.35 admits only that result for the declared architecture use. Cite an A.22.CGUS structure only when its positive A.22 and A.22.CGUS basis already obtains; retain an unestablished claim about an existing organization as unresolved, and a proposal as modal claim content.
The structure itself, when it exists, remains with A.22.CGUS, E.18.3, C.32.P2S, E.23, or another pattern defining the structure. If the encountered item is only a route card, narrative sequence, demonstrative slice, generated publication form, or presentation carrier, recover its claim-bearing result and represented object before making any positive structure claim.
When the encountered item is a narrative sequence, A.6.3.NAR governs only the selected-source carry-through, ordering and connective account, loss, reader use, and return.
Archetypal Grounding
C.35 is the pattern for admitting or rejecting a generated or discovered result before another architecture claim relies on it. The result may come from search, clustering, query, learning, transformation, simulation, or discovery. C.35 does not search, select, decide, or realize architecture. It asks what the result is, which organization it concerns, whether the claim about that organization is established, refuted or unestablished, or instead presents a proposal, what use it supports, and which conditions or limitations affect further reliance.
What a generated proposal establishes. An LLM returns proposal MedicalDeviceProposal-7 and diagram MedicalDeviceDiagram-7. An architect considering it for the candidate set records the following.
This is enough for the immediate candidate-use boundary. Add the generative-proposal branch only if the receiver relies on its constraints, omissions, or validation needs.
Treat MedicalDeviceDiagram-7 as a separate representation when graphical correspondence matters, using C.30.AD or C.30.ASV for its architecture description or view and C.29 when a mathematical-representation choice, transfer, or reliance question remains. A question about which requirements the designer used can arise in a new design. A claim to preserve the organization of an earlier design requires a baseline.
DSM and MDM clustering. A team groups the rows of a dependency matrix using co-change observations and interface data. Check whether the reported groups satisfy the grouping rule on that matrix. If the rule allows several groupings, the saved output identifies the one returned by this run. Reproduction also suffices when the procedure and its inputs determine the same choice, including any tie-breaking or random choices that mattered. To infer that the corresponding components perform one function, the team must relate the observed dependencies to that function and check what the matrix omits. A different decision is to reorganize the components into those groups. That proposal needs a reason the new allocation would serve the design. Use the discovery or proposal branch for the claim actually made.
Retain the matrix's mapping to the represented dependencies and the validity conditions of its operations; use C.29 when choosing, transferring, or relying on that representation remains unresolved. A comparison with an earlier modularization may add a declared baseline and preservation account, but clustering alone does not require one.
NAS result. A multi-objective neural architecture search returns a proposed neural organization and its graph. No evaluated alternative matched or improved on the proposal on every criterion and improved on it on at least one. Report that comparison over the evaluated set. To report the same conclusion over the full search space, establish coverage or a bound that excludes a dominating alternative. Keep the search constraints, hardware conditions and validation needed for deployment with the proposal.
The graph represents the proposal; C.29 applies when a separate mathematical-representation choice, transfer, or reliance question remains. The graph does not prove that the proposed relation occurrences obtain. C.35 does not invent a preserved-dataflow claim unless the run explicitly transforms or compares against a declared baseline.
C.32 may consume the proposal as candidate input; C.32.ACE handles evaluation results.
Graph grammar or model transformation. A graph-grammar Method is applied in dated generation Work and returns a claim-bearing result plus a graph representation for a product-line model. This is the transformation branch: C.35 names the Method, Work when performed-work reliance is current, source and result objects, rules, preserved interfaces, lost manufacturing constraints, and transformation trace.
If the resulting organization is only proposed, it remains modal content in an architecture claim and the graph remains a representation of that proposal, with C.29 used for an unresolved mathematical-representation choice, transfer, or reliance question. Source or result A.22 structures are cited only when their four discriminators independently resolve.
If the use additionally asserts an actual formal or world-side change, it cites the A.3.4 occurrence and the separately governed Work-to-change or A.15.PROD claim; otherwise model transformation remains the Method or operation-family label and no U.Transformation is inferred. C.34 may check preservation; C.32 may admit the proposal without actualizing it.
Bias-Annotation
Conformance checklist
Common Anti-Patterns and How to Avoid Them
Consequences
Positive consequences:
- Generated or discovered results can enter architecture work without becoming authority. Ordinary use can stop with a sufficient account; additional detail appears only for a receiving dependency.
- C.35 keeps the result-use return visible. If the result cannot support the next architecture claim, repair returns to that result, the organization it concerns and any unresolved or refuted claim about it, the missing condition, and the required rule or test; the return never turns the proposal into an A.22 member.
- C.32 continues to define candidate-palette admission. C.35 supplies the result-use boundary that C.32 may need.
- Search, query, transformation, and AI-assisted results become auditable without conflating claim-bearing content, represented objects, representations, publication items, and carriers.
- Reusable generator governance stays outside C.35 until opened, which prevents one-case result review from becoming a hidden method or mechanism-suite pattern.
Costs and trade-offs:
- Assessing a result takes attention. Reuse a sufficient account; obtain further detail when it can change the warranted next use.
- Some outputs will be useful but not yet admissible. Supply the missing branch-specific basis, bearer boundary, validation, or next claim plus its required rule; keep useful outputs and leave irrelevant preservation rows empty.
- The pattern is narrow. It does not choose among alternatives, manage archives, define eval programs, or authorize work.
Rationale
Architecture synthesis receives results from search, model transformation, LLM proposal, code-agent mapping, DSM modularization, NAS, simulation, benchmark and source discovery. A short account can identify the result, the organization it concerns and the supported next use, with any condition or limitation that affects reliance. The larger record is justified by a receiving use that depends on its additional distinctions.
The result can carry a proposal, describe an obtaining structure or serve as a representation. Its selected use determines which distinctions and supporting results matter: representation correspondence, feasibility, evaluation, publication or decision authority, for example. Obtain each required result under its own subject rule; retain a mathematical representation's mapping and validity conditions, using C.29 when a separate choice, transfer, or reliance question remains.
The source families explain why the branches differ. MBSE query practice, DSM and MDM work, and code-agent mapping expose discovery questions about extraction basis, observation, inference, unexplored regions, uncertainty, and validation. Graph grammars and model transformations expose the distinct need for source and result objects, trace, preservation, and loss. Multi-objective NAS and LLM-assisted design expose proposal questions about constraints, proposed organization, omissions, and validation without inventing a source structure. The GonzoML intake supplies neural operation vocabulary; resolve a source label when its meaning is unclear and changes the branch-specific claim or use.
SoTA-Echoing
Relations
- Builds on:
C.30,C.30.AD,C.30.ASV,A.22,C.32.P2S, andC.32. - Uses:
C.34when the transformation branch or an explicit baseline comparison must preserve selected source structure;C.33when capture and loss in the output are the current issue;C.29when choosing, transferring, or relying on a mathematical representation remains unresolved.
Coordinates with:
C.35 also coordinates with C.30.STRAT, C.30.TFS-REL, A.6.M, C.31, C.31.ASAP, C.32.ACS, C.32.ACE, C.16, C.25, G.5, C.18, C.19, E.18, C.32.PAD, and C.32.ADR.
- Boundary: C.35 governs admission of a generated or discovered result for architecture use before or around C.32 candidate admission. It does not build the candidate palette, select from alternatives, govern reusable generators, define eval programs, measure values, decide projects, supply evidence or assurance, authorize work, publish the result, or prove realization.
C.35:End
Cultural Evolution and Cultural-Evolution Engineering
Tech-name:
CulturalEvolutionEngineeringPlain-name: cultural evolution and cultural-evolution engineering Type: Conceptual and project-use pattern (C) Status: Stable Normativity: Normative unless marked informative
Problem frame
Use this pattern when the project question is about how a culture, style, tradition, discipline practice, method family, work family, canon, recognition regime, selection regime, or mediating system changes and can be deliberately influenced.
When changing participants, tools or questions puts a shared method out of reach, C.36.RP helps make its use and development obtainable: locate the missing contribution, arrange its use and retain the means of renewal.
Typical first-use situations:
- an engineering group treats its product family, toolchain, platform family, research program, or AI-agent framework as an evolving set of variants rather than one fixed system;
- a scientific, medical, pedagogical, engineering, music, dance, organizational, or AI-agent discipline is changing through related methods, work products, training forms, memory epistemes, recognition regimes, and selected variants;
- a music or dance steward needs to compare style, genre, technique, scene, canon, platform, or tradition labels without assuming that the label names one root kind;
- a project lead wants to influence the evolving practice—for example by changing how variants are generated, transmitted, recognized, selected, remembered, measured, or refreshed, or by changing a Method family, Work family, assignment, mediating architecture, or performed intervention.
What goes wrong if missed
The team treats culture as shared vocabulary, treats style as a genre tree, treats a platform as the cultural object, treats a QD archive as the decision, or treats one scalar popularity or quality score as cultural development. The project can then generate many variants but still lose the relations that make those variants transmissible, recognizable, selectable, retained, refreshed, or turned into work.
What this buys
The practitioner gets one small statement of the cultural relations now supported, what is changing or worth continuing, and the next action or stop. Add collective holons, local system-role kinds, classifications, assignments, Work and Method families, canon or memory epistemes, architectures, measurements, and refresh relations only when the claim needs them.
First useful move
State what someone needs to decide about the practice. A school choosing whether to keep an archive for its next class can use evidence that a teacher and learners retrieved a variant and performed it under the conditions that class will have. To credit a teacher with transmitting a specific change, establish what the learners received from that teacher and what they subsequently used. To estimate how widely the change spread, identify the eligible classes and how their use was observed. Use the smallest of these inquiries that answers the decision.
When a card helps retain or share the result, use:
@Context is part of the card's context-qualified retrieval name. CaseScopeOrModelUseBoundary names the project, discipline, scene, product-family, publication, or model-use boundary of this case. This boundary stops a local trend from becoming the whole culture merely by wording.
PublicationRefs is optional: when a publication distinction matters, name only the E.17 source-backed face or E.24.PUB publication occurrence, publication form, presentation carrier, audience-declaration episteme, bounded-use-declaration episteme, or availability claim needed by this case. The card does not require a complete publication record.
Actual access, reliance, use, and Work stay outside this field unless their own relations or occurrences are separately current.
Variants may be generated, retained, inherited, or observed. An archive or front claim still uses C.18 or C.19.
Expand the card only when later use needs more detail. Possible additions include:
- direct participation or position relations;
- local system-role kinds, separate System-classification judgments, assignment species and obtaining occurrences;
- Work and Method families;
- Method relation structures and descriptions;
- canon or memory epistemes;
- recognition and selection regimes;
- mediation systems or architectures;
- characteristic spaces;
- style or tradition term rows;
- publication relations;
- measurement; and
- refresh.
Each addition identifies its own object or obtaining relation; the card creates none of them.
The card is optional and keeps the case's claims together; their definition and testing remain with the patterns for those claims.
Working scope
Many current projects no longer develop one isolated object. They shape, for example, product families, methods, research directions, medical and pedagogical practices, AI-agent frameworks, artistic styles, engineering traditions, canons, archives, frontiers, and recognition regimes as these evolve. The project often generates variants cheaply, while the hard work shifts to the relations that determine what is produced, recognized, retained, selected, used, changed, or kept current. That work can include, for example, problem production, characterization, archive stewardship, comparison, selected-set result declaration, actual publication, local choice, performed Work, effect measurement, and refresh.
Cultural evolution is current when the question is how a collective or discipline generates, transmits, recognizes, selects, retains, or changes variants. Memory or canon epistemes, recognition and selection relations, comparison, platform or algorithmic mediation, and changing Method families may all matter.
When the case says that Work was performed, recover each performer through A.13 and let A.15.1 independently admit the dated Work occurrence and enacted Method.
Include A.2.1 assignment refs and F.6 when the case or receiving use represents precise assignment-bound attribution. The A.13 assignment basis remains required for actual Work even when those fields are omitted; missing or failed F.6 attribution alone leaves independently admitted Work intact while its independent A.13/A.15.1 grounds still hold. A local system-role kind, classification judgment, assignment species, assignment occurrence, Work occurrence, Method, effect claim, responsibility relation, and family description remain separate.
This pattern provides a first-use cultural-evolution case. The same pattern can serve engineering product families, scientific research programs, medical disciplines, pedagogy, music styles, dance styles, organizational cultures, and AI-agent framework evolution because it begins with existing FPF objects and relations rather than domain labels.
Problem
Culture, style, tradition, genre, scene, practice, platform, regime, technique, and developmental-machinery wording is useful in source and project prose, but one label can refer to different objects or relations. For example, it may name:
- a method family or method relation structure;
- a work family or family of performed works;
- a local system-role kind, its classification judgment, or a separately obtaining system-role-assignment occurrence;
- a discipline or collective holon;
- a canon or memory episteme;
- a recognition, selection, measurement, or visibility relation;
- a mediation system, product architecture, platform architecture, or algorithmic mediator;
- an archive, front, current pool, selected set, lineage, or edition set;
- a publication label or cross-context term bridge.
If the project accepts the word as ontology, FPF grows a second ontology beside method, Work, system-role kind and assignment, discipline, episteme, architecture, selection, publication, and refresh. If the project hides the case as an example inside open-ended search, the cultural-evolution question becomes invisible and the first useful move is lost.
Forces
Solution
First state the cultural-evolution case in ordinary language: which practice and variants are in play, which relations transmit, recognize, select, retain, or mediate them, what the available basis supports, and whether to continue, change, narrow, or stop. Keep the population, period, uncertainty, and receiving-use limits that matter to this conclusion. Then use the applicable FPF pattern only for a claim whose definition or test matters.
An admitted System may perform dated Work, and that Work may enact a Method. Work and Method families may organize comparison. Variants may be preserved, transmitted, selected, suppressed, or refreshed through canon or memory epistemes, recognition and selection relations, mediation systems or architectures, measurement or visibility relations, and publication forms.
These are separate facts. For every claimed Work occurrence, recover each performer through A.13 and let A.15.1 independently admit the Work. Add assignment refs and F.6 when precise assignment-bound attribution is represented; this does not make the A.13 assignment basis optional for actual Work. Keep families, assignments, Methods, epistemes, and selected structures distinct from the performers and Work.
Cultural-evolution engineering proposes or performs a deliberate change to one or more of these relations. Proposal, performed Work, actual transformation, measured effect, responsibility, authority, selected structure, and publication are different claims. Assert each as obtaining only when its own predicate is satisfied.
Ask what would count as the change for the decision at hand. A decision to adopt a method may already be enough to schedule its introduction. A first execution can settle whether someone used it on that occasion. To entrust another group with the work, establish what the recipients can do with the tools, instruction and help they will actually have. If the practice includes an adviser or a shared tool, successful use with that support can meet the commission. A claim about performing without that help calls for evidence under the changed condition.
To trace transmission, identify the operation or choice that a recipient retains, the source they obtained it from, and its use in their own work. Decide what may vary across performances while preserving that operation. To establish transmission from a particular source, identify something that links the receiving work to that source. One well-documented handoff can suffice. Similar performances can also come from separate invention or a common third source. A work record can identify an immediate teacher, peer or publication while leaving the first inventor unknown. Seek that earlier history when the claim depends on it. Evidence of use may already suffice to continue the practice.
For a claim about a group, specify whom or what you are counting, over what period and what qualifies as use. Count plays as play events. To count people or teams reached, count distinct recipients. For a count of users, establish which recipients performed the relevant work. If estimating a proportion from a sample, explain how the sampled work represents the population and how much uncertainty matters to the decision. Date the work as well as the arrival of its evidence. A growing file of confirmations may document earlier uses. When a network expands, the number of users can rise while the proportion using the method stays the same. Give the counts and their populations. To establish a change among the original participants, follow those participants over comparable opportunities to use the method.
For a target stated as a minimum, count the confirmed qualifying cases. Reaching the minimum settles that criterion. To decide whether it has been missed, include unreported cases that could have qualified during the period. For a deadline still ahead, include uses that could still take place. If even an upper bound falls short, the target is out of reach. Where several groups need the same remaining teacher or session, check which uses can take place together before promising the target can still be met. If the evidence settles neither outcome, report the confirmed count and what could change the decision.
A causal claim about an intervention needs grounds that distinguish its effect from other changes that would produce the same observations. Use C.28 for that inference. A project with the authority and a sufficient practical reason can decide its own next action while a claim about the wider population remains open.
When the question is how the practice may develop, retain credible hypotheses whose differences could change the continuation, forecast, or warranted use. State the uncertainty that matters; a fixed number of hypotheses is not a completion condition. Use B.5 and B.5.2 for a needed hypothesis account and its consequences. A bounded forecast or continuation may be complete on the available basis while a stronger explanation remains unresolved.
Select a new observation or intervention when its attainable result can improve that use enough to warrant its whole cost, delay, participant burden, and displaced work. Include the effort of designing and describing a proposed study, before treating it as an available option.
Use C.11.DUA when that connection between a demanded inquiry and its receiving value needs recovery, and C.11 when a named deciding System and an already formed bounded choice exist.
Use the DPF or field Method that governs the selected inquiry to design and perform it; C.36 is not a general research-budget method. An unavailable or unhelpful probe withholds only the claim that needs it, not an independently supported current account.
Use A.3.3 when the claim states a state space and transition law, C.28 when the current use relies on a causal claim, and A.15.7 for the next action during ongoing Work. Use A.10, C.16, and C.27 for the evidence, measurement, and time questions that arise.
Use only the smallest form the task needs:
CulturalEvolutionCaseCard@Contextkeeps a multi-relation case together;StyleTraditionTermBridgeTable@Contextkeeps a familiar local label connected to the recovered FPF value or relation;CulturalEvolutionInterventionCard@Projectretains an intervention when proposal, Work, effect, or later comparison needs explicit identity.
These forms assemble existing FPF values; cultural labels retain the object or relation meaning recovered for this use.
Style And Tradition Term Bridge
Retain a clear source-local meaning. When ambiguity changes the cultural claim, state which object or relation the label names. Use F.17 when a durable term account is needed, F.18 when a durable name needs restoration, and F.9 when cross-context substitution or claim transport requires a Bridge. The following optional table can retain those results:
The table records the term meaning and any independently established Bridge used by the case. It need not introduce a new name, term row or Bridge when an existing interpretation suffices.
For music and dance, a label such as prog, post-prog, contemporary, hip-hop, battle, TikTok dance, canon, school, or technique may point to different FPF values in different contexts. Resolve the meaning needed for this use before relying on an ambiguous label.
When the claim concerns a retained style, name what varies and what the comparison retains. Depicting a new scene in a retained visual style changes the represented content; depicting the same scene in different styles changes the style. These are different preservation relations. Say which works or variants are compared, by what relevant features, and for which use.
A stylistic similarity alone does not establish that an audience recognizes or likes the work, that a performer can reproduce it, or that its method has been transmitted; those conclusions need the corresponding recognition, response, performance or transmission relation.
Intervention Card
Work back from the change sought in the practice. Who would have to do what differently? Trace what participants can select, teach, adapt, use or retain. If they already receive it, establish what could change its use. Choose a relation the project can influence and explain how the change could reach the intended work. When the change depends on several contributions, explain how their arrangement and conditions can produce the required result together. Where one participant receives another’s contribution, trace what was preserved, changed or combined. Independent choices can also produce a common pattern under shared conditions. Identity of a variant matters where the claim requires it. An existing arrangement can already support the intended use.
Use an intervention card when the project needs to identify a proposed or performed intervention later. State the action, the relation expected to change and what would warrant continuing, revising or stopping. A festival might offer earlier jury feedback so instructors can use it in the next syllabus revision. An approved syllabus revision establishes which change the editor adopted. A draft produced from the feedback establishes what the proposed text says. To report adoption, identify the editorial action or the rule that put the revision into effect. To establish that classes used the selected variants, examine the teaching and work of those classes. To attribute a difference in use to earlier feedback, examine what else changed and what grounds the comparison. Use A.15.2 for the plan and A.3.4 for a claimed actual change.
Keep proposal and performance separate. The full card below is an assurance expansion, not a first-use form.
Open its Work, assignment, transformation, effect, architecture, and publication fields only when those identities matter. AffectedMediationSystemOrArchitectureRefs names actual mediating Systems or architectures only. Publication refs name only the objects needed by the intervention; omit them otherwise. Actual access, reliance, use, and Work stay outside this field unless separately at issue.
If actual performance is claimed, recover each performer through A.13 and let A.15.1 independently admit the U.Work. Add assignment refs and F.6 when the card or receiving use represents precise assignment-bound attribution. Required A.13 assignments still underlie actual Work; missing or failed F.6 attribution alone leaves independently admitted Work intact while its independent A.13/A.15.1 grounds still hold.
Add actual change and a Work-to-change relation only when each independently obtains. An effect can obtain without manufacturing a performer, assignment, or Work.
Recover unresolved claim-bearing role wording through E.10.ROLE; a local system-role kind and classification judgment remain independently optional.
@Project is part of the card's retrieval name. It establishes no project entity, composite-work identity, context, authority, viewpoint, or parthood.
When the card is used in a project, ProjectWorkOccurrenceRef identifies the composite U.Work, and InterventionCardProjectUseRelationRef identifies the relation by which that Work uses the card. The suffix or either reference alone establishes no project locality. The proposed intervention, card, and project Work remain separate.
Use the expanded identity fields only when a later claim or comparison needs them. For performed intervention Work, recover each performer through A.13 and let PerformedInterventionWorkRef name an independently admitted A.15.1 U.Work. PerformedInterventionWorkAttributionRefs, assignment-species refs and assignment-occurrence refs are optional fields used for precise assignment-bound attribution through the same obtaining A.13 assignment. The assignment basis remains required for actual Work when those fields are omitted.
A proposal omits Work and attribution fields. A local system-role kind and classification judgment remain optional and separate. Assignment establishes no classification, Work, capability, functioning, authority, or responsibility.
Responsibility and change. A positive responsibility claim needs an admitted domain predicate through TargetedRelation or EffectClaimOrRelationRefs; otherwise return A.6.RCD's missing-governor. ActualTransformationRefs may cite only changes independently identified under A.3.4.
Flow representation. TransformationFlowStructureRef may cite an E.18 transformation-flow structure selected under A.22. Membership or adjacency in that structure proves neither actual change nor a Work-to-change link.
Work-to-change. A positive link from intervention Work to an actual transformation or effect needs a direct predicate that obtains for those participants, an A.6.1 application binding when that declaration supplies the link, or an admitted A.6.RCD local claim. If none applies, return the reason-specific non-assertability result.
Effects and production. A.15.PROD answers only its production-work, entity-inception, or completion question; it does not supply the Work-to-change link. An effect that does not require Work stays on its own relation. Observing a value alone does not prove the effect.
The intervention card keeps the proposed intervention, targeted relation, and next applicable pattern together; it neither authorizes Work nor establishes that a targeted effect obtains.
For planning and performance, use E.18.1 for P2W carry-through, A.15.2 for work planning, A.13 and A.15.1 for actual performers and independently admitted Work, and A.2.1/F.6 only when precise assignment-bound attribution is expressly consumed. Use A.3.4 for actual change. A.15.PROD may answer one production-work, entity-inception, or completion question; the Work-to-change link still uses the direct predicate, A.6.1 binding, A.6.RCD local claim, or non-assertability result above.
For the next question, use:
Evolution Sense Split
When generic development or evolution wording still hides the changed or represented subject, needed continuity or membership, posture, direction or value basis, or direct owner, enter through E.10.DEV before this split. A recovered cultural-population or discipline-facing variant claim may continue here.
For a non-cultural population or lineage, use the domain-account branch of E.10.DEV: a sufficient supported account can finish the use; an architecture gap remains only when that FPF use needs a definition or relation the available account does not supply.
Open E.10.MOVE afterward only when a separately relied-on trajectory, route, path, ordering, posture, or representation ambiguity remains.
Then use this cultural split:
An engineering development loop may use C.36, but it does not automatically become cultural evolution. Use C.36 for that loop only when the question concerns the collective-holon or discipline-facing cultural-evolution relations above.
Platform, Regime, And Attractor Wording
Use clear domain meanings. Recover the object or relation when platform, regime or attractor wording leaves an action-changing ambiguity.
-
Platform, recommendation environment, visibility infrastructure, algorithmic mediator, or platform-regime wording may name a System, a System classified under a local system-role kind, another relation participant, a system or product architecture, recognition or selection relation, measurement or visibility relation, publication relation, model-use boundary, project scope, or source-currentness relation.
-
Measurement regime wording may name a characteristic space, measurement relation, visibility relation, publication relation, dashboard relation, source-currentness relation, or comparison setup.
-
For an attractor, basin, stable-dynamics, state-transition-law, or other mathematical-model claim, recover the model and the conditions supporting that claim. Use
A.3.3for the state space and transition law,C.27for a temporal-adequacy question, andC.29when a separate representation, transfer, or reliance question remains.A clear style metaphor can remain ordinary wording; use
F.17,F.18orF.9only for the unresolved term, naming or Bridge question.
Archetypal Grounding
For example, consider a dance school deciding how a festival’s feedback could change what students practise. Earlier feedback gives the editor an opportunity to change the syllabus. Instructors must then have access to the adopted revision in time to use it. What students practise depends on the tasks the instructors actually set. Each link suggests a different intervention if the preceding link already works.
The engineering-product-family and music-and-dance slices below show how unlike projects recover Methods, Work, variants, memory epistemes, recognition or selection relations, mediating Systems, and refresh without inventing one root culture kind.
The AI-agent slice separates a project's benchmark choice from performed change and from later evidence that a wider practice generated, copied, recognized, selected, retained, or lost variants. The neighboring-boundary table then gives reduced non-use cases for claims that stay with their direct patterns.
Engineering Product Family
An engineering lead has an archive of candidate cooling-module designs, a Q-front over energy use and maintainability, competitor product families, and a roadmap pressure to keep more than one line current. The first C.36 question is not "which module is best?" but whether the project is shaping a product-family culture: shared methods, work products, review criteria, memory epistemes, local system-role kinds and any obtaining assignments needed for Work attribution, architecture-candidate generation, selection regimes, and refresh rhythm.
If the question is only archive or front treatment, use C.18 and C.19. If the team is changing how the engineering organization generates, recognizes, retains, compares, and learns from module variants, explain those relations in a sufficient account, using the optional case card if helpful. Use E.18.1 when problem-to-work carry-through is needed.
When that work yields a generated or discovered result intended for architecture use, use C.35's one-sentence first result to state its supported use and any remaining condition before C.32 candidate synthesis; the cultural-evolution case remains in C.36.
Music And Dance Style Engineering
A dance community uses the same label for a battle practice, a theater style, a short-video platform format, a pedagogy, and a canon. C.36 first resolves the meaning needed for this case. A retained term row can state:
The term row identifies meaning. To consider a change to the style ecology, recover the relevant cultural relations. An expanded case card can show them:
Suppose the festival needs six of its eight partner classes to teach the variant before it commissions a shared rehearsal. The timetable lists all eight classes, the archive records 90 plays of the clip, and judges scored the festival performance highly. These records can support scheduling, a report on recorded plays and an account of jury recognition. To make the rehearsal decision, count distinct partner classes in which the variant was taught during the agreed period. If six class records document it, the festival can commission the rehearsal on that criterion. Two confirmations and six outstanding reports justify saying that at least two classes taught it. After the period closes, suppose three classes confirmed teaching and three established that they taught no session using the variant. Even if both remaining classes taught it, the total could reach only five. The target was missed. If the period is still open, check which classes have time to teach and which share an instructor or a session before committing the festival. To credit the clip with supplying a change, look for its use in teaching or rehearsal.
The project's next move may concern the archive (C.18), the pool of variants (C.19), a declared selection (G.5), teaching, recognition or platform mediation. Use E.24.PUB for a publication occurrence or audience availability and E.17 for a source-backed projection.
If the case also claims a new level, new holon, model-use or scope reframe, feedback-down relation, whole reidentification, cross-scope frustration residual, or interlevel ethical conflict, keep the C.36 result and test the additional claim separately.
For example, use B.2 or B.2.P for MHT and whole reidentification; A.1 or the applicable System or holon pattern for kind and boundary claims; B.2.5 for an obtaining supervisor-subholon feedback relation; C.30.ILC and C.29 for cross-scope architecture residual or mathematical-lens use; and D.2, D.3, or D.4 for value, harm, responsibility, or admissible sacrifice across levels.
AI-Agent Framework Culture
A team sees the same agent patterns rising in its benchmark rankings and appearing in more community projects. The team can report the observed co-occurrence within the rankings and projects actually examined. The rankings could have influenced the projects. A change in the work teams undertake, a common template, or a change in which projects become visible could also produce this record.
Suppose the benchmark rewards passing fixed examples and the team's own commission requires agents to adapt to changed inputs. The team can change its evaluation to test that commission directly. To claim that the changed dashboard altered other teams' choices, establish which teams could see it, what choices they made, and what comparison or other causal evidence distinguishes dashboard influence from the alternatives. A repository history can already establish one team's use of a particular template. It can leave open why the template became more frequent across the community. Record the project decision with C.11, and examine transmission, selection or retention across the practice here when that wider conclusion matters.
Neighbor boundaries
Continue the Supported Practice or Select a Useful Trial
In a constructed dance-school case, the same teachers and learners have repeatedly used the current archive examples in the stated class, and the available teaching observations support that bounded transmission claim. No participant, access, or protection condition has changed.
Continue the arrangement for that class. The teachers' remaining time is needed to run it; an additional comparison would displace useful teaching without changing the present continuation. The account need not invent a rival explanation or a study, and it makes no claim about other schools or long-term retention.
Now suppose the next class must learn from the archive without teacher assistance. The current observations do not support that receiving condition. The school has an authorized preparation session in which the learners can try a representative task, access and support are available, and its result will determine whether the archive-only arrangement can be used or teaching support must remain.
Select that bounded trial through the teaching Method. This completes the recommendation to test, not the trial or a transmission claim for the new condition. A later tested-change claim must use the actual receiving Work and observations; causal explanation remains a separate question.
Bias-Annotation
Scope: Limited to cultural-evolution questions about variants and their generation, transmission, recognition, selection, retention, loss, mediation, and deliberate influence across one named practice, population, collective, or Discipline boundary. C.36 is not a universal culture ontology, a project-authority rule, or a claim that every evolving engineering object is cultural evolution.
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Consequences
Positive consequences:
- practitioners can identify the cultural-evolution question and finish a qualified current account or warranted continuation, selecting a new inquiry only when it contributes enough to the receiving use;
- style, tradition, practice, platform, regime, and technique labels remain usable without becoming root kinds;
- engineering development loops, cultural-evolution cases, archive and front relations, selected-set result declaration, publication, and refresh stay distinct; practitioners apply the definition and test for each claim they make;
- music, dance, science, medicine, pedagogy, organization, product-family, and AI-agent cases can share one FPF modeling line.
Costs:
- first use must name more than one value; a cultural-evolution case is not captured by one label;
- projects must decide whether their question is variant-set generation and retention, cultural-evolution structure, architecture work, local choice, or refresh;
- durable style and tradition terminology may need F.17 term accounts; cross-context substitution or claim transport uses a Bridge when required.
Rationale
C.36 keeps a complex practical situation usable by naming a small bundle of existing FPF objects and relations instead of minting a root kind for every source word. This preserves the gain from cultural-evolution and open-ended-engineering sources while leaving Method, Work, system-role kind and assignment, discipline, episteme, selection, architecture, publication, and refresh claims with the patterns that define or test them.
SoTA-Echoing
Source-use currentness. One row's adopted move, rejected overread, and named field or boundary form the smallest source-use decision and stay current only until its cited edition changes, a materially newer cultural-evolution or algorithmic-mediation result challenges that transfer, the current QD/OEE line changes its archive, front, pool, selection, or refresh account, or a directly consumed FPF interface changes. At that trigger, recheck only the affected row and field or boundary; revise or withdraw an unsupported transfer and leave unrelated rows current.
Relations
Builds on: A.1, A.2.1, A.3.1, A.3.2, A.3.4, A.15, A.15.1, A.15.6, A.15.PROD, A.22, C.18, C.19, C.20, C.23, E.18, E.18.1, F.6, F.9, F.17, F.18, G.5, and G.11.
Coordinates with: E.10.DEV for generic development or evolution wording before the cultural case is known, E.10.MOVE for a remaining independent trajectory or path ambiguity, C.36.P for cultural-evolution wording repair, A.3.3, A.6.1, A.6.RCD, C.11, C.11.DUA, C.16, C.27, C.29, C.30, C.30.AD, C.30.ASV, C.32, and C.35.
C.36:End
Select and Use Representations for One Action
Type: Method pattern Status: Stable Normativity: Normative unless marked informative
Problem frame
Plain name. Selecting and using representations for one action.
Use this when. One person, team, organization, or other consuming System must take one action or make one decision, and several diagrams, tables, models, records, plans, descriptions, views, notations, or other results may each support only part of that use.
This is the one-use selection branch of representation work. Decide how the receiver may use independently governed candidate results for that action. C.37 does not decide what those results are, whether their subject-side claims obtain, whether an episteme is a view, whether a graph is mathematically admitted, whether evidence may be relied on, or whether the receiving action is authorized. Those answers remain with the patterns that govern those claims.
Primary working reader. A practitioner who has more than one plausible way of seeing or carrying information into an action and needs a selection without building a universal representation taxonomy.
First useful move. Name the receiver and what they must decide or do. Which fact permits the next action or rules an alternative out? Find a source for that fact. A second source may resolve a question the first leaves open or supply corroboration required by the governing rule.
First useful result. One complete use-bounded representation-selection account: one or more candidate claims with their material loss, basis and receiving action supported, declined or unresolved. It can be the immediate working answer; the row form and persistent record are not required when no later use needs them. Its use boundary is <receiving System, action or decision>.
If retained as a standalone claim-bearing object, it has ordinary C.2.1 identity, an EntityOfConcern and effective reference scheme; the use boundary does not replace episteme identity.
What goes wrong if missed. A readable diagram is treated as a conforming view; provenance is treated as approval; an evidence classification is treated as permission; several adjacent results are treated as one coherent structure; or a choice made for one action silently travels into another action with different loss, evidence, and decision conditions.
What this buys. The practitioner can say which candidate is selected, declined, or unresolved for one use; which claim is being carried; what remains hidden or transformed; what result supports the use; and what change requires return or reconsideration.
Not this pattern when. Use the pattern governing the candidate and stop when it already returns the complete one-result/one-use selection and limits. Use E.17.0 for view conformance, C.29 for a mathematical-lens use, A.6.3.RT for changing representation while preserving content, A.22 for selected structure, C.13 for a construction or collection question, E.24.PUB for publication, and C.2.P.DR for declarative-representation overread. Use C.37 only when the receiving action still needs the cross-kind selection account after those results are available.
Problem
Representations are useful because they foreground different things. A workflow diagram may expose order while hiding effort. A work plan may expose intended timing while saying nothing about actual performance. A work record may expose an observed breakdown while saying nothing about whether a proposed change will repair it. A graph may support traversal or calculation while omitting distinctions needed by a receiving decision.
The practical question is therefore not “which representation is best?” It is “which candidate result may this receiver use for this action, for which claim, under which limits, and what result makes that use available?”
Five recurrent shortcuts substitute other facts for the grounds the use requires:
- Object shortcut. A label such as diagram, view, model, graph, or record substitutes for the direct result that identifies the candidate and its subject-side claim.
- Classification shortcut. An A.2.4 intended first evidence-use classification is treated as evidence sufficiency or permission.
- Provenance shortcut. A source path, current carrier, or authentic publication is treated as a positive
RelianceDispositionor receiving result. - Decision shortcut. A selected row is treated as the choice, authorization, permission, or gate result without the direct receiving pattern.
- Composition shortcut. Several rows are treated as a collection, structure, integrated view, world model, or graph merely because one receiver reads them together.
Forces
Solution
Use one action spine:
- name the receiving System and action or decision;
- recover each candidate and its direct subject result;
- separate the direct subject result, optional first-use classification, bounded reliance, receiving conditions and result when needed, and auxiliary facts;
- state what the candidate exposes or preserves and what it withholds, loses, transforms, or leaves uncertain;
- mark the row
select,decline, orunresolvedfor the named use and give its return trigger; - for representations used in one action, connect the claims the receiver will rely on and establish that they concern compatible subjects and conditions.
Local mantra. Who must decide or act? Which fact settles what they can do? Where does that fact come from? What remains open? Which change would make us decide again?
The mantra is a recall aid, not a decision rule. The receiving pattern still emits the choice, gate, permission, authorization, or domain result.
Fix the receiving use before inspecting candidates
Write one sentence:
Recover the judgement or construction the receiver owes. For a decision under a given rule, start with that rule. Where the commission is to develop the rule, identify the possible cases it must distinguish and why they matter to the receiver. State which alternatives it admits, rules out or leaves for comparison. Connect each fact you use to the part of the rule it satisfies. A source can reliably establish a fact that explains why someone proposed an option. The decision may still turn entirely on another fact. Apply the rule with the first fact unknown or changed, keeping other grounds fixed where possible and varying linked facts together where the conditions require it. If the answer could change, identify the check that would settle it. If the answer holds across the remaining possibilities, you can give it now.
A second source may independently support the same premise. Losing one source then leaves the other available. If the governing rule requires two independent observations or a particular authorization, that condition is itself a premise. Where it matters for action or review, say which sources supported the decision, which the rule required and which answer a different question.
Every row in an account concerns the same receiver and action. A diagram used to select a proposed Method edition and the same diagram used later to tailor that Method have different use accounts. Reuse the supported facts that carry over and reconsider the ones on which the new action depends.
If one direct pattern already returns the complete representation–operation choice and limits for this use, take that direct exit. Do not add C.37 merely to rename its result.
Recover each row through five separate layers
Open only the layers required by the attempted use.
If a condition required for the candidate's stated use is negative, missing or unresolved, decline that use or mark it unresolved and name the missing fact or direct result. Preserve independently supported premises.
Use ready qualified inputs in the receiving comparison or construction; an unfinished choice or task result does not by itself create a missing input-admission result. Any action that depends on that later result still waits for its grounds.
State exposure, loss, and the row disposition
For each candidate, record what the receiver can establish with it. Keep the facts that affect the choice, its grounds or the permission to act.
- what it exposes, foregrounds, or preserves;
- what it withholds, omits, loses, transforms, or leaves uncertain;
- the claim for which it is selected or declined;
- the direct result and reliance basis that bound that claim, including an A.10 disposition when that account is needed;
- the condition that sends the reader back to the source or reopens selection.
Use exactly these ordinary row dispositions:
Selection is use-bounded. It does not make the candidate true, complete, current, published, conforming, relied on, assured, or authorized outside the claim and action stated in the row.
Use the smallest complete account
Use this shape when the result must be retained:
C.37 may use a single row when it still adds a needed layer separation. If the pattern governing the candidate already supplies the same complete one-result/one-use answer, use that result instead.
This is a logical claim group, not a universal record kind, U.Representation, RepresentationOf relation, taxonomy, manifest, view family, collection, or structure. Do not give it another schema merely because several domains use the same questions.
Realize the result once
First determine whether a later receiving use needs the selection basis retained. An immediate sufficient selection can finish with its grounds, material loss and receiving conditions, without a new record.
When retention is needed:
- If an owning domain result already carries the same receiving use, embed the complete needed claim group and action boundary there.
- Otherwise retain that basis as one ordinary C.2.1 episteme.
- Create no embedded and standalone duplicate for the same use.
A later reader who must reconstruct the selection receives its complete basis; a bare verdict or locator is insufficient for that use.
Retention preserves the needed separation: direct subject result, optional A.2.4 classification, reliance basis and any required A.10 account, receiving conditions and result when needed, exposure and loss, disposition and return. An immediate answer preserves the same substantive limits without instantiating every form field. A cross-use ensemble may later relate accounts under its own direct pattern; C.37 does not perform that organization.
Keep co-use local to one action
A receiver co-uses representations when they rely on claims from two or more of them for one action. Show how those claims work together. They can establish different premises or independently corroborate the same one. A list of documents that the receiver consulted leaves this question unanswered. Each source keeps its own limits and the conditions under which it must be checked again.
Co-use does not establish:
- one collection or selected structure;
- one multi-view family or mutual conformance;
- one integrated model or coherent world account;
- one constructional whole or composition relation;
- one shared representation language, graph, or correspondence;
- one assurance result or authorization.
For an additional claim, open C.13, A.22, E.17.0, C.29, a domain integration pattern, or another governing pattern only when the receiving action depends on that claim.
Use C.37.MM when the decision requires an inference across models or a successor must recover their governing conventions and the methods by which their results were obtained. A straightforward use of several sources can end here.
Recognition, reliance, assurance, and action remain separate
Recognition asks what the candidate is and which direct result or relation obtains. A.2.4 may add its first evidence-use or status-use classification. Relied-on claims need the source, support and use conditions that warrant them; A.10 develops the source-to-use account when it remains implicit or must be replayable.
Use B.3 only when the question concerns a named assurance claim; consequence or reuse alone does not require an assurance package. The receiving pattern owns the action result.
Use an already adequate source-to-use account. A new A.10 path or disposition is unnecessary when the claim and its warranted use are already explicit and no further replayable account is needed. A source gap, stale consumed premise or unsupported use still blocks the conclusion that depends on it; narrow the use, obtain the required support, or stop. Reuse an applicable A.10 disposition when one already supplies this boundary.
Direct exits and boundary cases
Archetypal Grounding
Qualified premises for a bounded-trial decision
MethodEngineer-ME1 must choose between keeping the current Method edition for Work item WP4 and preparing and running a one-hour trial of MethodChange-MC7. The commissioning rule admits a proposal only when its estimated total time fits a three-hour budget. The current edition fits. MC7 requires an estimated four hours of preparation before the one-hour trial. Both figures charge the same time budget.
Under C.11 the result CR7 is choose now: retain the current edition for WP4. The preparation estimate alone exceeds the ceiling. The planned hour gives the fuller estimate of five hours. This calculation settles eligibility under the stated planning rule. How long preparation would actually take remains unmeasured.
Raise the ceiling to five hours with the estimates unchanged. MC7 now meets the budget rule. If the engineer was asked only whether the trial fits, they can answer yes on the estimates. Choosing it over the current edition would require a reason to prefer it among the eligible options. To guarantee completion by the deadline, establish that every admissible duration meets it. To rule out timely completion, establish that every admissible duration exceeds it. A reliable upper bound at or below the deadline establishes the first claim. A lower bound strictly above it establishes the second. A five-hour upper bound against a three-hour limit leaves open the possibility of finishing sooner. An estimate alone settles neither claim.
Failed diagram use
A release team receives a polished architecture diagram and wants to authorize deployment. E.24.PUB establishes that the diagram edition is available through a current carrier. C.2.P.DR repairs one route-shaped arrow that had been read as operational authority. Neither result establishes view conformance, a representation correspondence, runtime structure, evidence reliance, or deployment permission.
Until the needed direct subject result, reliance basis, and permission or gate result are available, the row is unresolved. Use A.10 when the source-to-use account remains implicit or must be replayable. Visual polish and provenance cannot upgrade it.
If the direct release gate or permission pattern instead returns a negative result because its required basis is absent, the row is decline; classification, publication, provenance, and repair facts cannot override that direct result.
One decision now and a later reader's need
The engineer can finish the current comparison by giving the planning rule, the estimates and the resulting choice. The source of the four-hour estimate matters even though the workflow diagram and the earlier work record have identifiable editions. Neither supplies that estimate.
A later reviewer may need to reconstruct the choice. Give them the estimates used, what each covered, the budget in force and the alternatives compared. If they also need to reconstruct why MC7 was proposed, give them the earlier rework record and its comparability limit. Keep this basis once in the existing decision account when it can serve those uses.
Bias-Annotation (informative)
Conformance Checklist
Common Anti-Patterns and How to Avoid Them
Consequences
The gain is a small, repeatable bridge from heterogeneous results to one practical action. A practitioner sees why each candidate may be used, what it cannot support, and where the decision must return when sources, conditions, or reliance change. Domain results remain authoritative and can embed the claim group without duplicating it.
The cost is that some attractive candidates remain declined or unresolved because publication, provenance, classification, or visual form cannot supply a missing required result. That cost is preferable to an account that looks integrated while borrowing warrant across incompatible layers.
Rationale
The receiving use is the smallest stable boundary shared across domains. Representation kinds, correspondence relations, view predicates, plan claims, Work records, mathematical objects, and decision results do not converge on one ontology, but practitioners repeatedly need the same action sequence over them: recover the direct result, state the relied-on claim and loss, establish its warranted use and required entry conditions, and select, decline, or stop. Selected inputs contribute to the receiving operation; obtain its action result through the receiving Method.
Co-use is chosen instead of composition because the rows need not form a new whole. The same receiver may use them together while every candidate and relation retains its own identity, predicate, and return condition.
SoTA-Echoing (informative)
Reopen this source use when newer representation, provenance, view, or decision practice supplies a lower-effort way to keep the same direct-result and receiving-action boundaries, or when interoperability requires a common standalone account kind that cannot preserve them through ordinary C.2.1 identity and direct references.
Relations
C.37.MP relates several completed use-bounded selections when a cross-use Method-representation profile is needed. C.37.CF constructs a comparison of changed content and changed form for a stated reader and use. Neither profile nor comparison retroactively merges the one-action boundaries here.
- Builds on:
C.2.1for any standalone account episteme and for the identity of claim-bearing candidate results. - Coordinates with: the pattern governing each candidate's subject result;
A.2.4for optional first evidence-use or status-use classification;A.10for evidence-provenance and bounded reliance; and the choice, gate, permission, authorization, acceptance, or domain pattern for the receiving result. - Coordinates with:
E.17.0for view conformance,C.29for mathematical-lens use,E.24.PUBfor publication,A.6.3.RTfor representation transitions,A.22for selected structure,C.13for construction and collection boundaries, andC.2.P.DRfor declarative-representation overread repair. - Used by: domain patterns only when differently governed candidate results must support one receiving action and the direct one-result/one-use path is not already complete.
C.37:End
Maintain a Method Representation Profile
Type: Method pattern Status: Stable
Working result: a Method representation profile for one MethodDescription or candidate Method account. It relates complete use-bounded selections across Method actions, preserving source claims, cross-use correspondences, omissions, edition relations, keep-separate decisions and returns. A profile can be sufficient working prose; when later use needs the basis retained, its rows carry their complete applicable
C.37claim groups once.
Use This When
Use this pattern when a current MethodDescription or candidate Method account must support several unlike Method-related actions or decisions, and the practitioner needs to relate their use-specific representation selections without treating one text, diagram, table, model, or view as the Method for every user. Begin by asking: which action or decision must this receiving System perform, and which Method claim must become visible for it?
The first useful result is one complete use selection in a Method representation profile. Name the Method or candidate status and source edition, receiver, action, needed Method claims, direct basis, applicable reliance and receiving conditions, material loss, disposition and return. Retain the complete C.37 group in its owning profile row when a later use needs it; an immediate sufficient selection needs no new record or standalone duplicate.
Use C.37 or the applicable direct pattern and stop when only one action is current and no Method-specific cross-use profile is needed. Do not use C.37.MP to select representations for an arbitrary entity, format unchanged content, establish publication or access, or create a collection or selected structure. C.37.MP begins only when the missing result is the Method-specific organization of independently governed, use-bounded selections around one MethodDescription or candidate account.
Working Distinctions
Problem Frame
Method material serves unlike actions. A performer needs actions, inputs, results, guards, and stops. A method engineer compares contribution and variation. A support builder needs tool, resource, permission, and feedback positions. An assessor needs evidence and the qualifications, uncertainty and return conditions that affect the assessment. Project, process, and case viewpoints may expose different claims about the same independently admitted Method-related Work when the corresponding views supply or corroborate grounds for the same action.
One representation rarely serves every action without hiding something. The problem is to relate several use-specific selections while preserving Method identity, candidate status, description currentness, WorkPlan versus Work, the direct authority of every selected result, and the limits of every carried claim.
Problem
Teams often place Method representations on one page and call the result an integrated Method view. The page then hides which receiver and action each item supports, which result identifies it, which claim is relied on, what it omits, whether a reliance or receiving result exists, and whether the subject is a Method, candidate account, WorkPlan, or admitted Work.
The opposite failure declares one canonical Method representation for every user. Performers, assessors, and support builders then reconstruct missing distinctions independently, so inconsistent guards, stale evidence, borrowed authority, and hidden omissions enter action without a return condition.
Forces
Solution
C.37 completes representation selection for one receiver and action. Use that complete result directly for one use. The profile below is needed only when several different uses must remain connected: recover common referents, correspondence, omitted claims and change dependencies without promoting co-use into identity or conformance.
Build the Method representation profile from the current MethodDescription or candidate account. For each receiver/action pair, recover its subject results and complete the applicable C.37 selection. Then relate the uses through shared source claims, unlike omissions, edition relations, correspondences and keep-separate decisions.
Existing claims and matching observations can suffice; new user Work is selected only for a use-changing question worth its attainable whole burden. Retain the complete basis where later profile use needs it, without a second C.37 copy.
Local mantra. One receiver, one action, one qualified selection. Relate uses through the Method's shared claims and dependencies.
Pattern-Use Unfolding
-
Name one Method-related use. State the receiving System, decision about the Method or enactment-support action, and situation, with the qualification window or stop needed for that use. Another action has its own
C.37use account even when the user, carrier or Method is unchanged. Reuse unaffected grounds and reconsider the dependencies changed by that action. -
Fix the Method subject and source account. Name the admitted Method or candidate status and the current MethodDescription or candidate-account edition. Use A.3.2.DE when the needed Method claims are absent or not current enough.
-
Recover the claims needed by the action. Select only the needed purpose, input/result, performer, capability, tool or resource, action or Work specification, variation point, guard, evidence, authority, support, feedback, or stop claims. Keep their subjects and statuses.
-
Recover every candidate through its direct governor. Use
C.2.1for the claim-bearing episteme,E.17.0for a view with its own conformance relation, the applicable domain result for a mathematical object and correspondence, withC.29when a separate representation, transfer, or reliance question remains,E.24.PUBfor publication, andA.22for a selected structure. A title, layout, Method label, carrier, or profile row supplies none of those results. -
Complete the use-bounded selection. For the receiver and action, recover the applicable basis in §4.2 and resolve any missing required subject, reliance or receiving condition at its owner. Use qualified inputs to construct the receiving result; that eventual result is not a universal prerequisite to input selection. Use the direct exit when one owner already supplies the complete answer. Otherwise apply C.37; retain its basis once in the owning profile row only when later use needs it.
-
Compose the Method profile without flattening. Connect the completed rows to one MethodDescription or candidate account. Record cross-use correspondences, conflicting omissions, edition dependencies, and keep-separate decisions. Use §4.4 to construct a missing correspondence and follow its dependencies. Shared profile membership creates no composite Method, super-view, collection, selected structure, or new description edition.
-
Recover supporting subjects before relating their accounts. Preserve WorkPlan, Work, Method, result, and other subject statuses. Use
A.15.6when project, process, or case wording leaves the subject unclear. Accounts of one independently admitted Work retain that same subject; accounts of different subjects remain separate epistemes with only the needed correspondence relations.Each claimed view needs its own
E.17.0conformance result. Either arrangement may inform the same action underC.37; common subject identity alone neither permits co-use nor creates a multi-view family. Another action belongs in another row. -
Resolve the live user-action question. Direct claim inspection or an already matching observation may suffice for the profile. Select a new probe only when a usability, interpretation or loss uncertainty can change the decision and its full design, reader, interpretation and displaced-Work burden is warranted and obtainable.
For a selected probe, the named receiver retrieves the claim, distinguishes status or alternatives, performs the bounded action and applies its stop; report the result of that probe, not observed success from a plan. Return a defect to A.3.2.DE, C.37, the direct subject or receiving-result owner, A.15.8.MS, or the owning Method decision.
Record the Result
Use this shape when a later use needs the profile retained. Its applicable distinctions govern an immediate working answer too, without requiring a stored row for every use.
What Changes in Practice
Practitioners stop asking one artifact to be the Method for every user and stop treating a profile as authority by colocation. A performer can receive an action-and-stop representation, an assessor an evidence-and-limit representation, and a method engineer a comparison or variation representation. Each row shows its own direct basis and permitted use, while the Method profile shows how those unlike selections return to one current source account.
Build the correspondence that a receiving use needs
Begin with a claim that two uses must interpret together: an input identity, an operation, a condition for continuing, or a result passed from one participant to another. The first result can be a short explanation of that correspondence and its basis. A complete crosswalk of every document is unnecessary when the current action depends on only one claim.
- Recover what each expression asks its user to do or conclude. Read the expression with its legend, surrounding instructions and source edition. Identify the subject, input or result, operation, condition and stop that affect this use. A table cell such as “ready” may abbreviate several conditions; a diagram edge may assert that one condition is sufficient. Recover those meanings before matching words or shapes.
- Establish the common referents and interpretation. Determine which Method claim, input or result each expression concerns. Use the applicable naming, measurement or domain rule to resolve different labels, units, reference times or levels of detail. State the mapping and the evidence or definition that supports it. When an essential meaning cannot be recovered, name the missing interpretation and return to its source or responsible practitioner. Keep the other supported correspondences.
- Compare what the two instructions let their users do. Recover the action, reference time and conditions in each instruction. For one pending job, apply both to the established facts and compare the decisions. To replace one instruction with the other for future jobs, check the rules across the conditions in which the replacement will be used, including every exception that can change the decision. One admissible disagreement defeats that claim. Selected trials can reveal a disagreement. Equivalence needs a reason the comparison covers every relevant input or class of inputs. Observing a user or testing an implementation answers a further question.
- Separate correspondence from dependence. Two accounts may agree without one being derived from the other. For each result that could need refreshing, identify the source claim used and how it contributes: for example, a readiness cell abbreviates two conditions jointly, while a performer instruction states the actions needed to satisfy them. An independently maintained rule needs its own basis. A tool's label or a person's correct answer does not show which rule its calculation implements.
- Return a usable cross-use result. Relate the shared claim and its supported interpretation to each affected use row. Preserve useful omissions: a supervisor may need the readiness condition without the operator's detailed steps, provided that omitted detail is not needed for the supervisor's action. Record an unresolved correspondence and name its missing input rather than treating every row as equivalent or unusable.
- Follow a substantive change through the result it changes. Replace the changed premise in each known dependent explanation or calculation and determine which action, result or stop changes. Revise that owned result and reconsider the use that consumes it. F.0.2.RC:4.4 develops this refresh and the return when dependency reach is unknown; A.3.1.CC handles a contradicted maintained claim. Repeat the affected boundary case after repair. Preserve another result only because its own basis still supports its declared use, not because it is in another file.
Use C.37.CF when the question becomes whether the revised content or its form improves the reader's action. Use A.3.1.MV when the proposed instruction changes reusable operations, dependencies or conditions of the Method. These questions can arise from the same comparison but need different answers.
Archetypal Grounding - EC-417 Method Material for Four Uses
The EC-417 team holds candidate whole C-EC-Release-v2, its current candidate-account edition, and trial WorkPlan WP-EC417-B2-Trial-1. A.3.2.DE has supplied the bounded claims needed for the proposed B2 trial: purpose, A/B/B2/R alternatives, evidence entry, Work order, allocations, authority, confidentiality, recovery, and stops. C.37.MP does not admit the candidate as a Method or turn the WorkPlan into Work.
The four proposed uses have the following claim groups. Their subject and reliance results do not fill the missing receiving decisions:
C.37.MP relates these use-specific selections in cross-use profile MRP-EC417-CrossUse-1. The profile supplies the qualified input claims and identifies the unresolved view-conformance question. Actual choices, permissions, Work, and task outcomes retain their own required grounds:
The fourth row asks about one completed trial, so its three proposed readings concern that independently admitted Work. This case does not select the subject of every project, process, or case account. Each becomes a U.View only through its own positive E.17.0 judgment against a viewpoint edition, and their claims belong in the same decision account when each supplies or corroborates grounds for deciding whether to reopen the trial. The profile does not make an integrated super-view.
A timeline, dependency network, state-transition drawing, or other node-link material supports a mathematical use through its identified formal object, correspondence, preserved and lost structure, and inference conditions. Use an adequate domain result directly; use C.29 when a separate representation, transfer, or reliance question remains. Appearance or vocabulary alone does not establish that mathematical basis.
An adequate existing profile and a changed omission
In a constructed unchanged-use case, a method engineer relates an existing action-and-stop description for preparation to an evidence-and-limit description for assessment. Their results, source claims, receiving uses and earlier matching observations remain adequate. Inspection establishes the needed shared-source correspondence and shows that each omission is irrelevant to the other action's permitted contribution. The profile and choice are complete without scheduling a user study. Retain the basis only if the receiving use needs later reconstruction.
Now a proposed shorter preparation representation omits an uncommon stop. A fresh user's wrong action under that condition would materially change the choice. Suppose a capable receiver, permitted material and a protected preparation window make the discriminating task worth its whole burden: select that task and obtain the receiver's answer before claiming observed usability. If the task is unavailable, the omission can still be rejected by direct content inspection; neither the attractive layout nor the study plan establishes safe user performance.
One decision with accounts of different subjects
Suppose the engineer now asks whether a signed-pinout discrepancy requires revising the candidate's evidence-entry guard before another trial. Three available accounts concern different subjects: the WorkPlan states when signed evidence is required; the candidate Method account states the proposed guard; the trial account states which pinout edition was used in W-EC417-B2-1. Keep those subjects and statuses.
In this constructed case, the naming scheme defines how each account identifies the required H-17 pinout input. Recover whether the planned evidence requirement, candidate guard, and trial input record refer to that same required input, while preserving any different revision values. Carry their claims into this decision's C.37 selection only with the supported correspondences.
When those correspondences and the applicable reliance and conditions for using these inputs are supported, the three accounts can inform this one decision without becoming views of one entity. The selection answers which claims may be used; the governing Method decision still determines whether to revise the guard.
If the trial’s exact pinout edition is unknown, compare the admissible editions on the input to the guard. Agreement there can settle the guard decision. If they differ in a way that matters, identify the needed evidence. Identifying the exact trial edition can itself be a separate commission. If the mapping from the observed input to the guard is unknown, return that correspondence question. Neither gap makes the WorkPlan into performed Work or requires a new project/process/case relation definition. The existing accounts remain usable for questions whose premises are still supported.
Build a shared readiness meaning, then change one condition
This constructed case concerns material for an equipment-inspection Method. The operator prepares an inspection; a supervisor decides whether preparation is complete; an assessor later reconstructs what happened. The supplied local rules permit the named materials to inform those tasks. Completing preparation does not authorize operation of the equipment.
The current instruction says: “At the planned inspection time, use a calibration certificate no more than 30 days old. If its date is unknown or its age exceeds 30 days, withhold preparation and obtain clarification or a current certificate. Also complete the fixture check.” A supervisor's table has columns “calibration current” and “fixture checked”; its legend defines preparation-ready as both conditions holding. The controlled glossary defines “calibration current” by the instruction's 30-day rule at the same planned inspection time.
To relate the accounts, the method engineer first expands “calibration current” into the age condition and compares its subject and reference time with the instruction. Both concern the certificate for this equipment at this planned inspection time. The supervisor's two-column conjunction corresponds to the two preparation conditions, although the table omits the operator's recovery steps. That omission is acceptable for deciding preparation readiness; it would be insufficient as the operator's only instruction.
With the fixture checked, a certificate aged 20 days permits preparation under both accounts. That settles today's preparation if those facts hold. For future use, compare the definitions themselves. Both include day 30, exclude day 31, and withhold readiness when the age is unknown or the fixture check is incomplete. The shared definition and conjunction explain why the two accounts agree for readiness across those conditions. An operator learning how to proceed after a stop needs the recovery instruction as well.
The maintenance question needs one more distinction. The table's age test depends on the glossary definition, which points to the current instruction. Its fixture condition has a separate source.
The prototype displays green for today's 20-day certificate and completed fixture check. That is an observation about this run. A program that always displays green would give the same result. Before relying on the display for other inspections, inspect its rule or establish its behavior over the intended conditions with adequate coverage. Until then the supervisor can determine readiness from the instruction and the established job facts. The prototype's behavior on other inputs remains open.
Now the Method owner changes the allowed age to 14 days for future inspections. The changed reusable entry condition belongs to the A.3.1.MV variant or revision question; the new document alone supplies no effectiveness evidence. A 20-day certificate now distinguishes the two rules.
Revise the maintained definition, the table's derived age condition and the operator guidance, then repeat the preparation question under the 14-day condition. Reconsider any implementation claim that consumed the earlier rule. The fixture condition remains supported by its unchanged source; no change to it follows from the shorter calibration interval.
An assessor's archived account records that an earlier inspection used a 20-day certificate under the then-current 30-day rule. Preserve that historical claim and its evidence. It does not become a 14-day inspection by relabelling it. A teaching card used only to distinguish preparation from authorization states neither age limit and returns eligibility questions to the current instruction; its declared use remains supported. These retained differences are useful results of the profile.
If the glossary had instead said only “current means valid”, with no recoverable interpretation of validity, the initial correspondence would remain unresolved. Ask for the rule that decides the table's condition, including the relevant time and missing-data treatment. Do not infer 30 days from familiar wording. The operator instruction can still support its own action while the supervisor's abbreviated condition is clarified.
Bias-Annotation
Conformance Checklist
- The profile names one admitted Method or candidate status and one current MethodDescription or candidate-account edition.
- Every row names one receiving System, Method-related action or decision and situation, with any qualification window or stop needed for that use.
- Every row carries the complete applicable C.37 selection basis defined in §4.2.
- A later use that needs retention can recover the complete selection basis once in its owning result; an immediate sufficient answer requires no new record.
- Each representation kind is established by its direct FPF governor rather than by layout, title, carrier, or profile membership.
- Each row preserves the subject and status distinctions defined in §4.2; the carrier does not supply a missing governed result.
- Project, process, and case accounts retain their recovered direct subjects; co-use for one action has its own complete C.37 basis, and every claimed view has its own conformance result.
- A mathematical-use claim states the formal object, correspondence, intended use, and loss boundary under the applicable domain result;
C.29addresses a remaining representation, transfer, or reliance question. - Failed or missing direct results return the lower episteme and named gap rather than borrowing support from another layer.
- Current adequacy may be established by direct inspection or matching prior observations. A new user-action probe is selected only for a useful attainable question worth its whole burden; a claimed observed success or failure has supporting evidence and no wider capability, enactment, fit, transfer, worth or authority claim.
Common Anti-Patterns and How to Avoid Them
Consequences
Method material becomes usable for unlike actions without multiplying Methods or hiding the source edition. A practitioner can recover why each row is selected, declined, or unresolved, what claim it carries, what it omits, and which change requires return. The profile adds the Method-specific cross-use account without copying the common representation-selection move.
The cost is maintaining the rows and profile relations needed by later use. Some attractive artifacts remain lower epistemes or fail a direct view, lens, publication, reliance, receiving-result or structure rule. Another action has another selection boundary, not automatically another stored record. A changed Method claim may reopen A.3.2.DE or its source edition instead of being patched independently in several representations.
Rationale
MethodDescriptions combine action, purpose, inputs and results, variation, evidence, support, and stop claims that different users consume differently. C.37 governs transdisciplinary selection and co-use for one receiver and one action. C.37.MP connects these complete use-bounded selections to one MethodDescription or candidate account and maintains the cross-use correspondences, omissions, edition relations, and keep-separate decisions of that profile.
SoTA-Echoing
Project, process and case management can raise questions about one Method-related Work or about different direct subjects. Use A.15.6 when the management wording leaves the subject unclear. Test each claimed view and select its claims for the named action. The management labels establish neither common subject identity nor three Methods.
Reopen when a representative Method user cannot perform the named action without hidden reconstruction; when a Method-content role or MethodDescription practice changes the profile action; when a supporting account's direct subject or needed correspondence changes; when a row no longer carries the complete current C.37 claim group; or when a changed supplier result affects the consumed representation selection.
Relations
- A.3.2.DE supplies the current MethodDescription or candidate-account claims and their use boundary; C.37.MP relates complete use-bounded representation rows to the Method-specific uses that consume them.
C.37governs selection and co-use for one receiver and action. C.37.MP adds the cross-use Method profile; §4.2 defines the complete basis retained when a later use needs it.C.2.1identifies claim-bearing epistemes;E.17.0governs viewpoint conformance;C.29governs mathematical-lens use;E.24.PUBgoverns publication;A.22governs selected structures;C.13governs material collection treatment. C.37.MP cannot borrow one result from another.A.15.6recovers the direct subject from management wording;A.15.2governs WorkPlan; Work retains its A.13 assignment bases and independent A.15.1 admission. A Method representation may rely on either as a supporting subject without changing its status or making C.37.MP their general representation owner.C.11.DUAappraises a questionable user-study demand, including design, obtainable contribution and whole burden.- F.0.2.RC follows a changed source premise through the descriptions and uses that depend on it, preserving independently supported results and widening recovery when the affected reach is unknown.
- A.15.8.MS takes the direct exit when one owning result already supplies the complete one-result/one-use answer. It consumes C.37.MP only when unlike Method actions need the cross-use profile; no second standalone C.37 copy is required.
C.37.MP:End
Compare Content and Form Changes
Type: Method pattern Status: Stable
Problem frame
Use this when. You must choose between two versions of a Method description or establish why one works better for its reader. The change may concern instructions, the form in which they are expressed, or both. When a revision changes both, a better answer can follow from the new instruction even if its presentation adds nothing. A reading result can support a choice of description. A claim about the Method's effectiveness needs evidence from its use.
The working reader is a Method engineer comparing descriptions, pattern languages or other inspectable representations for a named reader and action. Start by asking: what answer or action should improve, and what changed in the material available to that reader?
The first useful result is a bounded comparison: the earlier and proposed answers, the content and representation differences, what the evidence supports, and the next repair or selection. One local contrast can be enough. If the problem is simply choosing a representation for one action, use C.37; if several unlike Method uses need complementary representations, use C.37.MP. Use C.11.MV when the question is the practical worth of the Method itself.
Problem
A new pattern language often adds examples, stops and source returns while changing headings and navigation. Readers then answer better, and the gain is attributed entirely to the language form. Conversely, a terse table can look worse because essential content was removed, not because tables are unsuitable.
The same confusion occurs with a checklist, diagram, executable description or AI-supported presentation. A comparison that changes subject, content, tools and task together can support a useful bundle choice, but it cannot isolate the contribution of one change.
Forces
Solution
C.37 supplies the use boundary and C.2.8 the selected structural-recovery question. The comparison below constructs contrasts that can separate changed claims from changed expression, records reader help and tests an affected use. It does not turn a small contrast into a causal effectiveness study; that stronger question uses the applicable research design.
Compare the description change at the receiving use, separate the available contrasts, and report only the difference that the observations warrant.
Fix the comparison question
Name the reader or consuming system, the Method or candidate account, the intended action or decision, and the result and stop the description should let the reader recover. Keep the task facts, relevant source conditions, tools and opportunity to act comparable.
State the content that must survive: the relied-on action claims, relations, applicability, uncertainty, result, stop and source-return information. Then state the representation difference, such as notation, arrangement, salience, navigation or interaction. Do not call a new condition or an unsupported diagram edge a form-only change.
Use A.6.3.RT when a change of representation language or reasoning medium affects the receiving operation, or the applicable same-regime wording, narrative or coarsening result. Its preservation and loss answer helps establish the comparison inputs; it does not demonstrate improved performance. If the Method's reusable semantics also changed, retain the A.3.1.MV identity/variant question and bound the description comparison accordingly.
Choose the smallest informative contrasts
When only content changed in a fixed form, compare those two versions. When only form changed and the relevant content is preserved, compare those two forms.
When both changed and the receiving decision needs their separate contributions, use:
The three-variant arrangement gives two local contrasts. It does not identify an interaction. Four cells make that question inspectable but supply neither replication nor causal identification by themselves.
If a cell cannot be constructed without changing another load-bearing feature, say so. A bundle comparison may still support a local choice. Do not describe it as an isolated content or form result.
Observe the receiving use
Ask the reader to do the work on which the comparison turns, through the explanation, decision or handover that the recipient needs. A first useful result suffices when the commission ends there. Recover the following facts.
- the action or decision called for, and the reader's answer;
- the result, stop or missing condition the reader identifies;
- unsupported inferences or lost distinctions that change the answer;
- the source returns needed to support or correct it; and
- the observed burden when time, source openings, attempts or other relevant effort were recorded.
When the question is how much selected structure the reader can extract, use C.2.8 ExtractableStructuralInformation. Identify the expressed account, the expression being tested and the reader, with selected relations, correctness criterion, prior knowledge, operations, help, access and budget qualified for this comparison. A qualitative account of recovered and missing relations often suffices. A number needs a declared structural scale and fixed units or denominator, not counts of words or arrows.
Keep a design estimate separate from an actual recovery observation or a mapped formal estimate. A response establishes what was returned in that trial. Attribute recovery or application to the reader when their observed work and the help available support that inference. A correct answer of unknown origin can still serve its intended use while authorship stays open. One response cannot establish maximum performance or reliability across a population.
Before interpreting a correct answer, reconstruct what each reader received. The text, prompt, examples, available tools or another participant may already have chosen the unit of comparison, the rule for matching two entries, or the action to take. Credit the reader with the choice they made. Following a supplied rule can be exactly the intended use.
A stronger claim may be that the reader can turn a general method into a rule for work in a discipline. Give the reader the general method and material from that discipline. Leave the consequential modeling choices open. Ask them to construct and justify a rule that another practitioner could apply. Where plausible rules remain and their difference matters, use a change that distinguishes them. A justified construction can already cover the variations in the commission. A request for an actual trial also requires its execution. An illustration of the principle or a fresh set of numbers for an already specified calculation provides different evidence. B.5.MR:4.3 explains how to test a claim that someone used a transdisciplinary principle (L3) to construct a repeatable rule for a discipline (L2).
A missing basis leaves the affected conclusion open. A result under changed conditions can support that use within the conditions examined.
These are candidate observations, not one universal score. Select those that can change this comparison. A source-opening count does not measure comprehension; shorter reading does not prove lower total effort; recognizing a prescribed action does not establish its correct enactment.
Choose readers and allocation to variants in proportion to the claim. A fresh probe withholds the other answers and the author's preferred conclusion when they would teach the task. Repeated-reader comparisons retain learning and order effects instead of calling the reader fresh.
In AI probes, keep the model, supplied context and available tools recoverable when they affect inference. Generalizing from a few AI answers to human performance is another unanswered question.
When testing a changed-condition return
Use this probe when the comparison needs to establish how a reader responds to a new fact. First retain the reader's initial answer. Then disclose the changed source, missing fact or altered receiving use. Ask what must change and what remains usable.
The reader should revisit the dependent claim rather than repeat the whole task or preserve an invalid assumption. Compare the reader's revised answer, its source returns and its stop. If the changed condition changes the comparison question or available action set, identify that new question; do not present the two answers as observations under identical conditions.
Return a bounded revision result
State the inspected variants, the content/form contrasts they support, observations, losses, uncertainty and next decision. Distinguish a demonstrated local difference, no distinguishing result in this probe, an unavailable contrast and a confounded comparison.
Compare extraction effort and receiving value separately from structural amount through C.11.CRC and C.11. Retain the sufficient incumbent when the proposed change supplies no worthwhile gain.
Return a content correction to A.3.2.DE, a contradicted maintained claim to A.3.1.CC, and selection evidence to C.37 or C.37.MP. Use A.3.1.WT for representative enactment evidence and C.11.MV for a Method-worth decision. A causal claim needs its own justified research design and inference basis; neither this table nor an attractive explanation supplies it.
An ordinary comparison can remain short prose. When another decision will rely on it, retain the candidate versions, receiving use, observations and inference limits as an identifiable account, with the source conditions that reopen it.
Archetypal Grounding
Added stop information versus a table
This constructed comparison uses the texts below; it reports no reader study.
The receiving action is deciding whether an earlier review may support a current description revision.
A, earlier prose: “Reuse the previous review for a display-only change.”
B, revised prose: “For a display-only change, reuse the previous review only while the reviewed claims, question and qualification window still match. If a required source argument cannot be recovered for a new recheck, stop that recheck and return to its source; preserve unrelated earlier results.”
C, the same revised content in a table:
Initially, the facts say that the change is display-only, the claims and question are unchanged, and the qualification window is current. All variants can support reuse. The later condition says that a new recheck needs an argument outside the available excerpts and that the retained dossier contains only its locator.
B and C state the local stop and preservation instructions; A does not contain that distinction. This establishes a content-coverage difference in the constructed texts. It does not show that an actual reader preferred or used C more effectively than B. A reader comparison would still be needed for that form claim. The first task lets the reader use a stated review rule. To find out whether they can devise a review rule for their own discipline, give them records from that practice and leave the criteria for reusing work for them to choose and justify.
For the revision, retain the missing stop information and withhold the table-superiority claim.
A graph that adds a claim
A proposed graph draws an unconditional edge from “review complete” to “publish”. The prose requires a separate publication decision. The graph has added a consequential claim; it is not another form of the same content. Repair the edge and its condition before testing a pure form contrast, or compare the changed bundle with that difference explicit.
Compare explanations of a pattern language
An engineer has a use-bounded description of an admitted Method for checking a claim against its source. The engineer now asks why the surrounding language separates source allocation, coherence checking and reconstruction: should those three entries become one review procedure? This is a constructed comparison. The engineer knows ordinary procedures and can consult the complete framework, but has no private authoring notes.
Use A.3.2.DE for the claims describing the checking Method. E.8.ML supplies the different answer about the language arrangement: its entries serve independent questions, with different useful results and source returns. The existing arrangement is candidate A. The author proposes a shorter paragraph, B, and a table, C, for the same receiving question.
B — Selected explanation in prose. Use F.0.2.MA to decide where a recovered source contribution belongs in the language. Use A.3.1.CC when a maintained claim contradicts its basis. Use F.0.2.RC when a promised source contribution cannot be recovered.
Each entry can finish the question it answers. If you also need to choose where to start, identify the result still missing. To establish what a publication actually contains, compare its text with the allocation. Grouping the entries in an explanation leaves open which Methods are constituents of a particular whole.
C — The same selected claims in a table.
Each entry can finish the question it answers. If you also need to choose where to start, identify the result still missing. To establish what a publication actually contains, compare its text with the allocation. Grouping the entries in an explanation leaves open which Methods are constituents of a particular whole.
Make the comparison. C.37.CF separates the available contrasts. B selects and rewrites content from A; A against B does not isolate a layout effect. C rearranges B's selected claims while retaining its common conditions. B against C can therefore address that presentation difference. The full E.8.ML account remains available for conditions or questions beyond this selected explanation.
For the stated question, explain why the entries can be used separately. A practitioner may need to allocate a recovered contribution while its claims are coherent, correct a contradiction while its allocation is known, or reconstruct missing content. Each result can therefore serve a different decision. For a further request to choose an entry, identify the missing result. With an adequate allocation and one contradictory maintained claim, A.3.1.CC can suffice. In a changed case, an argument needed for a new recheck becomes unavailable. Return the affected current-use claim and source relation, preserving the earlier answer and independently supported contributions. If the question becomes publication authorization, obtain that result separately.
A reading comparison would keep the engineer's preparation, task, source access, tools and permitted help comparable and retain the initial answer before giving a corrective cue. Record reading or source-return effort only when it is observed. A shorter paragraph or a table alone establishes no reduction in total effort, and a correct response establishes neither learning nor the effectiveness of the described Method.
The source comparison supports the selected claims and their returns. It supplies no observed advantage of B or C for this reader. Retain the existing explanation unless an obtainable comparison can establish a worthwhile change. If a needed relation is absent, restore that content with A.3.2.DE or the applicable language answer before crediting a new form. If the promise is instruction, obtain the applicable subject- and holder-specific learning contribution and its evidence conditions; a form comparison does not establish that learning.
Bias-Annotation
Authors tend to know the omitted condition and may answer correctly from memory. Readers who see all variants can learn it too. Separate text inspection, fresh reading, repeated reading and actual enactment evidence.
The constructed examples expose content and inference boundaries. They contain no measured time saving, error rate or Method-effectiveness result.
Conformance Checklist
- The receiving reader, subject, action, result and stop are fixed sufficiently for the comparison.
- Content changes and representation changes are stated, including added or lost relations.
- Every claimed contrast is supported by the available variants and conditions.
- The account identifies the consequential choices the reader made and those already supplied by the text, task, tools or help. Credit for constructing a rule requires evidence from work in which that choice was open.
- When a changed-condition return is tested, an initial answer precedes disclosure of that condition.
- Observed burden is distinguished from estimated or unmeasured burden.
- The result returns to the content, selection, trial or worth decision it informs.
Common Anti-Patterns and How to Avoid Them
Consequences
A useful revision can be retained without overstating why it helped. Content defects and representation defects receive different repairs, and later selection or worth decisions receive evidence with usable limits.
The cost is preparing informative contrasts and observing the reader carry out the task. When that cost cannot change the decision, take the sufficient direct result or a bounded bundle comparison.
Architectural Rationale
A.3.2.DE asks whether a description exposes the claims a use needs. C.37.MP organizes complementary representations across uses. A.3.1.CC checks agreement and repair ownership. None of those results alone answers which difference between revised descriptions explains an observed improvement.
This pattern joins established representation-preservation and validation Methods around that professional revision question. It does not replace experimental design, produce a universal representation ranking or merge description usability with Method performance.
SoTA-Echoing
ACAP analysis (Doellken, Nelius and Matthiesen, 2024) distinguishes attention, comprehension, application and performance when investigating design-method difficulties. Adopt that separation; the evaluated sheet-metal-design setting and interpretive analysis do not justify universal metrics or causal attribution to each content element.
NIST's factorial-design account supplies the established crossed-factor idea. Adapt it to the contrasts the Method-description decision needs; do not claim a new experimental theory. C.37, A.6.3.RT and C.37.MP retain their representation results.
The cheaper alternative is direct claim inspection plus the existing representation selection. In the review-reuse example it already identifies the missing stop information; no reader experiment is necessary for that content defect. Use the fuller comparison only when a material form, learning, burden or interaction question remains. Reopen the result on a changed task, source condition, candidate content, reader population or hidden confound.
Relations
- A.3.2.DE supplies the description and receiving use; A.6.3.RT supplies representation-preservation and loss.
- C.2.8 supplies the optional structural-recovery characteristic; C.11.CRC and C.11 keep extraction cost and marginal usefulness distinct.
- C.37 and C.37.MP consume bounded evidence for one-action selection and cross-use representation profiles.
- A.3.1.CC receives a contradicted-claim correction; A.3.1.MV distinguishes description revision from Method variation.
- A.3.1.WT supplies actual trial evidence; A.3.1.FT and C.11.MV separately judge transfer/fit and practical worth.
- E.8.ML can use the comparison when choosing a pattern-language representation; F.0.2.RC supplies reconstruction failures that may motivate another comparison.
C.37.CF:End
Coordinate Related Models for Use and Handover
Type: Method pattern Status: Stable
Problem frame
Use this when a decision depends on what several models establish together, or a successor must be able to interpret, reproduce or revise their results. The models may concern a system, a method, a relation, or several connected entities. A colleague can understand every diagram and still be unable to answer the question for which they received the set.
Start with the work the recipient must do and the conclusion on which it depends. If a single understood model settles the question, use it directly. C.37:4.6 covers the simpler case in which a recipient needs several representations for one action and each keeps its own supported conclusion.
Problem
One model can omit a condition another silently assumes. A recipient can also receive all the needed models and still lack the conventions for reading a blank entry, a shared designation or an estimate. Which models, relations and methods must remain available for the recipient's work?
Forces
Adding material from more specialists can improve coverage while making a contradiction harder to locate. The author may know an interpretation that a successor has never encountered. Recording every possible connection imposes work and can hide the few connections that determine the answer.
Solution
Establish what the models must support together
Name the intended inference, action or reproduction task and its possible outcomes. Identify which model supplies which premise. For each premise recover the modeled entity, conditions, interpretation and relevant edition. A local name for several models supplies a convenient reference. If the proposed multimodel is to be relied on as a coordinated set, establish the consequential connections that let the models support this use.
When several models concern the same entity, establish its identity across them. When they concern different entities, state the actual relation. A system, the work that produces it and a cost estimate for that work can all affect one design decision. The estimate's use depends on what work it covers. For an admitted holon, B.1 and the kind-specific method govern the inference from the constituents and their organization to a whole-level result. A modeled relation with independently identified participants needs no presumption of physical parts.
Choose and connect the required ranges
Start from the composition scales and model distinctions the decision calls for. If choosing a range is part of your task, use C.39.SA:4. If the intended recipient must choose, make the method and evidence for that choice available to them. In a multimodel, establish what each model contributes and how those contributions can hold together.
- Identify the types of entities and actions that can affect that result. For each relevant kind of holon, follow its own composition as far as the decision needs. A system has integrated components. A composite Method admits particular Methods under its construction rule. For composite dated Work, establish which actual performances did a constitutive part of that very job. An overlapping or earlier performance can help without belonging to it. For an episteme, identify the claims and grounds incorporated into the account and any external sources to which it appeals. For a discipline, identify the knowledge and practice its working arrangement integrates. A part, an available alternative and an external supplier can play different parts in the decision.
- For each inference, identify what its model can leave open. Which alternatives must remain distinguishable for this inquiry? Can a general rule settle it throughout the stated range? Use B.5.MR when comparing models towards greater abstraction or refinement. You can change the scope of composition while holding model abstraction fixed, or compare models of the same whole while holding its composition in view.
- Connect the results. Name who or what participates and the relation on which the inference depends. Work enacts a Method. A system participates in the Work. A report relies on an observation made during that Work. These links may be needed together. A rule at every eligible scale, an influence across scales and a result of several working together each need grounds for the connection they assert. A control layer adds its own functional organization when the decision depends on it.
- Trace the explanation forward to the recipient's decision. Could all the local claims stay true while a different connection changed that decision? If so, recover the missing connection when it is available. If it is unavailable, give the warranted partial answer and identify the condition that still decides the original question. A partial result can finish the commission when it already settles the recipient’s criterion. For an assertion covering a range, show why the grounds hold throughout it. One governing rule or a sufficient bound can sometimes settle the entire question.
For example, a receiving shift needs to know whether two language desks can jointly issue consistent stop instructions today. The desks are components of one publishing service. Comparing the stop clauses belongs to the release Method. Identify the reviews actually performed as contributions to this release. Check that each signed report covers the translation now ready to issue and identifies its applicable source edition. An earlier, independent review of those same texts can support their consistency when the release terms accept it. That reliance need not make the earlier review part of today’s Work. Both reviews can be correct while one source calls for a stop before the next operation and the other calls for it afterwards. Following the reports back to their sources reveals the conflict. If the sources differ only elsewhere, both translations convey the same applicable stop condition, and the commission accepts that equivalence, consistency is established. Suppose that both approved texts are ready, the desks have authority to issue them, the required staff and channels are available to both in the agreed release window today, and these are the remaining release conditions. The desks can jointly issue the consistent instructions in that window. If one desk has no usable route to the shift before the deadline, joint issue in that window is unavailable even when the instructions agree. A claim that the shift actually received them needs evidence of delivery.
The general result is a connected account of the conditions for one receiving action. The principle about agreement applies across disciplines. Editorial work makes the relation concrete through the applicable clauses, a local release procedure can further restrict which editions qualify, and one actual issue supplies evidence about identified texts and reviews. A larger service could still be examined on one occasion. The number or size of its desks assigns no level of model abstraction. Keep the account's declared types and ranges when applying the method to a different action. C.39.QR:4.1 gives the return from an example to that commission.
Test the joint conclusion
State how the claims from the models are to be read together. Align such matters as units, conditions, time, configuration and whether a value was stipulated, observed or predicted when the conclusion depends on them. Check whether the commitments can all hold in the same case. Pairwise compatibility can leave the whole set inconsistent. If no case satisfies all the commitments, trace the conflict to the models or their interpretation. If information is missing, identify the premise or correspondence needed to settle compatibility.
To establish that an outcome is possible under these models, show that at least one jointly admissible case has that outcome. You can exhibit such a case or prove that it exists. If the recipient needs a setting they can actually use, recover one and check it against the conditions. To guarantee the outcome under the stated conditions, establish that at least one case is admissible and that every jointly admissible case has that outcome. One admissible exception defeats the guarantee.
Ask whether the outstanding uncertainty affects the decision you owe. Two interpretations can yield different estimates and still put every admissible result on the same side of a decision threshold. That can settle the choice while the estimate remains open. If no fixed choice works throughout the allowed conditions, establish whether the recipient may choose after a distinguishing observation. The observation must be available before commitment and the action for each possible answer must work throughout the cases still compatible with it. If those conditions fail, return the unresolved choice or the demonstrated limit. An incomplete list of interpretations can still support a conclusion if every permitted interpretation must preserve a sufficient bound or other relevant property. Where no grounds cover the others, qualify the answer to the interpretations examined.
A claim about a candidate is distinct from a claim about how an actual system performed. To use a test result from an observed system, state what was observed and which target claim it supports under which conditions. B.5.MR supplies the test for a further claim that joint use strictly refines the earlier model. A gain in interpretation can occur while the set of eligible candidates stays the same.
Make the receiving use reproducible
Ask what the successor will need to do. To read an existing result, they need the governing types, meanings, constraints, editions and consequential links. A metamodel can supply conventions that determine which models are admissible. A megamodel, in this handover use, makes the related models and the relevant links to their governing models recoverable. A reference to an accessible, identified edition suffices when it gives the recipient what they need. If the governing conventions may themselves change, preserve the grounds for the relevant higher-order choices.
To repeat, adapt or examine the original derivation, identify its production method and the inputs and assumptions on which it depends. To check its conclusion independently, establish the common target, conditions and meaning of the result. A missing premise leaves the original derivation incomplete. The conclusion can still be established by another sound method. For a claim about every admissible case, one counterexample that meets its conditions refutes the claim. A glossary can settle what an estimate means while leaving the successor unable to recompute it. When one source supplies a well-understood method, name the applicable edition and the values fixed locally.
To act on the result, the recipient may additionally need the resources, capability and authority for the work. An assignment, where relied on, is a relation with a system as its holder under A.2.1. Handing over an account does not establish that anyone has performed the work.
Reopen only the dependent conclusions
When an observation, model edition or interpretation changes, follow the consequential relations to the conclusions and actions that used it. Tell their users what changed and which use must be reconsidered. Test the revised result under the conditions that made the change relevant. If an earlier model is retained as evidence of a past decision, keep its effective edition identifiable.
Archetypal Grounding
Two accounts bear on the same stipulated preference relation over three alternatives. The first reports that A is preferred to B, the second that B and C are tied. With a shared population, comparison conditions and the premise that these judgments are readings of one complete and transitive weak order, the two claims imply that A is preferred to C. If a blank in the second table means “uncompared”, the tie is missing and the conclusion is open. The two files can remain as they are. A recipient needs the applicable interpretation of that entry and the grounds for the common relation. Observed human choices require evidence for any claim that this weak-order interpretation fits them.
An engineering team receives a design and an annual maintenance estimate of €12,000 for one service visit. The contract file identifies two possible charging arrangements. One charges €12,000 per visit. The other charges €12,000 per year for up to two visits. The signed schedule that selects the arrangement is missing. Both account for the old figure. For two visits next year the estimates are €24,000 and €12,000. The team can confirm that the service charge fits a €25,000 annual allowance under either arrangement, provided those are the only possible terms and the allowance covers that charge alone. With an €18,000 allowance, one arrangement fits and the other exceeds the budget. The team needs to establish which arrangement applies. Reconstructing the original estimate or pricing a third visit would also require further information. A wider commission to account for the system throughout its life may require additional engineering coverage under SYSE.7.
Bias-Annotation
Readers commonly give precedence to the most precise-looking model. A model precise about functional behavior can omit a cost, construction or usage premise on which the recipient's decision turns. Recover the missing premise before preferring a format.
Conformance Checklist
- The intended result, the modeled entities, each model's contribution and the indispensable relations are recoverable.
- The account tests whether the models admit a joint case. A possibility claim has an example satisfying the joint conditions or a sound argument that such a case exists. A guarantee covers every jointly admissible case and at least one such case exists. An unresolved value can still settle the receiving decision if every admissible value yields the same choice. An inconsistent set is reported as a conflict.
- Material supplied at handover is sufficient for the recipient's actual use, including the production method for repeating, adapting or checking its derivation, and the grounds of comparison for an independent check.
Common Anti-Patterns and How to Avoid Them
A common title, shared file or new graph can hide a missing relation. Read its consequential links in ordinary terms and test the conclusion they are claimed to support. A publisher can also provide every governing convention while omitting an entire aspect the decision needs. Check coverage against the original assignment.
Consequences
A recipient can identify why a joint conclusion holds, which question remains open and which work must be repeated when a premise changes. One usable edition reference can support several models.
Architectural Rationale
The needed method concerns the relations required by the receiving use. C.13 governs a separate claim of collection or composition, and A.22 governs identity for a selected structure. The same work can be carried through in ordinary prose, a project account or a maintained registry.
SoTA-Echoing
Megamodeling in model-driven engineering makes models and typed relations, including conformance, available for navigation. Architecture-description practice uses stakeholder concerns to select views and viewpoints to interpret them. A model register can serve a successor who needs to retrieve the governing edition. For a decision that depends on several views, even a complete register leaves open which conclusions their commitments support together. Section 4.2 tests the joint inference. Section 4.3 distinguishes the inputs needed to repeat its derivation from the grounds for checking the conclusion independently. Revisit the selected views and links when a new decision needs a different premise or when a change of convention alters a premise already in use. Bézivin, Jouault and Valduriez, On the Need for Megamodels · ISO/IEC/IEEE 42010:2022
Relations
C.37 governs co-use for one action. C.37.MP relates different uses of representations of the same Method description or candidate account. C.2.1 supplies the identity of each model episteme, E.17.0 governs viewpoint and view when asserted, B.5.MR compares model commitments, and SYSE.7 applies this method in engineering.
C.37.MM:End
Construct Comparable Ways to Obtain One Result
Type: Method pattern Status: Stable Normativity: Normative unless marked informative
Problem frame
Plain name. Turn labels such as build, buy, reuse, provider, internal team, outsource, or AI into comparable complete ways of getting the same result.
Use this when. A person, team, organization, or other deciding System must choose how one named result will become available, but its candidate list mixes relations, performers, Systems, Methods, commercial labels, and enabling branches. The rows do not yet seek the same result or expose the same burdens, so sending them directly to C.11 would create a false choice.
Reader. The practitioner preparing decision-ready alternatives for one receiver. The pattern forms candidates; it neither chooses for the receiver nor realizes a retained way.
First useful result. A comparison establishes what the proposed ways can supply against the same receiving terms, which differences are material, and which premises remain open. Two labels can describe the same way. If only one or none of the named ways meets the terms, that finding can complete their appraisal. An order to construct additional ways stays open until supplied or their impossibility within the permitted range is established. When choice is current, hand the resulting OptionSet to C.11; its ChoiceResult is separate. For a retained way, name any first unsupported realization branch; if none remains, use the direct realization Method.
What changes in practice. Instead of comparing buy against AI or internal as if the labels were peer kinds, the practitioner asks how each whole way could make the same accepted result available. Duplicate labels merge, materially different ways under one label split, and omitted integration, evidence, support, capability, custody, resource, consequence, and exit burdens become visible before commitment.
When to stop. If the receiver can already obtain the ordered result on the agreed terms and no further choice or appraisal is due under the applicable commission, show how the existing provision meets those terms and omit the proposed new work. If one authorized way suffices and no comparison is due, use the direct domain Method. A domain comparison that already answers the decision can go straight to the chooser. Use C.11 when only the choice from a complete OptionSet remains.
Use another pattern for these questions:
A.15.9for one missing or unqualified bounded result from another practice;C.39for a missing obtaining explanation;C.40for developing and examining material, including its coupled entry when problems and ways must change together;C.18for archive, front or possibility-space claims;C.19for live-pool exploration policy;C.32for architecture-candidate synthesis;A.15.8for one actual-Work or present-WorkPlan support configuration; and- the direct domain Method for realization, production, provision, contracting, acceptance, or use.
Problem
Common option labels name unlike things. Buy may name a purchase relation, provider a continuing arrangement, internal a performer location, reuse an already available System, AI one possible Agent, and build many kinds of Work. Two labels can describe the same whole way, while two offers carrying the same label can differ in integration, custody, support, evidence, capability, and exit enough to reverse the choice.
The error is not merely missing detail. If candidate rows seek different results, use different acceptance bases, or silently switch situations and horizons, no amount of scoring restores comparability. If possible-future rows are written as facts, capability, authority, Work, provision, delivery, acceptance, and availability are asserted without their direct support.
The needed move is to fix one result question, construct complete-enough materially different ways on one parity basis, keep premise modality visible, and freeze the candidates for choice. It is smaller than a procurement or architecture programme and more complete than a make-or-buy label table.
Forces
Solution
Start from the receiving commission. When the answer depends on what present arrangements can supply, establish what the receiver can obtain if they continue. For an appraisal confined to named proposals, follow each proposal on its stated terms. Check access, fitness for the intended use and availability over the required period where these affect the answer. An item in a supplier’s catalogue or a file in an archive establishes availability to this receiver only if the receiver can use it on those terms. Where something is missing, name it. That may leave only a permission to obtain or an adaptation to make. Use C.42 if removing existing work would affect the provision on which the receiver relies.
If a choice is still due, give every way the same test. Continuing the present arrangement may itself qualify, with its ongoing costs and dependencies. A required appraisal may also use it to show what would happen without a change, even when it fails a condition for selection. An appraisal of a named change against that baseline can answer its commissioned question directly. If the choice concerns alternative ways to obtain the result, construct comparable ways and pass them to C.11.
Follow the seven-step same-result sequence
- Fix the result of the commission. Name the receiver, intended use, horizon, conditions for acceptance, and who can commit to a way of obtaining it. Separate unlike ends. If the commission requires both delivery now and the receiving team’s ability to do a later job under changed conditions, keep both in every way. State what will count as an acceptable delivery and, separately, what the team must decide or perform on the later job and what help it may use.
- Find what is already available and what is missing. Use earlier work when its conditions still apply. Prior independent performance can establish the team’s capability if it covers the change in question. When the team must exercise independent judgement, establish which choices the brief, tools or helper already supply and examine the choices the team makes on its own. If the order calls for constructing a rule for a class of jobs, examine an earlier work product or give a task that leaves the consequential choices open. Check how the team justified the rule and where it applies. When the order calls for applying an approved rule, give the team that rule. A supplier’s accepted delivery establishes the delivery. Credit the team with decisions evidenced by its own work. Seek a missing or unqualified result from another practice through
A.15.9. - Recover the proposed ways and construct alternatives when needed. Work back from the handover the receiver must be able to use. What must be produced, supplied or decided for it to occur? For a named proposal, follow its stated allocation, operations and sources forward through handovers, permissions and shared conditions to the result it could supply. If the recipient needs a different way, change an allocation, operation or source that matters and follow the altered arrangement through to the same receiving result. For each surviving way, describe the relevant proposed or available Agents, Work, Methods, Systems, values, provision or access relations, enabling branches, and conditions through which the same result could become available. Mark every decision-bearing premise as supported now, proposed, or unknown. Merge labels that resolve to the same decision-changing contents; split ways under one label when a content difference can change the choice. A way may combine delivery by a provider with work through which the team acquires and demonstrates the required capability. State how that capability will be checked and mark it as proposed until there is evidence. When the gap is an unexplained obtaining operation, use
C.39to develop a candidate explanation or identify the unresolved contribution. A single supported way can close a request for a way to proceed. For a request to find another eligible way, distinguish a way constructed, a proof excluding every further way in the permitted range, and an open search. Comparing specified ways can finish with one or none eligible. - Use one decision-changing parity basis. Ask the same bounded questions of every way: result and availability condition; performers, Work, Methods, and means; production, transfer, provision, access, custody, or ownership; fit, interfaces, integration, and configuration; evidence, uncertainty, and assurance; support, maintenance, change, exit, and recovery; capability consequences; resources; and consequences for affected Systems. Omit a group only when it cannot change this choice. Keep a decision-reversing unknown as a visible gap.
- Restore comparability without forcing totality. Keep the result, receiving use, acceptance conditions, situation, configuration boundary, evidence horizon, and burden categories shared. Use ordinary comparison when the ordering is direct. Use
A.19.CPMandA.19.SelectorMechanismonly when evidence gates, partial orders, incomparability, abstention, or set-valued retention matter. Do not hide protected conditions inside an unexplained score. - Return the comparison for its intended use. Deliver the appraisal when its grounds answer the question about the proposals. When a choice is due, hand a finite
OptionSetof complete-enough whole ways to the chooser, with the remaining gaps visible. ApplyC.11for that choice. Its result may choose one way or a retained tie-set, reject the current set, probe again, or reroute. The chooser retains authority for this decision. - Name the continuation and reopen. For a retained way, identify the first unsupported realization branch, if any, or the direct Method that can now proceed. Reopen only when a changed result identity, use, acceptance basis, candidate, specialist result, capability, interface, evidence item, burden, support condition, affected-System consequence, exit condition, or observation can change the comparison or choice.
Use a proportional parity view
There is no universal arrangement schema. For the current decision, a small table or a few parallel paragraphs are enough if another reader can answer these questions for every way:
Delete a row that changes no candidate comparison. Add a local question when omitting it could reverse the choice. A candidate is complete enough when its important means, dependencies, burdens, and gaps are visible to this decision—not when it fills a universal checklist.
Return the first result in plain language
For a short appraisal, the following form can carry the conclusion to its recipient.
Result and use — [What was requested, by whom, for which situation, horizon and acceptance conditions?]
Proposals examined — [What does each offer? Which names describe the same way? Which premises are supported, proposed or open?]
Basis for comparison — [Apply the same receiving terms to each way. Give the differences in provision, burdens and gaps that matter here.]
Conclusion for this commission — [Which distinct ways, if any, meet the terms? Give the grounds. Identify any open condition that could change the answer.]
Further work, when needed — [A choice can go to C.11. For a request for another way, give the way obtained, the grounds excluding any further way in the permitted range, or the search still needed. For a way already chosen, use the direct domain Method or develop its first unsupported realization branch.]
An appraisal of named proposals can finish with its conclusion. If that conclusion relies on an actual capability, authorization, delivery or use, identify the evidence for it. The responsible participants can use the comparison to choose, arrange and perform a way when the commission calls for those actions.
Retain the comparison without duplication
An immediate sufficient comparison can finish without a new record. When later use needs its basis retained, use the least burdensome arrangement:
- If a domain result already carries the complete same-use comparison, keep it there rather than producing a second account.
- Otherwise retain one ordinary
C.2.1episteme with the one result-obtaining comparison as its EntityOfConcern and an effective reference scheme. - If a direct domain Method already returns the complete useful comparison more cheaply, use that Method and do not add a
C.38account.
The episteme, table, document, or diagram that carries the claims does not, merely by carrying them, establish a world-side arrangement, selected Structure, architecture, plan, Work, or result availability. Use A.22, C.30, C.29, E.17, or another direct pattern only when its separate question is current.
Separate recognition from assurance
- Recognition. Start with the receiving use. If the ordered result is already available, establish whether anyone still has a choice to make. When several ways remain and their labels conceal differences that matter to that choice, two parallel descriptions can suffice for a reversible decision.
- Assurance. Consequential use checks that every row seeks the same result on the same basis, every actual premise has direct support, modal premises remain modal, protected conditions remain visible, and no omitted burden can plausibly reverse the choice unnoticed. Domain acceptance, assurance, authority, realization, and release stay with their direct Methods.
Worked cases
Greenhouse climate-control result
A project needs one accepted result: a named greenhouse configuration maintains the required temperature and humidity envelope through the next operating horizon. Its first list says build, buy, provider, and AI.
The team resolves those labels into three materially different ways: integrate internally owned equipment and control Work; obtain a configurable control System and perform local integration and operation; or obtain a provider-operated climate-control result with continuing sensing, access, support, and exit dependence. An AI Agent may perform forecasting or tuning Work inside any way; it is not a peer arrangement kind.
The same parity questions expose plant interfaces, commissioning evidence, response to sensor loss, maintenance, operating capability, data access, affected crops, support, and exit. The direct engineering pattern keeps the greenhouse-specific architecture and assurance. C.38 supplies only the same-result formation and comparison, then hands the finite set to C.11. After a choice, the first unsupported realization branch returns to the direct Systems Engineering Method.
A performance recording needs one usable artifact
A music project commissions a particular performance and a recording from which a later learner can recover when two groups entered and where they stood relative to one another. Before the performance it compares local capture, a specialist mobile team and the venue’s service for that same use. An earlier rehearsal can help test a proposed recording method. The commissioned performance still lies ahead.
For each way the team asks how the later learner could establish both relations, what they would have permission to use and which parts of the proposal still need testing. The team can send the alternatives to C.11 before the performance.
After the performance the learner can retrieve two time-aligned tracks identified with that run and labelled by group. Both start before either group enters, and the first musical entry of each group can be identified on its track. Their times on the common clock establish the order and interval. The recorder took the two groups’ feeds directly and saved no positions. The same pair of tracks is compatible with either group on the left. A cue sheet recording the actual placements for that performance may settle the remaining question if the learner can match it to the tracks and access it on the agreed terms. If a single annotated recording was ordered, use the sheet to complete that recording. If the existing material already meets the agreed terms, the order needs no further production. If placement cannot be recovered, record that limit.
The production Method governs capture, use rights and acceptance of the recording. If the commission also calls for a learner who can interpret a later recording independently, examine that learner’s work. When a tutor identifies both entries, the learner may still measure the interval. The measurement leaves open whether the learner could identify the entries unaided. If devising a rule for future recordings is also commissioned, the learner can choose how to distinguish an entry from a tuning sound and justify when the rule applies.
Direct Method-base stop
A Method Engineering team needs an enactment-support configuration for one bounded task set. Its direct Method already constructs candidate configurations on one basis, preserves evidence status and incomparability, tests retained routes, and returns a supported configuration, retained set, split, or gap.
Use that direct Method. Adding a C.38 account for the same decision would duplicate its action and result. Open C.38 only if a wider question appears—for example, several complete ways to make an independently named support result available to another receiver.
Precision restoration
Bias check. Compare the supported benefits, burdens and limits of each way for this result and situation, including any decision-relevant weakness. An adequate incumbent may remain preferable. Familiarity, novelty, popularity or institutional standing alone does not establish suitability.
Conformance and practical checks
A use conforms only when the checks needed by its claimed comparison pass:
- The result, receiver, use, horizon, acceptance conditions and decision authority are explicit. Existing provision has been checked against the same terms. If the commission includes the receiver’s later capability, the comparison identifies the work that would establish it and the assistance allowed.
- The alternatives address the same receiving result. Ground their material differences, what rules any out, and any claim that the permitted range is exhausted.
- Every decision-bearing premise is visibly supported now, proposed, or unknown; proposed or unknown premises establish no actual capability, authority, Work, provision, delivery, acceptance, availability, or use.
- The same decision-changing parity questions are applied to every way; an omitted group is justified by irrelevance to this choice.
- A decision-reversing unknown remains a gap or probe. Protected conditions are not hidden inside one unexplained scalar.
A.19.CPMorA.19.SelectorMechanismis used only when its evidence-gating, incomparability, abstention, or set-return contribution is needed.- Only complete-enough whole ways enter the finite
OptionSet;C.11alone makes the local choice and emits theChoiceResult. - The rule for retaining the comparison prevents a duplicate episteme when an owning domain result already carries the complete comparison.
- A retained way names any first unsupported realization branch or the direct Method that can proceed; the comparison itself claims no realization or later Work.
- The reopen condition names a changed premise or result that can alter the comparison or choice.
Recognition check. The practitioner can recover one common result from a build/buy/provider/AI list and identify a decision-relevant hidden difference or duplicate. An arrangement taxonomy is unnecessary.
Assurance check. For a consequential use, inspect each actual premise under its direct evidence and authority patterns, examine any plausible omitted burden that could reverse the choice, and verify that the selected domain assurance and acceptance Methods remain outside C.38.
Anti-patterns
- Labels as option kinds. Build, buy, provider, internal, reuse, or AI are compared as peers without resolving their contents.
- Different results in one table. One row supplies a System, another a service, another a capability, and another an episteme, while all are scored as if equivalent.
- Universal arrangement schema. A local parity view becomes a mandatory ontology, graph, taxonomy, or workflow for every domain.
- Modal-to-actual leap. A possible-future row is treated as capability, authority, Work, provision, purchase, delivery, acceptance, or availability.
- Score before parity. Numbers conceal different assumptions, missing burdens, or protected conditions.
- Endless generation. Open-ended candidate invention is kept inside
C.38. UseC.39for a missing obtaining explanation orC.40for feasible variation and examination; addC.18for archive/front claims andC.19for live-pool policy only when those contributions are needed. - Choice by formatting. The most detailed or first row is treated as selected before
C.11. - Domain takeover. The common Method absorbs architecture, procurement, contracting, finance, legal, artistic, operational, or assurance decisions.
- Duplicate account. A domain Method already returns the full comparison, but a second generic episteme is added anyway.
Consequences and trade-offs
Rationale and SoTA use
The practice question is how to construct several decision-ready ways to obtain one named result without implying a choice, treating possible futures as facts, or imposing a domain-specific arrangement schema. Begin with the receiving result and use, expand each candidate under consideration into a complete-enough possible-future way, and expose the same decision-relevant burdens and limits. The resulting finite OptionSet is an input to C.11 when local choice is current.
An initial table may contain labels such as internal, provider, buy, build, partner, procurement vehicle or lifecycle stage. Scoring those labels before recovering their contents can mix unlike alternatives and hide integration, evidence, support, capability, custody, consequence, recovery and exit burdens. C.38 repairs that defect by fixing one result and receiver, resolving every label into a whole possible-future way, applying a bounded parity basis, marking unknowns without inventing current relations, and stopping before local choice or actual Work.
The common comparison depends on its domain inputs. Systems Engineering supplies the engineered-System conditions, specialist criteria, realization and assurance; the recording Method supplies artistic, participatory and production requirements. Those differences determine what a complete way must include. When a domain Method already returns the sufficient comparison, use that result directly.
In formal appraisal, the expected result of continuing current arrangements supplies a baseline even when it falls short of the objective. In procurement, the buyer can commission knowledge and skills transfer alongside the immediate deliverable and specify how transfer will be assessed. These practices answer different questions in the comparison. HM Treasury, The Green Book, chapter 4 · Cabinet Office, The Consultancy Playbook, Knowledge and skills.
Reopen the affected candidate or parity question when a decision-changing premise changes. Reconsider the comparison approach when an available direct Method supplies the needed result more economically or when evidence shows that the current comparison hides a consequential burden.
Relations
- Builds on: the direct pattern for the sought result;
A.10for evidence and bounded reliance on actual premises;A.19.CPMandA.19.SelectorMechanismonly for comparison mechanisms that are needed; andC.2.1when a standalone comparison episteme must persist. - Hands to:
C.11only after a finite complete-enoughOptionSetexists.C.11owns choose, reject, probe-again, and reroute results. - Coordinates with:
C.39for a missing obtaining explanation;C.40for material development and its conditional coupled problem-and-solution entry;C.18for archive, front and possibility-space claims;C.19for pool policy;C.32for architecture-candidate synthesis;A.22andC.30only when a separate Structure or architecture question is current; and direct domain Methods for realization, production, provision, acceptance, assurance, and use. - Question-change boundary with A.15.9: use
A.15.9only for one missing or unqualified bounded result from another practice inside a way. Stay inC.38when the question is several complete ways to obtain the same receiving result. A contribution request or return is never the whole way by implication. - Keeps outside: universal arrangement kinds, graphs, taxonomies, schemas, workflows, authority transfer, open-ended search, local choice, and all actual Work or obtaining relations inferred only from possible-future rows.
C.38:End
Find and Develop a Way to Obtain a Result
Type: Method pattern Status: Stable Normativity: Normative unless marked informative.
Problem frame
Use this when you know what result is needed but cannot yet explain an adequate way to obtain it, connect it to another use, or change the way when a relevant condition changes. A diagnostic team may need to explain recurring errors; a recording workshop may need one preparation method to serve movement analysis and demonstration viewing under different timing requirements. “Use AI”, “get a specialist” and “synchronize” name possibilities without supplying the missing actions.
The working reader is anyone developing a way of acting for a practical use. No framework-author role or pattern format is required. The subject is the obtaining way and the connections needed for that use: how actions, inputs, conditions and support make the result possible. Developing a Method is one use; explaining the relations among independently usable Methods is another. They need not form one composite Method.
Begin with the sought change and what would distinguish its attainment. Reuse an adequate Method or result where one is available. Otherwise recover operations from relevant practice or theory, follow them on target material, construct the missing connection, and explain enough for the intended use. When that use includes later application or adaptation, keep the reasons, conditions and conditional returns available as well as the actions.
Methodological synthesis constructs or changes a way of working, including the conceptual work its operations require. Determine what each operation acts on, what its result means for the next operation and what counts as success. Use concepts that already settle those choices. If the sources attach different meanings to an input, result or condition, F.0.2 helps establish what their claims support. Use a common premise where the accounts agree. A supported difference can change which action applies. If a premise remains open, change how the evidence will be obtained or limit the conclusion that relies on it. A conceptual synthesis can also be the complete result of a separate inquiry into what the sources warrant.
The first result can be one sufficiently explained candidate, a recovered applicable Method, or a precise unsolved operation and a worthwhile next question. A connected account can extend that result when several uses need it. This transdisciplinary guidance does not supply missing professional knowledge, instruments, skill or permission, and a conditional construction does not establish empirical effectiveness.
Do not start this development when a qualified existing way already answers the current use. C.38 compares several complete-enough ways for the same result. C.40 develops and examines available material through variations. C.32.MWA supplies an architecture account when the question relates Methods to Work, objects, means, capabilities and development across several structures. If the intended result itself is too unclear to distinguish success from failure, clarify that local question first; a settled ultimate purpose is unnecessary for a sufficiently clear local result.
Problem
A desired result and a plausible label leave a gap. A found source can describe something successful without showing how to reproduce the relevant contribution under different conditions. Conversely, a useful description may be rejected because its vocabulary differs from the receiving practice.
The practitioner needs to construct and examine the obtaining explanation. Listing sources or drawing arrows does not do that work. A usable connection may need an input conversion, a choice, a shared condition or a return after failure; repeating each available operation can leave that connection unexplained. Detailing every action is also wasteful when one unknown join decides feasibility and the other operations are already qualified.
Forces
Solution
Recover how a relevant practice obtains its result, locate what fails or is missing on the receiving material, and construct the needed connection. Explain the resulting way for its intended use and, where change is part of that use, derive what a changed condition reopens. The following questions can return to one another; they are not mandatory stages for an already sufficient answer.
State the local result and try the cheap exit
Say what should change or become available, for whom or for which use, and which observation would distinguish an adequate result. Include a constraint when violating it would defeat that use. “Improve the recording” is too broad if the live question is whether movement and sound remain within a stated timing tolerance.
Establish what would make the answer sufficient before narrowing the work to a familiar operation. Follow the intended receiving work far enough to explain what someone needs to obtain, distinguish or do next, what makes that difficult, and how the proposed contribution would help. When the subject is still open, ask what could have the property or undergo the change in question. Use C.39.QR:4.1 to connect the material supplied to that subject before choosing the next work. An actual or stated case can supply a premise or reveal an alternative. A conceptual distinction can be the needed contribution: show which interpretation, inference or choice it changes. A description of a situation without that difference does not yet explain its usefulness.
Look first for an applicable Method or an available result that supplies this continuation. Inspect the operation and its qualifications, not just its title. A familiar adequate way can be used directly: this comparison need not become a separate document, experiment or narrated preliminary stage. An available partial result may also let the work stop. If only the next inquiry must be chosen, do not demand the explanation needed for full deployment.
Keep this receiving use as the basis of sufficiency while selecting a smaller operation, asking another contributor, or revising an example. The local question says which contribution is wanted; it does not silently replace what that contribution must make possible. A legitimate change of purpose changes this basis explicitly. Before concluding from the returned contribution, follow it into the receiving continuation: what can now be done, and does an indispensable transformation, distinction, coordination or ground still have to be invented? This is a way to discover the missing question, not only a check after someone has already named a defect.
Retain an adequate supplied contribution. When a connection is missing from the way itself, construct it through sections 4.2–4.3. If an adequate method is already known, recover the description or access that the receiving performer needs. For a job, obtaining an agreed product from a supplier can suffice. When a particular holder must gain the ability to do the work, identify the missing contribution and how the holder will use it under working conditions (E.23.CDI:4.2.1). When access, equipment or permission is known to be missing, obtain it from the responsible provider or choose a workable alternative. A.15.5 tests the full kit when the decision is whether this job can start. When several such contributions must fit together, the mantra in C.32.MWA:4 connects finding or developing a way to work with making it available, preparing the job and checking its actual use. Use C.40.JD if the method, the performer and their working conditions need coordinated development. C.11.DUA selects further inquiry when its attainable answer can justify changing the decision.
Follow a neighboring practice to the missing connection
Search through the sought result or unresolved change. In a useful source, recover:
- the input on which an operation acts;
- what the operation does and what it obtains;
- the conditions and support for that contribution; and
- the difference between the source situation and the receiving use that could defeat the transfer.
When the source offers a connected proposal, recover what the connection contributes before distributing its ingredients among familiar methods. Start with what someone is trying to make possible, follow the actions and intermediate results, and retain the choices, revisions and conditions that make the whole useful. A remembered term or a known defect can guide a search but should not determine everything extracted. Use B.5.PI when ordinary use has not yet yielded a question, and B.5.RC when the source's construction itself needs recovery. One passage may leave a needed operation unwritten; name that gap without treating the rest of its contribution as absent.
Keep three things distinguishable during this recovery: what the source proposes, what supports the proposal's use under its own conditions, and what the receiving situation additionally needs. A source can supply a useful way to generate or revise a proposal without supplying enough grounds to enact it. Conversely, a limitation in one inference can leave its constructive operations useful. F.0.1 recovers an action-changing source meaning; F.0.2 is needed only when comparing accounts can change the answer. Apply the direct reliance or subject method to a relied-on claim. Do not turn every extraction into a new synthesis, assurance study or framework.
Follow the described operations on one available piece of target material, by inspection or a permitted small application. Stop at the first needed input, operation, relation or result that the account does not supply. An inspected trace is not a performed application.
For example, a fault-investigation Method can compare error and non-error cases under comparable conditions. On the target records, “compare cases” may fail before comparison begins because the inspection cases cannot yet be connected to their setups. That missing connection is more useful to pursue than the broad request “find a diagnostic technology”.
Construct an explained candidate
Start from the recovered operation and locate what prevents the receiving result: an unavailable input, a case-specific choice, an incompatible intermediate result or a failed condition. B.5.RC helps recover a described construction; B.5.TU helps construct a calculation, prediction or other operation using an available theory.
Work backward from the needed result, then forward through a proposed way. Say what the receiving action must obtain and why. Follow that requirement back to what an available operation can supply. Where the two do not match, construct a transformation of the input or result, replace the failing operation, or change the arrangement. For example, a temperature result in Fahrenheit does not yet supply a Celsius threshold check: recover the units and use a justified conversion before applying the threshold. If the unit is unknown, arithmetic cannot repair the missing fact. If two records refer to different objects or intervals, unit conversion alone cannot join them.
Follow the proposed way forward on concrete material. At each consequential join identify what arrives, how its meaning, scope, units, state, precision and conditions are interpreted, what the receiving action does with it, and what permits that action. Explain the transformation where it is missing. A qualified supplied operation can remain intact when the practitioner can obtain its result under the required conditions. Compatible data types alone do not establish that two versions, intervals, assumptions or resource demands fit together.
Keep source-supported operations, proposed correspondences and new inferences distinguishable. Name what would defeat a consequential inference. If connecting inspection cases to setups requires a field that was never recorded, the candidate cannot justify the intended comparison. Recover the information, find another linking operation, or return a narrower result. A failed correspondence does not show that the source Method is useless in its own conditions or that the receiving problem is insoluble.
When several cases suggest a common way, compare the roles, relations and difficulty that explain their outcomes, rather than the words or objects they share. Propose the common operation and work it back into the cases: which means realize each role, which conditions are retained, and which difference requires a branch or defeats the correspondence? Remove a distinction only when doing so preserves the consequential actions and results; retain different concepts or methods when their differences change the work. A further unlike case can challenge the proposed generality. F.0.2 supports semantic synthesis when reconciling accounts is itself needed; C.39.RO develops a reusable operation from retained premises and variable inputs.
A faithful source extraction can finish without containing this generalization. When the receiving work asks for the general way, its construction remains authoring work; do not wait for a source to state the finished answer or present a proposed synthesis as that source's claim. Explain the common difficulty, the operation that addresses it, and the cases that support or restrict its use.
When learning is also required, derive practice from what the performer must coordinate, distinguish or control and the errors that defeat the result. Similar learning arrangements follow only as far as those demands are shared. Training is neither a substitute for explaining the way to act nor a compulsory addition to every method description.
Identify the connection's relation
Select only the structures whose relations change the use; C.30.ASV:4.5a develops that selection. An ordinary explanation can suffice. Use A.22 when a stronger structure claim is needed.
- Result use and continuation. For a required order, establish what makes that order matter. One result may supply an input for the next operation. A protocol may instead prescribe the order of two independent judgements. For work that can overlap, establish the independence actually needed and any conditions shared by the results or the performances. For alternatives, state the distinguishing condition and what happens when it is unknown or none applies. For iteration, say what the last result changes and when another attempt is unwarranted. A fallback needs a recognizable trigger and a result still useful to the receiver. A.22.CGUS develops the question of currently possible continuations; its formal qualification is needed only for the stronger claim.
- Composition and enactment. Call the connected operations one Method when the proposed whole has its own reusable action, applicability, result and constituent relations under A.3.1 and B.1.5. When an actual action performs part of encompassing Work, B.1.5.EW recovers that relation and the whole's constraints. Neither relation is just temporal order. A list of independently usable Methods can explain result uses without asserting one whole.
- Generality, variation and replacement. C.39.RO separates variable inputs from retained premises and develops a construction into an operation usable on another occasion. A permitted parameter change can remain a variation of one Method. Directed refinement between Methods additionally needs the comparison in A.3.1:4.7–4.8: participants, applicability, premises, action, intended result, constraints and acceptance conditions, with a stated direction and preserved meaning. A replacement also needs its affected uses and possible losses. Ordinary comparison suffices when no stronger identity or refinement claim is needed.
- Use across domains. A calculation, instrument model, observation method and access arrangement may support the same work. Explain the result each supplies and its conditions. That dependency neither makes one discipline a part of another nor assigns it a universal rank.
These relations can intersect. A calculation may constitute part of preparation while its result supports several independent uses. Name the relations separately. If a diagram calls them “vertical” and “horizontal”, state its orientation and represented relation under A.22:4.3a; the axes themselves supply no meaning. FPF, DPF and LPF name framework scope and basis, not a ready-made stack of Methods.
Repair the difficulty the case exposes
The following remedies concern recurring failures of construction or connection. Select the question that fits the difficulty, follow its dependencies, and stop at the sufficient result; this is neither a complete taxonomy nor a compulsory traversal.
Two diagnoses may concern one failure. An input conversion can be both unexplained and unavailable to the intended performer. Repair the relevant dependency and return to the original result. A second label does not require another pattern. A consequential difficulty absent from the table still calls for recovering the failed relation and constructing its remedy.
Keep an unavailable domain operation visible. General guidance can identify a required reset, calibration, joint movement or negotiation without possessing a means to perform it. A supported conditional construction with that gap may be the useful result.
Refine only what can change the answer
Which constituent operation needs further work?
Start from the decision about the Method as a whole. What must the organized operations accomplish for that decision? If one supplies an input to another, establish the condition the receiving operation relies on. The whole may instead require independently obtained judgements, a permitted ordering, a choice of alternatives or a shared condition. An established bound can already settle the decision even while the exact value or the way of obtaining it remains unknown. When the decision is still open, identify what is missing. If a suitable operation exists, its performer may need an input, access or an assignment. If the available operations cannot produce the required result under the stated conditions, develop the missing operation or change how they are connected.
To explain a composite Method, identify the repeatable action of the whole and the result it obtains. Show how the constituent operations together obtain that result under their operating conditions. Where one Method supplies an input to another and each has its own result, describing that exchange may be all the receiver needs.
To compare two models of a Method, name the question both must answer. What ways of carrying out the Method does the proposed model allow? Show that at least one is possible and that each satisfies the earlier model's assertions when read in its terms. A strict refinement also rules out an answer the earlier model allowed. Explaining the parts can leave the allowed answers unchanged. B.5.MR:4.1–4.2 gives the comparison for other modeled entities and inquiries.
Compare further inquiry with using the adequate current answer, narrowing the sought result and stopping. Include the effort and displaced work. A technically interesting uncertainty need not be resolved when none of its attainable answers would change the current decision.
Explain the way for the use that must follow
For a small use, a few sentences can state the candidate way, its supported and proposed parts, decisive unresolved condition and useful continuation. No special record is required. One explained way can finish the search. If complete-enough alternatives need comparison, use C.38; if feasible variations need examination, use C.40.
When another occasion or performer must use the way, choose the needed explanation by that use: performing an operation, learning it, comparing it, combining it, or changing it can require different detail. State the preparation and domain knowledge that the reader may bring. Explain the new action or inference that the account promises to supply; do not silently make the reader invent it as a prerequisite.
Connect the needed input to its source and the resulting action to its reason, expected result and continuation condition. Where “analyse” or “coordinate” leaves a consequential step unknown, recover that step backward from the required result and follow it forward on the case, as in section 4.3. Retain an adequate supplier when the reader can obtain and use it. Stop elaborating when the declared reader can do the intended work under the stated conditions; the explanation need not teach all professional competence.
For several independently usable Methods, explain where results pass between them, which conditions are shared, which uses can proceed independently and what a failed or changed condition returns to. Preserve each Method's direct use. An account of those relations need not assert a new Method or make the reader perform every described activity. A repertoire's selection guidance and compatibility can matter even when no composition is proposed.
Choose the description form after identifying that content. Ordinary prose, a worked calculation, a diagram, a procedure or a combination may suffice. A pattern is one possible form; its template is not the general Method-development process. A short reminder can point to a developed explanation when its reader can reach it. Keep any condition needed before the first action in the short entry itself.
Carry the explanation from the available inputs to the result the recipient needs, showing why the contributions work together and which conditions matter. Use a complete worked case when the recipient needs to see those decisions in use. For reuse or adaptation, explain what may vary and why the reasoning carries across the claimed range. A derivation can already cover that range. Try a changed condition under section 4.6 when it can distinguish an unresolved possibility, or perform the trial when it is part of the commission. When using a reader’s attempt, establish which choices the explanation supplied and which the reader made.
Keep three questions separate: whether the construction and its joins are coherent under their assumptions; whether it fits the intended result and situation; and whether its benefit warrants the effort, learning, support and change it requires. A derivation, a worked example, a prototype and an actual-use study support different claims. Select the needed evidence; these questions do not mandate three separate studies. A stipulated example explains a conditional construction, not empirical effectiveness or successful practitioner transfer.
Make the grounds recoverable in the same explanation:
- whose need is being addressed and on what basis;
- whether the proposal concerns a Method, a relation among Methods, a description or support;
- why consequential development choices were made;
- which actions and conditions the reader receives; and
- what supports each claimed consequence.
A final successful calculation cannot supply a missing reason for choosing the arrangement. Follow a gap back to the affected need, choice, instruction or claim and repair it or narrow that claim. This adds no compulsory record or fixed development sequence.
Include necessary training, equipment, assistance, access and recurring burden in the arrangement being considered. A tool can perform an operation without supplying a missing rule. Success with expert help does not establish unassisted use. Obtain missing support, develop the capability or restrict the promised use where necessary.
For an assurance-bearing use, distinguish a candidate account from an admitted U.Method under A.3.1 and its MethodDescription under A.3.2. A written sequence establishes neither formal composition, actual Work, capability nor permission. Apply the relevant rule only when the stronger claim is needed. E.4.CM applies this general construction and explanation to publishing connected Methods in FPF or a DPF; it is unnecessary for an ordinary practical description.
Change the affected way and retain what still works
Begin with the difference that matters: an input value, intended result, tolerance, available means, performer preparation or condition of use. Identify which action, connection or claim relied on the old condition. Follow its result uses and shared constraints far enough to determine what the change can affect. A changed whole condition may alter a constituent action; a changed receiving condition may leave another independent use intact.
Distinguish doing the same Method again with permitted inputs from changing the Method, and both from correcting its description or revising a claim about it. Use A.3.1:4.7–4.8 when identity or directed refinement affects the decision. Repeating an inadequate preparation does not by itself develop a better preparation method.
Construct the missing adaptation through section 4.3, or use a qualified alternative. For a constituent replacement, B.1.5.RS compares the affected encompassing uses, including shared or simultaneous demands. Do not infer that a local improvement preserves the whole or merely moves its burden elsewhere. If the changed result can alter which problem is worth pursuing, C.40.CD develops that question with the way.
Follow the proposed change through a complete affected use. Reuse a calculation, observation or earlier judgement only while its inputs, meanings, conditions and claim remain applicable. A numerical bound may remain valid while a tighter requirement makes the receiving use fail. Retain independently usable branches whose conditions still hold; neither reject them with the failed branch nor infer that every branch survives.
Update the explanation that later work uses: the changed action or relation, the reason, its supported conditions and what calls for another return. A reader should be able to reconstruct the changed case from that account. If no adequate replacement is available, name the unresolved operation, return a useful narrower result or justify stopping. Development can finish with that answer; it does not guarantee invention.
Archetypal Grounding
Explain recurring inspection errors
In this constructed case, the requested result is a source-traceable account distinguishing rival explanations of recurring errors. An existing service Method compares readings with configured limits and reports exceptions; it cannot answer the new explanation question. A stipulated neighboring fault-investigation Method compares error and non-error cases under sufficiently similar conditions.
Following that Method on the permitted local records reveals that timestamps alone do not establish a shared setup. The team proposes connecting cases to equipment, shift and setup records before comparing them. This could prevent a setup difference from being attributed to the inspected object. The connecting inference fails if the relevant setup cannot be recovered.
The resulting candidate is: recover setup-linked cases, compare eligible error/non-error cases under the specialist Method, and return what the comparison supports and leaves unresolved. Result checking can use an existing qualified operation. The remaining question is whether the permitted records can recover enough comparable setups.
That one explained candidate and its gap suffice to decide whether a bounded records inquiry is worthwhile. There is no diagnosis yet. A separate delivery or service comparison can later consider who could perform the investigation and its full burden; this search neither grants data reuse nor establishes that investigator's availability.
Preserve a performance's timing relation
A workshop wants to use a recording for timing feedback. Its constructed acceptance condition permits at most three milliseconds of movement–sound mismatch. If an available qualified recording Method already preserves that relation under these conditions, the workshop uses it and stops.
Suppose instead that the available picture and sound tracks have separate time readings. A supplied synchronization account aligns one event visible and audible in both tracks. The workshop follows that operation on the supplied marker readings:
Subtracting the initial 40 ms leaves mismatches of 0, 2 and 5 ms. Thus aligning the first event does not meet the stipulated limit at the last one.
The workshop first adjusts the constant shift using all three markers. Their offsets are 40, 42 and 45 ms; centering the smallest and largest gives 42.5 ms. Subtracting that shift leaves residuals of -2.5, -0.5 and 2.5 ms, each within the stipulated marker tolerance. A different constant shift therefore supplies a candidate correction without introducing a changing time correspondence.
The search returns that constant-shift candidate and the support question: does it preserve the required relation over the interval used for feedback, and can an available qualified editing operation apply it? A suitable calibration result or a narrower qualified interval can close the gap. Fitting the three markers does not establish adequacy between them or an achieved corrected recording. All numbers and professional inputs here are constructed, not device-performance evidence.
For repeated correction work, C.39.RO:5.1 develops this case into an operation on any nonempty finite set of marker offsets. It derives the correction and the smallest achievable worst residual, so a tighter tolerance can return an impossibility result and prompt a different correction method.
Build and change connected studio methods without a framework
Continue the constructed timing case. A studio practitioner must prepare a recording for two independently available methods: movement analysis needs at most 3 ms of movement–sound mismatch, and demonstration viewing permits 8 ms. A recording engineer separately supplies a justified bound: the relative offset stays within 40–45 ms throughout the relevant interval. A qualified editing operation can subtract a constant correction without adding timing distortion. These are independent professional premises; fitting three markers supplies neither.
Construct the needed preparation. Work backward from what each receiver needs: a recording with a timing qualification for its interval, not merely a correction number. The marker offsets 40, 42 and 45 ms from section 5.2 suggest a constant correction. With smallest and largest offsets L and U, any constant c has a worst endpoint residual at least (U − L)/2. Choosing c = (L + U)/2 attains that bound. Here c = 42.5 ms and the bound is 2.5 ms. C.39.RO:5.1 develops the reusable calculation and its retained assumptions.
To turn that calculation into preparation, connect source tracks and corresponding marks to the offset calculation, use the separately supplied interval bound to qualify the correction beyond the marks, apply the qualified editing operation, and retain the recording's interval and timing qualification for its recipients. Over that interval, subtracting 42.5 ms from an offset in [40,45] leaves a residual in [−2.5,2.5]. The mathematical reasoning explains the bound; the engineer's premise and editing qualification supply the professional connection. Without either, return that missing basis rather than call the recording qualified.
Calculation precedes editing. During a preparation episode, it also constitutes part of the encompassing preparation being performed. The order and encompassing relation answer different questions. A reusable preparation can qualify as a composite Method under B.1.5 when its applicability, action, result and constituent conditions are established; a four-step label list alone does not establish this.
Explain what a later practitioner does. For these supplied premises, the instruction is: identify the receiving interval and tolerance; obtain corresponding offsets and the justified interval bound; calculate the centering correction and its bound; proceed only if the bound meets the receiving tolerance and the editing operation is qualified; apply that operation and return the corrected recording with the interval and qualification needed by its receivers. Missing correspondence, interval support or editing qualification returns that question. The general explanation tells the practitioner why these inputs are necessary and why the stop is there.
The conditional bound of 2.5 ms meets both receiving requirements. Analysis can produce a movement judgement, and viewing can serve its different purpose. Their common input neither makes preparation a part of each receiving Method nor makes analysis and viewing one Method. They may use the same immutable recording independently under the stated access and resource conditions. Preparing a later recording is another application; developing another preparation method is different work.
Change the analysis requirement to 2 ms. The two marker offsets 40 and 45 already preclude meeting 2 ms with any constant correction: their separation is 5 ms, while two residuals bounded by 2 ms can span at most 4 ms. The calculation remains correct. Its result now rules out the analysis use under this correction family; repeating the preparation cannot remove that limit.
Viewing remains qualified at 8 ms under its unchanged conditions. Return only the unmet analysis requirement to development: seek another warranted correction, another acquisition method, or a justified change of the result sought. A varying correction is a candidate direction, not an available qualified solution. If a replacement is proposed, examine its interval basis, editing realization and affected uses through B.1.5.RS before relying on it. Method development can proceed while earlier qualified recordings are viewed.
The practitioner has obtained a conditional preparation, explicit joins to two receiving uses, an executable instruction subject to named professional prerequisites, and a selective return after change. No FPF/DPF publication or pattern form was needed. The example establishes neither improved movement judgements nor low learning cost nor the prevalence of this studio difficulty; those empirical claims require other evidence.
Obtain a usable way from a source discussion
Start with the receiving work. An author receives a discussion about taking the consequences of reasoning seriously in action. The needed result is an explainable way to examine a proposal and decide how to continue. The author starts with what the discussion makes possible, then follows the proposed action to its requirements. This example adapts the first section of A. Levenchuk's historical discussion Method Council, 19 October 2022, keeping its constructive proposal and the added receiving conditions distinguishable.
Recover the positive construction. The source describes a person proposing an unusual action, exposing it to reasoned objections, answering an objection or changing the proposal, and then acting when no further objection is found in the discussion. It also distinguishes deciding what is to be learned from designing how to teach it. Merely extracting the words “logic”, “criticism” and “action” would lose the revisable proposal and the connection to a next move. Treating the anecdotes as proof of a universally successful method would add support they do not supply.
The receiving author follows the proposed way on the present material: what would the next practitioner receive? The source supplies proposing, criticizing and revising. The instruction to proceed still needs the receiving action's applicable requirements and grounds. The absence of further objections does not show that the prerequisites for the proposed action are satisfied. Following the receiving use exposes the question: what further grounds would justify this action?
Construct the missing connection. Keep the proposal and reasons for revising it. Name the result wanted from the proposed action and recover the conditions on which its usefulness depends. Connect each consequential claim to the grounds on which this use can rely through A.10. Obtain a separate assurance answer through B.3 only when that stronger claim is required. A critique can refute one reason, suggest an adaptation, or reveal an unanswered question; determine which it does before accepting or discarding the proposal.
When complete-enough alternatives are available, C.11 compares them on the same basis. An available next inquiry is worth doing when its attainable answer could change that choice and warrants its burden. Otherwise choose under the adequate current basis or name the condition that remains unresolved. An absent indispensable condition does not become satisfied because inquiry is costly.
Return a usable instruction. For this use the author can now say: state the proposed result and action; expose the reasons that matter; examine substantive objections and revise the affected proposal or reason; establish the receiving action's necessary conditions; compare the available continuations; return the selected continuation with its grounds and the condition that would change it. A sufficiently qualified existing proposal can go directly to use. If the useful result is only a proposal worth testing, keep that narrower conclusion rather than claim completed performance.
Apply the instruction to a receiving proposal. Use the studio proposal in section 5.3: prepare one recording by subtracting a constant correction, then use it for movement analysis and demonstration viewing. Recover the two required results and their 3 ms and 8 ms tolerances. The independently supplied interval bound and qualified editing operation support a worst residual of 2.5 ms after the 42.5 ms correction. These are the grounds for using that preparation under the two stated requirements. If the interval basis is unavailable, the same three marker readings justify only their local residuals; the next useful inquiry concerns the interval, and “nobody objected” cannot fill it. This applies the source-inspired proposal–criticism–revision connection while the established mathematical and professional suppliers provide the receiving grounds.
The author has constructed an instruction and shown its conditional application. This does not show that a person has learned the method, that the recording was prepared, or that every proposal needs a formal assurance case. A content review asks whether the source contribution and added connection are faithfully and sufficiently explained; a claim of benefit in another practice needs its own evidence.
Change one receiving condition. Tighten movement analysis to 2 ms as in section 5.3. The 2.5 ms bound remains correct, but it no longer supports that use; the 5 ms span of the marker offsets rules out meeting 2 ms with any constant correction. Reopen the analysis proposal and develop another warranted preparation or return the unmet requirement. Viewing at 8 ms remains qualified under its unchanged conditions. The proposal–criticism–revision Method also remains available for examining the replacement. If a changed calibration instead undermined the shared interval bound, reconsider both receiving uses that relied on it. Whether the uses are independent depends on what each relies on, not on their different names.
Bias-Annotation
Familiar sources and impressive labels attract attention before their operations are inspected. Search by the needed change and include an ordinary way that could obtain it. A preferred analogy should face the target condition most likely to defeat it; otherwise the explanation becomes a story selected after the preferred answer. An author's successful private reconstruction can also hide what another performer still needs to invent. State that missing instruction or preparation.
Conformance Checklist
- The sought local result and a distinction between attainment and failure are recoverable.
- An adequate existing Method or result can finish the use without invented alternatives.
- Source operations and target differences are inspected before transfer is asserted.
- A proposed connection explains its inputs, contribution and defeating condition.
- Further refinement is tied to a result-, feasibility- or burden-changing gap.
- The returned explanation, unresolved support and actual realization remain distinguishable.
- The next inquiry has a use that can warrant its effort, or the search stops.
- For a connected use, result transfer, composition, enactment, variation and replacement are distinguished where their consequences differ; independent Methods retain their direct uses.
- An explanation intended for later performance or change supplies the needed inputs, actions, reasons, conditions and returns without making the promised construction an undeclared reader prerequisite.
- A changed-condition claim identifies affected connections and retained uses; a correct construction, practical fit and worthwhile application have the evidence their distinct claims need.
Common Anti-Patterns and How to Avoid Them
Consequences
The reader obtains a usable next action before qualification or comparison has a complete candidate to consume. When later use needs it, the explanation also makes connections and selective change recoverable without private reconstruction. Source recovery and explanation take effort; decision-relevant refinement limits unnecessary work. Some developments end with an identified gap or narrower result. This Method guarantees neither invention nor acquired professional capability.
Architectural Rationale
Construction, explanation and change can concern the same obtaining way at different moments. The content needed to use or adapt it belongs with that way; a separate publication role is unnecessary. The conditional branches in section 4 preserve a small sufficient answer while supporting connected use when the missing relation requires more development. A complete lifecycle for an organization's method repertoire, or an architecture of its performance and support, remains a different question.
Unknown-way search and same-result comparison can finish at different points. One explained way or a decisive missing contribution can answer this question, while C.38 needs materially different complete-enough ways on a shared basis. Joining them would make sufficient one-way results harder to recognize.
Material development also has a different immediate operation. C.40 can begin with a known editing or generation action and return an examined variation. It uses this Method only when an obtaining operation is missing. Neither relation imposes a universal order of search stages.
SoTA-Echoing
For construction and connection beyond retrieval, compare operations and assumptions. Stacey et al. (2025), §§2.4 and 4.4 explain method knowledge as explicit how-to knowledge that connects inputs, reasoning and outputs while relying on its users and context. Adapt that connected explanation in §§4.3 and 4.5; keep the FPF distinctions among Method, description and Work. The source's situated method-in-use terminology is not adopted as an identification of these objects.
OMG Essence 2.0 beta2 is a concrete alternative when practices represented in its language can be composed using its common elements and merging rules. Use such qualified domain machinery where it answers the receiving problem. The general branch here also permits independently sourced operations and ordinary descriptions; their shared words or source units do not establish valid Method composition.
Ralyté, Koutsopoulos and Stirna (2025, published online) distinguish correctness of a modeling method, fit to its domain and practical worth. Adapt those different questions in §4.5, without requiring separate studies or making a modeling-method procedure universal. Cash, Daalhuizen and Hekkert (2023) supplies the complementary connection from need and development decisions to content and claims; an end-result test alone can leave those grounds missing. Neither source establishes the effectiveness of the studio construction.
For the question “How can an unexplained way become usable enough for the next decision?”, the selected line is source-guided constructive explanation with target-condition testing and bounded refinement. Trace a relevant professional operation on target material, make the missing connecting action explicit, and refine the explanation while that refinement can change the next decision. This exposes what label generation or source retrieval alone can leave hidden. One sufficiently explained way can finish the search.
The alternative is retrieval of an already adequate Method; it wins when it returns the whole answer more cheaply. The Puranam, Sen and Workiewicz v1 abstract supplies bounded evidence for separating retrieval of a strategic analogy from matching its causal structure. Adapt that separation in section 4.2; reject treating successful retrieval as target support. The reported comparison concerns human participants and GPT-4 on the study's strategic-analogy tasks.
Reopen this line when a current direct Method supplies the entire use with less effort, when target applications expose a missing constructive operation, or when the explanation burden exceeds what the next decision can use.
Relations
C.39.QR returns work through a part, case, model or other intermediate construction to the original question. It distinguishes what the obtained result warrants about the subject from what it completes for the requested use, and brings a missing obtaining way here. C.39.WA develops the common grounds and organizing relations of a whole-subject account, follows them into choices about parts, and revisits its scope after consequential findings.
C.39.MC continues a connected Method-development question after a result or changed condition. It preserves applicable qualification, composition, description and support results while returning only the affected question to its supplier. Use it when that continuation is unresolved; an adequate direct MethodDescription remains the sufficient return.
- C.39.RO constructs a reusable operation from a working case, retaining input meanings, application conditions and the reason for its result.
- B.5.RC recovers a described construction; B.5.TU makes an available theory usable for the working question.
- C.38 constructs comparable whole ways for one result after sufficient obtaining content is available.
- C.40 develops and examines material variations; it returns here for a missing obtaining operation.
- F.0.1 resolves an action-changing source-local meaning; F.1 selects sources whose claims, methods or counterexamples can change the answer. F.0.2 compares source accounts when a semantic-synthesis question is current.
- E.4.CM applies this general Method construction, connected explanation and change to authoring framework content. E.8 governs the pattern form used in that application; neither is required for an ordinary practical description.
- C.30.ASV selects the relevant structural views; C.32.MWA relates Methods, performance, support and development for a practical architecture decision.
- B.1.5 qualifies Method composition; B.1.5.EW and B.1.5.RS develop enactment and replacement. A.22.CGUS develops the currently available continuations and their conditions.
- A.10 governs a stronger evidence or reliance claim. A.3.1 and A.3.2 distinguish a Method from its description when that distinction is at issue.
- C.11 supports a local choice about the next inquiry or use. The direct professional Method supplies realization and its subject-specific checks.
C.39:End
Return Results to the Original Question
Type: Method pattern Status: Stable
Problem frame
Use this when choosing what work would answer a question, or bringing a result back from a part, model, example, sample, summary or delegated contribution. A question can change in the first reply, before any partial result exists. A later handover can carry a sound piece of work into a conclusion about a different subject. In either situation, recover what the recipient needs to understand, obtain or do and how the proposed work can supply it.
Recover the original subject and required result, explain what the intermediate work establishes, and supply the grounds for returning that result to the original use. Where the return exposes missing work, develop that contribution or obtain its supplier. The first useful result is a warranted answer to the original question, or an account of what has been obtained and what still has to be done.
Use an already adequate direct answer. A local operation can finish under its own established criterion when the assignment is local. This Method is useful where a consequential return remains uncertain; ordinary continuation under unchanged, understood conditions needs no repeated reconstruction.
Problem
A team can produce a correct result and still lose the work it was asked to do. An example answers a question about that example. A component test answers its stated test question. A model calculation answers a question within the model. The contribution to the originating question depends on further relations and grounds.
Two judgements are involved. One concerns what a result obtained through a working construction warrants about the subject. The other concerns whether the work result answers the original question sufficiently for its intended use. A sound first judgement may leave the second unresolved: an accurate estimate can be useful while the requested explanation, decision or operational capability is still missing.
The loss can persist through review when later performers inherit the smaller question from the candidate, task breakdown or earlier verdict. Repair requires recovering the encompassing question and constructing its missing answer. Another correct local result can contribute to that work, provided its role and limits remain visible.
Forces
Solution
Begin with the work the question calls for. Which practice, object or relation could have the property the recipient needs to understand or change? What material gives access to it? Choose an inquiry whose answer would advance that work. When a contribution comes back, establish what it supports, develop the missing connection and use it in the original work.
C.39 supplies construction of an obtaining way; the subject practice supplies its inference and composition rules. C.41 supplies the attention aid and its use. The operations below connect those contributions for this recurring return.
Recover the question independently of the proposed answer
Read the initiating words and applicable clarifications before choosing a way to answer. Distinguish the material the speaker supplies from what the question asks you to establish or choose and who needs the answer. If the subject is still open, work back from the conclusion or change sought. To explain why a practice repeatedly produces a result, identify how that result could arise and what would sustain its production across the occasions in question. Follow the relevant contributions and their organization. Distinguish the evidence that an outcome occurred from evidence for how it was produced and why it recurs. If several explanations still fit, choose grounds that would tell them apart where the difference matters to the commission. One occurrence can motivate a question about a class. A principle can be supplied for a decision about one occurrence. Where composition matters, recover the kinds of holon, their own constituent–whole relations and the range fixed or left to choose in the question. Where model abstraction matters, recover the inquiry, its interpretation and the distinctions the requested answer must preserve. State how an argument from a different subject, scale or model bears on that answer. A bound or a proof about one constituent can sometimes decide the whole question.
Choose the question before developing the answer. Read the speaker’s words in their conversational setting. A remark may invite acknowledgement, a request may call for an explanation, and an open invitation may leave the subject to the responder. Keep the particular question a proposed answer would settle beside what the speaker is trying to establish. Could that answer remain entirely correct in two admissible situations which require different conclusions, explanations or actions for the original purpose? If so, identify the missing distinction and its grounds before treating the answer as sufficient. Both answers can use the same vocabulary and concern the same topic. An accessible passage, calculation or reader trial can be a sensible starting point. Before letting its availability decide the next work, identify what a possible outcome would establish for the original question and which outcomes would leave that question open. An intermediate result may supply a premise, narrow a bound or change the support for an answer. It may reveal which source, experiment or collaborator makes the next inquiry feasible or less costly. Identify how the result could advance the commissioned work. When that connection is missing, establish it or choose an inquiry with a justified contribution. The issue can be the scope of a text, the composition of a holon, the abstraction of a model, or the kind of judgement being requested. If the grounds already settle the question, give the answer. If a different question would serve the recipient better, explain why and make the proposed change visible.
Keep the commission when choosing the next work or using a result. A first restatement can already change the task. A later choice of plan, explanation or evidence can introduce the same loss. Compare what a colleague could deliver by following your words with the initiating commission and its clarifications. Identify who asked, who is to do the work and who will use it when these differ. Keep the receiving action, the recipient, the types of entities and actions in view, and any ranges the conclusion has to cover. A commission may concern several kinds of holons, different ranges of composition for each, and a range of model abstraction. It may ask about what holds at each level, a relation between levels, or something that depends on several together. Naming the end points leaves that last connection to be established. If the commission leaves someone to choose the ranges or connections, retain both the choice and the person or agent entrusted with it. The person entrusted with that choice can use C.39.SA:4 to make it and establish where to stop. An author teaching the choice supplies the method and preserves the reader’s opportunity to choose. The eventual recipient may simply need the resulting decision.
When a reply would settle a different question, restore the omitted work in the reply itself. An account of where the change occurred can finish a request for its location. For a request to explain a recurrence or change a practice, use that location as a possible source of evidence, develop the explanation or intervention and carry it back to the practice. When describing the change helps the recipient, name the particular loss. “The restatement changed the kind of action.” “The example shows what happens at one scale.” “We still owe an explanation of how decisions across those scales affect one another.” When a concrete example helps, introduce it as an example and state which part of the argument it illustrates. Return to the assigned types, ranges and receiving action before selecting the next edit, inference or handover. Set aside the choices fixed only for illustration and retain any general reasoning the example contains. Can a colleague recover what to do, for whom, over which ranges, and what would count as completion? If a whole family of decisions was reduced to one familiar decision, say so at that level and restore the family in the next instruction. Naming a different instance leaves the same omission. Use C.39.GA to restore the general argument in the reasoning or response when its example or a narrower class has taken its place. For a request confined to that one case, answering it can finish the job. C.37.MM:4.1.1 develops the joint work when several models are needed.
Distinguish the requirement from the author's proposed decomposition or solution. Derive an initial expectation from the requested use and competent subject knowledge. A proposal to explain a discipline needs its organizing problems, principles, Methods and relations at a suitable depth. A proposal to restore a service needs the relevant functioning and operating conditions. These expectations have different substantive bases.
Keep consequential uncertainty explicit. Obtain a clarification when materially different readings require an authorized choice; meanwhile continue the content required under the viable readings. A provisional expectation can change as the subject becomes better understood. Explain that correction and preserve the original commitment unless its authorized basis also changes.
A reviewer at risk of inheriting a narrowed question forms this expectation before relying on the candidate's outline or earlier verdict. The expectation states indispensable content or conditions, not a second complete solution.
Questions: What must become understood, decided or obtainable? For whom and under which conditions? What boundaries were assigned? Which expectations come from the subject, and which are only a proposed way?
Recover what “all” covers
Use this move when taking all, every, always, never or a similar expression literally would change a decision. Echo the word as a question to yourself: “Every single one?” Try one plausible exception. Who or what would count, under what conditions, and what is claimed, asked or required of each? A membership rule can identify the domain without a list of names. Establish separately the property or action claimed of its members.
Read the utterance in context. A literal “everyone” can cover exactly the people already identified in the conversation. An emphatic “everyone knows” can express impatience or a wish to move on. Preserve that purpose. To skip an explanation on the strength of an audience's knowledge, establish what the people who need to act already know. An explicit requirement holds on its agreed terms, including who owes an individual or a joint result and which exceptions it permits. A possible exception tests the reading. An established exception within the stated domain refutes a universal factual claim. An actual breach leaves the obligation in force under the governing terms.
For “at all levels”, recover what orders the levels in this question. Within any relevant kind of holon, part–whole relations can give scales of composition. A System and a Method each have their own. Control layers can be ordered by how one sets goals or constraints for decisions in another. Model abstraction compares what models admit for the same inquiry. Recover the relation the speaker uses and the ranges included, even if the field orders its levels in another way. Checking every immediate part of a whole can stay at one composition scale. Examining work at each of several scales can leave some parts unexamined. Recover whether the question asks what holds at each member, what carries between them or what they achieve together. “Each can use a method” can allow a different method for each. Promising one method usable by all adds a shared choice. Here the domain identifies the members. The logical scope of a quantifier identifies the part of the proposition it governs, including any choice made within it. Include combinations of ranges when the question or the organization of the work requires them.
Use the reading supported by the words, the agreed work and the situation. When several readings remain, give the answer they share if it suffices. If they lead to different decisions, state the difference or ask the speaker about the boundary that decides it. For a factual claim about every member, give an argument that covers the domain or adequate evidence covering each member. To determine what was promised to all, the authorized agreement can suffice. To judge feasibility, examine the promised work and its shared conditions. To report fulfilment, establish the outcomes the agreement calls for. Rhetorical emphasis gives no estimate of the proportion. For membership of a context slice in an FPF applicability scope, use A.2.6. The people, parts or models quantified over can vary within a single such slice.
Identify the relation used by the partial work
Name the original subject and the working construction in ordinary subject language. Establish how work on the latter can contribute to the former. Use the relation needed for the next judgement.
Several relations can occur in the same work. A model of a component is a representation of that component; the component's relation to a system is composition. Keep both when the conclusion traverses both. An example can be an instance of a kind and illustrate an operation, with different consequences for each use.
Tracing a relation back identifies its other participant. Inferring a property of that participant requires an applicable rule. A correspondence can warrant selected inferences while leaving others open. For a many-to-one abstraction, recover the omitted distinctions when the returned question depends on them. The subject practice determines whether a deduction, empirical inference, construction or qualified analogy supplies the needed warrant.
Questions: What are the participants and relation? Which claim is being carried across it? What permits that inference under these conditions? What has been left out?
Establish what the performed work obtained
Examine the contribution and its grounds. State the claim or result it supports, its subject and conditions, and what was supplied by another source, performer or tool. A reader's additional construction can be useful; identify it as that reader's contribution when deciding what the received material supplied.
Carry a conclusion to the original subject only through the applicable relation and inference from section 4.2. If that inference is unsupported, repair it, obtain the missing ground or retain the narrower warranted conclusion. Preserve sound local work.
Ask what could have produced the same result. To accept a model for use, check its assertions against the intended use. To credit someone with developing it, compare the original assignment, models and help available with that person's or agent's actual choices. A correct submission may reproduce a completed form. It can pass a content check while its authorship and even who applied the rule remain open. A calculation trace can show application. A dated design decision can show who selected and justified a rule. Credit the choices the evidence identifies. The distinction matters when commissioning a design, assigning work or assessing what someone has learned.
A single feasible design can establish that the specified performance is possible in the model. One verified occurrence under the stated conditions can establish that it was achieved. If the claim says every permitted design has a property, a single feasible design lacking it refutes the claim. One verified late delivery likewise refutes “every delivery arrived on time.” For a claim that at least 95 percent arrived on time, count it against the specified population and period. For a requirement that every delivery arrive on time, record the breach. The requirement continues to govern until the responsible authority changes it. A simulated failure can expose a defect in a design argument. Reporting a failure at an installation calls for evidence from that installation. In each case the next decision depends on which assertion the result bears on.
Then make the distinct comparison with the original task. Ask whether the result now available supplies the required answer, explanation or ability to continue. For example, a valid estimate of a system's load can support sizing while leaving a requested operating arrangement undeveloped. Conversely, an adequately supported narrow estimate may fully answer a request for that estimate.
Keep the two conclusions explicit when they differ: what is established about the subject, and what part of the assigned work is thereby completed. Truth, membership, fidelity, coherence, coverage, usability and beneficial effect require the grounds appropriate to each judgement.
Questions: What did this work establish, about what, and under which conditions? What is justified about the original subject? What required result remains unavailable?
Reconstruct the encompassing result
Compare the result now available with the expectation recovered in section 4.1. Examine the relations needed to obtain the encompassing result as well as the contributions themselves. Follow inputs and intermediate outputs, shared conditions, alternatives and returns at the level where they affect the requested use.
Use C.39:4.2–4.5 to construct a missing obtaining way or join. When the requested result is an explanation of an organized subject, C.39.WA develops its common grounds and organization, then brings those grounds back to choices about the parts. A whole can contain several independently usable Methods; a connected account need not force them into one sequence.
For promised source recovery, apply F.0.2.RC to the promised contribution and scope. For a system or service, use its engineering and operating methods to establish the relevant whole conditions. Where the missing contribution is capability, equipment, access or authority, obtain or revise that support through its responsible practice.
Synthesis connects contributions to obtain a result. Generalization widens the range for which a claim is warranted. Model abstraction constructs or selects a model that leaves some determinations open while retaining the conclusion the receiving work needs; apply B.5.MR to compare its assertions and admitted answers with the starting model. Changing what a description displays can leave the same model assertions recoverable and need not change its abstraction level. Choose the operation the gap requires, and explain its result. A wider field name cannot supply a missing connection.
Follow a changed encompassing condition to the contributions that relied on it. Follow a consequential finding in a contribution back to the encompassing account. Retain unaffected uses on the basis of their dependencies.
Questions: Which organization or connection does the requested result need? What governs the choice and change of the contributions? What missing content or support prevents the return?
Perform the useful continuation and settle its reach
Choose the continuation from the gap: develop content, correct the relation or inference, recover a source, obtain support, compare an alternative way, or seek an authorized change of the task. C.11.DUA compares further inquiry or change with continuing on the available basis, including the effort of learning, coordination, execution and displaced work.
Complete the selected contribution and return its result through sections 4.2–4.4. A changed original requirement can require a different comparison; make that change recoverable. A missing contribution keeps the corresponding requirement open even when another useful part is finished.
At a handover or acceptance decision, give the recipient the warranted result, the use it now supports and the consequential remainder. Reconcile several returns through their joins and shared conditions. C.41.SR supports this reconciliation. The recipient can accept a sufficient result, request the missing construction, retain a qualified partial result or decide a justified stop within its authority.
The work ends for this occasion when that decision is supported at the requested scope and the chosen continuation has been performed or transferred to its responsible recipient. A partial return can complete an assigned subtask while leaving the encompassing assignment open.
Put the return into ordinary work
Use the questions where a difference in their answers can change the work:
- Before a consequential decomposition or representation choice, recover the original question and the contribution expected from the smaller work.
- At a scope-changing inference or handover, establish what the partial result warrants and what the recipient still needs.
- Before a whole-result judgement, compare content and joint conditions with the original requirement.
- After an interruption or a changed condition, recover the affected connection and continue from it.
C.41.OF arranges the occasion, access and competent participation. C.41.UA connects obtaining an answer, deciding what to do and performing the action. Keep the subject material and these operations available at that occasion; a remembered slogan may be insufficient when several relations are being traversed.
Use another competent reader when independence or an unprompted initial judgement is needed. Give that reader the original question and necessary subject grounds before earlier verdicts or a suggested defect. Freshness can expose an inherited narrowing; domain competence supplies the grounds for the expectation. Select the arrangement by the judgement required, and reuse sound prior results within their scope.
For recovery after interruption, A.15.10 returns the purpose, operative method, established results, unfinished work and changed conditions to a useful continuation. For people and AI agents, examine separately recognition of this difficulty, performance of the relevant operations, and benefit in the receiving work. Reopen the aid when it repeatedly fails to expose a consequential loss or creates work without improving the answer.
Archetypal Grounding
A field question receives a true calculation
An assignment asks for an account of mathematics and an appraisal of its completeness. The contribution is a correct multiplication. Its membership in mathematics and correctness support conclusions about that calculation. The required field account has a different subject and adequacy criterion.
Recover the intended reader and depth, then develop an account of relevant mathematical objects, questions, forms of construction and proof, and relations among branches. The multiplication can illustrate part of that account. C.39.WA supplies the constructive return from the missing organization to an explanation. A review of the calculation remains useful within its scope while the broader assignment continues.
If the original request instead asked only for the multiplication, its correct answer can finish the work. The difference lies in the governing question. The same content has different adequacy in the two assignments.
An aggregate estimate leaves operating coverage open
An internal support team asks for an operating arrangement that keeps two qualified responders available throughout 09:00–12:00. A staffing spreadsheet sums the available person-hours and compares them with the six person-hours required. It finds 6.5 available. The sum is correct for the declared intervals. It establishes an aggregate capacity condition; the receiving question also requires simultaneous coverage, the needed access and an authorized allocation of the people.
In this constructed case, A is available 09:00–12:00 with a fixed break at 10:00–10:30; B is available 09:00–11:00; C is available 10:00–12:00. All three have the required working capability, but C initially lacks access to the service system. The service manager has authority to allocate their stated time. Recover the required whole condition: at every point in the interval, at least two assigned, qualified and equipped responders must be available. Total hours alone leave the time overlap and access conditions unresolved.
Working backwards from that condition exposes the first unsupported connection: B and C must cover A's break, and A and C must cover the final hour. The manager obtains C's role-appropriate access from its responsible administrator and verifies that C can perform the required service operation. The manager then allocates:
- A+B from 09:00 to 10:00;
- B+C from 10:00 to 10:30;
- A+B from 10:30 to 11:00; and
- A+C from 11:00 to 12:00.
C's remaining half-hour is unallocated. Each interval has the required two contributors, with the fixed break, availability and access conditions respected.
The completed contribution is this jointly feasible arrangement and its provisioned access. The manager communicates the assignments, obtains the performers' acknowledgement and hands the arrangement to the duty lead for the stated period. Within the case's supplied facts, the answer to the original coverage question is supported. The spreadsheet remains a useful capacity calculation; the interval construction and completed provisioning supply the previously missing grounds. Service quality during operation requires its own observation.
Now suppose C's access becomes unavailable from 11:00. The available person-hours and their arithmetic are unchanged, while the final interval has only one equipped responder. Return to that connection. If an already qualified D is available and provisioned for 11:00–12:00, the manager assigns D alongside A and confirms the changed handover. Earlier intervals remain valid. If no such contribution can be obtained, the duty lead is told which interval remains uncovered and the original commitment remains unresolved; changing that commitment requires its authorized recipient. This is a different continuation from recalculating the unchanged aggregate hours.
Obtain a warranted population claim and use it for the agreed decision
An audit lead is asked whether this month's completed records meet an agreed criterion for commissioning a cause investigation. The criterion is a defect proportion above 5%, supported by a two-sided 95% confidence interval whose lower limit exceeds 5%. The commissioned decision concerns that population and criterion.
An analyst supplies a correct calculation: 40 defective inspections among 400, or 10%. Before returning it to the decision, the audit lead establishes what the inspected records represent. The register is reconciled with all completed records for the month. The selection log shows 400 independent uniform draws with replacement from that register. Every selected record was accessible. Inspectors applied the agreed defect definition, resolving ambiguous classifications against the records. In this constructed case these conditions are established; together they supply the sampling and measurement grounds that the ratio alone did not carry.
Each draw therefore has the same defect probability: the unknown fraction of defective records in the fixed target population. The analyst applies the Wilson interval for a binomial proportion. For 40 of 400, its nominal 95% interval is approximately 7.43%–13.33%. The returned population claim is an estimate with that qualification, supported by the selection process and measurement. The confidence qualification belongs to the statistical procedure's repeated-sampling performance.
The audit lead then makes the second judgement. The agreed criterion compares the interval's lower limit with 5%; 7.43% exceeds it. The owner receives the population, estimate, interval, sampling basis and criterion comparison, and commissions the investigation under the existing authority. The statistical result and the task's decision rule have supplied different contributions to this completed answer.
Changing the decision rule would leave the population inference intact while changing its use for the task. For example, a rule requiring the same interval's lower limit to exceed 8% would not authorize commissioning on this evidence: 7.43% is below that threshold. The estimate and interval would remain the same. Conversely, a selection log covering only one office would change the population inference itself and require a return to the sample design. The two returns have different objects and useful continuations. The established relation, the permissible inference and the receiving requirement determine which work to retain and which to reopen.
Bias-Annotation
A visible, easy-to-check result can shape the next performer’s understanding of the question. An author can also supply missing connections from memory while appraising the text. Recover the original use and distinguish the received contribution from that added work.
The observed substitution identifies a failure in the reasoning. Its psychological or computational cause requires separate evidence; the repair can operate on the question, relation, inference and receiving use that are available for examination.
Conformance Checklist
At the relevant occasion in section 4.6, answer from the work and its grounds:
- What material did the speaker give, and what must the answer establish or leave to the recipient? If it says “every”, every member of which domain, and what is asked about them?
- What relation connects the working construction to that subject?
- What does the partial result warrant, and what grounds permit the returned inference?
- Does that warranted result supply the original use, including necessary connections and shared conditions?
- What constructive continuation follows from any gap, and has that contribution or responsible transfer occurred?
A maintained aid uses these questions with the needed subject criteria. C.41.QC applies when the questions or means of answering them remain doubtful.
Common Anti-Patterns and How to Avoid Them
Consequences
The recipient obtains a qualified answer and a usable connection between the contributions and the encompassing result. The operation can expose missing content early enough to change authoring, assignment or acceptance.
Its burden is recovery of the question, examination of the consequential relation and grounds, and any warranted construction. Keep an adequate direct connection and reuse valid results. The benefit and cost in repeated work depend on execution and should be observed at the receiving use.
Rationale
The Method separates two relations: the warranted use of a result to make a claim about a subject, and the adequacy of a work result for a question and intended use. Each can hold without settling the other. This separation permits retaining valid contributions while exposing the work still needed.
The encompassing question constrains decomposition and the choice of representations. Their results can in turn revise the proposed understanding and obtaining way. Making that exchange explicit gives the procedure its constructive part: a failed return shows which content, inference or support needs work.
SoTA-Echoing
Stanley and Szabó examine how the conversational setting can restrict the domain of a literal quantifier. McCarthy and Carter's corpus study distinguishes exaggeration in conversation through the surrounding turns and the speaker's purpose. Section 4.1.1 uses these observations to recover the commitment before testing it. Its echo question is a practical prompt. On Quantifier Domain Restriction · “There's millions of them”
The practice question is how to preserve an encompassing question through partial work and warranted appraisal at a burden suitable for repeated use. The selected line is the explicit connection in section 4 of question recovery, typed inference, adequacy comparison, constructive return and attention at consequential occasions.
An alternative is a competent practitioner's direct composition of C.39, the relevant subject methods, C.37 or C.29, C.41, and the receiving work's acceptance rule. That arrangement can supply the same valid result. Use it when its connections are already understood and available. The additional contribution here is a reusable explanation of the two distinct judgements and their constructive continuation, so the recipient can obtain that connection without devising it anew.
The trade-off is another description to learn and maintain. The selected arrangement keeps the detailed whole-account construction in C.39.WA and uses subject methods for questions in their domains. Section 4.6 places the short questions at consequential returns; it does not add an independent check to every local operation. Its logical distinctions and operations are inspectable in the text; reduced omissions, learning effort and useful effect require evidence from use.
Kahneman and Frederick describe how a readily judged attribute can supply an answer to a harder question in human judgement. In professional work, compare the question the answer actually settles with the one the recipient asked. Establish the origin of a recurring substitution from evidence of the practice where it occurs. Kahneman and Frederick, Representativeness Revisited
Frigg and Nguyen's DEKI and the Mislocation of Justification distinguishes derivational correctness from justified claims about a target. This distinction is adapted in sections 4.2–4.3; a semantic account of representation leaves empirical justification to the relevant practice. This Method adds the distinct comparison with the originating assignment and a constructive continuation.
Reopen the arrangement when a simpler available composition supplies the same needed connection at lower total burden, when the additional entry obstructs use, or when actual work reveals a consequential return that its operations fail to handle.
Relations
- C.39 develops a missing obtaining way and its explanation; C.39.WA develops an account of a whole subject.
- C.3/C.3.1, A.14, B.1.5, C.2.1 and C.20 govern the distinct membership, generality and composition questions.
- C.37, C.29 and A.6.3.CSC support representation choice, mathematical model use and scoped compression; E.13 returns a proxy to the value it serves.
- F.0.2.RC tests promised source recovery through a receiving explanation.
- C.41 and its use profiles support the attention questions and reconciliation of returns.
- A.15.10 supplies recovery of interrupted work; C.11.DUA appraises demanded inquiry and change.
C.39.QR:End
Develop an Account of the Whole Subject
Type: Method pattern Status: Stable
Problem frame
Use this when an explanation, proposal or review must answer a question about an organized whole or a broad field, and work on its parts can lose that question. The difficulty often becomes visible after a useful decomposition, a sequence of local corrections, a change of representation or a handover. An author may explain several operations accurately while leaving their common grounds and connections for the reader to invent.
Begin with the question the account must answer and the use it must enable. Recover the subject and required depth from that question and its substantive grounds. Then develop the organization that makes the contributions answer it together.
The result is an account whose relevant parts, relations, grounds and limits can be followed, together with a justified judgement about its adequacy for the original use. A review can instead identify the missing construction and the work needed to obtain it. Retain an already adequate account. A calculation or a local correction with an established question can finish under its own criterion.
Problem
A correct contribution can become the basis of a conclusion about a larger result through an unnoticed change of subject. “This is mathematical content” becomes “this explains mathematics”; a valid example becomes the explanation of a Method; a collection of individually useful Methods becomes an account of a discipline.
Three things have changed in that move: the subject being explained, the property being judged, and the scope supported by the grounds. A later reviewer can inherit the substituted question from the author's outline or earlier positive judgement. Further checking of the same fragment then leaves the original need unanswered.
The missing work is constructive as well as evaluative. The author has to obtain the relations, general grounds and choices that organize the subject, and explain how they govern its parts. The reviewer needs an independently grounded expectation of that account. Acceptance must preserve the relation between the received contribution and the encompassing request.
Forces
Solution
Keep the original subject available throughout construction, comparison and correction. Develop its organization, distinguish the relations used in consequential inferences, and follow the general grounds into choices about parts. Compare the account with the original question. Return a missing contribution to the operation that can supply it.
C.39 supplies general construction and explanation; C.41 supplies the design and use of an attention aid. This Method connects those contributions for accounts of a whole subject. The subject practice supplies its own knowledge, construction rules and evidence. C.39.QR supplies the broader return of partial work to an originating question. When that return exposes a missing whole-subject account, carry its question, scope, reader and available contributions into section 4.1 here; return the developed account or state what is still missing for the originating use. Use this Method directly when building the account is already the assigned work.
Recover the subject before accepting its decomposition
Read the initiating question and applicable clarifications. State in ordinary subject language what must become understandable or obtainable, for which reader or receiving action, and at which scope. Recover the source commitment when the task includes restoration of a body of knowledge. If the explanation must connect several holon kinds, ranges of composition and levels of model abstraction, retain that commission in the first restatement under C.39.QR:4.1. Construct the consequential connections through C.37.MM:4.1.1. A collection of sound local explanations can still leave the requested whole unexplained.
Derive an initial expectation from the question and the subject's organization. For an account of a discipline, ask which recurrent problems, objects, principles, Methods and relations the reader needs. For a proposed composite Method, ask what it obtains and which connections make that possible. For an existing system, recover its boundary, functioning, constituents and interactions relevant to the question. Different subjects require different grounds of unity.
This expectation is a proposal that substantive inquiry can correct. Keep its basis available so a change of scope or criterion can be explained. When depth is unclear, identify the consequential uncertainty and proceed with the content already required under the plausible readings. An unspecified depth in “explain mathematics” still permits recognizing that a single multiplication does not provide an account of the field.
The reviewer forms this expectation from the request and necessary subject grounds before relying on the author's readiness judgement. It can be a concise explanation of indispensable constructions and connections; it need not reproduce the author's final text.
Questions: What was asked about? Which question about that subject must be answered? What scope and limits were assigned? What would let the receiving reader continue?
Construct the organization the account must explain
When the explanation leaves open which parts and enclosing wholes to follow or which alternatives their models must distinguish, choose them for the reader’s task under C.39.SA:4. Recover or propose what connects the relevant contributions. Follow the required understanding or result backward to the objects, actions, relations and conditions on which it depends, then explain the proposed organization forward. C.39:4.2–4.3 supplies this constructive return when an operation or join is missing.
For a Method, explain inputs, actions, choices, intermediate results, common conditions, alternatives and returns. A branch is usable when the reader can distinguish its applicability; an iteration needs a reason to continue and a condition for stopping.
For a discipline, recover its field question and boundary, the knowledge contributions that explain or constrain reusable practice, and the methods by which practice tests and changes that knowledge. Explain how those contributions form the field's assembly and what their coupling makes possible. Distinguish a constituent from a supplier whose result the field uses. Retain competing explanations or methods with their conditions of use. A changed claim or method must reconnect to that assembly; C.20:4.1–4.5 supplies its construction and continuity criteria. Several independently usable Methods can belong in the account without being forced into one sequence. Explain which principles the families share, what guides a choice among them, and what their enactments can establish. Where one family uses results from another, explain that connection.
For an existing whole, establish its relevant constituents and organizing relations under its subject rule. For a proposed whole, explain how the arrangement could obtain the needed property and which premises remain to be established. A knowledge claim, a physical object and a reusable Method have different conditions of composition and identity.
Which whole must the reader be able to explain? When explaining a holon, identify it and the property or behavior at issue. Explain how its constituents and their organization bear on that property. A project can use a Method. Within a composite Method, B.1.5 asks whether two identified Methods stand in a constituent relation. The whole's construction must require the part's contribution or admit it as an alternative for a required contribution. A functional layer groups contributions by what they do. For a claim that one participant controls another’s action, identify how its decision can affect that action. If feedback is claimed, show how an observation enters a later decision. Use the relation on which the conclusion depends. Library.Reference:6.2 helps resolve familiar terms such as level and layer.
How far does the explanation apply? A record can explain what happened to an entire organization on one occasion. A principle common to several disciplines can govern just one decision within it. At a fixed level of model abstraction, ask whether the account of the constituent could remain true while the promised outcome of the whole fails. Keep the same extent of composition in view, such as an entire office in each setting. Ask which assertions hold when the discipline, local arrangement or individual occurrence changes. Extending an explanation to another setting needs grounds that its conditions still hold. If the task is to construct a more concrete model, use B.5.MR:4.2–4.3 to show that the new model has a feasible completion, that its permitted completions preserve the earlier assertions, and which previously admissible answer it rules out.
Choose the range of composition and the level of model abstraction. What must the recipient decide, for which occasions and by when? For a choice, ask which difference could lead the recipient to act differently. For an order to establish one value, start with that value. Trace it to the whole it concerns, the parts whose behaviour can change the answer, and any handover, shared resource or rule they depend on. Follow those connections far enough to settle the decision. At each one, choose a model that keeps distinguishable the cases that could change the answer. Explain which cases can be grouped, what makes the grouping adequate for this inquiry, and what change would require more detail. An existing model or an independently established bound can already suffice. This gives a reason to stop at different scales in different parts of the same whole.
Suppose the available records concern two stores and their transport service. One client wants the closing stock, another needs six crates delivered by noon today, and a third asks whether the service can fulfil that order on every shift. Choose what to model from the commission. For stock, consider which sites and goods belong in the balance. For today's shipment, trace the handover that can change the promise and retain its arrival time precisely enough to resolve the deadline. For all shifts, examine which other bookings and capacity limits the model must admit. A confirmed reservation may settle today's capacity without a model of the vehicle's internal construction.
When an explanation promises to work at every relevant level of composition, specify the eligible wholes. A stock balance can hold for each store and for a group of stores. Combining the accounts works when the stocks cover the group without overlap, the units and time window agree, and the two entries for an internal transfer cancel. Stock in transit must also be accounted for. These conditions explain why the balance survives regrouping. To explain whether a delivery can cross the same stores on time, follow the handovers and the capacity available for them. Separate balances leave that joint behavior open.
What may we leave out before combining the parts? If a known bound or direct observation already settles the decision about the whole, use it. To claim that a proposed summary calculation preserves a result, compare it with the result for the connected detailed arrangements. A derivation can cover all admitted arrangements without constructing them one by one. If two arrangements give the same summaries but call for different decisions, recover the missing connection, use an independent source that settles the question, or give the narrower answer the summaries support. A calibrated meter on a common supply, for instance, can give the present total without reconstructing its branch currents. An exact answer is recoverable when it is the same for every admitted arrangement with those summaries. Check that the proposed calculation actually gives that answer. A bound can already settle a limit even when the exact value is unknown.
Two stores can each report ten crates at closing time. For the consolidated stock count, twenty is correct under the balance conditions above. Now promise to deliver six of the first store's crates to the second by noon. Suppose stock and transfer capacity are the only limits. The same balances fit two configurations of the run that arrives by noon, one with room for eight crates and one with room for only four. Recover the available capacity before promising the delivery. For a network of stores, carry each intermediate result far enough to support the next handover. A conservation law can justify regrouping throughout its stated domain. A rule that works at each store alone leaves the handovers open. C.29.1:4.3–4.5 explains the comparison in both orders, and C.37.MM:4.2 identifies what arrangements the models admit together.
A useful general explanation supplies reasons for choices. It lets the reader say why these contributions belong together, what their connection makes possible, and what a changed condition would reopen. Where the available description says only “coordinate,” recover what must be made compatible, who or what supplies it, and how the resulting relation affects the whole.
Questions: What makes this the whole under discussion? Which relations explain the relevant property or result? Why this composition or organization? Which external conditions and alternatives matter?
Identify the relation used in each consequential transition
A mantra for reasoning with models and methods. What must we establish or do, and about what? If the kind matters, which one, and what tells us that an object belongs to it? What relates which objects, and how? From which observations or premises, by which rule, can we draw what conclusion, for which cases? By what method can we produce what is needed or bring about or maintain the required condition? What must one operation supply to another? If a premise changes, what must we reconsider?
For teaching, unfold these questions from the intended use. Supply the field's terms for the types and relations, and its operations for obtaining a result. Work through why the answers depend on one another. In practice, enter through what is available. A local record can lead back to the rule that gives it meaning. If that rule itself needs examination, make it the new subject and apply the questions to it. Later evidence about the same subject can reopen just the conclusions and actions that depended on the earlier evidence.
When teaching someone to design this comparison, give the commission and the ordinary materials from which the planner works. Let the participant choose the whole, the composition scales, the dependency to follow and which model distinctions to retain or merge. Ask what makes those choices sufficient and what would reopen them. Credit a different justified arrangement or a suitable existing model that the participant selected. If the task supplies the time resolution that decides the deadline, the participant can still demonstrate a choice of route and use of the given model. A correct calculation with all these choices supplied establishes application of the design.
Give each answer in the terms of the work. A field supplies criteria for recognizing instances of its kinds. To decide whether two records refer to the same instrument, use the instrument's identification conditions. Two instruments can satisfy the same kind criterion. A kind, a discipline and an individual instrument can each be what a claim is about. Name which one you mean. If a model loses a difference that the judgement depends on, amend the model or limit the judgement.
State the claim the evidence supports. In classification, apply the kind's criterion to the candidate. In a calculation, give the premises and the rule that yields the conclusion. To infer something about an operating system from a model, establish which correspondence and conditions permit the inference. Say which cases the claim covers. A.2.6 tests coverage when a proposed use requires membership in a declared claim scope. A Method has its own conditions of applicability. A contract can specify commissioned work, deliverables and acceptance conditions. Use the meaning of scope that the particular statement requires.
For an action, name the reusable way and what it must obtain or preserve under those conditions. Distinguish the conclusion reached, the report carrying it, its delivery, and any later change in the recipient's work. The report can be both the output of one operation and evidence used by another. State each relation with its own verb. A.6.P.WMR recovers the participants when calling something a “result” leaves its role unclear. If the goal is a change in the world, identify how to establish whether the change occurred.
For a move from one claim or subject to another, name the relation on which the inference relies, the kinds of its participants, what fills each relevant role here, and the direction of the inference. A multi-party relation can require more than two participants. If a boundary matters, identify what crosses it, between which places or organizations, and in which direction. Translating a record can leave the instrument it describes in the same laboratory. Use A.6.5 when declaring positions in a reusable relation signature. The following relations support different moves.
For a rule said to recur across composition scales, give grounds covering the eligible wholes and constituents. For a dependency between scales, identify the property that carries, its direction and the relation that licenses the inference. For a joint outcome, trace how the constituents are organized and which combinations can realize it. For example, consider the work that fulfils two orders in one service window. If a resource is available for four hours and three are already committed, at most one hour is free for additional work that needs exclusive time in the same window. A second order can still be fulfilled by the shared work itself when its result serves both recipients. The method used by that team determines what it can do in the hour. For a statement about Work already done, establish which performances occurred. For a claim assembled from reports, check which premises they support and whether they rely on the same observation.
A relation identifies how its participants are connected. A returned claim also needs an applicable inference rule. Separately, the resulting work must be compared with the original account question; C.39.QR:4.2–4.4 develops these two judgements.
These relations can coexist. A reusable calculation can be a constituent of one Method and supply a result to another. A discipline can contain other disciplines and use results from fields that are not its constituents. A representation can preserve some structure, omit some and introduce useful constructed objects. Explain the relevant correspondence before returning a conclusion to the represented subject.
Check a claim of strict refinement. When an account of a system, method or field treats one model as more concrete than another, keep the inquiry fixed. Show that at least one design, execution or other completion remains feasible, that each preserves the earlier assertions and that a previously admissible answer is excluded. B.5.MR gives the comparison for individual models and for models of their governing conventions. B.1 gives the grounds for a separate claim about constituents and wholes.
Identify the subject and relation before moving. Name what the claim concerns, the cases it covers and the use of the answer. “The laboratory's rule for releasing an instrument” identifies a starting point. If you need to reason from a particular record to the rule it cites, identify that record, the rule and the citation. A release record can direct you to the local rule under which a decision was made. Check the applicable edition and the issuing authority. To examine that rule in turn, compare it with the discipline's conditions for a valid verdict. To establish what actually happened, return to the identified instrument, configuration and observations. Carry forward only the conclusions for which that return supplies grounds.
A transdisciplinary distinction can guide reasoning within one discipline. Interdisciplinary work brings contributions from different disciplines into a common result. Ask which distinction guides the choice of types, or which contributions must be connected. Both questions can matter in one undertaking.
Locate the reasoning in terms the recipient uses. Identify the discipline that supplies the types, the local practice whose rule is being examined, or the occurrence to which a claim refers. A discipline can itself be the subject of a claim about its methods. To examine a field's type, ask what kind of distinction it makes and check it against the proposed higher-order criterion. In the other direction, propose a field type or relation, test its meaning against disciplinary practice and identify individual occurrences when the inquiry needs them. C.20 gives a test for the discipline as a whole. C.2.1 identifies the episteme stating a model or claim about the subject. A change of subject, a reclassification and a move to a constituent each license a different inference.
When the subject is a way of working, follow the result and the operation as well as the types. Start from the promised result, on what object and under which conditions. An agreed field procedure may supply the general operation with a domain-specific test, instruments or admissible evidence. Establish whether you have selected a setting within the same Method or changed the way and its applicability. One operation can use another's result through an external service. To claim a constituent Method, establish which reusable operations and connections make up the whole, how they jointly obtain its result and which replacements preserve it. Two comparable Methods can also be alternatives, each sufficient for the same use. A performed check on Tuesday establishes an occurrence under its own conditions. A.3.1:4.8 and B.1.5 give the grounds for these comparisons.
Grounding connects a case with an explained construction. Identify that construction and make its relevant objects, relations, operations and grounds available to the reader. Show which consequential moves in the case instantiate which operations, why the conditions select them, and what a change would reopen. In an account of a Method, distinguish the explained Method, its displayed performance and the subject worked on during that performance. For example, when explaining how to design an explanation, a calculation can be the material; selecting a contrast and revising the explanation are performances of the explained Method. E.8:4.2.1.1–1.3 and EXD.2 develop this correspondence for authoring explanations.
Choose meaningful variety from the account's declared subject and use. Differences in objects, relations, premises or available means can require different branches. Retain cases that make those differences understandable. A running case helps trace connected operations; add or change cases where it hides a consequential distinction, and return from each application to what it reveals about the general account. The choice of examples follows the work they must do, with no fixed number applying to every account.
When generalizing, state what elements and relations are being generalized, the wider class, retained premises and grounds for the extension. Varied cases can support construction, expose limits and contribute evidence under the relevant inference method. A general proof, a qualified empirical inference and a demonstrated performance have different warrants. Explain the general contribution and the role of the cases in obtaining or understanding it.
Questions: What was the subject before this transition, and what is it now? Which relation connects them, and what inference does it support? What explained construction is grounded by the cases? Which material differences do their variations reveal? What is generalized, over which range and on what grounds?
Return the general grounds to the parts
Use the account's general construction to choose, interpret, connect or revise its parts. Follow a consequential change of condition: which constituent action changes, which connection fails, which alternative becomes appropriate, and which contributions remain usable?
This is the constructive role of the whole explanation. For a family of Methods, a shared principle can explain why different branches use different means. For a composite Method, a whole requirement can constrain several concurrent actions. For a discipline, a changed knowledge claim can alter a practice's applicability and create a new inquiry. These are different returns with different grounds.
Then follow findings back into the common account. An unexpected case can reveal a missing premise, a new problem family or an incorrect connection. Revise the affected explanation and its uses. Keep the source's useful constructions while critically examining its claimed generality.
Distinguish the operations needed for that return. Synthesis connects contributions into an organized account or way. Generalization extends the range for which a claim or operation is warranted. Model abstraction leaves some determinations open while retaining the conclusion the receiving work needs; apply B.5.MR to compare the resulting model with the starting model. Selecting what an account displays can preserve all model assertions and leave the abstraction level unchanged. One piece of work can need several of these operations, with each contribution explained.
Questions: Which choices about parts follow from the common grounds? What changes when a consequential condition changes? What do varied cases reveal about the common explanation? Which general contribution still has to be invented by the reader?
Compare the actual account with the original question
Describe what the candidate explains, temporarily setting aside its title, field label and declaration of completeness. Compare that content with the expectation from section 4.1 across the assigned scope. Inspect the general construction and its relations as well as local claims. Recover what the cases concretize and which consequential variations of the declared subject they expose. Compare the knowledge and operations available in the general account with those displayed in its cases.
When the account solves a narrower problem and leaves the more general inference unwritten, use C.39.GA to find that step, restore its grounds and recheck the dependent conclusions. Keep the parts of the case that show how the inference works.
For an account of a principle, identify the relation the reader can recover and the decisions it supports when relevant conditions change. A single worked example can supply the whole explanation if it gives the argument for that range. If a completed case shows how someone used an already chosen rule, accept that demonstration. When the commission also asks how to construct the rule, return the account for the missing choices and their grounds. An explanation of why one particular event occurred can be complete once it establishes the requested connection for that event.
For each consequential conclusion, retain its subject, property, scope and grounds. Truth of a claim, membership in a field, coherence of a construction, coverage of an assigned account, usability by a reader and effect in work answer different questions. Select the conclusion the assignment needs and obtain its basis.
When source preservation is promised, F.0.2.RC tests whether the action, conditions, qualifications and returns can be reconstructed through the receiving explanation. Its source cut follows the promise: a commitment to a whole corpus requires that recovery basis. For a new synthesis, identify what the author constructed and how it is supported. A reader's own successful invention locates a useful contribution but does not make it content of the candidate.
Return one of the substantive outcomes the evidence supports: an adequate account for the named use; a statement of which construction is missing; a contradicted inference; an unavailable ground; or a changed, explicitly agreed question. Preserve a useful fragment with its qualified conclusion. Keep the original task open while a required contribution remains unavailable.
Questions: What does this text explain? Which required constructions and connections are supplied? Which are missing, unsupported or still unexamined? Did the subject or criterion change, and on what authority?
Arrange independent appraisal and complete the return
Select independence according to the intended conclusion and the observed difficulty. A reviewer who must judge the author's construction keeps authorship separate. A fresh reading context is useful when earlier narrowing or a disclosed expected answer could shape recognition. Supply the original question and necessary subject grounds before the candidate and earlier verdicts. A domain specialist can still be needed; freshness does not supply subject competence.
The reviewer compares the account with an independently grounded expectation and explains the missing relation or contribution in subject terms. The author develops that content or obtains its missing supplier. Reconsider the affected conclusion and dependent uses after repair. A coordinator accepting several contributions compares their coverage and joint conditions with the original result; C.41.SR supports that reconciliation.
Use the six question groups as an aid at the work occasions:
- Before decomposition, read-do: recover the subject and develop the expected organization in the explanation.
- During construction, read-do: follow the questions at a consequential join and construct or correct the missing content.
- Before an adequacy claim, do-confirm: use the performed work and its grounds to answer the questions about the complete relevant account.
- At independent review: recover the expectation first, then compare it with the full relevant candidate.
- At acceptance: connect the warranted contributions and unresolved conditions to the original requested result.
C.41.OF arranges access, timing and competent participation. C.41.UA separates obtaining an answer, deciding what to do and performing the resulting action. A negative or unresolved answer must have a useful continuation. Put the relevant outcome in the explanation, review or handoff already needed by the work.
Use C.41.QC when the question set or checking means is itself doubtful. A useful appraisal must distinguish a correct fragment from an account that supplies the required whole. A criterion that asks only whether the fragment is correct cannot support that decision.
After interruption, recover the original question, the relevant whole construction, the last supported contribution and the unresolved return. Reuse earlier grounds while their subjects, conditions and conclusions remain applicable. For an AI agent, keep the operative questions and subject material available in the working context. For people, provide the necessary time, access and help. Observe understanding, performance of the operation and the resulting benefit separately when deciding whether the arrangement works.
Archetypal Grounding
The following constructed cases expose different transitions in the Method. Their purpose is to make the common operation and its returns inspectable.
Mathematics replaced by a multiplication
The request is to explain mathematics and judge the completeness of the account. The submitted text is “2 × 2 = 4.” The reviewer establishes that the equality is true and concerns mathematics, then accepts the request as completed.
Recovering the subject immediately changes the judgement. The submitted text explains at most a calculation. An account of the field needs a chosen depth and reader, but it must develop mathematical objects and questions, ways of constructing and proving, and relevant relations among its branches. The breadth of a future overview remains a design question; the absence of that account in the submitted equality is already apparent.
The wrong inference joins two different relations: the statement's membership in the field and the account's adequacy for the requested subject. Retain the valid calculation as a possible illustration. Return the missing field explanation to authorship.
Suppose the author next provides ten correct calculations. This supplies more instances of calculation. Their contribution to the field account depends on an explained organization: what questions the operations answer, which structures and rules support them, how proof and construction differ, and how the selected branches connect. Diversity of examples can support that explanation while its organizing claims still require their own grounds.
The approving coordinator repeats the comparison with the original question. The previous reviewer has supplied a local truth judgement. Acceptance of the requested account remains unsupported until its substantive construction is provided and assessed.
Rebuild an explanation around the conditions of recovery
A technical author is asked to explain how a specified service design can resume within 60 seconds of a primary-server interruption while preserving acknowledged updates. The reader must be able to assess the proposed recovery and revise that assessment when the time requirement changes. The available draft describes the primary and backup servers and reports a successful backup start-up test.
The author first tries to order the account by those components. Returning to the reader's question reveals what that order leaves unexplained: how the acknowledged state reaches the new writer, and when clients may safely use it. The design supplies three relevant facts: acknowledgement follows replication of the log entry; the old writer's authority is removed before the backup receives authority; the client route opens after restoration of the required state. The author connects these facts through the conditions that each transition needs and produces. This gives the account its organization: preservation of acknowledged state, transfer of authority, and restoration of usable service.
That organization changes the text. The replication rule moves next to the claim about preserved updates, where it supplies the claim's premise. Authority transfer is explained between detection and recovery by the new writer. The start-up test stays with the narrower claim it supports. The new connecting explanation tells the reader why an available process, correct state and permission to write must coincide before the service can resume. A list of servers could not explain that coincidence.
The author then develops the timing paragraph from the same dependencies. In this constructed design the bounds cover successive elapsed intervals: at most 8 seconds from interruption to detection, 12 more until authority has transferred, 15 more until restoration is complete, and 5 more until clients can use the route. These intervals include waits and handovers between the milestones. Adding their bounds covers the entire interruption and gives at most 40 seconds. The text attaches the conclusion to that order and to the replication, authority and timing premises. A reviewer can now follow both questions in the commission: what preserves the acknowledged updates, and what supports the recovery bound. These premises support the bounds for the specified design. An account of recovery at an actual installation would need evidence about that installation and the event in question.
A new requirement of 30 seconds changes the author's comparison. Forty seconds no longer supports it, while the upper bound alone leaves the original design's feasibility unresolved. The author obtains the designers' alternative: maintained warm state reduces the elapsed interval from completed authority transfer to completed restoration to at most 4 seconds, with continuous compute and state-maintenance provision. The account is revised at the affected join: 29 seconds is supported under the new provision, and the state and authority explanation now includes that provision's role. Detection's established bound is retained. If the provision is unavailable, the explanation exposes the unresolved timing or support question for the designers.
The revision follows the account's organizing grounds. It changes the premises, connection and qualified claim that depend on restoration, while preserving the still-supported contributions. This is the author's construction the reader can reuse when another whole-subject account consists of accurate component descriptions with its decisive relations missing.
Build a disciplinary account through knowledge and practice
An author is preparing the common entry to an operations-management guide. Its reader needs to choose and revise ways of obtaining continuing production or service under competing demands and changing conditions. The supplied chapters explain admission of work, capacity allocation, development of operating methods and review of results. The initial entry groups them by department. That arrangement helps locate a chapter but leaves the reader to discover why a finding in one chapter changes work in another.
The author recovers the field question and reads the knowledge–practice connections in OPS.Reference:1.1–1.6. Two different grounds in the source guide the author's construction. With a consistently identified work population, opening unfinished work plus entries minus exits gives closing unfinished work. A completion forecast additionally depends on premises about capacity, work mix, variability and dependencies. The author uses this difference to construct the entry's organizing explanation: operating knowledge both constrains the arrangement and supplies revisable grounds for anticipating its results; performed work can reveal errors in the account, failed premises or a need for a different Method.
This explanation reorganizes the connections among the four chapters. Admission and allocation now return a jointly supportable arrangement. Review compares its predicted and observed consequences, identifies which ground is affected, and returns that question either to the operating account, to provision or to development of the Method. The author writes the reasons for those different returns alongside the connection. An arithmetic discrepancy requires recovering the event account; a defeated forecast premise requires revising the forecast or arrangement. The shared field question holds these contributions together while their different grounds explain why they remain distinct Methods.
To check whether this connection is available in the guide, the reviewer follows a supplied report of growing unfinished work. The report counts only items awaiting technicians and explains delay by technician capacity. Completion also requires specialist approval. The author recovers that omitted transition and revises the account's work population and exit condition. The guide's diagnosis paragraph now explains how more technician output can increase the approval queue, and directs the reader to establish the approval contribution or revise admission. These details are needed to show the author's repair: changing the represented operating unit changes the explanation and the Method choice together.
A second supplied situation concerns readiness for an unknown future incident. The initial entry treats every operating result as an item that can be completed and stored. The author discovers that this common assumption cannot organize the readiness chapters: the required capability must remain available during the covered interval. The entry is revised to explain the relation between receiving result, operating unit and commitment. Its production branch follows completed items; its readiness branch follows maintained availability. Their shared principles concern obtaining usable results under joint conditions, while their different objects select different operating Methods.
The completed entry contains that field explanation and the connections at the four chapter entrances. The reviewer reads those connections against the original commission, including how a changed knowledge claim returns to practice and how a practice finding revises knowledge. The guide now supplies the joins that its departmental outline had left to the reader. Further chapters receive their own place through the same field question and their knowledge–practice relations.
Organize a theoretical account by the questions its structures answer
A mathematics author is commissioned to explain finite-dimensional linear maps to readers who can solve elementary linear equations. The account must connect maps, equations and matrix calculations so that readers can reason about existence and uniqueness of solutions. This scope is part of the commission. The available notes define a map, introduce a matrix, and then present successful calculations in that order.
The author recovers what must be explained before choosing a more elaborate calculation. Two questions organize the subject: which outputs can a map attain, and which inputs it treats alike? In the standard theory, the image answers the first question. Linearity gives the second a common construction: $T(x)=T(y)$ exactly when $x-y$ lies in the kernel. Consequently, $T(x)=b$ is solvable when $b$ is in the image; after one solution $x_0$ is found, all solutions have the form $x_0+k$ with $k$ in the kernel. The UCL account of linear maps supplies the definitions and structural grounds.
These relations give the author a reason to change the exposition. The new opening states the two questions and derives their connection through linearity. The kernel and image definitions now help answer those questions. A solved equation is placed after the solution-set construction, so the reader can see which features of that calculation arise from the common relation. The matrix section explains how selected bases represent the map and make calculation possible. A change of basis can change the matrix while retaining the map and the two structural questions.
The reviewer then examines a transition the draft had left implicit: the claim that a map reaches every output. Two maps can have the formula $T(x,y)=(x,0)$ and different codomains. With codomain $\mathbb{R}^2$, some outputs are unattainable; with the first coordinate axis as codomain, every allowed output is attained. The formula alone leaves this judgement underdetermined. The author repairs the general explanation by including the declared codomain in the attainable-output question, and updates the affected statements and examples. The kernel argument remains valid. This local finding has changed the common account and selected its dependent revisions.
The reviewer returns to the commissioned use. Existence is explained through image membership; freedom among solutions through the kernel; matrix computation through the chosen representation. The text now gives their connecting grounds before asking the reader to interpret particular calculations. These grounds also let the author select useful contrasts: a different codomain tests the scope of the attainable-output claim, whereas changing only a matrix entry may exercise the same calculation again.
Here the whole account is organized by mathematical structures and justified implications. The author obtained that organization from the questions and theory, used it to choose and arrange the parts, then revised it when a judgement exposed a missing participant.
Bias-Annotation
An author can carry the missing general explanation in memory and see it in a text that supplies only cases. A reviewer can adopt the author's decomposition as the scope of the assignment. A coordinator can extend a warranted local conclusion when combining reports.
Restore the subject from the initiating question and subject grounds at the relevant decision. Track what the text supplies and what its reader has added. Preserve meaningful examples and partial results while constructing the missing explanation.
Conformance Checklist
At a claim about the whole account, answer the following from the account's content and grounds:
- What was the original subject and required use?
- Which constructions and relations organize that subject at the required scope?
- What relation supports each consequential transition between kinds, parts, levels, representations and claims? What general construction do the examples ground, which significant variations do they reveal, and what warrants any generalization?
- How do the common grounds govern the parts, and how do findings revise the common account?
- What does the candidate supply, and what remains missing or unexamined?
- Which result of appraisal or repair permits the proposed use of the whole account?
Arrange these questions at the work occasions in section 4.6. Preserve an unresolved answer and its continuation without presenting it as a positive one.
Common Anti-Patterns and How to Avoid Them
Consequences
The account can supply the connections and grounds that let a reader use and change its parts. Appraisal preserves the assigned subject across handovers and qualifies local successes within their scope.
The work costs subject recovery, construction, connected reading and repair. Reuse adequate explanations and valid earlier grounds. Select further inquiry by its attainable contribution and total burden through C.11.DUA. Ordinary use provides evidence about whether the arrangement reduces consequential omissions and rework.
Rationale
Adequacy is a relation among the account, the question, its intended use and the available grounds. Accurate local content is one possible contribution to that relation. The required organization depends on the subject: Method composition, discipline composition, system functioning and representation each supply different rules.
The Method joins construction with appraisal because missing content must be obtained, while an adequate account should remain usable. Independent expectation helps expose a substituted subject; the two-way return between whole and parts gives the repair its constructive operation. C.41 makes the questions usable at consequential work occasions.
SoTA-Echoing
The practice question is how an author or reviewer can develop an account of an organized subject that answers the whole assigned question. The selected line is the connected operation in section 4: independently recover the subject and expected organization, construct its grounds and relations, bring those grounds into choices about parts, and compare the account with the original use.
An alternative is a competent author's direct use of C.39 with the relevant subject patterns and C.41. That arrangement can supply the same account when the author already knows how to connect its operations. This profile supplies that connection as reusable content, including the discipline/Method/system differences, the whole-to-parts return and the question groups at their use occasions. It accepts additional reading and maintenance in exchange for making that construction available to readers who would otherwise have to develop it themselves. Use the direct arrangement when it is already adequate; the profile adds no further work by virtue of being available.
Paul E. Newton's 2026 Ofqual account of qualification design explains construct underrepresentation and how systematic omissions from assessment can narrow enacted teaching. Adapt that distinction to section 4.1 and section 4.5: derive the expected account from the subject and intended use, then inspect which constructions the appraisal elicits. The report concerns qualification design; it does not evaluate this authoring Method.
NASA's Systems Engineering Handbook, section 2.4, supplies the distinction between conformity to requirements and suitability for the intended purpose and environment. Adapt it to the separate questions about a local criterion and the whole use. Section 5.2 makes those different conditions visible in one conditional service construction.
C.39.QR owns the wider return from any partial result to its originating question. This profile supplies the detailed account-building branch. The difference from direct supplier use is the available explanation of their connection, not a new inference rule. Its usefulness to ordinary readers and its effect in work require evidence from those uses. Reopen the arrangement when an existing account supplies the same connection more simply, when the additional description causes avoidable work, or when an important subject cannot be handled by the stated operations.
Relations
- C.39.QR returns partial work to its original question and uses this Method when the missing result is a whole-subject account.
- A.6.H recovers the subject and relations of a holism claim.
- A.14, B.1.5, C.2.1 and C.20 govern different constructions of wholes.
- C.3/C.3.1 and A.3.1 distinguish kind relations and Method variation from composition.
- C.29/C.29.1 govern mathematical representation and returned inference.
- C.39/C.39.RO supply construction and generalization; E.4.CM applies them to framework publication.
- F.0.2.RC supplies source-coverage reconstruction and affected refresh.
- C.41 with OF/UA/SR/QC supplies selection, fitting, execution, reconciliation and qualification of the attention aid.
- F.19 repairs wording whose meaning or contribution obstructs the account.
- Library.Reference:6 explains the different structures needed to select the relevant levels and layers.
C.39.WA:End
Restore a Missing General Argument
Type: Method pattern Status: Stable Normativity: Normative
Problem frame
Use this when the grounds in a pattern, guide, recommendation or line of reasoning cover only some of the conditions its conclusion claims. Also use it when an author promises to explain how to reach a decision or construct an explanation but makes a necessary choice for the recipient and omits how to make it. Even for a single case, showing that a solution works can leave the promised explanation of how to construct it missing. Start with what the recipient must be able to explain, construct or decide. Two accounts can recommend the same action while giving different explanations of how the whole works. The text might explain it completely in a single example. If the commission asks only for a decision about one occurrence, its supported answer can finish the work.
Problem
An author knows why a choice worked here. In explaining the method, the author carries that choice into the instructions as if someone else had already made it. Every sentence about the case can be true while the reader has no way to choose under different conditions. The same loss occurs without a story when an argument for one permitted arrangement stands in for all of them. The given argument can even be valid throughout its range. An author who promises to teach how to build an explanation may demonstrate why a finished design works and leave the reader to invent how to choose its parts or develop the explanation.
Forces
A named part, action or event makes reasoning easy to follow. The decisive premise may be introduced there. Replacing its name with “an entity” can hide that premise just as effectively as copying the example’s answer. A repaired explanation keeps the working vocabulary and shows which changes its reasoning allows.
Solution
A mantra for restoring a general argument. What must the reader explain, construct or decide, and which choices are theirs? Which kinds of holon, over which composition scales for each kind? What connects these kinds? What holds at each scale, follows from part to whole or from whole to part, or depends on the parts working together? Which levels of abstraction of the relevant models does that work need, and which alternatives remain open at each? Which conclusion carries between models, in which direction, or needs several models together? Which promised choice or operation can the reader still make only because it was settled in the material? Where does the argument rely on a condition fixed only for a narrower use? Restore the way to make that choice or the grounds for the inference. Which later conclusions follow, under which conditions, and what is still owed? Use the questions about composition and models when the work calls for those relations.
- Recover the promise. From the commission, state the result the recipient must obtain or explain, which choices they must make and what can vary. For a claim across scales, identify the whole for each relevant kind of holon. Does that result need a rule at each scale, a link between scales or a result of several working together? For a range of model abstraction, ask what the model may leave open.
- Locate the lost step. Work back from that conclusion to the premises it needs. Read forward from what the passage, its accessible sources and the audience’s preparation supply. Where does an inference first depend on a choice already made for this case or subclass? Could two rules give the same answer here and different answers under another permitted condition? Keep any general argument made inside the example. There is a gap if the recipient needs an inference and has no adequate grounds for it in the account, its accessible sources or the agreed preparation. Also start from each operation promised to the recipient. Which inputs and ordinary help would be available to them before that choice was settled? A sound argument for that answer can leave the promised way of constructing an explanation wholly absent. Two different ways can both be adequate when the recipient is free to choose either.
- Repair the inference where it is used. For a transfer, name the property to be established and the direction of the inference. Trace a permitted situation through the relation to the receiving situation. Show why the property is preserved, or identify a permitted exception. When several premises are needed, establish the relation between their subjects. A record may concern the very job being assessed, an earlier test may cover the new operating condition, or two actions may compete for the same resource. When a shared constraint restricts combinations, establish which combinations it permits. One feasible combination can establish possibility. A promise that must hold throughout a range needs grounds covering every admissible combination in that range. For an explanation of how to choose or build something, identify which needed choice the material already settles. Starting from the information the recipient actually has, develop a way to find alternatives and distinguish the ones that matter to the promised use. If the promise is to explain an organization, trace how its constituent contributions fit together and what their arrangement makes possible. Two arrangements that justify the same immediate action may require different explanations. A known working construction can yield a reusable operation under C.39.RO. Develop a missing construction in its subject practice. Keep any worked case and show which part of the reasoning it illustrates. If the grounds run out, give the supported answer and identify what is still owed.
- Recheck the reach. Follow the repaired step through the conclusions that depend on it. A derivation for an arbitrary member, a sufficient bound or an exhaustive division into cases can cover a class. An empirical claim needs observations and grounds for extending them to the population and conditions claimed. When the missing work was a choice or construction, follow the repaired procedure from what the recipient has before that choice. Check whether the recipient can now make the entrusted choice, obtain the result and give the promised grounds with the help available there. Another admissible construction can be equally adequate. A changed case can expose an error. One sound argument can already cover the range, even when it gives different answers for different inputs.
Archetypal Grounding
A guide promises to teach when an earlier inspection can support a later acceptance decision. It recounts how a team corrected the drawing title and accepted a component on yesterday’s inspection. “Reuse the inspection after a document edit” gives the right answer in that episode. The same wording would also cover an edit that changed the allowable load. At the first step from “the title changed” to “the previous inspection suffices,” the guide relied on what this particular edit left unchanged.
A repair the recipient can use is to identify the property that acceptance depends on, the item inspected, the conditions the inspection covered and the evidence the acceptance agreement calls for. Compare these with the changed order. If the earlier evidence establishes the required property under the conditions now in force and satisfies the agreement, use it. If the load changed, an inspection that already covered the new load can still suffice. If the earlier result leaves performance there open, obtain the missing grounds before accepting. An order for a fresh inspection creates a further obligation. The title correction illustrates one way the earlier grounds can remain sufficient. The reason for reuse also covers changes that leave those grounds sufficient in other ways.
Bias-Annotation
A technical audience may supply a familiar unit conversion unaided. A reviewer who invents a new criterion for choosing the unit has contributed something different. Credit the passage, the intended preparation and the reader’s own construction for what each supplies.
Conformance Checklist
- The result supported by the actual reasoning answers the commissioned question about its subject, including any explanation or construction entrusted to the recipient.
- The recipient can make the promised inference or perform the promised construction at the stated reach, or identify which consequential part remains open.
- Where the conclusion crosses composition scales or model abstraction levels, the explanation supplies the link it uses. A change of values within one model can test an assumption while keeping the model level fixed.
Common Anti-Patterns
Calling the subject “a holon” leaves a rule for a system component unchanged if the rule still presupposes a physical interface. To use it for a constituent Method, establish how the whole Method admits that constituent and what its contribution makes possible. A common word alone gives no grounds for the transfer.
Consequences
An author can give this repair to the person writing one paragraph or reviewing a whole guide. For a whole publication, follow each promised conclusion to the place where its grounds become available, including consequential uses in diagrams, assignments and keys.
Rationale
Agreement on one instance leaves open which premise produced that agreement. Restoring the premise lets the recipient work out where the inference holds, where it fails and where the evidence is insufficient. Changing the example alone can preserve the same omission.
SoTA-Echoing
Mason and Pimm explain how a particular example can stand for an arbitrary member in a mathematical argument. Keep that argument when the reader can recover why it applies throughout the class. In experiments with negotiation strategies, Gentner and colleagues found that comparing cases helped participants recognize a relation they could use in another negotiation. Such a comparison can reveal the relation worth explaining. Its validity in the intended practice still depends on the subject grounds. Mason and Pimm, Generic examples · Gentner, Loewenstein and Thompson, Learning and Transfer
Relations
When the composition scales or levels of model abstraction still need to be chosen or revised, C.39.SA supplies that selection. With the reach established, the four steps in C.39.GA:4 complete the ordinary repair. C.39.WA develops an account of a whole or field. C.39.RO derives a reusable operation from a known working construction. C.37.MM works out connections among kinds of holons and models, and B.5.MR tests a claimed refinement. EXD.Reference:3 develops the subject account for explanation design. Use those methods when the particular missing step calls for their work.
C.39.GA:End
Choose Composition Scales and Model Abstraction for the Task
Type: Method pattern Status: Stable Normativity: Normative
Problem frame
Use this when you must decide how far to follow the parts and enclosing wholes of a system, Method, Work or other holon, or which differences its model must preserve for someone’s work. A commission can specify some of these choices and leave others to you. You can make the selection before a model or explanation exists, or revisit it when a recipient cannot use what you supplied. If the relevant whole and model already settle the commissioned task, use them directly.
Problem
An account of an entire enterprise can still describe just one occurrence. A rule that applies across enterprises can concern a single constituent Method. “Consider the relevant factors” settles neither the composition scale nor the distinctions a recipient must make. A detailed description of one arrangement can be just as unhelpful when the recipient must choose among arrangements. How can the author determine where to widen the account, where to make it more concrete and where to stop?
Forces
A total can answer a question that no individual entry settles. The same total can conceal a dependency that determines whether two parts will work together. A name or drawing can make a case recognizable at little cost while quietly making the very choice the recipient must learn. Each additional distinction takes work to obtain, explain and maintain.
Solution
Start with the work and its criterion. Who will use the result, to do or explain what, with which choices still theirs? To find the subject, ask where the property or change to be explained could arise. For example, in a practice that produces and reviews instructions, a Method can prescribe a choice, a particular performance can make it, and a review practice can repeatedly admit the resulting work. One passage from that practice may give access to several inquiries. In your own work, choose the subject that can answer the commission. Connect other subjects only as far as their contributions matter to the explanation. Recover the permissible conditions and the information available for that work. Supplying one workable design allows the designer to choose its arrangement. Teaching someone to design for several settings requires a way for that person to choose. Deciding whether a threshold is met and giving an exact value can also require different evidence.
-
Follow the composition needed for that work. Start where the recipient’s question starts. It may name a constituent, a whole or a connection between them. If you must choose where to begin, ask which difference in their arrangement would change the decision or its grounds. Could the same part serve a different enclosing whole? Could the same reported outcome come from different arrangements of the parts? Follow the relation in the direction the answer requires. A fact about one part can settle the question when you can give the inference to the requested conclusion.
When several kinds of holon matter, establish what counts as a constituent for each. The steps a Method calls for and the performances that made up a particular job can have different compositions. The job enacts the Method. An account of that job is an episteme whose claims can supply grounds for what happened. Trace each kind across its own scales and explain any connection between kinds that the conclusion needs. For a functional control layer, distinguish the responsibility assigned to it, the Methods that implement it and the controllers that perform the work where these distinctions matter.
-
Find which alternatives the model must distinguish. To form a first model, start with the action or explanation the recipient owes. Identify information that could distinguish two required answers and keep that distinction in the model. Then consider the situations the resulting model still treats alike. Could the recipient carry out each assigned operation with the information that remains? If indistinguishable situations require incompatible decisions or different grounds that the recipient must explain, preserve the difference or supply an observation available before that decision. For one action guaranteed to work, establish an admissible situation and a single action that works throughout the whole indistinguishable set. Compatibility in each pair need not give an action common to all. When a later choice is allowed, a rule based on information and authority available then may suffice. For a risk limit, retain the frequency, probability or justified bound the decision uses. The same list of possible outcomes can conceal very different risks. A new observation may settle the work within the same model. For a claim that the model itself became more concrete for the same question, compare its admissible answers under B.5.MR:4.2.
-
Try releasing a choice fixed by the account. What happens if a different permitted arrangement is used? If choosing one workable design is your job, deliver the chosen design with the grounds the order requires. If the recipient is to choose, show how they can recognize and use the permitted alternatives. A worked example can keep its particular design. Keep details required by the order and details that make the explanation intelligible. Identify which ones belong only to an illustration. One way can cover the entire promised range. Several ways can depend on the same unstated restriction. Where the recipient must explain a selection, show which information lets them make it and why the alternatives matter.
Apply the same test to the work you are about to propose. If you carried it out perfectly, could the original question still have different answers? A list of places to look or cases to examine can guide an inquiry. Use an unverified possibility to guide inquiry. Before describing it as part of the practice, establish whether it actually operates there. Distinguish that finding from a possibility suggested by you or someone else. To use it as the full range, derive its coverage from the question or the organization of the subject. When the choice remains open, retain other possibilities and specify what would distinguish the ones that change the answer.
-
Establish what follows at each scale, between scales or from their joint use. Does the decision need a rule at each eligible scale, a directed dependence from one scale or model to another, or a result that depends on several together? Identify the relation and give grounds for that claim. Check coupled choices under their joint constraint. A newly distinguished case may reveal a missing constituent or enclosing whole. Reopen the affected composition choice. Conversely, a shared constraint may change what the model has to distinguish. A subject rule, a sufficient bound or a justified empirical estimate can settle the task across its declared range. State its applicability. When the available grounds leave different required answers possible, identify the unresolved distinction and what would settle it.
The result is a choice of composition scales for each relevant kind and of model abstraction for the intended use, with the connections and stopping grounds the recipient needs. A range need not contain every intervening description or form a rectangle with another range. When several selections give the required result, effort, access and the consequences of an error can decide between them.
Archetypal Grounding
A cooperative asks how many crates it owns at close of business. Crates in transit between its branches still belong to it. A certified total for that time, covering every owned crate exactly once, settles the question. If the total is assembled from branch returns, count the stock in transit under the cooperative’s ownership rule. Two correct counts on the branches’ shelves can miss it. To combine changes in stock, take disjoint holdings, use the same unit and cover the same interval from opening to close. Count the departure and the corresponding arrival at each handover, including entry to and exit from stock in transit. Every internal handover then contributes equal and opposite changes to the combined holdings. These changes cancel at each regrouping. This is why the same rule can carry from branches to districts and to the cooperative. A crate omitted in transit breaks that reasoning. C.39.WA:4.2 develops the inference across composition scales.
Another recipient must tell a customer how many crates can be issued at one branch today. The cooperative could own the same total with the crates ready at that branch, at another branch, reserved, or on a vehicle. Preserve location, reservations and availability at the promised time. A reliable count of available stock at that branch can suffice. If a transfer will be needed before collection, follow the relevant transfer work and the stock it makes available. The size of the cooperative has stayed the same. Its ownership question and the branch’s issue question select different composition ranges and model distinctions.
A guide that supplies both counts can show how to use them. If it promises to teach how to design the explanation for the next order, the reader needs the step the author took first. Identify who owns the stock when ownership is asked about, where and when it must be usable when issue is asked about, and who may choose the means. From that commission derive which totals may be combined, which differences the recipient must retain and what grounds would be sufficient. The counts illustrate the reasoning.
Bias-Annotation
A specialist may supply a missing dependency from experience. Establish what the intended recipient can recover from their declared preparation, the passage and the materials they can actually use.
When an incident prompts a question about a recurring practice, ask what makes the substitute work attractive or allows its result to count as completion. At the first divergence, recover what the contributor could choose, what they knew then and what would count as completing their work. The same fault at the same point can have different causes and require different changes. For example, an author might choose a question that the available material already answers. An editor might accept the resulting explanation because it is clear and accurate. Keeping the recipient’s question visible when choosing the work can expose the first substitution. Trying the answer against the recipient’s intended use can expose the second. Look for an instruction, decision or consequence that distinguishes the live explanations. If the evidence only locates the substitution, its cause remains open. The two roles are illustrative. One person can make both choices, and the explanation can involve relationships across several parts of the practice.
For a proposed repair, identify a substitution that could previously pass and show why the changed practice would catch it while accepting an equally specific answer to a genuinely specific question. Whether people use the change and whether the fault recurs are questions about subsequent work. Locating an error can itself settle a request confined to that incident.
Conformance Checklist
- The recipient’s task determines which kinds of holon and which constituent relations matter, the required range for each and the decisions that remain open.
- The proposed model preserves the differences, grounds and freedom needed under the promised conditions. A claim spanning scales or models has grounds for the connection in its stated direction.
- The author can explain why a coarser account suffices or which required use it would lose, and which extra detail would change a decision or fulfil a separate obligation.
Common Anti-Patterns
Labelling an example L0 and its introduction L3 can leave the reader unable to construct the promised disciplinary rule. Find the choice they must make and give the relation and grounds they can use. Counting additional examples at the same scale leaves that choice open. Requiring every scale and every model level in each passage creates the opposite burden. Follow the dependencies of the commissioned work.
Consequences
An author can provide a short answer when one supported total decides the order. An editor can preserve a worked case and restore the choices needed to explain a family of cases. A joint decision can require additional work at several composition scales, even when every local conclusion is correct.
Rationale
Merging situations in a model is adequate for a decision when the resulting information still supports that decision under its permitted variations. Fixing one arrangement is adequate when the commission entrusts that choice to the designer. Constituents can vary together under a constraint. Following their separate variations alone can then misstate what the whole admits. A choice over scales and a choice of model abstraction therefore have separate tests and can require revision together.
SoTA-Echoing
Parker distinguishes success on one occasion from adequacy for a type of use and relates model evaluation to the user, purpose and circumstances. NASA’s modeling practice asks engineers to specify the intended use and the inclusions and fidelity it needs. Levin analyzes how a scientific question selects observational scales and how processes at different scales can interact. Describing every known detail can help when a recipient must reconstruct the whole record. For a decision with an established bound it can add work without changing what follows. Matching a previous successful case leaves the suitability of the model open when the purpose or conditions change. The constitution of a Method, a piece of Work or an episteme is established in its own practice. Parker, Model Evaluation: An Adequacy-for-Purpose View · NASA, Models and Simulations Use · Levin, The Problem of Pattern and Scale in Ecology
Relations
B.1:4.4 follows the composition of a holon. B.5.MR:4.2.2 constructs an abstraction of a model for a specified inquiry. C.37.MM:4.1.1 connects several models when their joint use matters. C.30.STRAT:4.2 resolves what a layer denotes. C.39.QR returns partial work to the original commission. C.39.GA repairs a missing general argument or explanatory step when the intended reach is already established. F.19 uses these methods within its ordinary reading when the choice affects the reader’s work. E.10:7 governs the generality and domain anchoring of names.
C.39.SA:End
Turn a Construction into a Reusable Operation
Type: Method pattern Status: Stable Normativity: Normative
Problem frame
Use this pattern when a worked way of obtaining a result should become available for changed inputs, another case or combination with other operations, and its reusable part is unclear. A correction works for three timing markers but its numbers have been copied into the next recording. A partition has been found for one set of conflicts, while the next project has different participants. A physical action succeeds once, while the conditions for repeating it remain implicit.
The subject is the obtaining operation being developed: what it takes, what it does, what it returns and the conditions under which that result follows. Start from a construction whose useful steps can be recovered. B.5.RC and B.5.RA help when those steps or their reasons are missing.
First useful move. Choose one foreseeable change of input. Identify which values belonged to the earlier case and which relations made its construction work. For a timing correction, the offsets can change while the operation remains “center the smallest and largest offsets.”
The result is an operation that another use can instantiate, together with its meaningful inputs, result and application conditions. A formula, procedure, diagram or demonstrated action can express the needed contribution. Select the expression that the receiving practitioner can use.
Use an existing adequate operation. Develop only the part whose reuse is missing. When the question is what happened in several observed performances, A.3.1.MR supplies Method recovery. When a formal U.Method or MethodDescription claim is needed, A.3.1 and A.3.2 govern that claim after the obtaining operation has been made clear.
Problem
A successful case mixes reusable relations with its values, available resources and incidental choices. Copying the whole case can fail when any of those particulars changes. Replacing all its nouns with variables can fail as well: the result may have depended on a fixed geometry, an allowed operation or a physical condition that the generalized description has erased.
The practitioner needs to expose what another use may vary while retaining how the result is obtained. This work can also reveal a useful limit: the generalized operation may return an obstruction or an unattainable tolerance for some inputs.
Forces
Solution
Recover the working construction, choose the intended variation, expose an operation under retained conditions, and establish what it returns over the intended range of use.
Recover the case → choose what may vary → construct the operation → justify its result → resolve any consequential change of use → retain, narrow or develop it.
Recover the useful construction
Follow how the existing case obtained its result. Identify the material or values used, the operations performed, their dependencies and the reason the result answered the question.
Select a receiving use that needs more than the original answer. It may need the same transformation on another input, an intermediate result for another operation, or a way to adjust the construction. If the original result suffices, use it and stop.
A trace can reveal that the apparent reusable step depended on an unreported contribution. Recover that contribution through B.5.RC or the relevant subject Method before building on it.
Separate case values from application conditions
Choose which parts should vary in the receiving use. Give each variable part a meaning, such as a finite set of offsets in milliseconds, a collection of activities or a requested volume. Retain the relations among them.
Identify the conditions that stay in force: a common time reference, symmetric pairwise conflicts, an allowed constant correction, or a physical rate over the relevant interval. A case value becomes a parameter only when the proposed operation explains how it uses different admissible values.
Examine a boundary of the proposed input. A procedure that takes a minimum and maximum needs a nonempty set. A partition procedure needs the conflict relation that it handles. State a useful failure result when an input can be understood but the requested outcome is impossible.
Construct the reusable operation
Replace case-specific choices with operations on the selected inputs. Preserve their order and dependencies wherever changing them would change the result.
Three common constructions can help:
- Parameterize a choice. Replace “subtract 42.5 milliseconds” with “subtract the midpoint of the smallest and largest observed offsets.”
- Compose available operations. Convert the receiving data into a form an existing operation accepts, perform that operation, and interpret its result for the receiving use. Check the intermediate meaning at each connection.
- Replace a constituent operation. When the intended variation invalidates one step, find a substitute that supplies the result its successor needs. Retain the other steps whose conditions hold.
Use the subject practice to construct a missing constituent. C.39 helps find such a way; B.5.TU helps make an available theory operative. If the desired result is unsupported, name the unresolved step and the smaller result already obtainable.
Explain what the operation supports
Recover the reason for the output over the proposed input range.
For a deductive construction, follow the argument with the case values left variable. State the premises used and inspect the branches, exceptional inputs and termination conditions that matter. A short bound can establish both an operation's result and its optimality for a stated criterion, as in :5.1.
For an operation proposed from observations, identify the observed support and the conditions under which repeatability is expected. Use an available subject model when it supplies that connection. A changed physical setting may leave the mathematical operation intact while invalidating its application to the setup.
Keep the conclusion at that supported scope. Choose further proof, model development or observation when the receiving use needs it and the obtainable contribution warrants its cost.
Examine a changed use
Use the argument for the commissioned range. Where its assumptions leave a consequential alternative open, choose a changed input that distinguishes it and interpret what the operation returns. Work through an additional use when the recipient needs the demonstration or has commissioned a trial. With stipulated physical conditions, a calculation can give a required setting, a bound or a prediction. To report what an apparatus actually delivered, use evidence from that run, such as a calibrated measurement of the delivery.
Compare the obtained result with the receiving question. If the operation returns a failure witness, use it to change the proposed action or question. If an assumption fails, repair the dependent part or narrow the application conditions.
A worked changed case can expose a defect in the proposed operation. The scope of a general conclusion comes from its argument or empirical basis. B.5.RR supplies local revision of that reasoning; B.5.QD develops a new question when the result calls for one.
Make the operation usable at the receiving point
What will the next practitioner do with the operation? To apply it, the practitioner needs to know what the inputs mean, which conditions to check, how to obtain the output and what decision the output supports. Sometimes a justified bound or an impossibility result already settles that decision. Give the practitioner the grounds and the conditions under which they hold.
To adapt the operation, the practitioner also needs to know why it works, which assumptions the new use affects and which parts can still be used. If the assignment is to develop a reusable operation independently, identify which construction choices the practitioner made and what was already supplied by the example, a tool or a colleague. Observing the practitioner correctly apply a supplied operation supports a claim about that performance. A claim about independent construction needs evidence of the practitioner's own choices and their grounds.
Archetypal Grounding
Generalize a timing correction and obtain an impossibility result
A recording has picture-to-sound offsets of 40, 42 and 45 milliseconds at three matching markers. The available correction subtracts one constant from the sound times. Centering the extreme offsets gives 42.5 milliseconds and residuals -2.5, -0.5 and +2.5 milliseconds.
The receiving use needs that construction for any nonempty finite set of marker offsets, expressed in one common unit and reference. Let m be the smallest offset and M the largest. Construct:
- correction c = (m + M)/2;
- largest residual magnitude r = (M - m)/2.
Every marker offset lies between m and M, so subtracting c leaves each residual between -r and +r. Any constant correction leaves the two extreme residuals separated by M - m. At least one therefore has magnitude at least (M - m)/2. The proposed correction attains that lower bound: it minimizes the largest residual magnitude at these markers.
The reusable operation takes the marker offsets and returns c and r.
For a stipulated tolerance T, r at most T supplies a feasible constant correction at the markers; r greater than T shows that no constant correction can meet that marker tolerance. One marker gives r = 0. An empty set supplies no extremes for this construction.
Now add a marker with offset 48 milliseconds. The operation gives c = 44 and r = 4. For a tolerance of 3 milliseconds, every constant correction fails. The two extreme offsets differ by 8 milliseconds, while two residuals each within 3 could differ by at most 6.
This failure witness lets the practitioner stop searching among constant shifts and consider another time correspondence or a changed use.
The derivation concerns the supplied marker offsets. Applying a correction throughout a recording also needs the relation between markers and a qualified editing operation; C.39:5.2 retains that receiving question. Unknown drift between markers remains unresolved by this finite calculation.
Turn a partition construction into an allocation operation
A workshop has activities A, B, C and D and two sessions. Its stated constraints are pairwise incompatibilities: AB, BC and CD cannot share a session. Activities have no other scheduling or capacity constraints in this constructed case.
B.5 supplies a construction that returns either a two-colouring of a finite undirected graph or an odd-cycle witness. To make it usable for workshop allocation, construct the connection:
- Make one vertex for each activity and one undirected edge for each incompatibility.
- Apply the two-colouring operation.
- Interpret the two colour classes as the sessions. If an odd cycle is returned, read its edges as the incompatible pairs preventing a two-session allocation.
For the supplied path, the operation returns sessions {A,C} and {B,D}. Every incompatibility has endpoints in different sessions.
Generalize from these four activities to any finite set with the same form of pairwise constraint. The encoding preserves the stated prohibition on sharing a session. A two-colouring therefore yields an allocation satisfying every such prohibition. An odd cycle cannot alternate between two sessions all the way around, so it gives a failure witness.
Add incompatibility AC. The triangle A-B-C-A is now an odd-cycle witness. Changing only the session labels cannot repair it. The receiving choice concerns another session, a changed incompatibility or a different activity arrangement.
The reusable contribution is the composition of encoding, the available graph operation and interpretation. If session capacities or precedence constraints are introduced, this operation supplies only the pairwise-compatibility part of the allocation. Those additional constraints need their own construction.
Retain a physical condition when reusing a dispensing operation
A dispenser supplied 50 millilitres during a 2.5-second open interval. A proposed reusable operation would dispense a requested volume V by opening for V/20 seconds.
A subject model can justify that operation under a constant flow of 20 millilitres per second over the chosen interval, with the opening and closing effects already accounted for. Under that model, a request for 80 millilitres gives 4 seconds.
Suppose the receiving setup instead has a stated constant flow of 10 millilitres per second. Keeping the old four-second action would supply 40 millilitres. Exposing flow q as an input repairs the construction to duration V/q, for positive q under the stated constant-flow conditions; 80 millilitres then takes 8 seconds.
When flow changes during discharge, V/q with an unqualified constant no longer supplies the volume-duration relation. C.16.MR can construct the measurement relation, and A.3.3.TR can supply the changing-state account. The useful outcome here may be the identified missing relation and continued use of the earlier operation where its original conditions hold.
The numerical cases instantiate supplied models. The single 50-millilitre observation supports that observed outcome; a repeatability claim for the real apparatus needs its corresponding physical basis.
Bias-Annotation
Under E.3's five Principle-Taxonomy lenses (Gov, Arch, Epist, Prag and Did), this Method favors constructions whose steps and conditions can be recovered. Its mathematical and programming examples show how to recover those steps and conditions. Skilled bodily or collective practice may need demonstrations, observation or other descriptions to expose the same operational dependencies.
People and computational agents can contribute different parts of this work: recovering a case, deriving a relation, testing an implementation or supplying a physical condition. Their division of work follows the operations and available capability.
Conformance Checklist
- CC-RO-1 - Receiving variation. The intended changed use identifies which part of the earlier construction needs reuse.
- CC-RO-2 - Meaningful inputs. Variable inputs and retained application conditions are distinguished.
- CC-RO-3 - Obtaining operation. The proposed steps explain how the inputs produce the result, including any composed or replaced contribution.
- CC-RO-4 - Supported scope. The argument or empirical basis supports the scope claimed for the operation.
- CC-RO-5 - Changed use. The grounds cover the variation claimed. Where a changed use remains uncertain or a trial was commissioned, its result or useful failure is available.
- CC-RO-6 - Affordable completion. The receiving use can continue from the operation or its supported limit; further inquiry is selected for what it can change.
- CC-RO-7 - Local repair. A changed condition reopens its dependent operations and conclusions.
Common Anti-Patterns and How to Avoid Them
Consequences
A useful case becomes an operation that further work can instantiate and combine. Its exposed conditions also make local revision possible. A failure result can save work by excluding an unattainable request under the operation.
The added effort is recovering the construction, choosing the intended variation and establishing its scope. It is justified when another use needs the operation. A broader library, a new publication or a fully automated implementation may remain unnecessary for that use.
Reopen the construction when an input meaning, application condition, constituent operation or receiving result changes in a way its account cannot handle.
Architectural Rationale
Case recovery and reusable-operation construction have different results. Recovery makes an existing construction available; this Method changes how it can be instantiated or combined. Stating the receiving variation prevents expansion into a general library-design task.
Meaningful inputs and conditions hold the proposed scope together. Parameterization exposes variation, composition connects available contributions, and replacement repairs a constituent that no longer serves the use. Their subject arguments explain why the resulting operation works.
The timing case derives a family of conditional results. The allocation case composes a representation change with an existing algorithm. The dispensing case shows where a mathematical rule depends on a physical account. Together they connect mathematical construction, computational performance and physical application while retaining the different work each requires.
An operation can be described in different forms. Its use depends on recoverable meaning and available performance, while formal Method and description classifications are governed by their own patterns.
SoTA-Echoing
For the question “How does a useful construction become an operation another use can vary?”, the selected line combines constructive generalization with application conditions and examination of a changed use. It exposes the obtaining step and its scope before investing in a larger reusable library.
For the timing case, compare carrying the earlier 42.5-millisecond correction with carrying the midpoint operation. Both use the same available marker data. Copying the constant is sufficient for the unchanged case; after the 48-millisecond marker is added it leaves a largest residual of 5.5 milliseconds.
Computing the new extremes costs a scan of the markers and yields the best constant correction, with a 4-millisecond residual bound and a proof that the 3-millisecond target is unattainable. Adopt the operation when changed data or a feasibility question makes that additional work useful.
Grand et al., LILO (2024), §3 and §4.1–4.2 connect reusable program components with further synthesis. Their comparison also shows that components can be hard to use when their names obscure their operations; generated explanations can introduce semantic errors. Adapt this contribution in :4.3/:4.6: expose the operation and describe its meaning. Their experiments concern three programming benchmarks and earlier language models.
Ahmed et al., TheoryCoder-2 (2026 preprint), §2.2–3 connect abstract operators with predicates, a lower-level world model and executable actions, then reuse the operators across related game environments. Adapt the connection between proposed effect and obtaining operations in :4.3/:4.4. The reported transfer uses those representational assumptions and game settings.
The source synthesis supports constructive reuse without requiring the whole automated learning system for an individual operation. Reopen the selected approach when another available Method already supplies the intended reusable result more affordably, when the operation's explanation obstructs its use, or when the receiving variation exposes an unsupported condition.
Relations
- C.39 finds and develops a way to obtain a needed result; this member develops a construction for reuse.
- B.5.RC, B.5.RA and B.5.TU recover constructions, arguments and operational theory use needed by the development.
- B.5.RR and B.5.QD revise affected reasoning and develop a next question from the result.
- A.3.1.MR recovers candidate Methods from Work sources. A.3.1 and A.3.2 govern Method and MethodDescription claims.
- A.6.3.RT and A.6.3.RT.OE change how content is represented and construct expressions usable for the selected operation.
- C.29.2 and C.29.3 construct computational formulations and their physical realizations.
- C.40 examines variations and useful continuations. C.38 compares obtaining ways when that choice is needed.
- C.11.DUA selects worthwhile information work for the receiving use.
C.39.RO:End
Choose and Reopen a Method Development Focus
Type: Method pattern Status: Stable
Working result: a Method-focus result that selects a Method, a proposed Method to investigate or construct, an established family or local grouping, a set of Method relations, or a non-Method return for one project decision. The result preserves every Method or candidate-account status and names one observation that would reopen the focus.
Use This When
Use this pattern when a project says that its “methodology”, “process”, “workflow”, or “way of working” must change, but the decision subject is ambiguous. The concern may be one Method, several related Methods, an established Method family, a temporary comparison grouping, or something else such as a tool, capability, support arrangement, description, resource, or project System.
Begin with the result that is missing, late, unsafe, or too costly and the Work in which it matters. Identify the Methods already used there and any accounts that still need examination. What could change the decision about that result: improving one Method, changing how Methods work together, or securing the tool, capability or service the work requires?
The first useful move is to compare the plausible focus options: one Method, an established family or project-local grouping, named Method relations, or a non-Method return. A project can choose a proposed whole as the subject of design or investigation while leaving its Method status open. To rely on it as an established way in current work, recover the grounds for that status and use. This gives later repertoire, qualification, or architecture Work a supported subject and the relations it needs to address, without inventing a Method or a composite.
Do not use C.39.MF merely because one tool failed, one practitioner lacks capability, one document is incomplete, or one project, process, or case view is inconvenient. Return that question to its subject unless changing a Method would change the receiving result.
Working Distinctions
When the consequential work or worthwhile improvement has not yet been found, C.39.WI supports its discovery and comparison. Use its result here to settle an ambiguous Method or non-Method focus; an already adequate focus needs no repetition.
Problem Frame
Project teams encounter Methods through handbooks, standards, stage models, issue trackers, training, job titles, toolchains, and remembered practice. These can reveal useful Methods, but their visible boundaries rarely coincide with Method identity.
A project schedule can describe intended Work, a process account a reusable Method, and a case account the condition or acceptance of a named result. These descriptions can reveal connected Method questions without sharing a subject. When several accounts concern one admitted Work, preserve that identity across them. Recover the claim and subject before selecting a viewpoint; changing a representation creates neither another Work nor a Method.
Problem
A focus chosen too broadly turns every artifact and supporting System into a Method part. A focus chosen too narrowly optimizes one operation while the receiving result depends on several Methods or a support relation. A familiar label can also turn a project-local shortlist into an alleged professional family.
The project then asks the wrong downstream question. It builds a repertoire for a tool problem, qualifies a candidate account as though it were an identified Method, or compares “architectures” that are only differently drawn views of the same Work.
Forces
Solution
C.39 supplies the receiving-result and missing-way inquiry. Use this narrower Method-focus construction when it is not yet clear whether changing a Method, a relation among Methods, or its support would answer that inquiry. An adequate focus supplied by C.39 or A.15.6 is an input, not work to repeat.
Select a focus whose subject and status are supported. Compare what addressing each plausible focus could contribute to the receiving decision, together with all the work and resources it requires, using C.11 or a domain decision Method.
Retain the dependencies needed for that contribution, including relations among several Methods when they affect the result. Keep the rejected focus options and one observable reopen condition in the result.
Pattern-Use Unfolding
- Name the receiving result and decision. State the result at risk, who will use it, the relevant situation and interval, and what decision the focus will enable.
- Recover representative Work. Identify the Work occurrence or intended Work that exposes the difficulty. Use
A.15.6to keep the project System, use, Work, Method, support, and development subjects distinct. - Recover the subjects of the accounts already in use. For each project, process, case, lifecycle, stage, or workflow description, state what its claims concern, their actual or intended status, the question exposed, and the material omissions. Use
A.15.6when the label hides the subject. Keep unlike subjects separate; a claimed view additionally needs its viewpoint andE.17.0conformance result. - Recover Method-status candidates. List Methods already identified under
A.3.1, candidate Method accounts, established family facts, local groupings, related Methods, and support or capability alternatives. Preserve their statuses. - Form different focus options. Include each plausible one-Method, plurality, relation, and non-Method branch. A renamed package or differently drawn view is not another focus option.
- Test family and relation claims. Apply the rule that defines the claimed family and the applicable relation pattern for each relation. Use
G.5when the question needs its Method-family registry or selector-facing result. For use in current work, claim a composite Method only on grounds for the whole and its constituent relations. For design or investigation, a proposed whole can be the focus. State what would constitute it and what still has to be developed or established. - Choose and record. Use
C.11or a domain decision Method to select the subject for the receiving work. State whether the project is choosing an established way to use or choosing what to investigate or construct. State rejected options, evidence limits, next useful result, and the observation that reopens the focus.
Record the Result
What Changes in Practice
The project stops asking which named methodology should replace another. It first decides whether the subject is one established Method, a possible whole to design or investigate, a governed family or local grouping, relations among several Methods, or a non-Method condition. Downstream Work can address the selected subject and the relations on which its result depends. A useful plurality can remain plural without being packaged as a fictitious composite.
Archetypal Grounding — EC-417 Release Focus
In the EC-417 release scenario, eight of twenty releases reopened. The package is commonly called the “release methodology”, so the first proposal is to replace it as one Method.
Three management questions expose different subjects in this scenario:
The status inventory contains four identified Methods: M-HW-Verify, M-SW-Integrate, M-Supplier-Approve, and M-Release-Authorize. It also contains candidate accounts C-Evidence-Reconcile-Internal, C-Evidence-Reconcile-Supplier, and C-AI-Trace-Review. C-EC-Release-v2 is only a proposed-whole account.
The PLM, CI, test rig, AI provider, safety capability, and supplier responsibility are support, System, capability, access, or assignment subjects rather than Method candidates by position.
The schedule puts the planned signature after integration. If integration requires that signature first, the plan conflicts with the required order. If the applicable method permits provisional integration and reconciliation before release, the dates alone leave feasibility open. Recover which evidence rule governs this case. These connections support a Method-relation focus without making the three subjects one Work.
The release team can now investigate the Method relation that affects integration. If a later commission asks whether these contributions could form one whole Method, that proposed whole can itself become the design subject. Its constitution and adequacy then have to be established.
Reopen to a test-capability focus if two of the next three comparable delays occur while required evidence is complete and the rig is unavailable.
Bias-Annotation
Conformance Checklist
- The selected subject and its status answer the receiving question. The focus can be an established Method to use, a possible Method to investigate or construct, a relation among Methods, or a condition such as access or provision that can settle the work directly.
- Every Method and candidate account keeps its prior status.
- An established family cites an independent classification or membership basis.
- A project-local grouping states its criterion, use, and non-family status.
- Project, process, and case accounts preserve their subjects and intended or performed status; a claimed view has its viewpoint-conformance basis.
- The options include a non-Method return whenever a tool, capability, support arrangement, description, resource, or System could be decisive.
- The selected focus states rejected focus options, uncertainty, next useful result, and an observable reopen condition.
- Method, family, relation and composite claims rely on their identifying evidence; package position, a view or shared use alone is not that evidence.
Common Anti-Patterns and How to Avoid Them
Consequences
Method development begins with a decision-usable subject instead of a fashionable label. Later repertoire and criteria Work can remain narrow, while a relation focus exposes coordination questions without pre-committing to a whole.
The cost is a short recovery of Work, views, statuses, and non-Method alternatives. Some projects will stop before changing a Method because the useful answer concerns a support System, capability, description, or resource.
Rationale
The focus identifies the subject and relations that later Method development must address. Preserving the difference among a Method, family, local grouping, relation focus, and non-Method subject prevents downstream qualification or architecture prose from upgrading an unsupported object.
C.39.MF connects questions about Methods, their descriptions, Work, capability, tools, variants and Method change. Project, process and case accounts help discover those questions. Recovering each claim's subject before selecting its representation preserves both useful accounts of one Work and useful connections among different subjects.
SoTA-Echoing
Reopen the affected rule when a current source or representative use exposes another materially different focus class, or when an account's subject, status, or omitted claim changes the Method question or non-Method return.
Relations
A.15.6distinguishes project subjects and routes a Method-of-interest question here.A.3.1governs Method identity;G.5supplies Method-family registry and selector-facing results when its entry conditions hold;C.11governs the bounded focus choice.- C.39.RP may organize a repertoire for a one-Method, plurality, or relation focus. C.39.RQ may state criteria at that same level of focus.
- A.3.1.QM receives identified Methods and candidate accounts without changing their status. C.32.MWA.PP receives a several-structure question; B.1.5.CM receives one proposed whole only when that question is live.
- An applicable research Method may explain an identified Method, established family inclusion, or named relations. A local grouping label is not its explanandum.
C.39.MF:End
Discover Worthwhile Method Improvements
Type: Method pattern Status: Stable
Problem frame
Use this Method when work consumes substantial resources, causes avoidable errors or rework, or leaves valuable results unattainable, and you need to choose where to improve the way of working next. It also applies after a successful change makes a different costly limitation visible. The sought result is a warranted improvement undertaking, a useful retained possibility, or a reasoned decision to continue current practice.
The practitioner may maintain a professional repertoire, an organization's practice, or a framework used by many others. Start from the result people and agents need to obtain and the work needed to make it usable. The size of a document, popularity of a topic and number of complaints are clues rather than measures of the value of its improvement.
This is the discovery and selection branch of Method development. It connects evidence of costly work to a useful change and later returns from its effects. C.39.MF resolves an ambiguous Method or non-Method focus; C.39 develops a missing obtaining way; C.11.MV compares the worth of complete-enough Method options. An already understood, worthwhile and authorized remedy can be performed. Do not require another improvement survey before correcting a clear error, satisfying a binding requirement or protecting someone from harm.
The first useful action is to follow one consequential work result far enough to identify what consumes resources or prevents its use. Compare a plausible improvement with continuing the current way and with addressing a more consequential limitation. A full inventory of the organization or framework is unnecessary.
Problem
Improvement work gravitates toward visible defects, easily edited descriptions and familiar techniques. Expensive coordination, repeated reconstruction, poor decisions and missing professional explanations can remain unnoticed. Conversely, an expensive activity can be necessary and well performed: its cost alone supplies no remedy.
Two shortcuts distort selection. Optimizing local speed can leave the whole's accepted output unchanged or make it worse. Rejecting every improvement outside the current constraint can preserve avoidable payments, human burden, failures and lost future possibilities. A third shortcut identifies an operation in a handbook and concludes that the practice is adequately supported, although the practitioner must invent how to select, join, adapt or use it.
The useful question is which attainable change improves the receiving work enough to warrant its complete burden. Discovery must continue as work and capabilities change, without becoming a standing demand for more measurement, more publications or more activity.
Forces
Solution
Discover a consequential limitation from work, estimate its relevant consequences, distinguish what needs to change, and compare feasible remedies with the current arrangement. Develop and use the selected contribution, then return what happened to the next choice. These operations can use existing results and return to one another; they are not a compulsory cycle after every task.
Find consequential work from the current position
Use an occasion already available: planning an undertaking, reviewing a service result, renewing a resource commitment, examining repeated failures, or completing a useful change. Follow the result into its receiving use. Ask where participants repeatedly spend effort reconstructing a way, obtaining missing inputs, waiting, negotiating incompatible commitments, checking, correcting or repeating work.
Look beyond reported failures. Examine a successful result whose preparation or support remains expensive, a recurring difficulty treated as normal, and an important contribution nobody can yet obtain. The last case may have no hours recorded because the work was abandoned or never attempted. State that opportunity and its basis; do not assign it an invented cost.
Recover the affected work population and horizon. For a recurring occasion, separate how often the work occurs from how often the difficulty occurs within it. For an uncommon but consequential choice, identify the later work or loss it can influence. Use existing traces, invoices, examples and competent accounts. A sample can reveal a mechanism or supply a rough bound without establishing its prevalence.
Collect only enough serious possibilities to make the present choice. Do not convert this into a taxonomy exercise or an exhaustive audit. A broader survey earns its cost when plausible unseen alternatives could change an important allocation and the narrower basis cannot answer it.
Make the burden and possible gain comparable
Describe the incumbent and proposed alternatives over the same receiving result and horizon. Retain any material difference in quality, uncertainty, accepted demand, service conditions or participant protection. A cheaper result with a different acceptance boundary is a different offer, not automatically a more efficient way.
For recurring human effort, an initial estimate can use occasions multiplied by participants and effort per occasion. Split roles when their participation differs. A repeated-failure estimate can separate occurrence frequency, failure fraction and additional recovery effort. Label assumptions and use ranges where variation matters. These calculations locate a possible improvement; they do not prove which share is avoidable.
Keep distinct:
- human effort, including learning, oversight and work displaced elsewhere;
- elapsed delay and its effect on a commitment or opportunity;
- payments, receipts and investment;
- usable capacity released, including when and for whom it becomes available;
- accepted output, reliability, exposure and other consequential results;
- the future capability or reusable means acquired.
Count each burden once. The same hour may appear under a goal review, coordination and capability development; these are overlapping descriptions, not three hours.
Do not add an opportunity-cost deduction when the comparison already includes the foregone contribution. Distinguish gross work from work added by the change: a new conversation may replace an older meeting.
Trace every claimed gain to its consequence. Released salaried time is not an avoided payment merely because it has a labor rate. It may enable another useful result, protect recovery time or remain unused. Reduced per-use charges or a contract not renewed can change cash even if throughput stays the same. A less burdensome method can be valuable on human grounds without fabricating a monetary return.
Use C.11.DUA for the value and burden of further inquiry and C.11.MV for a consequential Method comparison. OPS.14 supplies an operating financial comparison when that is the current use; the appropriate professional practice supplies other cost, service or risk accounts.
Relate local savings to the encompassing work
Identify the whole result whose accepted output, completion time or conditions might change. Ask whether the proposed remedy acts on a currently limiting relation, changes the load reaching it, protects it from rework, or only changes another contribution.
The operating constraint for sustained output, the conditions determining a deadline and the resource-constrained critical chain of a project need not select the same action.
When accepted demand and a limiting relation are adequately supported, compare the useful effect of relieving it. A faster nonlimiting operation does not by itself increase the whole's output. Do not count extra intermediate products as accepted throughput or remove protective capacity solely because it is not fully occupied.
Also compare consequences that do not require more output: avoided variable charges, a smaller future resource commitment, less rework, safer conditions, lower fatigue-producing load, or a reusable means that makes later change attainable. Obtain the appropriate basis for those consequences. A claim of safer conditions needs its professional grounds; a cost calculation supplies none.
Include what improvement takes from the whole: scarce professional attention, interruption, transition failures, review, qualification, support and displaced delivery. A useful nonconstraint improvement can be postponed, performed by another capable participant or made smaller when its current implementation would displace a more valuable contribution at the constraint. Conversely, an independently attainable saving need not wait for every constraint initiative to finish.
Use OPS.9 when the operating limiting relation is disputed, B.1.5.EW for the relevant encompassing Work, and C.32.MWA.PP when changing the arrangement across performance, provision and development is the live question. Do not impose a throughput model on every purpose. Binding service, authority and protection conditions retain their force even when a proposed alternative has an attractive economic comparison.
Distinguish the missing way from missing means and failed use
Before investing in a better way to perform the work, recover why the receiving result needs it. If a simpler contribution could suffice, compare removing or reducing the work or changing the demand that creates it. Preserve the acceptance conditions and obligations; changing a requirement needs its responsible authority.
If the work remains necessary, examine its way and support.
Follow a representative difficulty using the complete available Method and the support available there. Preserve what the practitioner or agent must bring, obtain elsewhere or invent. A plausible description is not performed work; an observed success with expert assistance is not unassisted capability.
Use the distinction that changes the remedy:
Several rows can apply. For example, an agent may recover a correct rule yet select another action because its task directs it toward a narrower result. Repairing retrieval alone will not change that action.
Conversely, a newly available agent can make a professional result obtainable without teaching every recipient its production Method. Preserve an intended human mastery result when that is what the undertaking promises.
For a source or framework adequacy question, recover the source's substantive difficulties, remedies, conditions and constructive organization before distributing them among familiar owners. Compare those contributions with the available explanations and intended reader preparation. Derive distinguishing cases from that recovered scope.
A few successful cases support those cases; they cannot establish that an assigned field or source repertoire is covered. Known constituent Methods also do not settle whether a new domain framework or a better organized explanation would improve use.
C.39.MF resolves the focus when the subject remains ambiguous. Do not turn the table into an obligation to investigate every possible cause. Return the supported remedy or precise unresolved distinction, and preserve unaffected results.
Repair a failure that survives correction
Use this branch when a consequential failure survives a correction, several results suggest a shared failure of method use, or an earlier repair leaves dependent work relying on a rejected basis. Restore the receiving work by changing the conditions that reproduce the failure and revising the affected results. A known sufficient correction can proceed. Protect a recipient who needs an immediate correction while resolving the remaining question.
The failed use identifies what needs restoration. Evidence about how it was produced, checked and previously repaired selects the intervention. Dependence on the changed basis determines which earlier results need reconsideration.
What happens after the intervention can reopen the explanation or the repair's reach. These connections let a repair finish with a usable result while retaining causal uncertainty that does not change the action.
Recover the episode and the result it had to supply. Start with what the next participant needed to decide or do, the applicable Method and the condition that was not met. Establish that condition on the receiving practice's grounds. A useful result can be rejected by a mistaken local check; the checking criterion then needs examination.
Follow available evidence from the input and its interpretation through the consequential choice and action to the result and its use. Recover the assignment, material and edition, available support and relevant constraints. A connected account from the performer can locate a missed transition; compare it with the passage, action or result that bears on the claim.
Distinguish an observed operation, a participant's account and an inferred cause.
Opening a document establishes access; a retrospective account from an AI agent does not expose its hidden reasoning.
Keep unavailable history unknown and obtain a comparable present observation when it can distinguish the remedies.
Locate the earliest supported loss while preserving later contributions that remain valid on their own grounds. A correct comparison can use an incorrectly narrowed assignment. Repairing the final comparison alone would leave that narrowing in place. When the assignment or acceptance criterion is disputed, recover how it was selected and what supported that choice.
Find what directed production, checking and the earlier correction. Compare the governing operation with the instructions, examples, tool behavior and criteria available at the consequential decision. Trace a contributing input to its maintained source and to the working form used.
A source that permits the wrong move and evidence that this source directed the observed move support different claims; retain that distinction in the diagnosis.
If the result passed a check or returned after repair, include those interactions. Recover what the checker or previous repairer had to establish, the information and feedback they received, the criterion they applied and the relevant condition they changed.
A shared mistaken criterion can direct both production and checking. A corrected source may never reach the working template. A previous intervention may reach its user yet address a different cause.
Ask what would have interrupted the failed transition and whether the earlier correction could provide it under the observed conditions.
When repeated authoring and checking share a disputed interpretation, obtain a separately formed account of the original question and relevant subject grounds before supplying that interpretation as the answer. This can reveal an omitted operation or an unjustified criterion.
Verification of a proposed repair then uses the diagnosis and its evidence. Select the needed independence and reader preparation for the judgement being made; a second participant using the same unsupported basis supplies no independent ground.
Distinguish explanations when their difference changes the repair. Use C.28.CM to construct or challenge an unresolved causal account: state the outcome contrast, proposed relations and serious alternatives, then derive what would differ in available evidence or an obtainable case. A missing operation, an unavailable current edition, a contrary criterion and a resource conflict can produce similar omissions while requiring different interventions. Several causes may contribute.
Select further inquiry through C.11.DUA. Existing evidence may settle the choice. A contradiction between an operative criterion and its governing rule can be corrected on substantive grounds while its contribution to past recurrence remains uncertain.
If the surviving accounts support the same sufficient repair, preserve that uncertainty and proceed.
If they require incompatible actions, obtain the discriminating contribution or state which choice remains unsupported. The recurrence makes the earlier repair a matter to explain; it does not require a survey of every possible cause.
Change the contributing condition and make the change usable. Choose the object of repair from the supported account. Use C.39 for a missing way or explanation; A.15.8.MS for method-use support; and the responsible provider for missing capability, inputs, resources or authority. A.15.10 helps restore lost continuation meaning, and A.15.11 brings an applicable Method into consideration. Use an already sufficient contribution.
Correct a wrong criterion where it is maintained and in the working forms that supply it at the decision. Remove the obsolete instruction when leaving it alongside a reminder would preserve the conflict.
A correct criterion that cannot be performed may instead need different provision. Preserve the useful operation while changing its support.
When the remedy changes the Method, identify and qualify that change. A.3.1.MI connects introduction with the comparison of delivery, mechanism and context; IS.2 and IS.3 supply diagnosis and strategy within a helping undertaking. Their adequate results can enter this repair without repeating their work.
Choose how to obtain the needed contribution. Qualified provision can supply an immediate result. When human mastery is promised, preserve the learner's own required operation and use the applicable HCD practice, feedback and assessment.
For an AI system, distinguish changes to instructions, retrieval, external state, tools, an evaluation criterion and model parameters, and obtain the corresponding technical method. E.10.LRN explains why those changes support different learning claims.
Make the corrected contribution obtainable at the relevant occasion, with its needed means and authority.
Select verification that can expose the remaining failure: does the corrected criterion reach the user, is the intended operation performed, or does the proposed cause defeat it? A changed condition is useful when it tests the relation being repaired.
Reuse checks whose grounds apply. Include diagnosis, transition, verification and continuing support in sections 4.2–4.5's comparison.
Restore dependent work and return what happened. Establish the affected set from reliance on the changed basis: instructions derived from it, conclusions using it and results consumed later. Use existing source and use records, the work and relevant participants to establish this reach.
Shared wording can locate candidates; their dependence determines inclusion. A sampled failure can expose a cause while leaving the affected population unknown.
Use B.5.RR to revise the affected reasoning. A conclusion may survive on independent grounds, need a weaker or conditional form, or require withdrawal or reconstruction.
Carry each changed result into its receiving use. Correcting a review criterion can require revising both the verdict and the published instruction altered to satisfy it.
Restore every result in the established affected set and identify any unresolved boundary. A newly discovered dependence extends that set; several successful repairs do not settle its unexamined remainder.
Keep source correction, restored outputs, changed ordinary performance and worthwhile effect distinguishable. An actual use establishes the supported performance under its stated conditions.
A claim that the intervention reduced recurrence or improved the encompassing work needs evidence about that consequence, including relevant changes in demand, other causes and added burden. Obtain that evidence when the intended decision or claim requires it.
If use contradicts the proposed explanation, return to the affected relation and preserve repairs and findings whose grounds hold. A sufficient receiving result can close the bounded repair while a further effectiveness question remains open.
Choose a worthwhile and attainable contribution
Compare materially different remedies for the same receiving need. Include continuing current practice where admissible; removing or reducing work that the receiving result does not require; a repair or better use of existing means; developing a missing explanation or Method; changing the performance/provision arrangement; and deferral where timing changes the comparison. Include only real alternatives.
Estimate the whole burden of reaching and sustaining the useful result: discovery, source recovery, construction, qualification, introduction, learning, coordination, publication where needed, recurring support and later refresh. The relevant unit is a usable change with its coupled dependencies. Two lines can require a costly release; a large coherent explanation can remove repeated reconstruction. Text size settles neither value nor the release boundary.
Compare expected and adverse consequences at the strength the evidence permits. Where ranges reverse the choice, identify the uncertainty and ask which affordable observation could discriminate it. Where the same choice survives the plausible range, avoid a measurement exercise that cannot change it. For noncommensurable gains, state the trade-off and the responsible decision instead of hiding it in an arbitrary score.
Select a contribution whose obtaining conditions can be supplied now. State what it should make possible, who will perform the needed work and use its result, what must remain protected, and what observation would change or stop the undertaking. Use the existing programme or decision account when continuity requires one. A priority label without an attainable next action is not an implemented selection.
Follow the change into use and renew the search
Develop, introduce and qualify the chosen contribution through the Methods governing its subject. Keep three conclusions separate: the recipient can recognize the applicable Method; the recipient or provider can perform it under the intended conditions; and the receiving work obtains a worthwhile effect. A readable publication or completed training can support the first two without establishing the third.
Observe the effect at the promised scope, including new burdens and displaced problems. Reuse existing service or work observations where adequate. Do not require another study merely to repeat a known result, and do not declare success from a proxy that improved while the receiving result deteriorated.
Return a missing operation, explanation, provision or execution mechanism to its supported owner. Keep a successful local remedy local when transfer is unsupported. When a reusable contribution is warranted, carry its reasons, limits and consequential variations into the maintained repertoire. E.4.CM supplies the framework-publication form when a public pattern is the selected receiving product.
Reconsider the next improvement from the changed position. A constraint may have moved; a recurring payment may have disappeared; a previously unavailable agent may change the attainable division of work; or a new practitioner may reveal an explanation gap. C.40.NP develops the next worthwhile problem from an achieved result. An unresolved opportunity can remain in the existing programme with a meaningful reconsideration condition. Adequate ordinary practice can continue without another improvement task.
Fit the recurrence to change in the work, meaningful accumulated loss, an existing planning occasion or a qualified signal. “Continuous” means retaining the capacity to notice and act over time. It does not require a permanent survey, constant monitoring of every method, or a fixed quota of changes.
Archetypal Grounding
Choose between a payment saving and more accepted output
In this constructed case, an owner uses the service's ordinary planning occasion to examine regeneration invoices, unserved requests and reconciliation effort. The service prepares thirty accepted packages each month. Review capacity limits output, while preparation has spare capacity.
Following a package shows that an avoidable regeneration step costs 15 monetary units. The provider already offers a qualified way to reuse valid material at a charge of 5 per package. Introducing that way would preserve the accepted packages, review load, quality and timing. The available qualification supports those premises for this package class.
Over twelve months, the differing use charges are 30 × 12 × (15 − 5) = 3,600. External implementation costs 1,200 and additional support costs 600 over that same horizon. No other payment or receipt differs from continuing the incumbent arrangement. The first-year net cash difference is therefore 1,800 in favor of the change. Throughput is unchanged. The supported remedy is to introduce an available way, not to invent a missing Method.
The change also needs twenty hours from an employed engineer for configuration and qualification. Those hours are available without displacing committed delivery; salary remains unchanged. Only one improvement fits that engineering allocation now.
A second qualified option introduces review aids and enables one additional accepted package per month. Supported demand exists, and each additional package contributes 250 after all its additional operating payments. Obtaining and supporting the aids costs 1,400 over the year; the same twenty engineering hours are required, with no other differing payment. Its first-year net cash gain is 12 × 250 − 1,400 = 1,600.
Both changes meet existing service commitments and participant-protection conditions. Under the stated comparison, the owner selects preparation: it gains 1,800 against continuing current practice, or 200 against choosing the review aids. The greater output of the second option does not make it financially preferable by itself.
The review-aid option remains available for a changed demand, contribution or implementation cost. If the options have other material consequences, those also enter the choice rather than being silently priced at zero.
A third possibility automates reconciliation that already meets its quality and timing requirements. It needs the same twenty engineering hours and would release forty salaried hours a year, without a payment change or a presently warranted further use. The people concerned identify no material reduction in harmful or onerous work. The owner retains this possibility for a change in its receiving use rather than counting forty hours as a cash saving.
The owner commissions the provider's reuse operation and allocates the engineer's twenty hours. Existing invoices and acceptance records can show whether the charge changed while quality, timing and review load were preserved; a failed acceptance condition returns the change for repair or withdrawal. The stipulated implementation and support costs already include this introduction and qualification.
At the end of the case's first year of ordinary use, those records show 360 accepted packages at the new charge, the stated change payments and preserved acceptance conditions. The owner retains the change and closes this undertaking.
Now change one premise. The twenty hours would instead displace a committed contribution that brings 4,000 in additional net cash over the same horizon, with its own avoidable costs already deducted. The salary is unchanged.
Relative to retaining that contribution, the preparation proposal changes net cash by 1,800 − 4,000 = −2,200.
Compare another qualified performer, a later implementation window or deferral. Do not count the foregone contribution twice or turn this changed scheduling result into a claim that nonconstraint savings are always useless.
Consider instead a different starting situation: the incumbent regeneration service already charges zero. Continuing the incumbent arrangement and introducing the preparation change both keep that price. If the only operating gain is freed salaried preparation time and the salary stays unchanged, that gain supplies no cash saving by itself.
Implementation or support payments caused by the change must be deducted; if they remain 1,200 and 600, the first-year cash difference is −1,800 before any displaced contribution.
A worthwhile next use of the time or reduction in human burden may justify the change on its own grounds.
Return to the successful first arrangement. A later request for thirty-four accepted packages per month makes review capacity the next consequential question. The retained aids would raise capacity only to thirty-one. The owner compares additional qualified review provision and a more capable review method before accepting the larger commitment, while ordinary production at thirty continues.
The earlier saving supplies resources and a changed starting position; it does not establish which new remedy is adequate. This is a new useful comparison, not a scheduled repeat of the completed improvement survey.
A costly goal cycle exposes a methodology question
A framework maintainer receives repeated reports that goal preparation and review occupy substantial time. In a hypothetical organization, a scenario calculation places those activities between 3,847 and 7,793 human hours a year. The range comes from chosen participation, frequency and duration assumptions; it is neither a measured organizational total nor proof that the work is waste.
The first proposal is to shorten review meetings. Following a representative goal into work instead reveals disputed outcome indicators, unprovided shared contributions, ambiguous aspirational versus committed goals, and revisions made after feedback. Shorter meetings could remove necessary decisions while leaving their causes intact.
A search finds general goal-setting and monitoring Methods. The maintainer can use their adequate operations, but this does not settle the requested professional account. The source literature may explain how to negotiate cross-team commitments, distinguish learning from delivery goals, change an inadequate measure, introduce the practice gradually and separate development conversations from remuneration decisions.
Recover those contributions and their conditions from the assigned sources, then compare the full receiving work with the current framework. Do not make the four initially visible difficulties define all useful content in the sources.
The possible results differ. An unavailable joining operation calls for construction. A privately understood but unwritten way calls for a developed explanation. An adequate public method with no allocated participant time calls for provision. A mandated but unnecessary weekly update calls for examining the authority and rationale of that demand. A new framework grouping can improve independent use and maintenance even if many constituent Methods remain unchanged.
A proposed 10% reduction of the scenario's hours means 385–779 gross hours before transition and continuing support. It is a sensitivity calculation, not an expected effect.
If a fully supported change requires 200 hours, its hour-only break-even reduction is about 2.6–5.2% under the corresponding endpoints. Quality, displaced work, availability and other consequences decide whether to proceed. One poorly formulated goal can also misdirect much more work than the goal-setting hours.
The figures establish the scale at which an improvement could repay its cost. To choose the improvement, establish which difficulty actually interferes with the practitioners' work and how the proposed change would reach it. If the needed operation exists and people can use its account but have no time to perform it, the next decision concerns provision. If the operation is missing, construct it. The savings estimate can inform either decision.
Less agent activity, better development work
An AI-assisted author repeatedly starts new source searches after it already has a supported answer to a bounded question. The search and subsequent reconciliation consume time and context, delaying an assigned explanation. A proposed remedy shortens the instruction to “do less research.”
The decision trace instead shows that each returned source was treated as opening an unbounded obligation to inspect every related topic. The needed change preserves the original source scope while linking each optional additional inquiry to an attainable result that could change the receiving conclusion. A binding source-recovery assignment remains binding. An unsupported field-wide conclusion needs its missing basis.
The next ordinary task is the relevant observation. It should complete its assigned substantive scope, use adequate existing evidence and stop optional inquiry whose possible answers do not change the decision. Merely obtaining the phrase “I applied DUA” proves none of those actions. A shorter turn that omits required content is not an improvement.
This is a proposed mechanism and observation design. Actual traces and receiving results are needed before claiming reduced recurrence or a measured saving. If the failure instead arises because the agent cannot recover its instruction after a context change, retention and re-entry are different necessary remedies.
Choose the repair from what kept producing the failure
In this constructed case, a software provider maintains five customer guides. Its writing method requires a caution when a credible mistaken use would change the customer's action or result. Two guides describe imports that overwrite stored settings; customers need that condition before importing. Three guides concern navigation and display layout, and the available use evidence supplies no comparable caution for those instructions.
The review template nevertheless requires a caution in every guide. Revision records show that authors added generic cautions to the three navigation guides to obtain approval. Review accepted them by checking presence, without establishing the cautions' contribution. An editor later removed those cautions, improved the examples and explained the writing rule at a team meeting. The next review again rejects the three guides for lacking cautions. The provider needs usable customer instructions and a warranted release judgement.
The maintainer recovers the rejected passages, authors' recorded revisions, review decisions and the template used. The reviewer identifies the template item that governed the rejection. The earlier editor's changes and meeting account show what the correction supplied. The template remained unchanged at the required decision point.
These observations connect production of the defect, its acceptance by review and the ineffective repair: improving the examples left the criterion that had induced and approved the extra text in place.
The maintainer changes the criterion at its maintained source: inspect whether a credible mistaken use changes the reader's action or result, and require the corresponding caution when it does. The release form displaying that item is regenerated and compared with the source.
Another meeting would add effort while preserving the known contradiction; deleting every caution would remove the two independently warranted overwrite instructions. The supported change repairs the criterion and its working form while retaining the useful subject judgements.
Release history establishes that these five guides are the complete set produced under that template edition. The maintainer follows each review decision into the resulting customer text. The three unsupported rejection reasons are withdrawn; their guides keep the useful examples and omit the generic cautions. The two overwrite cautions and their supported acceptance decisions remain.
The corrected five guides reach customer distribution. An additional guide used another criterion and remains outside this repair on that checked basis.
At the next release in the constructed case, the reviewer applies the corrected criterion to a navigation instruction and to an import that overwrites settings. The first needs no caution; the second retains its overwrite instruction. This establishes a corrected source, restored affected guides and use of the new criterion in that release.
Whether later releases require less rework remains a further question for ordinary release observations.
Change the source-use relation. Suppose the maintained template already agrees with the rule, but a pinned link opens an obsolete copy. The same visible demand for generic cautions calls for repairing that access route and following the uses of the obsolete copy. Rewriting the correct source criterion would leave the failed relation unchanged. The affected set now follows use of that copy, so the earlier five-guide boundary must be reconsidered.
Change the performer's preparation. A new editor reaches the current criterion, but the editor's comparison of two instructions treats every imaginable misunderstanding as credible misuse. A qualified helper uses that comparison to locate the missing distinction. If learning is the promised result, the HCD route supplies practice, feedback and assessment of the editor's own operation. If the immediate undertaking only needs a corrected guide, expert provision may suffice. The receiving promise selects between those contributions.
For an AI-assisted review, inspect the instructions and evaluation material supplied. A corrected context can support an adequate review under that configuration. Performance after that context is replaced requires its own basis; the observed review establishes no parameter update.
Across these changes, the reusable move is to return from the visible failure to the relation sustaining it, choose the repair that changes that relation, and restore the work that relied on it.
Bias-Annotation
Measured time favors visible repeated work and can undervalue prevention, informal coordination and opportunities never attempted. Complaint counts reflect reporting conditions as well as difficulty. Automation proposals can omit verification, recovery and supervision; familiar human allocations can also preserve unnecessary duplication after agent capability changes. Compare arrangements, including their supported capabilities and limits.
The examples simplify demand, cost and effects to expose consequential distinctions. They establish no universal savings rate, organizational timetable or model ranking. Work involving human health, rights or other protected conditions retains its relevant professional and authority basis.
Conformance Checklist
- The selected work and receiving result are recognizable, with a meaningful population and horizon.
- Observed burdens, estimates, source claims and hypotheses remain distinguishable.
- The comparison follows local change into encompassing work, including constraint, payment, capacity, human and future-use consequences that matter.
- Overlapping effort and already counted foregone contributions are not added twice.
- The remedy distinguishes a missing Method or connection, explanation, access, provision and execution as the evidence warrants.
- When a failure survives correction, the repair addresses the supported conditions of production, checking and the earlier intervention, reaches the working forms, and restores the established dependent set while retaining independent grounds.
- A source-based adequacy conclusion covers the assigned source and practitioner scope; successful examples do not redefine it.
- Selection includes the full change and continuing burden, feasible alternatives, an attainable next action and a reconsideration condition.
- Recognition, execution and useful effect receive their own supported conclusions.
- The search returns from changed conditions without manufacturing another mandatory intervention.
Common Anti-Patterns and How to Avoid Them
Consequences
Method improvement becomes answerable to work that matters and to the cost of obtaining the improvement. Adequate Methods and useful source contributions survive, while missing explanations and provision are less likely to be mistaken for each other. A saving outside the output constraint can be selected on its merits.
The work needs professional judgement and enough evidence to discriminate its serious alternatives. Discovery and measurement consume resources. A useful result can be a justified stop, a better provision arrangement or a retained possibility rather than a new Method or publication.
Rationale
C.39.MF selects the subject of Method development once a relevant difficulty is visible. C.40.NP continues from an achieved result. The recurring contribution here connects discovery across actual or credibly anticipated work, whole-burden comparison, diagnosis of the needed change, implementation and return. That connection lets a practitioner decide what deserves methodological attention before selecting a familiar repair.
The same construction can support an enterprise practice, a domain repertoire or a transdisciplinary framework. Their objects, professional Methods and maintenance duties differ; framework level is not a rank of the work's importance. Developing a framework is worthwhile only when its contribution reaches the intended users under obtainable conditions.
SoTA-Echoing
Compare the developed toil method at the receiving use
The SRE Workbook's Measuring Toil, Toil Management Strategies and Conclusion provide a serious existing route. The comparison concerns their explained operations; it makes no independent claim about the case studies' effects.
For a service-toil problem adequately answered by that route, using WI again would add unnecessary reconstruction.
A framework maintainer choosing among a goal-setting practice, a rare engineering decision and recurrent service work needs the wider comparison: what useful change is missing, which work it should change, and whether a new Method, a developed explanation or different provision supplies it. This is the receiving extension, not a claim that toil discovery or project comparison was absent from professional practice.
The trade-off is that WI cannot prescribe one measurement regime or service rule across those uses. It supplies the common choice and return, while professional suppliers retain their calculations and conditions.
Reopen this boundary when an obtainable treatment supplies the same wider entry, choices and return at lower total burden, or when an actual use exposes a missing selection operation. Do not retain a second comparison merely because it belongs to a framework.
Reopen a relied-on empirical or capability premise when the proposed performer, tools, demand or receiving work changes. A recent source date alone does not establish superiority. The chosen actual-use comparison supplies the evidence for the local effect.
Connect causal repair with restoration of dependent work
C.28.CM supplies causal alternatives and discriminating consequences; B.5.RR revises reasoning whose grounds changed; A.3.1.MI:4.5 separates delivery, mechanism and context when selecting an adaptation. Use their sufficient results. Section 4.4.1 connects those results for a failure that survives correction: explain production, checking and the earlier repair, change the operative condition and restore the results dependent on the rejected basis.
The RCA² guidance, Appendix 6, p.39, requires a causal explanation beyond naming an error or procedural violation. The NHS Safety Action Development Guide, pp.13–19, develops action selection and testing, separates introduction, actual use and effects, and includes unintended consequences and observation cost.
These are established professional contributions. Their clinical investigation and reporting arrangements remain specific to that practice.
This branch connects a supported cause repair with revision and delivery of results that relied on the defective basis. A sufficient specialist repair can finish. The wider connection is useful when changing a source leaves its working forms or earlier products unchanged. Its general explanation and constructed case support examination of that choice; effects in a practice require that practice's evidence.
Relations
C.39 supplies development of a missing way; C.39.MF settles the Method or non-Method focus; C.39.RP preserves a relevant repertoire; C.39.MC returns changed facts to the affected development question. C.40.NP continues from a qualified result to the next worthwhile problem.
C.11.DUA governs the burden and useful contribution of additional inquiry. C.11.MV and the appropriate professional comparison govern Method worth. B.1.5.EW keeps encompassing Work conditions, and C.32.MWA.PP compares performance, provision and development arrangements. OPS.9, OPS.14 and OPS.16 supply operating constraint, financial and operating-change results where those questions apply.
C.28.CM constructs and challenges a needed causal distinction; B.5.RR revises dependent reasoning while preserving independent grounds. A.3.1.MI introduces the selected change and examines an adaptation. A.15.8.MS, A.15.10 and A.15.11 supply method-use support, recovered continuation and recognition when their respective conditions select that repair.
E.4.CM gives a reusable contribution its developed framework explanation. A.3.1 and A.3.2 retain the distinction between a Method, its candidate account and its description. This pattern's proposed remedies, plans and worked examples establish no performed Work, acquired capability or authority by themselves.
C.39.WI:End
Organize a Repertoire for Method Development
Type: Method pattern Status: Stable
Working result: one inspectable Method repertoire for a named use that keeps identified Methods, candidate Method accounts, MethodDescriptions, source contributions, family or local-grouping claims, variants, supported lineage relations, evidence limits, source editions, and missing positions distinct.
Use This When
Use this pattern when a one-Method, plurality, or Method-relation focus needs alternatives, variants, evidence, and provenance that another practitioner can inspect and reuse. Typical uses include comparing several Methods, forming a project proposal, locating a variant, or checking whether a familiar source contribution supports the current question.
Begin with the named comparison or proposal use, the focus result or equivalent subject statement, and the practical result the project needs. Search the practice that owns that result; add primary Method-construction and reuse sources when their contribution can change the repertoire. Keep editions and status visible while assembling a repertoire sufficient for the use. Reuse an adequate repertoire whose grounds apply.
The first useful result is a set of status-preserving entries with claim-bearing source material, supported relations, evidence limits, currentness, and missing positions. A gap is part of the repertoire when it changes comparison or proposal work.
Do not treat a bibliography, repository, package outline, observed Work, or list of popular names as a Method repertoire by itself. Co-listing supplies neither Method identity nor an established family, and documented lineage supplies neither superiority, causation, composition, nor present applicability.
Working Distinctions
Problem Frame
A project rarely receives a clean inventory of Methods. It receives old procedure editions, package sections, conference descriptions, local checklists, tool configurations, retrospective accounts, and claims that one practice “came from” another. The most familiar names can hide the best-supported contribution, while a well-stocked repository can hide a decisive gap.
Method Engineering sources encourage reuse from method bases and situational construction. Engineering-design sources add intended use, representation, procedure, tool, and ecosystem questions. Those source lines improve search and recovery, but their local elements are not automatically FPF kinds. The repertoire must therefore preserve what each source contributes while keeping Method, account, description, Work, family, grouping, and lineage claims separate.
Problem
Three shortcuts make a repertoire unusable. First, every package item is called a Method. Second, co-listing or a common ancestor is called an established Method family. Third, the newest source edition replaces older evidence even when the current use relies on an earlier claim or the changed edition has not been checked.
The resulting list appears complete but cannot answer which alternatives are identified Methods, which are only accounts, why a relation is believed, what changed between editions, or which missing position blocks the next decision.
Forces
Solution
A repertoire here serves a named Method-development use. G.5 supplies a Method-family registry or selector-facing result when that question is current; G.11 governs a refresh that its entry makes necessary. The construction below recovers the source contributions, alternatives and practical conditions that those results do not supply merely by listing Methods.
Build the repertoire around one receiving use. Recover subjects and source contributions, qualify every relation by its evidence, and record unsupported positions as gaps instead of filling them with guesses.
Recognition is cheap: a source or repository entry is worth inspecting when it may supply a relevant alternative, contribution, or missing relation. Assurance begins only when a decision relies on the row; then its subject status, source edition, evidence, relation basis, currentness, and limit must support that use.
Pattern-Use Unfolding
-
Bind the repertoire use. Name the focus, receiving result, decision, situation family, and what the next comparison or proposal must inspect. State what would make an entry relevant.
-
Search the contributing source lines. Search the problem-owning practice for Methods, accounts, variants, evidence, and current operating constraints. When constructing, adapting or reusing ways of working is unresolved, include primary Method Engineering sources for the relevant construction and source-organization contribution. A sufficient domain repertoire needs no additional literature pass solely because its subject is a Method.
Record source editions or dates and the claim-bearing passages or artifacts used.
-
Recover subjects before grouping them. Cite the existing
A.3.1identity result for every identified Method. Keep provisional material as a candidate Method account. Attach a MethodDescription only to its already identified Method. Keep tools, roles, capabilities, support arrangements, evidence, and Work occurrences under their kinds. -
Extract contributions with limits. For every useful source, state the contribution used, the subject it concerns, its evidence or status, its scope, and any material limit on its use. A package position or source vocabulary does not determine the FPF kind.
-
Record relations independently. Separate established family membership, project-local grouping, variant claims, input/result relations, and supported lineage relations. For lineage, name both subjects, the relation claimed, and the source that supports it. Leave same-Method identity, causal effect, superiority, and composition open unless their governors settle them.
-
Pin editions and use-qualified currentness. Distinguish publication or artifact edition from the claim relied on for the current use. Keep its relevant conditions and evidence, permission or qualification windows; identify the changed premise that would require reconsideration. A new edition or review date can invite attention without defeating that claim.
Use a broader
G.2SoTA pack only when the use needs refreshable multi-source synthesis rather than a small question-relative source set. -
Expose missing positions. List absent alternatives, unknown editions, unsupported ancestry, missing evidence, unresolved family status, and contributions whose source meaning cannot yet be recovered. Say which next use each gap blocks or merely weakens.
-
Return the bounded repertoire. Make the subjects, source contributions, relations, applicable conditions and any gaps available for the named use. Reuse an adequate existing account; publish a separate record when later receivers need one. For a choice among the named entries, their relevant differences and status may suffice. For a complete repertoire, recover its membership rule and establish that every eligible way is included at the granularity the recipient ordered. For an exclusion, establish that the item falls outside the membership rule. Grounds for preferring one eligible way to another support a recommendation and leave both in a complete list. A record of the ways found can serve a bounded review while completeness remains open.
Record the Result
The following positions can organize a record when one is needed. Include only relations and gaps relevant to the use; the table does not require an unsupported relation or missing position to exist.
What Changes in Practice
Teams stop searching a repository as though every file were a reusable Method. They can compare identified Methods alongside promising accounts without promoting either, trace a proposal to contributions without asserting composition, and stop with a useful gap instead of inventing a complete lineage.
Archetypal Grounding — EC-417 Release-Evidence Repertoire
The EC-417 project needs a repertoire for a Method-relation comparison around one safety-relevant controller release. The relevance criterion is narrow: an entry must contribute to verification, integration, supplier evidence, release authority, or evidence reconciliation for the named receiving result. The following identifiers and editions are local scenario records; they are not population evidence.
For the D-21 through D0 use window, the project relies on the contributions and conditions in the table.
If a new export or edition changes only their layout while the relied-on claims, configuration, access and use remain the same, retain the repertoire without a new search, trial or renewal record.
If FI-4.8 instead loses coverage for a changed controller or toolchain, or a permission or qualification window ends, reopen or suspend that affected use even before the planned review date.
Keeping the old edition identified does not authorize continued reliance on defeated premises. Reconsider only the affected source-use row, preserve the earlier claim for its supported conditions, and keep the needed limitation with the result for later receivers.
The existing project-local locator LG-EC417-ReleaseMethods contains only the four identified Methods for the bounded release comparison. Its criterion is contribution of a verification, integration, supplier-approval, or authorization result to EC-417.
The three candidate accounts remain adjacent repertoire entries with their statuses; a wider source-inspection list does not add them to the Method grouping. Family status remains unresolved. No family membership, fifth Method, or composite Method follows.
Description-side material stays separate. Existing A.3.2 results identify HV-6, FI-4.8, H17-3, and RC-17.3 as MethodDescriptions of the four already identified Methods.
The stage table, bundle records, AI prompt, and evidence grades remain other claim-bearing epistemes or artifacts. They are not additional Methods.
The two reconciliation records and the AI review record remain candidate accounts, so no A.3.2 MethodDescription is asserted for them.
Proposed-whole account C-EC-Release-v2 is recorded outside the repertoire membership table as an architecture subject. It cites contributions from the four Methods and three accounts, but that source relation establishes neither a whole Method nor methodPartOf.
The repertoire stops with four missing positions: no established Method family covers the local grouping; no source establishes a Method-lineage relation among the four identified Methods; the same-Method or distinct-Method relation between the two reconciliation accounts is unresolved; and the AI account lacks transfer and effectiveness evidence.
Those gaps do not prevent A.3.1.QM from qualifying individual subjects or C.32.MWA.PP from comparing named structures. The repertoire alone does not support family, ancestry, superiority or composite claims; a claim needed by the receiving use requires its basis.
Bias-Annotation
Conformance Checklist
- The repertoire has one named comparison or proposal use and a relevance criterion.
- Every Method cites an existing
A.3.1identity result; every provisional subject remains a candidate account. - MethodDescriptions attach only to already identified Methods and remain distinct from those Methods.
- Every source contribution names a source or artifact edition, subject, contribution used, evidence/status, scope, and limit.
- Established family relations, project-local groupings, variant claims, and lineage relations are recorded separately.
- Every lineage claim names both subjects, the relation, and the supporting source without implying causation, superiority, or composition.
- Publication/artifact edition and current project reliance are distinguishable. A new date alone neither defeats nor renews support; changed evidence, applicability and time boundaries govern the receiving use.
- Missing alternatives, provenance, relations, editions, evidence, or source meanings remain visible.
- The result supports the assigned use of the repertoire. A claim to list every eligible way has grounds for coverage.
Common Anti-Patterns and How to Avoid Them
Consequences
The project gains reusable alternatives with recoverable provenance, bounded evidence, and visible gaps. Later qualification and architecture comparison can inspect the same subjects without repeating search or inheriting false family and composition claims.
Recovering subjects and provenance takes work. Some lineages remain partial, and a new source whose changed meaning or applicability is unresolved may be unusable for the affected claim.
Rationale
A repertoire becomes reusable when its users can distinguish statuses and trace provenance. Source lineage can explain where an edition, account, or variant came from, but it cannot answer whether a Method exists, whether two Methods form a family or whole, whether one caused an outcome, or whether the latest variant is preferable. Those questions have different subjects and governors.
The repertoire distinguishes Methods, performed Work, descriptions, variants, capability and tools, including cases where several representations concern the same Method. These distinctions keep a source document's organization from determining the kind of thing each passage describes. Method Engineering literature supplies search and reuse practices; FPF supplies the status and relation boundaries needed to use their contributions across projects.
SoTA-Echoing
Reopen when a relied-on source contribution, its support or its applicability changes, a missing position becomes available, an A.3.1 result changes a subject's status, a family or lineage relation gains or loses support, or the named comparison requires an alternative outside the current relevance boundary. Age or edition succession alone establishes none of these changes.
Relations
- C.39.MF or an equivalent result supplies a one-Method, plurality, or Method-relation focus and its statuses. C.39.RP may also start from another receiving use that meets the same conditions.
A.3.1governs Method identity;A.3.2governs MethodDescriptions. Repertoire inclusion changes neither.G.2supplies a broader refreshable SoTA pack only when that heavier source result is needed;G.11governs an applicable currentness question without requiring refresh Work or a waiver for continued applicability;G.5supplies governed family or selected-set results when its entry conditions hold.- A.3.1.DR may recover decision-relevant entries from a heavyweight package. An applicable research Method may supply scoped candidate accounts. Use either only when its result is needed.
- C.39.RQ may use the repertoire to expose situational criteria; A.3.1.QM receives identified Methods and candidate accounts with their source limits; C.32.MWA.PP receives only named structures and relation questions.
- B.1.5.CM may cite repertoire contributions for a proposed-whole account, but citations create neither a whole Method nor obtaining composition.
C.39.RP:End
Construct Situated Method Requirements
Type: Method pattern Status: Stable
Working result: one situational Method-criteria result that states the receiving Work and result, required Method contributions, performer capabilities, technical and organizational conditions, allowable variation, non-negotiable conditions, burden limits, evidence needs, and acceptance or stop observations without selecting or admitting a Method. A question about one disputed criterion returns a supported recommendation to retain or change it, together with the criterion's current force and available amendment route.
Use This When
Use this pattern when a Method, an established-family or project-local grouping choice, or a proposed Method/Work/support structure may fail in the project's situation. Enter when the practical question concerns the receiving result, performer capabilities, technical or organizational conditions, variability, evidence, authority, or acceptable burden and those conditions are not yet explicit enough for individual qualification or architecture comparison.
Begin with the recurring situation, the intended or actual Work and the result its recipient needs. Identify the decision that the criteria will serve. Selecting one Method and arranging several Methods to run together can impose different conditions on the same work. State what must be achieved before selecting a candidate. Attach each criterion to what it constrains, such as the Method, a performer's capability, an assignment, permission to act, the authority to decide, the support available or the result delivered.
The first useful result is a bounded set of criteria with subjects, allowed variation, evidence needs, satisfaction observations, and stops. A criterion can be ready for later use even when no candidate currently meets it. Enter also when the present question concerns the justification of a criterion. That use can finish with a supported requirement recommendation and the criterion's current force or amendment limit, without selecting a Method.
Do not use this pattern as a generic product-requirements template. Criteria do not identify, admit, select, qualify, or compose a Method or establish present or future fit. Product acceptance, Method identity, individual qualification, and architecture choice remain separate results.
Working Distinctions
Problem Frame
Situational Method Engineering asks practitioners to fit a way of working to its context. Engineering-design research adds useful questions about a Method's goal, procedure, rationale, framing, mindset, intended use, representation, tool, and ecosystem. A project also faces domain constraints that do not belong inside the Method: protected information, performer capability, authority, shared capacity, evidence timing, reversibility, and the acceptance conditions of the receiving result.
When these unlike claims are placed in one “method requirements” list, description content becomes a Method part, culture becomes a performer property, authority becomes skill, and product requirements become proof of Method fit. A useful criteria result keeps the connected situation visible while returning each claim to its subject.
Problem
A universal checklist invites premature scoring. It may reward a well-written MethodDescription even when performers lack capability, declare a Method unsuitable because a provider arrangement fails, or hide an architecture-level peak-load conflict inside an individual candidate score.
The project then appears to have objective fit criteria, but nobody can say what must change: the Method account, the description, the support System, the covering Work assignment, a permission or decision-authority relation, the Work arrangement, or the receiving result's acceptance rule.
Forces
Solution
C.11 governs a choice among available options, and C.11.DUA appraises a disputed demand and worthwhile inquiry. The construction here precedes that choice when the required Method contribution is unclear: connect the desired result and situation to criteria on the Method, its performers and its provision without treating them as one subject.
Build criteria from the receiving situation outward. State required contributions and conditions, assign each criterion to its subject, preserve allowed variation, and defer every fit or selection verdict to the pattern that evaluates the relevant subject or structure. For a disputed criterion, start with its existing context and the appraisal in §4.4; build or update other criteria only where that question needs them.
Recognition is cheap: recurring Work, a changed constraint, or one plausible capability, access, authority, evidence, or burden failure is enough to expose a criterion question. Assurance is later and row-specific: the named observation, edition, and window must support satisfaction before qualification or architecture work may rely on that row.
Pattern-Use Unfolding
-
Bound the situation family. Name representative Work, receiving result, affected subjects, decision window, recurrence expected, important variants, and ordinary exclusions. Use
A.15.6when a project, process or case claim leaves its subject unclear; attach a criterion to the plan, Method, Work, result, or other subject it constrains. Several accounts concern one Work only when their claims support that reading. -
Recover required contributions before assigning candidates. Start from the professional result, available plans or instructions, and a relevant performed or anticipated case. Follow the receiving operation backward: what input, intermediate result or preserved condition must be available for it to succeed?
Then follow an available way forward to find where it supplies that contribution, needs support, or fails. C.39:4.2–4.3 develops this source recovery and backward/forward construction when the connection is missing. State the resulting requirement without prescribing a familiar candidate's procedure unless that feature is independently non-negotiable.
Recover a binding limit from the receiving commitment, applicable rule or authorized allocation that sets it; use available provision to determine what can meet it. Observed duration or resource demand can expose a mismatch, but does not set an acceptable limit by itself.
For example, a signed-approval requirement before closure makes obtaining that approval or returning a missing-approval stop a required contribution; it need not prescribe a particular review tool.
If the allowed deadline, protection or allocation is unknown, retain that criterion question and identify the agreement, authority or professional basis needed to settle it. Reuse an adequate existing criterion; seek additional information only when its attainable contribution warrants the burden.
-
Use source prompts without importing kinds. Ask what goal, procedure, rationale, framing, mindset, intended use, representation, tool, scope, and adaptation conditions a source makes visible. Place the answer with its subject. A stated mindset may be description content; an obtaining capability or cultural relation needs separate evidence.
-
Recover conditions around enactment. State performer capabilities, technical and organizational conditions, inputs, support/provider access, responsibility, covering Work assignments, permission and decision-authority relations, evidence timing, reversibility, and other Work or result constraints. Keep these conditions distinct from one another and from the performed Work and its decision result.
-
Separate variation from invariants. Name lawful branches, substitutions, timing ranges, and local adaptations. Mark non-negotiable result, safety, confidentiality, authority, or evidence conditions.
If a criterion's justification is the question, use §4.4 to appraise it while retaining its current force. An already justified criterion needs no new appraisal.
-
Identify who bears each burden and when. Attach time, attention, capacity, delay, meeting, or cost limits to a participant, Work scope, and window. Keep individual burden distinguishable from combined peak demand and burden transferred to another participant or period.
-
Name evidence and observations. For every decision-changing criterion, state what record or observation a later evaluation needs, its edition or window when material, and what would count as satisfied, failed, or unknown. Do not turn the requested evidence into a fit verdict.
-
State acceptance, stop, and reopen rules. Say when the criteria set is adequate for its next use, which missing fact stops that route, and which situation or source change requires rebuilding it. A stop may route to capability, support, authority, product, or Work redesign rather than Method selection.
-
Return without selecting. Could an arrangement satisfy every listed criterion and still fail the receiving use? Could the criteria exclude an arrangement that meets the terms? Add the missing relation or relax the unwarranted restriction. An inquiry about one disputed criterion can finish with its effect on the decision. An adequate existing account can carry the criteria. Use A.3.1.QM to qualify one identified Method, retain or exclude a candidate account, or examine a connection that can be assessed on its own. Use C.32.MWA.PP when combined work, provision or shared constraints can change the decision.
Record the Result
Use the rows that carry the present result. For a disputed criterion, retain the relevant existing context and return the recommendation described in §4.4.
What Changes in Practice
Teams stop asking whether a methodology “fits the context” as one opaque question. They can see whether the issue is a Method contribution, a description gap, missing capability, covering assignment, permission, decision authority, provider access, Work timing, receiving-result condition, or combined burden. Later qualification and architecture comparison receive explicit conditions instead of a precomputed winner.
Appraise a Disputed Criterion
Use C.11.DUA §4.3 when the practitioner needs to decide whether a Method criterion or protective requirement is justified. Recover the person, system or receiving result it protects, the failure or exposure, the governed activity or quantity, the evidence and threshold basis, and the protection the requirement adds under these conditions. Compare that contribution with the burden, delay and displaced protective work. Use the relevant domain judgement and measurement or causal analysis where the question needs them; a label such as safety does not supply that basis.
Return supported retention, tightening, revision, replacement or removal only among meaningful alternatives. Keep the requirement's current force and the authority and time needed to change it explicit where they determine the feasible continuation.
A recommendation can be complete while amendment remains unavailable.
Continue under the current allowed options until the governing authority or agreement changes the requirement; if none meets the receiving need within the window, return that limitation. A later amendment does not restore an opportunity already lost.
Carry the useful reason and limit in the same criteria or recommendation result. Finish when it answers the present question at a supported strength. Choose a further study only when its obtainable result could change that answer enough to warrant the full burden.
Archetypal Grounding — EC-417 Situational Criteria
The situation family is a safety-relevant controller change combining firmware and supplier-originated harness geometry under a fixed release calendar. The intended Work produces one released change with traceable affected requirements, implementation revisions, verification results, evidence status, and human release authority. Some cases have signed supplier evidence before integration; others have only versioned provisional evidence until later. AI support may suggest trace links but may not receive confidential geometry or decide release.
Three accounts inform the release decision. The project plan gives dates, allocations and decision authority. The process account shows how supplier evidence, integration and verification support a release decision across changes. The case record shows which evidence is available for this release and how it affects the next decision. Use each account to identify criteria for what it describes.
Required Method contributions are stated without selecting an architecture: produce the affected hardware verification result; integrate the implementation against a versioned evidence state; produce signed supplier approval or a missing-approval stop; and return release, withhold, or next-slot authorization under named human authority. Evidence reconciliation and trace review are additional candidate contributions, not pre-admitted Methods.
The alternative labels refer to the timing and capacity comparison in C.32.MWA.PP §5.3. In A, the team waits for signed supplier evidence before integrating the software. In B2, it integrates from a versioned provisional edition and reconciles it with signed evidence before safety closure. Most preparation of the changes between those editions goes to the supplier-configuration role, keeping the safety engineer within the peak limit. A is a choice before integration; recovery R applies when closure fails after B2 integration.
The case admits three human Systems and keeps their decision Work, assignments, permissions, authority, and results explicit:
Assignments, permission relations, and authority relations each need their basis. Each asserted Work occurrence has an A.13 assignment basis and independent A.15.1 admission; its occurrence and decision result need evidence beyond an assignment, permission or authority relation.
F.6 tests precise assignment-bound attribution when that claim is needed; a failed attribution alone leaves independently admitted Work intact. If the same finding invalidates a premise used by A.13 or A.15.1 admission, recheck that premise and its dependent conclusions.
Responsibility, access, capability, assignment, permission, authority, performed Work, and decision result therefore remain separately testable. The AI provider is a separate System and is holder or subject of none of these assignments or relations.
The criteria leave serious alternatives open. A signed-first alternative may wait for supplier evidence; a provisional-first alternative may integrate earlier and reconcile later; preparation may sit with the safety engineer or with the supplier-configuration role. C.39.RQ does not choose among them.
A later individual qualification may apply contribution, capability, access, and evidence criteria to each identified Method or candidate account.
A later architecture comparison must inspect combined peak demand, timing, covering assignments, permission and decision-authority relations, provider access, and burden transfer. Passing one row or staying below 0.40 does not establish fit of the whole proposed structure.
In this case the criteria are sufficient for use when they distinguish the arrangements that meet the receiving terms, including the relations among the required contributions. A test for every listed row can still miss the requirement that the evidence and the integration decision concern the same configuration.
Apply each stop to the route that depends on it: missing signed evidence stops release; confidential geometry exposure stops AI-supported use; assignment, permission, authority, capability or verification gaps stop the contributions that require them. Rollback beyond one hour excludes early integration; excess peak demand or board duration excludes the proposed arrangement.
Preserve independently supported alternatives. Reopen when the change class, supplier information boundary, evidence timing, toolchain, performer assignment, release authority, capacity window, or receiving-result acceptance rule changes.
Retain Approval and Question a Duplicate Trace Record
Consider two disputed requirements in the same release situation. The supplier's signed approval establishes which pinout the supplier endorses; the provisional file alone leaves a material possibility of implementing an unendorsed connection. That protective contribution supports retaining the approval condition for release. A cheaper trace procedure does not supply the missing supplier decision, and the present release authority withholds release when the required approval is absent.
The other requirement is a second manual copy of every trace link. In this constructed case, the versioned authoritative trace already exposes the same links and failed correspondences to the review, and the duplicate copy adds no check or independent information. Producing it consumes time needed to inspect a changed safety requirement.
On those supplied facts, the criteria author can finish with a recommendation to remove the duplicate-copy obligation while preserving the trace and review that detect the defect.
A different case in which the second check detects a consequential omission can support retention.
The recommendation names the governing trace-record rule and the authority needed to amend it. It does not treat the release decider's authority as authority to change that rule.
If timely amendment is unavailable, the rule remains in force and the release arrangement must use a permissible continuation or a later slot. The present recommendation and its limit are complete; a new experiment is selected only if an obtainable result could change the appraisal enough to warrant its full burden.
Bias-Annotation
Conformance Checklist
Apply these questions to the criteria being established or appraised and to the present use of the result. Existing relevant context can supply the answer.
- The criteria name a situation family, receiving Work and result, decision to be made, scope, window, variations, and exclusions.
- Required Method contributions are stated before candidate assignment.
- Every criterion names its subject or relation rather than hiding it under “method fit”.
- Method/account content, MethodDescription, performer capability, responsibility, support/access, covering Work assignment, permission relation, decision-authority relation, performed Work, decision result, culture, and receiving-result claims remain distinct.
- Capability, assignment, permission, and decision authority imply none of one another; an assignment or authority relation does not prove performed Work or its decision result.
- Allowable variation and non-negotiable conditions are explicit.
- When a criterion's justification is disputed, the recommendation compares its protective contribution and burden and distinguishes that judgement from current force and the feasible amendment route.
- Every burden limit names its participant, Work scope, applicable time window, bound and basis, and how satisfaction can be determined. A measured burden retains its measurement method and conditions; a stipulated or derived bound retains its governing or derivation basis.
- Every decision-changing row names evidence and
satisfied | failed | unknownobservations. - The criteria support the intended evaluation, including allowed alternatives and joint conditions. Selecting a Method requires the subsequent judgement.
- Individual and combined structure questions are routed to different later results.
Common Anti-Patterns and How to Avoid Them
Consequences
The project obtains criteria that say what must hold and where, rather than a generic context score. Missing capability can trigger preparation, an access failure can trigger support redesign, a product stop can withhold release, and an architecture burden can remain open without blaming an individual Method.
The cost is more explicit subject placement and fewer easy rankings. Some criteria remain unknown until evidence or a candidate account is available, and a complete criteria set makes no selection.
Rationale
A fit claim is bounded by a use, subject, situation, evidence basis, and time. Criteria are inputs to that later claim, not the claim. Returning every condition to its subject prevents a well-written description from standing in for capability, a provider arrangement from standing in for a Method, or one management view from standing in for the Work.
Method Content Theory and the design-method ecosystem line are complementary because they ask different source-side questions. This pattern connects situational criteria for Methods, Work, descriptions, capability, tools, culture and variants while keeping their subjects distinct. Project, process and case accounts can expose different requirements; their subjects determine where those requirements belong, whether the accounts concern one Work or several related subjects.
SoTA-Echoing
Reopen when a source model changes a decision-bearing prompt, a later evaluation exposes a criterion applied to the wrong subject or used for the wrong decision, a recurring architecture failure can be prevented by one affordable criterion, or the situation family, receiving result, variation, evidence window, performer, support, authority, or burden regime changes.
Relations
- C.39.MF or an equivalent result supplies the Method, family/local-grouping, or relation focus. C.39.RP may supply inspectable alternatives and source limits. When equivalent content already exists, C.39.MF and C.39.RP need not be applied first.
A.3.1governs Method identity and applicability. C.39.RQ criteria identify or admit no Method.A.15.6supplies project-subject distinctions. When building situational criteria, distinguish the Method, Work, description, capability, tool or support arrangement, and viewpoint.- A.3.1.QM qualifies one identified Method or candidate account against applicable rows without turning the criteria into admission or whole fit.
- C.32.MWA.PP compares combined Method, Work, allocation, support, authority, description, subject, capability, or cultural structures when several-structure relations change the decision.
C.11or the applicable domain decision Method may consume later qualification and architecture results. The C.39.RQ criteria set itself selects nothing.
C.39.RQ:End
Continue Method Development from Changed Conditions
Type: Method pattern Status: Stable
Working result: a bounded Method-architecture continuation: the result needed now, the relevant pattern knowledge, the available, blocked, or unknown continuation, and the smallest return after a decision-relevant change.
Problem Frame
Use this when. You are developing a Method or its relations, several questions remain connected, and current facts can change which question should be answered next. A pattern language can help, but its contents page does not tell you whether to qualify a candidate, resolve a whole, compare arrangements, repair a description, or stop.
Start by asking: Which Method-architecture result do we need now, and what fact could change that need? For example: “We can review the candidate description, but cannot yet claim that its proposed Methods form one whole. Which result should we obtain, and what stays valid when the missing evidence arrives?”
The first useful answer can be spoken: “Review these unchanged claims through A.3.1.QM; keep the whole claim unresolved under B.1.5.CM; publication remains blocked by its separate condition. Reopen the whole claim when its missing relation basis arrives.” During ongoing Work, the deciding System uses A.15.7 to choose the next action and names the intended performer and any feedback condition needed for that continuation.
The practical gain is continuity across Method-development questions. Each new fact returns to the result it can change; the practitioner need not restart qualification, comparison, description, and support work together or reconstruct their boundaries from a reading sequence.
Ordinary non-use. If one current MethodDescription already states the applicable action, conditions, result, and stop, use it. If one existing pattern or result answers the whole current question, use that answer and stop.
A single action does not need a continuation account, alternative set, CGUS, or durable result.
Before Work begins, use the architecture, planning, or work-entry question; A.15.7 governs only an ongoing-Work choice.
This is the situated-use branch of Method development using pattern-language knowledge. It does not recover and reconcile an entire source corpus or construct the complete language.
Problem
A Method engineer receives knowledge in different forms: candidate accounts, source dossiers, qualified contributions, architecture proposals, descriptions, and review results. A convenient pattern sequence can hide the distinction between a result that exists, a claim awaiting support, and a next action that somebody must choose.
When one condition changes, two errors follow. The practitioner may carry a stale composition or authority claim into the next action, or reopen every result because “the situation changed.” Both lose useful work: the first by relying beyond its basis, the second by repeating unaffected judgements.
Forces
Solution
C.39 connects a missing way to its receiving result. The construction here uses that connection across distinct Method-development questions when a changed fact can alter the useful continuation. It retains already supported results and returns only the affected question to its supplier; an adequate description remains a sufficient exit.
Keep the current Method question connected to the result that answers it. Obtain the narrow pattern contribution, expose only the continuations whose conditions matter now, and after a substantive result check which earlier claims or uses changed.
Find the Missing Method-Architecture Result
Name the receiving problem and the result it needs. Distinguish an unavailable result from an available result that is not currently applicable. Use the following returns only when their question is present; they are alternatives, not a prescribed sequence.
Retrieve the selected pattern through E.11.PUA/PUR and read its Solution and boundary. Use a sufficient current result in the receiving work. Stop when it settles the commissioned continuation. When it settles the supplier’s question alone, carry it forward and develop the contribution still owed. A title match or search result is not a substitute for that check.
Make the Present Continuation Usable
-
Carry the result boundary forward. In ordinary language name the Method, candidate account, proposed relation, description, or other subject under decision. Retain which claims are established, proposed, contradicted, or unresolved and which source or result supports them.
-
Expose only relevant continuations. For each alternative state the result it could obtain, the condition it needs, the facts known now, and whether it is available, blocked, or unknown. Name the first missing fact or rule. Do not add alternatives merely to make the account branch.
-
Keep material relations readable. State a prerequisite result, local order, alternative, conflict, support condition, check, or return when it changes this use. Two nearby patterns can remain an unordered neighborhood. Use each relation's governor; a connecting arrow, contents order, or shared profile supplies no relation by itself.
-
Return the choice to A.15.7. During ongoing Work, the deciding System uses the available facts and the domain Method's limits to choose the next action. That result names the chooser and, for a selected action, its intended performer, with any stop or feedback condition needed for the continuation; establish capability, access, permission, or authority only when the action needs it.
If Work is blocked by a missing performer, support, or continuation-state relation, use A.15.8's actual-Work configuration branch before returning to choice. An assistant's recommendation is input to that choice.
Before Work, use the architecture, planning, readiness, or fixed-option decision instead.
-
Observe a result before updating its consequences. An expectation remains an expectation until the result is obtained or independently supported. Use the patterns governing result and Work/evidence claims when those claims matter. Choosing an action, changing a plan, performing Work, and learning a fact remain separate.
-
Check the decision-relevant delta. After a substantive result, or before relying on possibly stale information, use §4.3. Reopen the smallest affected question, retain results whose defining conditions match, and stop when the receiving use has its result or blocker.
“No decision-relevant change; continue using this result within its stated window” is sufficient when a continuity statement is needed. If no later use needs that statement, no additional record is required.
Localize a Situation Change
Compare the new fact with the claims and uses currently being relied on. Check applicability or preconditions; alternatives; constraints or interfaces; resources or capability; assignment, access, permission, or authority; evidence/currentness; expected result, stop, or risk; and the decision-relevant situation boundary. Inspect only positions whose possible change could alter the next result or action.
These are different change questions, not a universal event classification. For example, learning that evidence is missing does not prove that the claimed relation is false. If the reach of the change cannot be bounded, name that uncertainty and widen the relevant question; do not claim unaffectedness from the absence of a visible text edit.
Retain Only What Another Use Needs
An ordinary continuation answer states the current difficulty, selected knowledge, available/blocked/unknown action, any missing fact that matters, and the applicable stop or return. Retain an existing result by reference when it already carries the needed claims.
Only when another turn must cite, compare, audit, or rely on the connected answer, keep a Situated Method-Architecture Result as an ordinary C.2.1 episteme. Its content includes the architecture claims and statuses in use, their source/result returns, the current continuations and conditions, the affected-delta answer, and the next return. This is a descriptive result name, not a new root kind or mandatory register.
Its EntityOfConcern is the architecture question or proposed set being considered, or an independently qualified selected structure when that is the subject. Do not use a fictitious CGUS as the concern of an ordinary proposal.
A changed ClaimGraph, EntityOfConcern, or effective ReferenceScheme changes episteme identity under C.2.1; any continuity between results needs a meaningful basis.
The profile/language episteme, a description of one Method, a selected structure, a continuation judgement, a next-action decision, a WorkPlan, performed Work, and evidence retain their separate identities. One result may cite another; that does not merge their subjects or claims.
Open Formal CGUS Only for Its Own Use
The plain alternatives-and-conditions answer often suffices. Use A.22.CGUS when the needed claim concerns a selected structure with potential continuation branching, and that claim must be qualified, persisted, compared, published or more strongly relied on.
A structure requires its constituents, selected obtaining relations, applied constraints, and named use frame; its CGUS membership requires at least two potential continuations across allowed cases. Zero or one currently enabled continuation is compatible with that potential branching.
For a Method-architecture continuation use, the local frame name Situated-Method-Architecture-Continuation denotes this question: which continuation can obtain the next bounded Method-architecture result or stop? Its admissible actions are to select a continuation, retain alternatives, request the first missing fact, or stop. The stop is the first decision-usable result or a missing condition that prevents it. Naming this frame does not establish the structure or choose the action.
Use A.22.CGUS's complete basis for each formal judgement, retaining condition evaluation separately from an obtaining relation.
The formal continuation outcomes are enabled, disabled, unknown, or error. A completed negative condition evaluation has the outcome notSatisfied; derive its continuation effect from the declared required polarity.
An execution or evaluation failure remains error; return the affected test before relying on its result. Missing evidence can leave the condition evaluation unknown. Keep structure identity, membership, case judgement, description adequacy, and later reliance separately checkable.
Archetypal Grounding
Continue a Release-Method Proposal Across Two Turns
This constructed case concerns ongoing Method-architecture work on proposal P. Whole Method W and participant Methods M1 and M2 have independent identity results, and the participant qualifications remain current. P proposes a construction of W using both participants, but the claimed methodPartOf(M2,W) is unresolved: its cited procedure has not been checked for a required contribution or an admitted alternative in W's construction. A description review can inspect the independent participant claims now. Publication is blocked because no publication decision has been obtained.
The Method engineer needs the description-review result now while keeping the whole claim visible for the next meeting. The continuation answer is:
Using A.15.7, the engineer selects description review as the next action and commissions it. The review lead compares their participant-identity claims with the existing identity results and finds support for W, M1 and M2 as identified Methods. The material examined still supplies no basis for the proposed inclusion of M2 in W. The lead gives the engineer those findings now. The first is usable in the meeting, and the second states which construction rule still needs examination. Because the next meeting will rely on these distinctions, the engineer retains one situated result about P's architecture question, citing the existing A.3.1.QM results rather than copying them. No formal CGUS is needed for this use.
Before the next choice, the relevant procedure is inspected. Its construction rule places M2 outside W: M2 supplies optional advice but is neither required nor an admitted construction alternative. That changes the proposed composition claim, not the independently established identity of W, M1, or M2.
The engineer returns the affected composition question to B.1.5.CM. The proposed relation set is not supported as W's obtaining composition; retain the useful support relation and return P for correction or another proposed arrangement. The engineer can use the completed comparison with the existing A.3.1.QM identity results at the meeting. Publication still requires its own decision.
The new situated result changes the composition-claim and return content. It does not backdate a new Method, change independently established completed review Work, or announce a new structure from one new fact. If another proposed arrangement changes W's reusable semantics, that later proposal needs its identity question.
The newly inspected rule decides the composition question and changes the uses that depended on it. The completed identity comparison remains available for its original use.
A Sufficient Description
The current approved description already says: inspect the named evidence field; if the signed supplier result is absent, retain the release hold and return to the supplier. The field is absent and no competing action is admissible. Use that description and its stop. There is no extra continuation result, choice dossier, formal structure, or invented second alternative.
An Assistant Recommends from Stale Facts
An assistant retrieves a relevant review pattern and recommends sending the candidate to the reviewer. Its recommendation uses yesterday's access statement. The current question is whether that reviewer can inspect the restricted evidence today; the assistant has neither that access nor publication authority.
Keep the recommendation as a recommendation. Ask the access holder to refresh the decision-bearing fact. If access is unknown, the evidence-dependent review continuation is unknown; if access is denied, it is blocked. A permitted review of unrelated public claims may remain available if the same review Method allows that narrower result.
The deciding engineer uses A.15.7 for the next action, the access holder supplies the access answer, and the permitted reviewer performs any selected review. Confidence in retrieval supplies none of those relations. If no safe narrower review is available, stop at the missing access result. No performed review or publication follows merely from the recommendation.
Bias-Annotation
Method engineers may favor explicit architectures and overlook a sufficient description. Start with the non-use test. Pattern-language enthusiasts may attribute capability to the language itself; retain the acting and deciding Systems only where those claims matter. Automation can amplify old source assumptions, so inspect currentness before a confident recommendation enters action.
The cases emphasize documentary Method Engineering and review. They establish no empirical advantage across domains, human readers, or AI systems. A high-consequence decision may need stronger evidence than these constructed continuations supply.
Conformance Checklist
- The current question is about a Method, its relations, or its description; a sufficient result is used without a profile wrapper.
- The next needed result and its supplying pattern are explicit; established, proposed, contradicted, and unresolved claims retain their statuses.
- Every exposed continuation has a relevant condition and current basis, or a missing fact.
- During ongoing Work, A.15.7 supplies the action choice; the chooser, intended performer, and any material authority/access/capability relation remain distinguishable.
- A substantive change reopens its affected Method-development result and dependent uses; unaffectedness is checked against the retained basis.
- Formal CGUS, persistent results, and Work claims are opened only for a use that needs them and satisfy their rules.
- The answer states the first useful result or blocker and the next stop or return, without treating reader order as Work order.
Common Anti-Patterns and How to Avoid Them
Consequences
The practitioner can continue Method development without rereading the whole repertoire or treating every local change as a new Method. Missing basis remains visible at the result that needs it, so independent useful work can continue without claiming that blocked work has become possible.
The cost is maintaining a few explicit result-to-use connections when they matter across turns. A poor or stale pattern language cannot be repaired by this continuation method; the missing content must return to source recovery, qualification, architecture, or authoring.
Architectural Rationale
General pattern use and live Work steering already exist. The Method-development contribution is the connection between a current professional question, the distinct result that can answer it, and the smallest affected return when that result's basis changes. The table in §4.1 and the two-turn case make that connection usable without copying the full qualification, composition, representation, or steering Methods.
A MethodDescription is preferable when it already supplies the action and stop. A fixed route is useful when its dependencies remain sufficient. For a changing plurality of questions, reconstructing the returns from separate patterns can hide an unresolved whole or repeat unaffected qualification. This method keeps those returns visible at a cost proportionate to the present decision.
A pattern language supplies problem-oriented knowledge for the deciding practitioner. Local action order follows the applicable prerequisites and causal assumptions; a contents order does not establish them.
SoTA-Echoing
Use these source contributions for the stated questions and limits. Reopen the affected source or action when a current alternative supplies the same result with less reconstruction, a source change alters a load-bearing distinction, or use evidence reveals a missed dependency or unnecessary burden. Claims of human or AI superiority require comparative evidence supporting those claims.
Relations
- C.39.RP and A.3.1.DR supply documentary knowledge; A.3.1.MR and, when needed, an applicable research Method supply occurrence-evidence recovery results.
- A.3.1.QM, C.32.MWA.PP, and B.1.5.CM keep individual qualification, alternative architecture, and whole resolution distinct. A.3.2.DE, C.37.MP, A.15.8.MS supply description, cross-use representation, and support results; A.3.1.CC and A.3.1.MV carry the applicable correction or variant return.
- E.11.PUA/PUR govern the pattern-use and recommendation questions, including reuse of a matching earlier result. A.15.7 governs live next-action choice; A.15.2/A.15.5 and C.11 retain their other entry conditions.
- A.22.CGUS governs structure and continuation judgements. C.2.1 governs a retained result episteme. A.3.1/A.3.2/B.1.5 and A.15 keep Method, description, composition, plan, Work, and result claims distinct.
- E.11.PFP and E.4.PFIP supply publication-form and publication-preservation work when that question becomes current; a situated continuation grants neither publication nor release.
C.39.MC:End
Develop Branching Search from Reusable Material
Type: Method pattern Status: Stable Normativity: Normative unless marked informative.
Problem frame
Use this when you need to develop alternatives from reusable material, to design how someone can find and examine them, or to establish what a search actually tried and found. A sound workshop can alter its own recording; an investigation team can change a view of permitted records; a coding team can develop a working program. The eventual contribution may already be selected or remain unsettled.
An evolving construction, a working Method or a retained intermediate result can be material for search. The commission determines whether the recipient needs a new variation, a way of finding one, or knowledge of what earlier work produced.
When commissioned to vary the material, make a feasible change, examine the difference and retain what the next use needs. When commissioned to design the search, explain how future practitioners will generate, distinguish and use possible results. For an inquiry about earlier search, work from the evidence of the actual attempts. Each return has its own grounds for stopping.
Sections 4.5–4.12 develop conditional uses:
- opponents that respond to a candidate;
- a set of challenges whose selection changes as candidate ways improve;
- ways compared through the results of their application;
- policies developed through models of their consequences;
- development guided by pairwise comparisons;
- human contributions that change the continuing development;
- representations built together with their useful changes; and
- materially different continuations made available for further development or use.
None is required for an ordinary material variation. Use C.40.CD, also reached from section 4.3, when problems and ways need development together. Ordinary variation does not require inventing a new problem. Reuse a sufficient existing result or way without entering a search. If the missing operation is how to obtain a local result at all, C.39 supplies that constructive explanation; a settled ultimate goal is not the discriminator.
Problem
Search can become an inventory of interesting possibilities without any performed development. It can also discard everything that loses today's comparison, including material that makes a useful next attempt possible.
The opposite excess is endless novelty or indiscriminate retention. A changed object, a higher score or success in a source problem does not by itself justify more work. Productive branching needs feasible operations, examination under the current question and an affordable continuation. Coupled search additionally needs validly constructed problems and target-tested transfer.
Forces
Solution
Begin with what the recipient needs from the search. For a usable variant, make a feasible change and examine the result against that use. To design a search, work back from the differences it must reveal. Which material and operations can generate the alternatives? What examination can distinguish them, and how will its result change what is retained, tried next or returned? A search can be well designed before it runs. For an account of past search, recover the attempts and outcomes from evidence of those runs, including failed or discarded attempts when the question covers them. One retained success can answer a request for a usable result. Its selection alone leaves the history of the other attempts open.
Make one small development pass
- Recover usable material and the local question. Identify the editable or executable input and the conditions for using it. State the difference you can examine and any selected receiving contribution. Keep an original when comparison or recovery requires it.
- Select and perform a feasible variation. Use the available professional operation to change, recombine or use the material differently. State what is changed and what remains protected. If the operation is unknown, use C.39 for that local contribution. When several changing operations need selection from their continuing results, use section 4.11. An unperformed variation remains a proposal with its missing prerequisite.
- Examine the resulting difference. Compare what was intended with what can now be observed or otherwise established under the relevant Method. Include a useful unexpected difference. Distinguish a result that helps one question and fails another.
- Preserve the material needed for the justified next use. Keep the changed artifact or result and the inputs, settings or explanation needed to continue. Retain a failed or mixed result when it changes the next question. A descriptor or favourable evaluation alone may not let another participant reproduce or develop the result.
- Choose continuation, retention or stop. Ask what another attainable result could change, what producing and carrying it would cost, and what work that would displace. Continue only the branches that warrant that burden. A sufficient result or a lack of worthwhile continuation can close the search.
Several variations can share one pass when they answer the same local question and their outcomes remain distinguishable. No population size, fixed branching factor, archive, new hypothesis or exhaustive trial of every variant is required.
Continue without losing the question or the material
For a selected contribution, examine variations against its requirements while retaining an intermediate result only for a stated useful continuation. A lower present score can coexist with that reason; retention does not make the variation the preferred option.
Before a receiving contribution is selected, examine a local change that is already intelligible: for example, what reversing a playable recording does, or what a complete time-ordered view exposes. A newly suggested use becomes a question to investigate, not evidence of demand or value. Stop if no continuation warrants its burden.
Preserve the permissions, confidentiality limits, means and participant capabilities relevant to the next use. A colleague who receives the recording or code may still lack the skill or right needed for the proposed continuation. Retaining material is different from funding further work or preserving an exercisable option.
When a single present winner is eliminating other worthwhile possibilities, section 4.12 develops the choice of useful differences, local comparison, renewed variation and use of the retained material. It also explains when changed observations require rebuilding that collection. A small sufficient set remains a valid result.
Coupled problem-and-solution search
Enter here when problems and available ways need development together. C.40.CD supplies the full Method: recover a sought result and usable construction, form a consequential changed question, adapt the obtaining operation, establish what follows under the target conditions and use the outcome to choose the next question or use.
Keep the original receiving requirement connected to any narrower or exploratory problem. A partial target result may permit one useful comparison while another needed comparison remains unsupported. The next inquiry follows that remaining condition or another worthwhile use of the new construction.
Use section 4.1's material-changing and examination operations where needed. A sufficient existing answer can finish the work. B.5.QD constructs a missing question; C.39.RO constructs an operation for changed or combined use. C.40.CD connects those results through target application and preserves the conditions for a useful continuation.
When a construct's unfamiliar response must become a useful reproducible application, C.40.CU connects the needed probe, application, arrangement and receiving use. It can stop at a bounded response or conditional construction, and returns to this pattern for further worthwhile variation. Ordinary material development does not require that larger question.
Use stronger claims only when they matter
A few explained variations need no formal archive or front. Use C.18 when retained exploration value, lineage, possibility-space change or a non-dominated front is claimed. That account records the relevant relations; it does not perform generation.
C.19 supplies continuation policy for a live search pool, while C.11 supplies a local choice. C.38 supplies comparison of complete-enough ways for one result. None converts material retention into current selection or permission to use the result.
For assurance-bearing use, identify the material, performed operation or still-proposed operation, examination basis and receiving conditions precisely enough for the claim. An observed difference, its evaluation, a proposed explanation and a claimed later effect need their own support.
Develop against opponents that adapt to the current way
Use this branch when a useful way must respond to another participant's actions and a fixed collection of challenges no longer exposes its consequential weaknesses. A simulated opponent, a test generator or an authorized exercise team can search for those weaknesses. Prefer fixed representative trials when they already answer the question; maintaining adaptive opponents adds work and can reward behavior irrelevant to the receiving use.
Begin with the interaction that matters. Specify what each side can observe, when it acts, which actions are permitted, how the interaction ends and which outcomes receiving work values. Give the challenger a way to produce a meaningful difficulty within those conditions. A challenger that wins by exceeding the permitted load, seeing private future choices or exploiting a simulator error has exposed a setup defect unless that condition belongs to the intended use. Developing another participant's behavior changes the conditions faced by the candidate; it need not change what counts as a useful result.
Build an informative first interaction. Start with a challenge the available way can engage with, so different variations can produce distinguishable outcomes. If every attempt fails before the operation under development can act, simplify a named condition or supply the missing prerequisite. Retain the original receiving requirement while using that easier development case. Choose the challenger's objective separately from the candidate's objective: negating a candidate score can reward early termination, wasted effort or another event that teaches nothing useful.
Alternate development with an intelligible comparison. Hold the candidate fixed while finding a permitted response that defeats or strains it. Examine the failure mechanism. Then hold a selected set of challenges fixed while changing the candidate to address that mechanism. A search algorithm, a subject specialist or a team can construct the variation; the representation and changing operation must permit it. Several fresh starts, different initial conditions or different allowed strategies can expose weaknesses that one continuing opponent misses. If teams participate, examine the interaction of their strategies as well as individual performance. Alternation is an affordable default, not a requirement that every interacting population update in this order.
Keep challenges for what they reveal. Preserve executable opponents or sufficiently complete exercise instructions, initial conditions, observation rules and meaningful failure cases. Include earlier challenges that expose different weaknesses, even when they no longer defeat the incumbent. Recheck proposed replacements against them. Sampling from a growing collection limits recurring cost, but rare failures can disappear from the sample; retain protected cases when their loss matters. Remove redundant cases only relative to the candidate set and conditions used to establish redundancy. A later way can fail on an old challenge that today's way handles, so a presently weak opponent is not universally dispensable.
Choose on a common basis. Compare the incumbent and proposed replacement against the same declared opponents, conditions and outcome rule. For a stochastic interaction, use enough repeated trials to distinguish a useful change from the uncertainty relevant to the decision. Keep individual outcomes or failure classes visible beside any aggregate. A weighted average answers a question about that weighting; a worst-case requirement needs its own comparison. Cyclic advantages can make different candidates useful against different opponents without providing one best candidate.
Separate success against the latest opponent from success against retained opponents. Establish why the evidence or argument supports the broader claim. When empirical generalization remains uncertain, use opponents or situations outside selection that distinguish the relevant possibilities. A reserve repeatedly used to select replacements becomes part of selection evidence. It is no longer an untouched final test. Broader success supports the examined family and conditions; it does not establish superiority over every possible opponent or indefinite progress.
Use the result and revise locally. Return the usable candidate, the challenges needed to reproduce its strengths and failures, the comparison and its limits, and the next justified use or development question. A specialist can be retained for its supported situation while a more broadly useful candidate is selected elsewhere. Include challenge construction, repeated interactions, retained material and independent examination in the cost comparison. Stop when the receiving result is sufficient, no informative permitted challenge can be obtained at worthwhile cost, or the interaction has become irrelevant to the receiving need.
If the required outcome or evaluation procedure changes, use E.23:4.1a and A.19.ECS:4.5 for the affected comparison. If only the opponent changes, rerun the candidates against a common relevant opponent set before calling the score change improvement. If the challenger defeats the task formulation rather than the candidate, return that condition to C.40.CD instead of treating impossible demands as productive pressure.
Develop a set of challenges and ways of handling them
Use this branch when choosing the next challenge and finding a way to handle it need to inform one another. A fixed exercise sequence may repeat what is already easy or jump beyond every available way. Improving one candidate in one situation may also miss a construction already useful elsewhere. The result sought is a small set of worthwhile challenges with usable candidate ways, examined transfers between them, and a justified next development or receiving use.
Start from one usable challenge and the way being tried on it. Recover the challenge's inputs, required result and conditions, the candidate operation, and how an attempt can be judged. Make a small feasible variation through §4.1. Keep the original challenge and candidate available until the new comparison settles what to retain. A domain operation must supply the changed material: for example, a programmer changes an input grammar or an exercise designer changes an available tool.
Choose a challenge worth developing. Examine proposed changes on three different grounds. First, is the task meaningful and is its required result assessable? Second, is there a plausible attainable development step with the available ways and means? Third, would learning to handle this case add a consequential distinction, construction or possible use beyond what the retained cases already supply? An attractive but currently inaccessible task can be kept for a stated later question; a trivial variation can remain a regression case without receiving more development work. Use C.17 for a stronger novelty or usefulness claim, keeping its comparison basis explicit.
Estimate attainability by trying the inherited way or another inexpensive candidate. In a setting with a graded outcome, a declared band can help avoid spending all effort on trivial or wholly inaccessible cases. In another setting, use a worked partial result, known prerequisite or failure mechanism instead. Failure by current candidates establishes their limit in that attempt, not impossibility of the task. Preserve the unresolved premise when the attempt cannot distinguish inadequate means from an invalid challenge.
When task generation also creates its judge or reward, examine the connection before using the score. Recover what successful action must accomplish, whether the generated situation permits it, and what observations would distinguish accomplishment from exploiting the scoring rule. A successful execution check shows that the task runs under the checked conditions; whether its reward or success check measures the described task needs separate grounds. Use an independently grounded contrast case, a domain model or a supplied requirement for that distinction; A.19.ECS develops a missing evaluation. If no adequate ground is obtainable, keep the task and evaluation as a proposal rather than call its score success. If another task is preferred because a model calls it interesting, retain that model judgement as a selection aid and identify whose preference or which intended use warrants relying on that judgement.
Separate three returns from this examination. If the task cannot be executed, repair its construction or discard it at the chosen effort limit. If it executes but adds no worthwhile distinction, propose another task. If it is worthwhile but not yet handled, examine the missing prerequisite, ineffective feedback or candidate limitation before making it harder. A task made easier by supplying a prerequisite can focus development on the next operation; do not credit the candidate with obtaining what the setup already supplied. A progress reward can guide attempts while a separate result condition establishes success. Both still need grounds.
When there is a fixed receiving use, compare progress on development tasks with results in that use. High local success with low receiving performance calls for a transfer or representativeness inquiry. Low local success despite adequate receiving performance can expose an unnecessary or malformed exercise. Change a named condition on that evidence and recheck the connection; automatic escalation after every success misses these alternatives. Keep representative receiving trials in view during development, and retain a separate examination for any stronger final claim. A repertoire of interesting tasks and improved performance on one receiving task are different possible results.
Improve locally, then test useful connections. Develop each selected way in its current challenge far enough to obtain an interpretable result. Keep unlike successful constructions and informative failures available. Try a transfer when a retained operation, artifact or starting configuration could supply a missing contribution in another challenge. Test it against that recipient's incumbent and result conditions. Compare direct reuse first when meaningful; try a bounded adaptation when its possible benefit warrants the work. Passing the source task alone does not choose the target's replacement. When the transferred object is a description for a person, allow for acquisition, practice and support instead of assuming that copying the description transfers performance.
Use the target comparison to keep the incumbent, replace it for that use, retain several situation-specific alternatives, or return a missing contribution. A transferred result that makes a proposed challenge already easy can change the earlier decision to spend development effort there: recheck that decision after transfer. The challenge may now serve as a regression case or as a parent for another variation. Preserve its solved result even when it leaves active development.
When one way must retain the whole repertoire. Choose this arrangement when the receiving use needs the same candidate to handle several tasks. Continue developing that candidate across them rather than treating separately successful specialists as its combined ability. Keep earlier and current outcomes on the tasks whose retention matters, under comparable conditions. After work on a new task, examine whether the same candidate improved there, lost a previously useful result elsewhere, or merely produced a noisy fluctuation. Use repeated trials when uncertainty could change the decision, or an exact result or discriminating case when that establishes the change. A success rate near one half can reflect random difficulty without any improvement; task difficulty alone is not learning progress.
Use that comparison to choose the next development work. Return a still-needed task on which performance was lost to the training, practice or repair inputs. Include a promising new task when its possible acquisition warrants the effort, and keep enough work on the receiving use to prevent local gains from displacing it. Recently mastered tasks can receive less attention while protected comparisons still check them; a stale result needs renewed examination before it justifies omission. Interestingness can help choose among feasible acquisitions, but does not excuse losing a required old result. The mixture and frequency depend on cost, uncertainty, forgetting or regression risk and the result required; no fixed proportion applies to every use.
Perform the available learning or repair operation with those chosen tasks, then examine the changed common candidate on the affected old and new tasks and on the receiving use. Update the next allocation from that evidence. Retain the previous candidate when it is needed for comparison or recovery. If a change repeatedly trades one required result for another, investigate the conflicting requirements or the candidate's means; alternatives include a richer common operation, situation-specific dispatch, or a justified narrower use. Further resampling alone supplies none of those missing constructions. Stop when the shared result is sufficient or the next acquisition and its retention burden are not worthwhile. An independently reserved final examination remains separate from cases repeatedly used to select these changes.
Compare challenges by what matters to continuation. A fixed task parameterization can describe relevant differences cheaply. When appearance or generator parameters miss them, examine how the same available ways perform on the tasks. A task on which one way succeeds and another fails can reveal a different opportunity from one where all behave alike. Recover the indexed ways and common evaluation conditions behind that response profile. It is relative to that repertoire: replacing a member can change which tasks look alike. Update the affected comparisons rather than treat an old descriptor as a permanent property of the task. Different generating encodings and different characterizations can serve different purposes; changing one does not automatically change the other.
Retain enough to continue, and choose where effort stops. Keep the task material, its evaluation grounds, the candidate way, outcomes and any adaptation needed for the useful next attempt. Preserve a failed task when its diagnosed obstacle could become tractable with another operation. Feed that obstacle and the useful differences from solved tasks into the next task proposal. An archive of separately successful specialists does not establish that one way handles every task. For that stronger requirement, establish how one candidate covers the tasks. Its construction or a bound can settle the claim. Examine its performance where material uncertainty remains. Separate cheap retention from continued optimization; C.18 supplies stronger archive and lineage claims, and C.19 supplies a decision about funding several continuing lines. Bound the number of live tasks and the work spent trying transfers by the contribution sought and total burden. A cheaper fixed suite, direct specification or a known curriculum can finish the work when it already supplies the result.
This development need not reformulate the meaning of the receiving problem at every step. Use C.40.CD when a failed condition or new construction changes the answer that is now worth obtaining. Use §4.5 when a challenge must respond to the candidate during interaction. Keep evaluation repair under E.23:4.1a/A.19.ECS:4.5. Those returns allow one useful result to continue through a different operation without making all branches compulsory.
Develop a way by what its use produces
Use this branch when the material being varied is a way of obtaining a result: a learning rule, a constructive heuristic, a training loss, a procedure with adjustable settings, or executable code. The useful difference concerns what the way can produce when applied. A derivation may establish that consequence before execution. Choosing a good finished schedule and discovering a rule that makes good schedules on further inputs are different searches. A sufficient known way, direct derivation or small complete comparison can avoid the larger search.
Construct the two connected operations. Give the candidate a usable representation and a permitted changing operation; section 4.11 develops that connection when it is missing or obstructs a needed variation. Parameters permit tuning within a form; changing a formula, operation or connection can enlarge the forms reachable. Include a known usable incumbent. Start with variations whose application can be performed and examined; a larger representation is useful only if it exposes a consequential alternative at an affordable search cost. C.39 develops an unavailable obtaining operation, and C.38 compares complete-enough alternatives.
Establish what each candidate can produce on the stated inputs. Use an applicable derivation or prior result when it settles the comparison. To observe an uncertain response or perform a commissioned trial, apply the candidate and examine what it produced. In a learning application, this includes initialization, training and subsequent use of the trained model. In a scheduling application, it includes constructing and examining a schedule. Recover the candidate way, its starting state, the resulting artifact or state, and the outcome separately. The outer development changes the way; the inner application obtains the result used to compare ways. These are roles in this construction, not a universal hierarchy of Methods.
Specify which inputs the candidate may use and which observations its assessment will use. A learning rule can see training examples while its performance is assessed on other examples. For a reusable way, establish why the construction works over the required range. An applicable rule, bound or exhaustive account may settle it. Where the inference depends on an observed response, choose cases that distinguish the consequential alternatives and retain their actual reach. Protect important failure classes beside any aggregate. Keep cases repeatedly used to select changes distinct from a later examination supporting a stronger receiving-use claim. Repeatedly consulting a reserve makes it selection material.
The arrows show result use and a return for development. The forecast selects work; the produced result supports the corresponding outcome claim. When the use is prediction alone, C.29 supplies that different bounded claim.
Choose what is allowed to change together. A useful form may require its own tuned parameters. Compare a form with parameters fitted by the allowed inner operation rather than condemn it under settings suited only to its rival. Conversely, adding adjustable parameters is not automatically an improvement. Compare fixed forms, tuned forms and changed forms where those alternatives can settle the question. State the allowed tuning work for each; count it in the development burden. If the changed form needs another initializer, data preparation or support, compare that complete arrangement and preserve the dependency in the returned way.
Changing a training loss changes what guides updates. It does not by itself change what the receiving use values. An outer comparison can favor a loss whose numerical value is larger or whose early learning is slower because the resulting model performs better in the intended use. If the evaluation itself is defective, return to A.19.ECS and E.23; optimizing the candidate against that defect is no repair. A change of the receiving requirement instead reopens the affected question through C.40.CD.
Identify which work could release the constraint. If useful proposals have stopped appearing, compare how the current way generates, examines and passes them on. Give an already qualified proposal to the receiving operation. If it can use the proposal, inspect whether generation reaches proposals of that kind and whether the assessment would retain them. If assessment discards a proposal that withstands a relevant trial, repair the comparison or the reference collection. If it is retained but the receiving work cannot start, identify the missing knowledge, access, authority or other means. A successful intervention locates one constraint under the tested conditions. Several contributions can constrain the same continuation. Compare their combined change with the least costly sufficient repair.
Keep three grounds for allocating another attempt distinct. Difference from an archive depends on its contents and the means of comparison. Suitability for the receiving work depends on a criterion for its intended result. Promise for further development depends on what could be done next with the candidate. A favourable judgement of promise can warrant a bounded trial. To credit the candidate with opening a continuation, establish what the trial or a justified construction makes attainable and what provision it requires.
Make search affordable by choosing an approximation. First obtain enough applications to make the comparison intelligible. Then choose among the following arrangements according to what is costly and what evidence the decision needs. They can be combined when their combined losses remain acceptable; they are not compulsory stages.
For prediction-assisted search, choose features available at the time the decision must be made. Relate them to completed outcomes from the relevant regime; a feature measured only after completion cannot save that completion. Fit the prediction rule, use it to propose the next costly trial, perform that trial, and update the rule with the obtained outcome. An early trace can support a forecast of later performance but can miss a late improvement. Keep a way to investigate such misses when they matter: complete a discriminating delayed case, reserve some effort for uncertain regions, or broaden the calibration data. Stop relying on the forecast for a changed regime until its relevant relation is supported.
An early stopping rule similarly rejects further work, not the possibility of eventual success. If it is calibrated only on fast learners, a repeatedly discarded slow family can remain invisible. Recheck that omission against the receiving question. A cheap filter for malformed candidates is different: it can reject a proved type error or inadmissible operation without predicting final quality. Reusing an earlier result for an equivalent candidate requires the relevant equivalence; agreement on a few sampled outputs provides only that sample's evidence.
For asynchronous evaluation, retain the identity and starting conditions of every pending candidate. Queue enough independent work to use the available workers. When a selected batch of results returns, compare those results under their applicable conditions, select material for variation, and submit replacements without discarding the still-running candidates. A small return batch permits prompt adaptation; waiting for a larger batch can reduce how strongly a few fast lineages determine the next proposals. Retain each result's task, starting state, work allowance and candidate version. Check whether older parent material or faster return changes what is selected, and refresh the material used for variation when that is the cause of lost progress. Compare resulting quality and total work as well as elapsed time before retaining the asynchronous arrangement.
Choose a fresh-start comparison or continued development. With independent applications, each candidate receives a defined starting condition and work allowance. This supports a claim about the candidate's way under those conditions. With continued development, a successful state can be copied or retained, its settings changed, and work resumed. This can produce a better final artifact cheaply even when the last settings would be poor from the original start. Return the trajectory or reproducible continuation, including the state and earlier operations it depends on. A claim about the final settings from a fresh start needs grounds for that start. A derivation, adequate prior result or a fresh-start trial may supply them.
For example, in a stipulated update toward a known target 10, the rule is x ← x + a(10 − x). Two updates with a=0.5 from x=0 produce 7.5. Two updates with a=0.1 from an inherited x=8 produce 8.38, but the same rule from x=0 produces only 1.9. The continuation ending at 8.38 has an error of 1.62, smaller than 2.5 for the other run; this supports that continuation. It does not show that a=0.1 is the better two-step rule from zero. Directly setting x to the already known target would be cheaper in this toy problem, so the arithmetic illustrates the comparison distinction, not a need for evolutionary search.
State reuse needs a legitimate realization. Software may permit copying model weights and compatible optimizer state. A changed structure may make that state unusable. A person's acquired skill, fatigue or experience cannot be cloned by copying instructions. For development involving people, obtain the preparation and learning conditions for those people and use A.3.1.WT for the selected Method trial. If a comparison requires identical acquired states, keep that unavailable condition explicit; copying instructions supplies no such state.
Develop persistence and recovery from states the process reaches. A repeated rule can produce a wanted result and then destroy it, or work from its original start but fail after a disturbance. Use this construction when the receiving demand includes continued useful operation or recovery. Distinguish what must be obtained from the original start, what must remain true during use, and which disturbances must be recoverable within the available time and resources. Continued functioning can involve changing configurations; a fixed picture need not be the target.
Begin with a sufficient known rule or complete small comparison when available. Otherwise, one arrangement is to run candidate rules for longer and judge the relevant result at several times, using that feedback to change the rule. Checking only the first attainment of the target leaves later destruction unexamined. Checking only one later endpoint can miss an intervening failure. In gradient-based training, retaining a long computation for a backward pass can be expensive; other forms of development can have different costs.
When restarting from retained states makes the work practical, construct a small pool of continuation states. Each entry contains what the process needs to resume, including necessary hidden state and relevant conditions. It must be copyable or reproducible by available means; a picture or score alone may not provide it. Use long trials or obtain the missing continuation support when such starts cannot be realized. The rule being developed, one state it produced and the useful whole obtained through execution remain distinct.
Initialize the pool with realizable permitted starts for the receiving claim. For a development trial, take a manageable selection from the pool. Run the current rule for an agreed interval, inspect its produced state and receiving result, and use that feedback to update the rule through the selected learning operation, whose required inputs and update conditions must be obtainable, or to compare executable variations. The feedback can concern a whole later configuration or function; it need not supply the correct local action at every intermediate step. Return suitable output states to the pool as starts for later trials. Retain reached, incomplete or failing states whose continuation matters. When the ability to begin remains required, include enough original-start trials to expose its loss. Merely keeping a seed in storage gives it no influence on development.
The returned states make the next trial begin where an earlier application ended. The rule must maintain or improve an already formed result; include creation from the original start when the receiving claim also requires that ability. Limit the pool and select its entries by the needed state range and affordable work; keeping only easy successful states can hide a recurring failure. Record which rule and conditions produced an entry. If a revision changes what its stored state means or requires, reconstruct compatible starts or examine a supported conversion before reusing it. Restarting from a saved state avoids storing its entire preceding computation; it does not assert that the resulting learning update is equivalent to training through that whole computation.
For recovery, apply a relevant permitted disturbance to some suitable states before continuing their trials. Keep undisturbed and original-start cases where those abilities remain required. Perform the continuation, judge recovery within its allowance, and keep running long enough to examine whether the recovered result functions and persists. Use consequential failures to change the rule or the conditions it needs, then repeat from the affected starts. Match the training feedback to the receiving result: resemblance to a desired body may support a shape claim while its subsequent movement remains poor. Compare functioning when that is what the work needs. States and disturbances repeatedly used to choose changes are development material; examine further relevant conditions before making a broader claim.
Damage can also remove the information that identifies the intended result, or remove the means to restore it. Obtain a known target, retained distinguishing information or qualified assistance when different required restorations are compatible with the same remaining state. Obtain the required actuation, material and energy for a physical repair. If those contributions are unavailable, return that limit; additional training cannot perform an absent operation. Stop with a sufficient directly chosen rule, a supported growth-and-continuation arrangement, or a statement of the missing contribution. Finite trials support their examined conditions; persistence for an arbitrary duration needs an additional ground, such as a preserved invariant.
Learn from the histories the developing way causes. Use this construction when a learned action rule or fitted component helps determine the inputs it will encounter next. A small initial error can take it outside the histories demonstrated by a competent source. Keep a fixed dataset when it already supports the intended behavior; use a direct rule when that obtains the required response more cheaply.
Run the candidate inside the permitted interaction, retaining the observations and actions it causes. At a consequential reached history, obtain a teaching target from a suitable expert, source model or other qualified feedback operation. The target can be a corrected action, a distribution over continuations or a component output that supports the acting way. Pair it with the information the learner will receive. An observation or outcome following the old action belongs to that action; obtain the response to a corrected continuation through a new permitted application or supported model.
Fit the selected learner with a procedure appropriate to its representation, available training material and receiving criterion, then run the changed candidate in the interaction again. Use the new histories to obtain further needed targets, retaining relevant earlier examples and protected behavior. A changed learner changes what will be encountered; repeating the original expert demonstration alone can miss the new error. Stop when the required behavior is supported at its receiving conditions or the remaining learning no longer warrants its collection and use costs. More collected histories are useful only through the learning and later behavior they permit.
Training may use information unavailable during action, such as a simulator's hidden state, to obtain targets. Keep that information out of the learner's operating input unless it will be available there. Test whether the available history can support the required response. If indistinguishable histories require different actions, return to the observation, retained state, supported assistance or narrower use; inconsistent targets cannot repair the missing distinction. Preserve this information-and-timing condition in the constructed policy and in the qualification of its receiving use.
Check whether the teacher can supply useful targets at the learner's reached histories. If it needs adaptation, obtain several teacher continuations from a learner-produced prefix. Judge their completed outcomes with an available check, use that feedback to train the teacher on its continuations, and return the updated teacher to the learner's next training. Repeat as the learner changes when benefit warrants the cost. This requires access to change the teacher and useful outcome feedback; otherwise retain another supplier or an unsupported continuation. Improved completed outcomes do not establish correctness of every local instruction. Include data collection, target preparation, fitting and repeated interaction in the complete development comparison. A source-model trial supports that model's setting; physical or other receiving use still needs its own permitted examination.
Return a usable discovery and its supported claim. Keep the executable construction or sufficiently developed description, necessary initialization and support, settings or adaptation rule, outcome comparison, and consequential failures. If the sought result is the final trained artifact, return that artifact rather than claim a new reusable learning way. If a general way is sought, establish its reach over the commissioned inputs. Investigate a dependency on the search setup when it could change that result. Separate changes that preserve the computation from simplifications or decouplings whose effect needs a new comparison. Removing an apparently redundant instruction can expose a useful compact method; a component that merely looks strange may carry the improvement. Establish that the changed construction meets the receiving terms before replacing it.
Compare development cost, recurring use cost and time to a useful result separately. Count candidate construction, partial and full trials, tuning, repeats, prediction-model construction and updating, retained state, and the receiving examination where applicable. Fewer full trials can coexist with expensive feature collection; less waiting can coexist with more total work. Stop with a sufficient incumbent or retained conditional alternative when further improvement does not warrant that burden. Changes to task inputs, training horizon, allowable support or receiving criterion reopen only the comparisons they can invalidate.
Search for a way to act through a model of its consequences
Use this branch when actual trials of many action rules are costly, slow or harmful, but an obtainable model can help select a smaller number of worthwhile candidates. The search develops a policy: a rule choosing an action from information available when acting. A predictor instead estimates what follows from an action in a context. Choosing the prediction with the best number is not yet constructing, performing or validating that action rule. A small direct comparison, an adequate known policy or an exact solution can finish the work without this arrangement.
Connect the action rule to the consequence account. Name the contexts, allowed actions, outcome meanings and receiving choice. A context contains observations or remembered information available before the action. Distinguish controllable settings from conditions the practitioner cannot choose. For repeated decisions, preserve the transition, observation timing, resource use and accumulated consequences in the policy's state and update account. A fixed sequence and a rule reacting to later observations can produce different results.
Write the two roles as a = π(c) and o_hat = P(c,a). The first returns an action; the second returns a modeled outcome, distribution or bound. Give π a representation that can be applied and varied: a table, conditional procedure, rule set, parameterized function or another suitable construction. Include an incumbent whose operation is understood. The professional Method must supply how to perform each returned action; changing a policy description does not make new equipment, access or skill available.
Choose what P must return for the intended comparison. A one-step response can be iterated only with a sufficient update of context and relevant uncertainty. A predictor of total return already summarizes a continuation under its training conditions; repeatedly adding its outputs can double-count that return. A terminal score cannot supply an omitted trajectory constraint. For several outcomes, retain their trade-offs and any protected conditions. A penalty permits violations; an action restriction excludes them only insofar as its enforcing operation works. An aggregate improvement can conceal a loss in a context or for an affected party.
Obtain a model for the receiving use. Recover observations as matched context, action taken and resulting outcome, including timing, selection, censoring and changes of regime. Train or fit P with an applicable learning or modeling Method, retaining its required input meanings, fitting criterion and qualification conditions. When replacing an expensive simulator or source model, construct or obtain a cheaper response that preserves the comparison needed; state the substitution limits and the condition that returns an unsupported query to the source. Keep a source call when no adequate substitution is available. Preserve this distinction: agreement with a simulator supports the simulator's response, while agreement with observations addresses the modeled phenomenon. Neither automatically identifies what an untried intervention would do. Use C.28 and the corresponding subject model when that intervention claim matters.
Keep the inputs available during policy use separate from information used only for assessment. For a human or expert-generated proposal, store a proposed action separately from the action performed. Fit on the latter's outcome. An action chosen only in favorable conditions may appear better because of those conditions; merely adding more such records can preserve the error. Where the intervention consequence is unidentified, the useful output can remain a conditional model comparison or a question for a discriminating trial.
Assess the model where the search will use it. Compare relevant response errors, constraint crossings and ordering of candidate actions, not just an overall prediction average. Include withheld contexts, histories or regimes according to the claimed further use. For an iterated model, test the horizon and feedback operation as well as one-step error. Optimization can drive π toward poorly supported combinations even when P predicted historical behavior well. Restrict that use, obtain an informative source response, or retain its uncertainty; repeating the optimizer does not repair the missing relation.
Develop and compare complete policies. Fix a common set or distribution of contexts and a common model version for a comparison. For each candidate π, compute its actions and feed those actions with the corresponding contexts to P. In sequential use, carry the modeled state forward and apply π to the information then available; include the chosen horizon and terminal consequences. Calculate the agreed outcome profile and action costs. Retain nondominated candidates when the receiver has not selected one trade-off. C.18 supplies the stronger front and archive claims; a finite search usually returns the best alternatives found, not an exhaustive optimum.
Vary promising policies using the representation's feasible operations, and repeat within an affordable allowance. A rule for one context can be combined with a rule for another; a joint action can combine operations inside the same context. Those are different search spaces. Check their joint conditions: two separately allowed actions may share a resource or interfere. Keep a sufficient incumbent and useful diversity so that the next model update does not leave only closely related candidates. Direct enumeration is preferable when the candidate family is small enough; evolutionary variation is one obtaining mechanism for a large family, not a condition for model-assisted development.
These arrows show result use. The return through actual use is conditional on permission, feasibility and the needed evidence; a recommendation can be the current endpoint. A simulator test supplies a source-model outcome, not the observed-world data represented by D.
Use uncertainty to change the continuation. Decide whether the uncertain comparison warrants a source query, a limited trial, a conservative alternative or an unresolved answer. With supported response bounds, propagate them into the policy comparison. With fitted predictive distributions, compare calibration and proper scores on relevant observations and retain the population, horizon and selection conditions. A nominal interval at one fixed query is not a simultaneous guarantee over all policies searched. Common model omissions can survive an ensemble, and a narrow fitted interval can be wrong outside its supported regime.
For a retained point predictor, a residual model is one possible addition. Pair its predictions with known outcomes, form residuals r = y − P(x), and fit a model for those residuals using the context and the original prediction. The corrected point is P(x) + estimated residual mean; the residual model supplies a conditional uncertainty account. RIO realizes this with a Gaussian process whose kernel is the sum of an input kernel and a prediction-output kernel, fitting its parameters to residual data. It leaves the original predictor unchanged. Its primary method, §§3–5 and Algorithm 1, supplies that numerical construction and its assumptions. A practitioner using it must distinguish uncertainty of a latent response from variation of a future observed outcome, choose the corresponding prediction, and assess the resulting intervals. Near-zero fitted training residuals do not establish small further-use error. A simpler empirical correction, a justified bound or a direct source call can be preferable.
For a multistep forecast, propagate the joint uncertainty through the update. Sampling a modeled next outcome and feeding it into the next step is one possible rollout; sampling independently at each step would lose a persistent shared disturbance unless that disturbance is retained. Quantiles of such rollouts summarize that model, not all possible failures of it. When the result changes a consequential action, examine the omitted dependence or regime before relying on numerical precision.
Use expert material without claiming to transfer expertise. This optional branch is useful when several people or existing programs already supply diverse workable policies. First agree on the input meanings, available observations, action interface and outcome comparison. Obtain their prescribed actions on contexts that expose both ordinary behavior and consequential differences. Keep permission and confidential information conditions in the collection arrangement. The submitted program or action record is reusable material; the person's capability is not copied.
When the representations are incompatible, fit each policy's behavior into a shared evolvable representation with an applicable fitting procedure, or translate it exactly when the finite cases and semantics permit. Compare the original and translated actions, the resulting outcome profiles and important exceptions before using the translation as a seed. If the original uses unavailable information, repair the common inputs, keep that expert as a separate callable supplier, or preserve the failed approximation. An average imitation score cannot establish preservation of a rare safety condition. Ask the expert or the receiving use which distinctions must survive; use those in the comparison. When the learned translation changes the histories it subsequently encounters, use :4.7 to obtain and fit targets on those histories, then compare the resulting behavior again.
Seed the search with the usable translations and preserve the originals for comparison and possible return. Recombination can retain different specialists in different contexts or combine useful action parts within one context. Mutation can explore beyond the submitted behavior. A mixture that only chooses whole expert outputs cannot express every such combination; it can still be the cheaper adequate alternative. If direct translation already returns a sufficient policy, stop there. Track derivation when the receiver needs to inspect origins, but re-evaluate the child: its ancestry does not establish its behavior, safety or the causal value contributed by an expert.
Return to use, then revise what changed. Return an applicable policy or a qualified set, the observations it needs, its action realization, supported outcome profile, material model assumptions and fallback. The decision maker may choose a trade-off or change an action because of a fact absent from the model. Recheck the changed action's feasibility and modeled consequences where useful. Preserve the reason for the change so that the next learning pass does not mistake it for the original recommendation.
Choose an actual trial or deployment only within its permitted conditions. Observe what was done and what followed; compare that result with the prior prediction and incumbent under the claim being tested. A failed source-model approximation returns to the operation that constructed the substitute and its comparison with the source; a failed account of the phenomenon returns to the model construction and its subject evidence. A changed objective returns to the receiving choice, not automatically to retraining. If a model is revised, re-evaluate retained contenders on a common relevant basis before comparing their scores across versions. Continued policy evolution can reuse material; it cannot make the old and new scores commensurate by itself.
Count model construction, expert querying and translation, policy search, uncertainty calculations, actual trials, oversight and recurring use. Stop with a sufficient known policy, a conditional recommendation or a statement of which action remains unsupported when the remaining search cannot justify its cost or risk. This arrangement develops policies and their consequence models together; repair of what counts as a valuable outcome remains the separate evaluation-development question under E.23 and A.19.ECS.
Try a continuation rule against a bounded recorded history
Use this branch when earlier search has left costly, reusable observations and the immediate question is which continuation or stopping rule deserves a new actual trial. A recorded history can supply a cheaper, restricted response mechanism for :4.8. It does not predict an unrecorded outcome. Prefer a direct comparison when a few affordable trials settle the choice, or when preparing and maintaining the replay would cost more than the trials it could usefully avoid.
Construct the history that the rule can encounter. Recover the initial material and the attempts, with their predecessor states, performed actions, observations, examination results and costs. Preserve the conditions that gave each result its meaning: the performer and available means, evaluation, observation interface, relevant environment and timing. A final leaderboard without this connection cannot reconstruct sequential choice. Mark what was available before each decision separately from what became known only after an attempt.
Choose a finite action interface whose responses the records supply. One simple construction opens root branches in their recorded order and permits a branch to continue only to its recorded next child. The new rule may choose among already revealed branches, request the next root branch, or stop. This changes the allocation and order of recorded attempts, not their content. Other recorded structures can support a different interface, but every permitted transition needs its own corresponding response and conditions. Do not manufacture a result for an action merely because a similar action occurs elsewhere in the archive.
Implement the replay as successive revelation. Start each run with the same initial observations, undisclosed outcomes and reset rule state. Ask the candidate rule for its next action using only the information then available; reveal the matching recorded response, charge its recorded trial cost, update the visible history and ask again. The person or program selecting the action must not consult future outcomes, final branch ranks or a summary derived from those outcomes. Where a recorded limit itself reveals information unavailable in use, do not expose that limit to the rule as a predictive feature. A rule may have been developed using other histories, but its decision inputs still obey the declared interface.
An action outside the supplied transitions, an exhausted branch, or a changed material condition returns no supported response. This is not a failed trial, a zero payoff or proof that continuation is unpromising. Stop the supported replay there, narrow the question, or obtain the missing response through an allowed actual trial or a separately justified consequence model. Do not let one candidate exploit an archive boundary as a free oracle. A partial run cannot be ranked as a completed equal-horizon run without a justified comparison of that missing continuation.
Check what can be reordered. A saved branch state can support independent continuation when the changed order leaves all consequential conditions fixed or represented. If earlier actions change a shared resource, another participant, a future observation or the performer's learned state, simply replaying the old outcome in the new order drops that dependence. Restrict the permitted order, supply an adequate transition model under :4.8, or return to actual trials. Likewise, a batch of separately recorded attempts is not a recorded joint attempt when its members can interfere. Changing the working performer, its instructions, tools or evaluation usually changes the response question; the old history then needs a new justification, not just a new selector.
Compare at the scope the history supports. Fix the history set, initial information, trial allowance, horizon and outcome comparison. Include an adequate incumbent, and run every candidate against the same restricted response mechanism. Keep useful trade-offs between outcome, recorded trial cost and other protected conditions. Charge replay construction and execution, development and selection of rules, examination and the intended actual trials separately from the trial costs taken from the history. Reducing recorded trial calls does not by itself reduce total development cost.
Keeping the incumbent among candidates can ensure that the selected recorded value is no worse on that same fixed comparison. It says nothing by itself about a new search. Repeatedly selecting against a known history can fit its accidents. Use histories held out from rule development when the intended claim needs them, and return the selected rule to a bounded new actual search before claiming improved real performance. An actual result can extend the history; once responses or evaluation change, compare retained rules on a common relevant basis again. E.23:4.1a governs a changed evaluation and the evidence that its earlier choices helped produce.
Worked construction. An organization has archived four independent, unit-cost prototype tests. Opening the first root branch produces A with value 4; continuing A produces A2 with value 9. Opening the next root branch produces B with value 6; continuing B produces B2 with value 5. Values use the same stipulated evaluation, and the useful result of a run is its best tested prototype. Root openings must retain the order A then B. The table below is the replay designer's complete record; a candidate rule sees a row's value only after requesting that row.
With allowance 3, both candidate rules first open A and B. The incumbent then continues the higher-valued branch, B, and returns best value 6 at cost 3. A candidate rule reserves its third test for the earliest opened branch, A, and returns 9 at cost 3. That is a reproducible difference on this record. The candidate's instruction must be expressible without reading the hidden 9; choosing A because the designer already knows that number is not an information-respecting decision. Even an expressible rule may have been fitted to this example, so this result selects a further trial rather than proving its advantage on future prototypes. Stopping after two tests returns 6 at cost 2 and remains an alternative when that result is sufficient. Requesting A3 has no recorded response and cannot be charged as an observed failure.
If the prototypes compete for a shared rig whose temperature changes between tests, the independent unit-cost construction no longer answers the reordered-use question. Retain a permitted order or obtain the affected outcomes. A lower replay score cannot settle that missing premise.
Human self-development. A learner can use an old problem-solving trace to practise choosing the next diagnostic question before its recorded answer is revealed. The usable result is an explicit choice rule, a justified comparison of its choices on the supplied trace, or a question for new practice. Remembering the solution leaks future information; a different trace or an external participant may be needed for that exercise. The replay cannot restore the learner's former knowledge, fatigue or learning trajectory. It therefore does not show how that person would have learned under a different sequence. Test the proposed way on new appropriate work, and keep its support and acquisition costs under C.36.RP rather than treating the person's capability as a saved branch state.
Guide development with pairwise comparisons
Use this branch when comparing two candidates is obtainable at useful cost, while predicting each candidate's numerical outcome is unnecessary or unreliable. A workshop can compare two trial products, a computational search can learn which of two configurations its simulator favors, and an author can ask which of two explanations better serves one reader. The comparison can guide what to develop or examine next. It supplies neither an absolute outcome nor assurance that the preferred candidate is adequate. Use a small direct comparison or an already sufficient way when learning a comparator would add more work than it saves.
Obtain the comparison the next move needs. Fix the candidate meaning, context, criterion, judge or evaluating operation, and intended use of its answer. Comparing two complete policies under a common collection of situations differs from comparing two actions in one situation. Specify which is needed; one favorable local action does not establish a better policy. Preserve separate criteria when their trade-off is unsettled. Keep requirements for permitted trials and sufficient receiving results outside a relative preference.
Obtain initial comparisons from compatible existing observations, an evaluable source model or a competent judge. When numerical outcomes are available, derive the pair labels from outcomes obtained under comparable conditions and retain those outcomes for later magnitude or threshold questions. When only preference is elicited, retain whose preference, about which consequence and under what presentation. A preference is not automatically an observation of effectiveness. When these meanings still need work, obtain the values and consequence account for the receiving choice: whose consequence matters, what difference is preferred and what grounds connect each alternative to it. Retain an unresolved comparison when that required account is unavailable.
Distinguish a decisive preference, a supported tie, supported incomparability and an unanswered comparison. Preserve material disagreement between judges instead of treating it as repeated observations of one common preference. If presentation order may change a judge's answer, compare the pair in both orders. An inconsistent answer exposes that dependence; discarding it from a binary fit does not establish a tie or remove the missing information. Repeated agreement by the same fallible judge can leave a shared error untouched.
Construct and examine an applicable comparator. Represent an input as the two candidates together with the conditions that change their comparison. Obtain an effective learning operation for the obtained labels, chosen family and fitting criterion. For a small finite family of comparison rules, evaluate each on the labelled pairs, minimize the declared classification loss and retain tied rules when their disagreement affects the next choice. A parameterized classifier instead needs its fitting procedure. When the labels come from an expensive source model, qualify any substitute for the response and receiving comparison, and retain a return to the source where that qualification is insufficient.
Choose the output the caller can use: a predicted preference, a probability with a stated interpretation, or an unresolved relation. For a binary no-tie model, exchanging the inputs should exchange the alternatives' probabilities; enforcing this symmetry prevents one inconsistency but does not establish transitivity or accuracy. Add a tie model or keep ties unresolved when they matter. A probability near one half can express uncertainty; it does not by itself establish equal outcomes.
Examine errors on the pairs and conditions the search is likely to encounter, including contenders and consequential exceptions. Hold out candidates, contexts or judges when the further-use claim concerns those new objects: separating pairs at random can leave the same candidates in both fitting and assessment. Check whether the learned relation remains informative after variation produces unfamiliar candidates. A classifier's confidence and overall accuracy are insufficient grounds for dropping an unexamined family.
Turn comparisons into a development choice. A reliable transitive comparison can support ordinary sorting. An incomplete or noisy relation needs a different construction. Keep the obtained and predicted comparisons available while choosing among these useful arrangements:
- For a small set, obtain the missing decisive comparisons directly and retain an unresolved set if they cannot be obtained. Do not convert absence into a loss.
- A binary tournament can cheaply propose a parent or next trial. State how pairs and ties are chosen; with cycles, the order of encounters can change the survivor. A tournament survivor is not thereby better than every candidate.
- When a common reference set is meaningful and pair predictions are affordable, state and use an aggregate to prioritize investigation. For a finite set S of n candidates and binary predicted win probabilities p(i,j), one such score is t(i) = sum of p(i,j) over j in S, divided by n, with p(i,i)=0. It describes modeled wins against a uniform draw from that set. Ordering these scores is transitive even when the pair predictions cycle. This creates a different ranking rule; it neither corrects the pair predictions nor recovers the outcome's magnitude.
- When comparisons themselves are costly and sparse, a fitted preference model can guide which pair to ask about next. A Bradley–Terry model assumes p(i,j)=1/(1+exp(-(u(i)-u(j)))) for a common latent score u. Obtain the scores by fitting the observed comparisons with an explicit identification constraint or prior, using a solver appropriate to that model and the selected identification constraint or prior. Its transitive score structure is an assumption. A prior can produce a numerical ordering across disconnected comparison groups without supplying evidence between them.
For the aggregate construction, give every candidate the same stated reference basis. With m unknown entries in a row, treating each as ranging from zero to one yields the arithmetic interval [s/n,(s+m)/n], where s is the sum of that row's available entries. Overlapping intervals can leave the score order unresolved. These intervals describe missing entries in the proposed scoring rule, not uncertainty about the truth of its known predictions. Changing S or its sampling weights changes the question; recompute affected scores rather than call their change improvement in the candidates.
Locate the reason for a cycle. If the target is the strict order of one fixed numerical outcome, a cycle contradicts that target and calls for a data, condition or approximation repair. If different participants or context-dependent preferences are being combined, a single latent ordering may be the wrong target. Retain the separate accounts or an expressly chosen collective rule. A fitted total order is not a discovery that the disagreement has disappeared. Likewise, equal aggregate scores do not establish equal outcomes.
For several criteria, compute and interpret comparisons separately before any admitted trade-off. A partial nondominance result under the modeled criteria can guide search; it cannot supply an omitted constraint or the distances used by a numerical diversity metric. Preserve different constructions or use a declared diversity rule when those distances are unavailable. A.19.CPM applies only when its declared comparator and input conditions hold; it does not establish a lawful relation from inconsistent predictions.
Choose an informative return and revise the comparison. Name what the next comparison or actual trial could change: the proposed survivor, a decisive missing relation, a model failure, or the supported receiving use. Include promising candidates, disagreement and poorly covered but consequential cases according to that purpose. C.11 supplies the worthwhile local choice among feasible inquiries and current actions. Exploring every pair is unnecessary when the next decision is already supported; excluding every presently disfavored candidate can instead make a shared model error permanent.
Perform the selected comparison or permitted trial and retain what it supplies. Another judge answer adds evidence about that judge's preference. A source-model evaluation supplies that model's response. An observation from actual use can challenge the model of the phenomenon. Feeding the comparator its own previous predictions supplies none of these new grounds.
Add the new compatible observations to the learning material, revise the affected comparator or its target, and compare retained contenders on the same updated basis. If a source observation contradicts a prediction, preserve both the observed result and the obsolete prediction's identity rather than averaging them as equal-quality votes. If the criterion, participant or context changed, separate or reweight the old comparisons only on a stated relation to the new use. A change of the sought result returns to C.40.CD.
Use the chosen parent or retained set to make the next feasible variation through :4.1. A new candidate needs comparison grounds; its parent's rank is not inherited performance. Return the usable material, the supported relative result, the chosen next use and any missing absolute condition. Recognition of this branch needs only an affordable informative comparison; reliance on a learned ranking additionally needs a qualified judging or source operation, applicable learning and comparison assumptions, and evidence for the receiving claim. Stop with a sufficient result, a bounded comparative recommendation or a named unresolved condition when further work is not worthwhile. Exhausting a search budget can justify stopping work without establishing that its best-ranked candidate is sufficient.
Make human contributions usable in continuing development
Use this branch when a person can recognize, demonstrate or explain something useful that the current search does not obtain affordably. The contribution may change an editable candidate, the examples used to construct it, the examination conditions or which branch deserves continuation. It must reach that operation. Use a sufficient existing construction, a direct professional correction or an ordinary comparison when that finishes the work more cheaply.
Choose a contribution the person can supply. Show the candidate in a form that exposes the relevant behavior and permits a consequential response. A person who can perform an example need not be able to write a complete rule; a person who can explain one exception need not be able to implement the whole policy. Name the missing contribution and make its effect inspectable:
These are alternative inputs, not successive stages. Combine them when their results support the same development. A selected demonstration can suggest a rule; a failed rule can expose a missing observation or a better training task. If the required input is unavailable to the later performer, obtain it there or change the candidate. Training cannot manufacture an unobserved fact.
Construct a representation that supports the intended change. For a policy, keep :4.8's context-to-action meaning and available-input boundary. When people need to inspect and change conditional behavior, an executable rule set can be a useful alternative to an opaque parameterized function. Define the input features and units, permitted conditions, actions and evaluation rule. Conditions can compare a feature with a threshold, two compatible features or a remembered value; introduce products, powers or time offsets only when their meaning and cost are understood. A small table or hand-written procedure may already express the needed behavior.
Specify what happens when several rules match and when none matches. Executing the first matching rule, selecting the strongest action and performing all matching actions are different procedures. Preserve priority, aggregation, ties and fallback when they matter. A printed coefficient called confidence can be merely an evolved decision weight; it has no calibrated probability meaning without its own basis. Obtain an executable interpretation and any needed translation, including binding, control and a preservation argument for the relied-on behavior. When access and update costs matter, compare representations under those operations and costs instead of choosing by their appearance.
Obtain a candidate by direct construction, finite enumeration or search over those admitted expressions. In a rule search, start with a usable incumbent and any compatible advice. Vary a threshold or feature, add or remove a condition, exchange whole rules, change priority or combine conditions for the same action. Combining conditions with AND narrows where that rule fires; it does not automatically combine the useful behavior of its parents. Reject malformed expressions and evaluate each child's complete behavior, including interactions and fallback. The local operations must preserve any protected conditions or return their failure; a desirable parentage cannot supply that result.
Separate a hypothesis from a requirement that later variation must preserve. Advice supplied as a mutable seed permits departure. A binding action restriction needs an effective operation that excludes prohibited outputs, a protected part of the representation or examination sufficient for the required claim. A large penalty or repeated demonstration can still permit a violating candidate. If advice conflicts with the current requirement, resolve that conflict with the responsible practitioner before treating either as a training target.
Choose what the rule is being fitted to. Fitting recorded outcomes answers a prediction question; matching an existing program answers a behavior-copying question; obtaining useful outcomes by acting answers a policy question. Use a fitting procedure that constructs the selected learner from the stated inputs and criterion. For an opaque incumbent, query it on relevant inputs and fit a rule set to its returned decisions. Compare agreement with the incumbent separately from correctness against available subject evidence. Queries generated within individual feature ranges can still describe impossible joint situations; keep the query domain and later use meaningful. If original data are unavailable, the incumbent can still supply target outputs, but it supplies no independent ground truth about their quality.
Retain complexity as a separate consideration when a shorter rule set would help inspection or use. Counting conditions or penalizing their number can expose a fidelity–size trade-off. It does not measure whether a person understands the rules. Compare plausible simpler alternatives, and retain consequential exceptions even when they are infrequent. Removing a rule that never fired in a sample preserves that sample's answers at most; the rule may handle a legitimate unseen case. A claimed behavior-preserving simplification needs the corresponding domain argument or bounded comparison. For example, deleting a false condition from a conjunction can make a formerly impossible rule fire.
Return the behavior to the contributor and revise what changed. Present a relevant input, the rule or construction used, the returned action and its supported consequence. Invite the contribution that can change the next move: correct an input meaning, supply an exception, show another action, choose a trade-off or challenge the objective. A natural-language explanation of a rule must preserve its conditions and action selection. A convincing explanation of an approximate copy is still not an explanation of every original decision.
After an edit, distinguish three returns. A change meant only to make the expression easier to read returns to preservation of behavior. A changed action rule returns to comparison of the policy and its consequences under :4.8 or direct examination. A changed purpose or valued outcome returns to the receiving question, including :4.3 when the problem and available ways now need development together. Compare the revised candidate with the retained one on the affected cases and on still-required earlier behavior. Retain a qualified improvement, repair the failed change or restore the earlier candidate. When actual use is appropriate, observe the action performed and its result; the person's approval of a displayed proposal is not that observation.
Prepare material that permits a useful contribution. If the initial variants are too trivial or alike for a meaningful human choice, first obtain better starting material through an available generation or variation operation, suitable reusable examples or a short assisted construction. Automatic preparation may use cheap distinctions such as novelty or structural complexity to escape a nearly uniform initial set. These measures guide preparation; they do not establish what the person values. Present a small prepared set and examine whether the person can now identify a consequential difference or make a useful continuation. If not, revise the preparation or use a simpler sufficient construction rather than spending more comparisons on the same uninformative material. When voluntary participation cannot supply this preliminary work, a bounded paid contribution may be an available alternative. Agree what it must produce and examine the result in the receiving use; payment alone does not establish suitability or shared preferences. Count this preparation and human work in the comparison of complete arrangements.
Make continued participation affordable. Ask for a contribution where it can change the continuation: an unresolved comparison, a poorly supported case, a disagreement between candidates or a failed attempted correction. Show enough context to make that response interpretable. Small distinguishable batches, recovery of an earlier version and controllable variation can let the person inspect differences without repeatedly starting over. Count the person's preparation, inspection and correction alongside computation. A fixed adequate rule, a direct question to an expert or an unassisted search can be the better complete arrangement.
When several people continue development, share the editable or executable material and the conditions needed to use it. Let a recipient resume from a useful intermediate result and branch for a different question while preserving the original. Different preferences need not be averaged into one objective. Usage counts, repeated playback or clicks can suggest an inquiry, but exposure, defaults and interface behavior can cause them; qualify the preference inference before using them as selection evidence. If a person becomes unavailable or fatigued, stop, use a qualified substitute within its scope, or return the missing contribution. A new rater or a learned proxy may change the meaning of the examination.
The useful result is an applied contribution with an examined outcome: changed material, a justified retention of earlier behavior, an improved examination or a supported choice of continuation or stop.
Build the representation together with its changes
Use this branch when the material's form makes a needed variation difficult or unavailable: a shared rule changes too many parts, a grammar cannot express the proposed construction, or repeated changes never leave the same family. Also use it when choosing material for further development requires examining which useful variations it can produce. The useful result is editable material, a way to obtain its candidate result, and changes that expose worthwhile continuations while preserving the conditions that matter. A familiar direct edit or sufficient construction can finish this work without a population search. The operation-choice construction below also applies when the representation is adequate but choosing useful changes remains costly or unreliable.
Compare what can change from the same present result. State the next difference the work needs, which properties must survive the change, the permitted operations, and the available resources. Two representations can yield identical current behaviour and admit different changes. Every permitted operation may preserve a relation that the proposed target violates in one representation. A legal sequence in the other may reach that target within the available budget. The target and the protected properties now supply a reason to choose between the representations. A different target can reverse that choice.
When a change is expressible, distinguish its existence from the ability of the current search to find it. Try a directed change of known effect and inspect the target property and the protected behaviour. If the intended change is reached but protected behaviour disappears, vary the update or its size before concluding that the representation excludes it. If a previously attainable behaviour seems to have vanished, compare access and activation conditions where they can affect observation (E.23.CAE). Reconstruct the needed response under a qualified intervention. Restoration under changed access establishes that the response remains attainable there. It leaves the original cause open if other conditions changed with access.
Connect the stored material to what is examined. Start with one current example, one consequential difference to obtain, and the properties that must survive. Follow the material through the operation that constructs or executes it to the result used in comparison. For a drawing, a parameter record drives a generator and produces a figure. For a procedure, an expression is interpreted on inputs and produces actions or another result. State the inputs, initial state and support that affect that passage. When learning or development is part of it, include that work before judging the resulting candidate. An encoded description, its constructed object and the behavior of that object have different change conditions.
Construct the simplest adequate connection. Store independently adjustable parts directly when local changes matter. Use a shared parameter or generating rule when coordinated change is useful: equality can follow from using one value twice instead of independently varying two values and repairing disagreement. Use a grammar or construction procedure when admissibility depends on how parts fit, giving its operations the bindings, input and output kinds, and stopping conditions needed to produce a usable result. Trace one production through to the result; a short description can still require expensive construction and examination. A fitted or learned generator is another possible obtaining operation, with its training and use conditions. None of these forms establishes the quality of what it produces.
Derive changes from their intended effects. For each important difference, identify which stored part controls it and what else that part changes. Alter that part in a copy, obtain the candidate, and inspect both the intended effect and protected properties. Start with a small change whose effects can be followed. A shared rule may be the right operation for a coordinated change and the wrong one for an exception. Add an independently controlled part or a bounded exception only when that new freedom is needed; retain the relation that keeps the remaining parts valid. For a mathematical family, construct admissible parameters and their joint conditions before choosing how to vary them. A valid parameter value and a change that stays within the valid set are separate contributions.
For recombination, decide what makes two parts corresponding and what lets a transferred part fit its receiving construction. Match by role, binding or retained construction history where it supplies that correspondence, rather than by an accidental position in a list. Carry the required connections with the part, resolve clashes, then construct and examine the whole child. For example, moving a conditional action requires its inputs and action meaning, not just its text. Alignment enables a meaningful combination; it does not establish a competent combined behavior. Keep an unchanged usable parent when the proposed combination still needs examination.
Find where the useful continuation is lost. Attempt a witness construction before enlarging the search. Can the needed result be expressed at all? If so, can allowed changes from the retained material produce it? Exhibit a sequence for the case, or a reason every permitted change preserves a relation the target violates. A finite complete traversal can establish a bounded impossibility; an unsuccessful sample cannot. If the sequence exists, ask whether it is affordable and likely to be attempted, and whether its intermediate material survives long enough to be developed. Finally, check whether the examination distinguishes the useful result. These questions locate different repairs: the representation, changes or starting material, allocation of attempts and retention, or the evaluation. Return only as far as the supported cause requires. When the cause is unresolved, compare a small discriminating case rather than declare the family incapable.
Do not infer that an intermediate candidate is useless because it first performs worse. A new component may require adjustment of its partners before its contribution appears. Give it a bounded opportunity to be tuned or keep a separate branch when the possible continuation warrants that work; specify what result would justify continuing it. This consumes resources and can still fail. Conversely, a hard admissibility condition cannot be waived merely to keep an interesting lineage. A complex construction is worth retaining only through the opportunities and results it supports; neither growth nor minimality is a universal objective.
Change the representation without losing the needed continuation. Identify what must carry over: a particular usable result, a family of results, the means to reconstruct them, or particular future changes. Construct the receiving material from the old one and check the required correspondence. Adding an exception with an initial neutral value can retain the old family while admitting a new difference. Expanding a shared generator into separately editable parts can enable local repair, but may lose the compact operation that changed those parts together. Keep that operation, retain the generating material, or accept the loss explicitly according to the intended continuation. If exact transfer is unavailable, retain the original and compare the consequential loss before replacing it.
Equality of current outputs is insufficient when future changes matter. Two encodings can produce the same drawing while one keeps a shared construction and the other keeps only flattened parts. They can then admit different next operations. Likewise, a temporarily inactive part can carry a future option. Retain such material for a named continuation and its affordable recovery, rather than retaining every duplicate. If the decoder or generator itself changes, re-establish what the retained inputs now produce; unchanged parameter bytes do not preserve their result automatically.
Improve the generating arrangement only when that is the useful next change. A collection of useful examples can support fitting a generator that proposes related material. Fit it to the selected examples, use it to construct new candidates, examine those candidates in the receiving question, and return useful outcomes to the collection. Preserve another affordable way to introduce differences that the fitted generator excludes; learning only from its own familiar outputs can confine later search. Compare the obtained continuations and total work with direct variation. When a candidate carries its own change settings or change procedure, treat those as editable material too and examine their effect through descendants. The procedure that varies material and the rule that decides what survives remain distinct even when both are implemented by one system.
Compare starting material through a trial of further development. A good current result need not be a good starting point for the changes ahead. Use this construction when that difference could change which material to retain or how to generate from it. Begin with the further use: a family of changing demands, worthwhile behavioral differences to explore, or an inner search whose results must improve. A known adequate edit or direct construction can settle the question without learning a more adaptable generator.
Identify the object being compared. It may be an editable starting candidate, inherited settings that govern its changes, a distribution from which candidates are sampled, or a mapping that constructs results from encoded inputs. Keep that basis distinct from the descendants or search results used to judge it. Two bases can give the same present output and different subsequent opportunities.
When a material difference between the bases remains open and worth testing, construct an affordable trial. Preserve or reconstruct its initial material and state; name the permitted changes, relevant inputs or tasks, means of obtaining and examining the results, and work allowance. When comparing different generating mappings, vary their encoded inputs through the declared operation. When comparing distributions, draw the declared samples from each. Make the trial comparable on the receiving question, including generation and examination costs: equal numbers of evaluations can conceal very different work. Retain the starting basis while its trial is in progress.
Decide what continuation the claim requires. Examining one-step changes establishes a local offspring profile. To report what a longer search actually did, recover its selection, retention and subsequent attempts under the declared allowance. Reachability can instead follow from the permitted operations or a limit that no sequence of them can cross. If learning is allowed, perform that learning before judging the resulting behavior. Use :4.7 for this comparison through application and :4.12 for the chosen diversity and retention construction. A successful cloud of immediate descendants does not establish that a longer search can leave that region or adapt to a new task.
For each obtained result, keep its consequential difference together with its quality and admissibility under the receiving conditions. Summarize their joint behavior: which useful kinds were reached by acceptable results, which were reached only by failures, and which were not obtained. Separate repeated outputs and unchanged results from newly obtained differences; a neutral change may still deserve a later continuation under :4.12. Do not combine a favorable average quality from one part of the sample with diversity supplied only by unusable descendants. Include failed trials and consumed work. A complete finite enumeration and a limited sample establish different reach; a region absent from a sample is not thereby impossible.
Use that result to compare the starting bases. If the question is obtaining different acceptable results under a common allowance, prefer a basis that supplies the needed kinds while preserving the required quality and cost conditions; retain distinct alternatives when their supported advantages differ. If the question is one particular adaptation, compare whether and at what burden that adaptation was obtained. Section 4.7 supplies the outer cycle for changing and trying a mapping or other obtaining way. Return the observed result to that basis or distribution before the next trial. Keeping only its best child answers a different question and can discard the very construction that made useful variation available. Useful descendants may separately join a retained collection under :4.12.
Reopen the affected comparison when tasks, admissibility, behavioral distinctions, mapping, variation operators, initialization, retention or allowance changes. Reuse observations only where their conditions still answer the question. A selected basis carries evidence for the tested continuations, not a context-free capacity for future innovation. Stop improving it when a sufficient fixed basis or direct change supplies the required result at lower total burden.
Choose changing operations by their continuing results. Use this construction when several feasible operations can develop the material and their useful effects depend on what happens after the initial change. A sufficient fixed operation or simple fixed mixture can avoid the cost of learning another selection rule.
Decide which alternatives will share accumulated experience. Separate individual operations when their different effects matter. A group of related operations can share an estimate when examining every member separately is too costly. Keep its executable members and a rule for choosing one after the group is selected, such as uniform sampling. Retain the member in the trial observation so that a consequential difference can later separate the group. Splitting an operation by context is another construction; use a context available before choice and enough relevant trials to support the distinction. More categories can leave each estimate too poorly supported to improve the next choice.
Specify what one trial includes: starting material, operation and settings, allowed subsequent tuning, examination conditions, and the outcome available when that allowance ends. A structural change may need its new parameters or partners adjusted. Compare it after the allowed development, retaining its intermediate state while the result is pending. Keep an immediate observation separate from the later result. A completed unsuccessful trial, unfinished work and an inadmissible change call for different responses.
Associate each completed result with that operation and its starting and continuation conditions. Choose the useful response measure: for example, improvement from the parent at a common work allowance. A high child score alone can reward an operator for receiving an already good parent. When later refinement updates the same trial, replace its earlier provisional observation or account for the added contribution; do not count repeated observations of one developing candidate as independent completed trials. This association guides further development without isolating the operator's causal effect from its partners and continuation.
Give the eligible alternatives affordable initial exposure, then state and use a rule to select the next operation and update its comparison after examination. For a comparable numerical gain, one simple arrangement keeps the mean completed gain of each alternative, whether an operation or a group. For each proposal, with probability 1/4 it samples an eligible alternative uniformly; otherwise it uses the largest mean, resolving ties by a stated fixed rule. Execute the operation, choosing its member first when a group was selected. When a completed trial has gain g and that alternative has n earlier completed trials with mean m, set its mean to (n*m + g)/(n+1) and increase n by one. This is a usable heuristic, not an optimal allocation or a guarantee that a rare useful operation will receive enough trials. Its exploration share and trial allowance must fit the search; a large repertoire can exhaust the budget before its members become distinguishable. Other supported selection procedures can replace it without changing the trial-to-result connection.
Reconsider those estimates when parent material, representation, tuning opportunity, group membership, member-selection rule or examination conditions change. Use relevant recent trials or keep distinct comparisons when they answer different conditions; an old average is not automatically useful in a new regime. Remove an unavailable operation from the current eligible set; repair the affected representation or operation if its capability is still needed. Compare the adaptive arrangement with the fixed alternative by the receiving result and total development cost. More promising proposals or more test-passing variants can leave the useful final results unchanged. Continue only while the additional selection and examination work earns its place.
A generator that uses the current input or state can produce different structures during use. Preserve those inputs and the relevant state/reset conditions in the candidate; a single generated structure is not a complete substitute for that way of producing it. Count repeated generation, interpretation, learning and examination when comparing cost. Fewer stored parameters need not make the complete arrangement cheaper.
Return the current material with its construction or execution means, the supported changes, the protected properties, and the grounds for the next use or stop. Section 4.7 supplies comparison through application when the material is a way of obtaining results. The assurance needed for a stronger preservation or reachability claim depends on the receiving use: a proof, exhaustive finite examination and selected trials establish different scopes. If the obstacle is an unavailable observation, operation, permission or capability, obtain that contribution or change the proposal; a richer encoding cannot supply it by itself.
Develop and use materially different continuations
Use this branch when improving one kind of result is eliminating other worthwhile possibilities, or when retained variants still provide no usable next action. The result is material that supports different examined uses or further development, together with the operation that makes those differences usable. Start with a few explained variants when that is enough. Maintaining a large collection, measuring novelty or building an ensemble adds work and needs its own reason.
Choose which differences to preserve. Begin with the use that requires alternatives: handling different situations, exploring ways whose eventual use is unsettled, preserving an intermediate construction, or combining complementary behavior. Recover the conditions and observations that make a difference consequential. Two implementations can produce the same behavior now yet permit different future changes. Conversely, very different encodings can produce the same behavior. Describe the property that matters at the point where the next operation uses it; inspect applications when the property depends on their outcome. C.17 supplies a qualified characterization, and :4.11 supplies the connection between representation and available change.
For example, an image's encoded byte length need not distinguish the visual forms the designer wants to retain. A controller's parameter distance need not distinguish its responses in the situations that matter. A behavior observed on one input supplies that case; it cannot characterize unseen situations by itself. Use trajectories, responses to selected inputs or another obtainable observation when the receiving difference requires them. Retain a consequential rare failure beside an aggregate. If the chosen observations cannot expose the difference, change that examination before using its score to protect diversity.
Then choose how much variety to keep and how to compare like with like. A few named kinds can suffice. A grid over descriptors is useful when its cells distinguish relevant behavior and its resolution is affordable. A distance-based neighborhood can avoid a rigid grid but needs a distance and scale that express the intended similarity. Adding coordinates or finer cells can multiply the material and examination needed without changing a useful choice. Empty regions remain unproduced regions unless a separate argument establishes impossibility.
Compare keeping separate ways with one way that takes the relevant condition as an input and produces an adequate response. A parameterized construction, a goal-conditioned controller or direct enumeration may answer the same need with less retained material. Include its training, planning, observation and use costs. The smaller description is a saving only if that complete arrangement supplies the required differences.
Connect production, local comparison and renewed use. When local competition is worthwhile, construct the following loop with the available professional operations:
- Obtain initial feasible candidates and retain their editable or executable material. Include a sufficient incumbent when one exists. State the behavior distinctions, comparison conditions and quality criterion used in this pass.
- Select retained material and perform a permitted variation or recombination. Selection can draw from another behavior group: useful changes need not remain in their parent's group. If no existing material can support the attempt, obtain another feasible starting candidate.
- Apply or inspect the new candidate to establish its behavior and quality on the declared basis. Assign its group from that result. A requested descriptor is a target for generation, not an observed descriptor.
- For an empty relevant group, retain a usable candidate. For an occupied group, compare with the incumbent there and replace it only when the resulting retained set better serves the declared use. Protect essential conditions and material needed for another continuation. In the simple one-winner arrangement, retain the better local candidate and resolve ties without assuming that equal scores establish interchangeability.
- Use the retained material to choose the next variation, application, reexamination or stop. Keep enough construction, inputs and conditions for that chosen operation to be performed.
This is the useful connection in a MAP-Elites-style construction. A higher global score no longer expels every other kind of behavior. The group's winner remains the best retained candidate under that comparison, with its stated limits. The original MAP-Elites algorithm, §3, supplies that grid arrangement. The same receiving question can warrant neighborhoods, several retained candidates or a direct small set instead.
Selecting parents and retaining results answer different questions. Uniform parent selection gives each retained candidate an opportunity but dilutes that opportunity as the collection grows. Bias toward parents whose recent variations improved the collection can focus work, but that success depends on the current collection and attempted operators. Preserve an affordable way to examine an uncertain region when its possible continuation matters. Retention alone supplies no opportunity if a parent is never selected or all its descendants are filtered out.
Preserve exploration without confusing it with present quality. A low-scoring intermediate may support a useful later change. Protect its material and a bounded opportunity to investigate that continuation. When the eventual use is unsettled, the reason can be an intelligible exploratory question about a region or kind of behavior; it need not assert the value of a particular unknown descendant. Stop when the question no longer warrants its burden.
Novelty-guided search instead favors behaviors different from a declared reference set, which can include the current population and selected history. Specify what observation, distance and reference update make the difference meaningful. Recompute a novelty judgement when its reference changes. A history used to guide exploration can need different retention from the compact collection returned for use. Local quality comparison may be combined with novelty, but maximizing novelty alone supplies no quality guarantee.
Crowding control and separated subpopulations are other ways to limit premature concentration. Their effect depends on the similarity account, opportunities for variation and exchange between groups. A rule that repeatedly replaces a boundary candidate with a slightly better neighbor nearer the center can erode the very range being preserved. Examine the resulting collection, not only each replacement's score. If a promising path is lost, distinguish an unavailable change, an early filter, parent neglect and a retention rule that removes its intermediate. Repair that connection rather than merely increasing the number of attempts. Sections 4.7 and 4.11 develop those returns.
When a population has become too similar to provide useful further variations, restarting around a retained useful result can reopen the search. Preserve that result as a fixed base and construct a fresh population of permitted changes relative to it. For numerical material, obtain each candidate as base plus change, then examine the resulting candidate; a small change value is not itself a good result. Include the unchanged base for comparison and choose the range of changes for the question, with a way to explore beyond small adjustments when useful. After a justified improvement, save the resulting material as the new base before another restart. For other material, supply the edit-application operation instead of assuming numerical addition. This delta-coding construction can restore variation; it cannot repair an unavailable operation, uninformative evaluation or an exhausted useful question. Compare it with an independent fresh start, a different representation and stopping.
Use repeated contraction and regrowth only for a justified exploration question. Another arrangement changes which lines survive by periodically reducing an active experimental population and growing it again. It is worth comparing when existing search concentrates its descendants into too few useful kinds. A restart around a fixed incumbent, continued search without contraction and the direct continuation trial in :4.11 remain alternatives.
Construct the full cycle before attributing a benefit to the reduction. Define the behavior groups, active population, permitted reproduction or variation, population bound and affordable interval between events. One survivor rule chooses a small stated number of occupied groups uniformly without replacement and keeps all their members. This differs from choosing individuals uniformly: group size no longer sets the group's chance of selection. Other rules, such as keeping one representative from each selected group or always retaining a champion, change the opportunities and must be specified as such.
After contraction, produce new candidates from the survivors, examine them and admit them under the stated quality and diversity conditions. Allow the population to refill: a replacement rule that cannot increase its size needs a temporary refill rule, with its admissibility conditions retained. Continue variation for the declared interval, then apply the next survival event to the groups now occupied. If the survivors cannot reproduce or admissible variation cannot refill the population within the allowance, return that failure rather than treating planned regrowth as achieved. The interval must leave an opportunity to generate a spread before the next event; no universal event frequency follows from the construction.
The connection to future variability is conditional. Suppose four groups are occupied and the rule keeps all members of one uniformly selected group. A line with descendants in three groups has survival probability 3/4 at that event; a line confined to one has probability 1/4. Wider occupation changes survival under this rule. For repeated events to favor a continuing ability to spread, the survivors must carry material that can produce another spread during regrowth. One lucky placement, one deletion, or better current fitness does not establish that ability. The argument also changes when group definitions, survival rules or inherited changes change.
Compare the complete cycle by the useful results and total effort obtained, including lost lines and rebuilding work. Where the work permits it, retain useful delivered answers outside the active experimental population; deleting an experimental parent need not destroy an already usable result. Reintroducing saved material or protecting elites changes the cycle and its comparison. When a learned descriptor changes after contraction, recharacterize retained candidates through the following construction before using the new groups. Contraction can change the data used to learn that descriptor, but a changed map alone is neither new behavior nor evidence of improved future variability. Stop or change the arrangement when destruction and regrowth cost more than the continuations they obtain.
Reexamine the collection when its basis changes. Chance can affect both quality and which behavior is observed. One lucky trial can then create a false local winner or apparent coverage. Choose the further observations that could change the retention or use decision and reserve enough effort for them. Keep evidence for the candidate and conditions. Expected performance and repeatability are different receiving requirements; a mean score cannot settle both.
When reexamination changes a candidate's descriptor, withdraw its old current membership, update its supported characterization and place it under the current retention rule. Reconcile any collision with another candidate. Retain the old observation as history when useful, rather than counting it as an additional currently supported behavior. Temporary alternatives within a group can preserve a fallback when an apparent winner fails. Their carrying and rechecking costs count alongside new trials. Repeated observations improve the stated estimate; they do not remove a shared measurement defect or establish behavior in an unexamined regime.
A revised descriptor or learned representation creates a related return. Keep the material and observations needed to characterize retained candidates on the new basis; recompute their descriptors and rebuild the affected grouping and comparison. Merely changing the threshold for future additions leaves old members under old conditions. If the required observation was never retained and cannot be recovered, mark that part unresolved or examine the candidate again. A rearranged map alone is no new behavior or coverage achievement. Keep a means to explore beyond the regularities in the material used to learn the descriptor or generator.
Make the retained difference available to the next operation. Select the receiving branch:
- For a choice among alternatives, return the relevant candidates, their differences and supported conditions; make the receiving choice through its own Method. C.18 supplies an archive or front claim when required. C.19 supplies policy for continuing, retaining, narrowing or retiring live lines.
- For another generation, supply the retained construction and the changing operation. Preserve relationships the next change needs. If two binary choices must sum to one, independently sampling them from the previously useful pairs (0,1) and (1,0) can produce inadmissible (0,0) or (1,1). Sampling a whole pair preserves that known relation; producing a useful new combination requires a further relation-preserving operation. A learned generator similarly needs its fitting and generation Method through :4.11, followed by examination of its proposals. Using more historical material can reduce one loss while increasing fitting, storage and stale-condition costs.
- For one combined way, construct how the components receive inputs and how their outputs produce an action. A vote, numerical average, selection rule and combination of action parts are different operations. Check the resulting behavior: averaging two actions that pass on opposite sides of an obstacle can send the combined action into it. Train or develop a selector through :4.7 when its rule is not already obtainable, then compare the complete arrangement with a sufficient single way under common conditions.
- For learning from retained behavior, supply the relevant situation, actual action and outcome, the rule for selecting useful examples, and an available learning operation. A weak overall performer can supply an excellent local example. The recipient still needs the observations, action meanings and capability to learn from it. Apply the operation and examine the recipient's resulting behavior. Sections 4.8 and 4.10 preserve the additional conditions for expert and human contributions; a transmitted example does not itself transfer a capability.
When learning is part of a candidate's examination, choose what goes into subsequent candidates. One arrangement uses performance after learning to select the original inherited material: descendants constructed from it still need their own learning operation. Another writes compatible acquired changes into the material that descendants inherit, such as trained weights in an editable network encoding. Preserve the original, learned state and inherited material distinctly, and perform the chosen copying or reconstruction. A learned state that cannot be encoded there requires another construction. Section 4.7 supplies the fresh-start and continued-state comparisons; include the learning needed by descendants rather than crediting them with the parent's acquired performance before they can produce it.
For a combined stateful way, follow the history performed. A component that was idle may have stale state; one updated on its own imagined actions may have the wrong history. Supply the real history or a supported state transfer, redesign the component to use available state, or retain the resulting gap. Reinitialization is sufficient only when it preserves the intended use. Test the switching and combined trajectory, not merely each component used alone. Availability of the selector's inputs at the moment of action is part of this construction.
If the components themselves need further development, use their participation in complete ways to guide it. Assemble compatible combinations, perform and judge each whole, and associate its outcome with the participating versions, partners and connection. Compare a proposed component change with the incumbent in common relevant combinations; try other partners when a complementarity question warrants it. A component that performs poorly alone can be useful in such a combination. Conversely, a better isolated component can make the whole worse. An average over its sampled participations is evidence for those combinations, not an intrinsic value or an isolated causal contribution. Reexamine the affected wholes when partners or the selector change. This feedback can favor complementary parts, but does not guarantee that diversity will maintain itself.
Return the material and supported continuation, including a sufficient direct answer when the larger arrangement adds no worthwhile result. Count generation, repeated examination, learning or selector development, storage, recovery and recurring use. Retiring active development can leave inexpensive useful material retained; an obsolete or unaffordable continuation can justify retiring that retention too. The difference made in practice is that a developer can reproduce how useful alternatives survive, how they are used again and which connection a changed condition reopens.
Archetypal Grounding
Develop a recording with or without a composition in view
A constructed workshop has a permitted 100 ms percussion clip, a qualified editor operation for reversing it and playback suitable for examining the resulting sound. The original and editable copy are available. The exercise protects the sample format and permits modification for this workshop.
The practitioner reverses a copy, plays both versions and inspects their envelopes. In the supplied observations, the original has a sharp early peak near 10 ms; the reversal swells toward a peak near 90 ms. The changed recording and reversal setting, rather than the description “interesting sound”, are the reusable result.
If the selected purpose is a sharply marked beat, the reversal fails that immediate use. The workshop can still retain it cheaply for a specified next attempt at a swelling transition; it need not call it the better beat. If no composition is selected, the same examined change can support a question about transition material. It establishes no audience demand or future musical success.
For a target requiring a peak within the first 20 ms, the unchanged reversal fails. Trimming or otherwise adapting it is a new performed operation to examine against the target's duration and sound requirements, not a successful transfer by assertion. If a qualified existing clip already meets the target, use it. Another participant needs the file, its usable settings and the capability and permission for the next operation.
The observations and resources in this case are stipulated teaching inputs. They demonstrate the distinctions, not a field test of the search Method.
Develop an investigation view
A constructed investigation uses six separately permitted synthetic records and a supplied specialist rule that admits comparisons within the same known setup. An available operation groups those records and can report unmatched cases:
A matched-pair view returns E1–N1 and E2–N2. A performed output variation also retains E3 and N3 with their times and missing-reference flags. Its additional review burden is five minutes in the supplied estimate. Retaining that small output is worthwhile for the next question of whether setup records at 09:40 and 11:15 can recover the missing conditions. It has not diagnosed the errors.
If the receiving contribution is unsettled, a complete time-ordered view supplies another small pass: it exposes one unmatched case between the known setups and one after the last known setup. An operator-support use suggested by that result still needs a recipient and need inquiry. The local examination does not have to wait for that answer.
Let a mixed transfer change the target question
A target investigation requires comparison cases to share both setup and temperature regime. A source operation groups by setup alone. C.40.CD:5.1 carries the complete four-record case and the changed construction.
The original grouping returns T1–T2 and T3–T4. The target's temperature condition rejects T1–T2. Adding temperature to the grouping retains T3–T4 for S2/warm and leaves no eligible S1 pair.
The result supports the narrower S2 comparison if useful. An S1 explanation needs comparable cases or a revised question; repeating the old grouping leaves that gap. The source operation retains its earlier qualified uses. The grouping establishes comparison eligibility, with causal explanation remaining a separate question.
An adaptive challenge changes an inspection procedure
This constructed model develops a procedure for inspecting experimental coupons. It is not a production safety rule. A coupon can have a surface defect S or an internal bond defect B. A supplied ideal test A detects S and a supplied ideal test B detects B; neither detects the other defect, and neither rejects a sound coupon. Each test consumes the coupon, so the development budget permits one test per coupon. The provisional aim is a detection probability of at least 0.45 for each admitted defect class, rather than a high average over an unspecified mixture.
The challenge generator prepares a defective coupon after seeing the inspection procedure but before seeing its private random draw. Its permitted responses are S and B. In a physical exercise, specialists would have to make and verify those defects and establish test sensitivity; here those facts are stipulated. The generator seeks the lower detection probability. It may not change a coupon after inspection or invent a defect outside the stated classes.
Start with procedure P-A, which always uses A. A generator producing S yields detection 1. Letting the generator respond to P-A produces B, with detection 0. Changing the procedure to P-B, which always uses B, repairs the latest failure, but a generator now choosing S defeats it. Keeping only the newest challenge makes this oscillation look like repeated improvement.
Retain both responses and compare on those same columns:
The new operation is a private random choice before the test. If A is chosen with probability p, detection is p on S and 1-p on B. Their minimum cannot exceed 0.5 and reaches 0.5 at p=0.5. Thus P-M maximizes worst-class detection among these randomized one-test procedures. This argument supplies the model result; a short lucky run would not establish the probabilities of real tests. An equal-mixture average is 0.5 for all three procedures and would conceal the difference that matters here.
The adaptive search produced two failure witnesses and led to a revised procedure. Once the complete two-class model is known, direct enumeration and the argument above are cheaper than maintaining an evolving population. Retain S and B as distinct regression challenges and retain P-A and P-B only if their single-class uses or construction history warrant it. A physical adoption decision still needs independently prepared specimens, sound controls, actual test sensitivity, costs and evidence about the receiving defect population. Passing those trials would remain bounded to their conditions.
Now change the information available to the generator: it can observe the selected test before preparing the coupon. It can choose B after A and S after B. P-M's detection then becomes 0, even though its unchanged old table still says 0.5. The failed premise is the timing and privacy of the choice. Restoring that separation could restore the model result; if the receiving process cannot provide it, P-M has no supported fit there. Merely adding more copies of old S and B outcomes does not repair the interaction.
A different change admits a third defect C, detectable only by a third destructive test C. Under a single-test budget, the three selection probabilities sum to 1, so at least one class has detection at most 1/3. The original 0.45 per-class aim is now unattainable in this model. The next question concerns another testing operation, additional independently testable specimens, or an authorized change of the aim. It is not another round of replacing the current winner. Earlier two-class evidence remains valid within its old scope.
Transfer a parsing repair and reconsider which task is informative
A publisher develops a checker for a miniature link format. A link is a filename, one literal #, and an exact anchor. The rule is to split at the first literal separator, percent-decode the filename once, then look up the decoded filename and anchor in an independently supplied complete catalogue. The catalogue contains guide.md/intro, Guide One.md/intro, C# notes.md/usage and C%23 notes.md/archive. Directories, external URLs and encoded anchors are outside this model.
Prepare four challenges. Each contains the valid link below and a second link with the same path and anchor missing. Correctly accepting the first and rejecting the second gives two correct decisions. The catalogue, rather than the candidate's output, determines the answers.
These are exact results of the stipulated parsing operations on the eight inputs, not measurements on a production manual. Rejecting every link scores only one on each challenge; the valid/invalid pair prevents that shortcut from looking successful.
Suppose Space is being developed from A while Hash is being developed from B. Both incumbents score one. Inspecting Hash explains B's failure: decoding first creates a # inside the filename, and splitting there changes the target. The programmer constructs C by moving the split before the single decoding operation. Hash now scores two. Test C directly on Space; it also scores two and can replace A there without another independent repair. That target test, not the Hash result, supports the replacement.
A proposed new Plain challenge is already solved and adds no needed distinction, so keep it only if its regression use is worthwhile. Percent is also solved by C. That success can remove it from active repair, but deleting the case altogether would lose a useful distinction: the plausible repeated-normalization variation D fails on it. Retain Percent when such a variation is a live possibility.
The response profiles also explain how the comparison basis changes. Under A, B and C, Space and Percent both have profile (1, 2, 2). That profile alone cannot distinguish them. Including D changes their profiles to (1, 2, 2, 2) and (1, 2, 2, 1). The tasks did not change; the set of tried operations did. Recompute that comparison and retain the difference where it affects further development. Surface novelty such as another filename would not necessarily reveal this distinction.
Now suppose one shared checker C must handle all four challenges, and a proposed change produces D. Retaining C as a specialist elsewhere does not make the changed common checker adequate. Rerun the common checker on the retained tasks: Percent falls from two correct decisions to one while the other three stay at two. That exact counterexample establishes a regression without a noisy-rate estimate. Return Percent to the repair inputs, restore single decoding and rerun the four challenges before replacing the common checker. If restoring it prevents a required new use, identify that conflict and construct another operation instead of alternating two inadequate releases. This demonstrates the shared-candidate return for a software repair; it does not demonstrate learning dynamics or an effective training mixture.
For this fully specified small grammar, directly implementing C and retaining the four discriminating challenges is sufficient. A population of parsers and an automatic task generator would add burden without supplying a missing answer. The construction nevertheless explains what a larger search would have to connect: independently assessable tasks, informative variation, local repair, target-tested reuse and reconsideration of retained challenges. If the catalogue becomes incomplete, a larger search over these same binary answers supplies no missing target fact; return that information condition through C.40.CD. If the grammar changes, reconsider the affected rules and challenges while preserving results within the original scope.
Develop a scheduling rule rather than one schedule
A constructed workshop model has one machine, all jobs available at time zero, no interruptions or setup times, and known processing times p and due times d. For an order, completion time is the cumulative processing time; total tardiness is the sum of max(0, completion − d). Lower total tardiness is the receiving criterion. These are stipulated teaching conditions, not an operations recommendation or measured workplace effect.
The material under development is a rule that constructs an order from job data. Two executable incumbents are S, shortest processing time first, and E, earliest due time first; ties use job number. A third candidate H uses S if any job has p > d, otherwise E. This condition can be evaluated on new inputs, unlike retaining a literal order for one job set. The outer development proposes a rule; each inner application sorts the supplied jobs, computes completion times and returns an order plus total tardiness.
On A, S completes jobs at times 1, 3 and 7; only job 2 is late, by 3. E completes them at 4, 6 and 7, with none late. On C, E's individual tardiness values are 4, 3 and 3; S's are 0, 0 and 6. Across equally weighted A–C, the mean totals are 11/3 for S, 10/3 for E and 2 for H. Thus H is the best of these three on this selection set. No search over an unspecified larger space or evidence about future workloads follows.
Now apply the selected H to a previously unused set D: (4,3), (1,4), (2,7). Its p > d condition selects S, giving order 2,3,1, completion times 1,3,7 and tardiness 4. E gives order 1,2,3, completion times 4,5,7 and tardiness 2. The earlier improvement did not transfer. Once D is used to repair H, it is part of development, not an untouched final examination.
Inspect the failed condition: knowing that one job must already be late does not establish which ordering minimizes total tardiness. A feasible repair Q constructs both S and E orders, computes their total tardiness with the supplied model, and returns the better order, choosing E on a tie. Q returns totals 0, 0, 6 and 2 on A–D. It costs two order constructions and comparisons per use. Its result is guaranteed to be no worse than either supplied order under this model, because it compares those same two exact values. It is not a globally optimal scheduling rule for all inputs. A complete enumeration of the six orders is another affordable alternative for three jobs and can settle each of these cases directly; a larger population search is unnecessary here.
Suppose search is accelerated by judging only jobs completed by time 3. On A, S has completed two jobs and E none, but the final tardiness comparison favors E. That cheap progress signal cannot replace the receiving criterion. It may select trials only with a justified relation to final outcomes and a return when ordering reverses. In this small model the full calculation is cheaper than developing such a predictor. In an expensive simulator, the same distinction can justify a calibrated predictor and occasional complete trials instead.
If execution has already completed job 1, Q cannot select an order that places another job before it. A continuation comparison must hold that performed prefix fixed and order only the remaining jobs. It answers a different question from starting the rule at time zero. If processing times become uncertain or depend on order, the exact calculation above no longer supplies the receiving result; obtain an adequate scheduling model and trial basis before selecting on its predictions. Preserve the deterministic result within its original conditions.
A discovered learning component depends on how it was trained
A neural-network developer can vary the formula used during training while keeping the receiving question about the resulting model. In TaylorGLO, a vector specifies a polynomial training loss; each trial trains a model with that loss, and validation accuracy guides the outer search. The vector's own loss value is not the outcome compared across candidates. Short training makes initial trials cheaper, while later full training examines the selected result. The polynomial representation supplies a smooth search space, not every possible learning rule.
PANGAEA makes a different coupled choice: evolution changes an activation function's computational form, while gradient training adjusts parameters within that form along with the network. Returning only the form and omitting its parameter initialization and training changes the proposed application. Its source comparisons include fixed functions and parameterized functions, with transfers that succeed on some architectures and fail on others. This makes fitting a complete candidate and checking its receiving architecture consequential operations.
EPBT instead continues partially trained models while changing settings and training loss. Its selected final model carries earlier training history. Treat that as a continued-development result; replaying only the final loss from a fresh initialization is a new test. AQuaSurF selects expensive activation trials with a fitted predictor based on function behavior and network-derived features, then updates from evaluated outcomes. Constructing those features also costs computation. These alternative realizations of §4.7 differ in what they vary, what they retain and what their evaluation costs. The linked primary sources in §11 supply their neural mechanisms; the scheduling construction shows what can be retained without those mechanisms.
Combine processing rules, expose an interaction and revise the policy
A constructed document-processing service chooses between two optional operations: a preflight check P and an alternate parser R. Context S denotes a standard template; F denotes a fragile template. The template type is known before processing. Each row below describes a 100-document test batch with reference answers. Time includes the selected operations; errors are remaining incorrect records. The two context types have equal weight in this illustrative comparison. The numbers are stipulated test results, not measured production performance.
The receiver seeks mean time at most 20 minutes and mean errors at most 5 for the two test contexts. These are conjunctive requirements, not weights in one score. Expert A recommends P in either context; expert B recommends R in either context. Query both at S and F and represent each as a two-row action table. Here that translation is exact for the entire declared finite context set. A real expert's behavior outside that set has not been captured.
The first model keeps the measured times and estimates the missing F interaction by adding the separate error reductions. P reduces F errors by 5 and R by 12, so the additive prediction for both is 20−5−12=3. The same construction gives 8−4−3=1 for S, agreeing with its observed joint case. Agreement at S motivates a candidate assumption for F; it does not establish it.
A policy chooses one action at S and one at F. There are only 16 such policies, so enumerate them; an evolutionary algorithm would add needless overhead here. The enumeration expresses the same representational choices a larger search must make: combine specialists across contexts, combine operations within a context, and compare the resulting whole. Five useful candidates are:
C improves both measures relative to B by combining the experts across contexts. Neither selection of A nor selection of B as one unchanged whole obtains C. D adds an untested within-context combination. Under the first model D meets both requirements and dominates E, so it is the tempting selection. E already supplies one admissible policy on the two test batches from observed responses. Return E for that commission. To decide whether the faster D qualifies, compare D with E, or recover every qualifying policy in the named family, the unknown joint response on F now matters.
Suppose the permitted test of P and R on the F batch returns 13 errors. Inspection shows that preflight's rewrites disrupt this alternate parser's handling of the fragile template. The additive prediction missed an interaction of 13−3=10 errors. Add that interaction for F while retaining the already established S responses. D now has (4+13)/2=8.5 mean errors at 19.5 minutes and is dominated by C's 6 at 17 minutes. E, whose two component responses were already tested, meets the requirements at 20 minutes and 4.5 errors. Re-enumeration of the 16 policies shows E is the only one meeting both stated limits. Return E for these batches, with the type-dependent rule and the source test; do not call this production effectiveness or a learned general law of parser interaction.
A direct test of all eight context/action combinations would also have been inexpensive in this toy setting. The surrogate is useful here for exposing where prediction-guided selection needs a return, not for claiming an economic saving. In a large processing family, test cost, the number of context/action combinations and the consequences of a miss can favor different mixtures of modeling and direct trials.
Now suppose template type will not be observable until after the action has been selected. E cannot be applied as written. If the available policy must choose one fixed action for both contexts, neither P, R, both nor neither meets both limits: their respective mean pairs are (15.5,9.5), (18,6.5), (22.5,7) and (11,14). With no permitted extra observation or changed processing operation, return the unmet requirement. The change invalidates the policy's information condition, not the earlier batch measurements. An applicable choice Method supplies the choice conditioned on information available before action; obtaining a timely type check is a new practical contribution to compare with relaxing a limit or changing the service.
Preserve the receiving claim in land-use prescription
In the source application of Young and colleagues (2025), the predictor approximates long-term committed carbon emissions supplied by the BLUE model, using land-use and geographical inputs. A prescriptor maps a cell's context to a new allocation. Its output construction protects primary land and urban area while redistributing the allowed land fractions with their total preserved. Search compares predicted carbon with the fraction of land changed, a proxy for disruption. It returns trade-offs for a decision maker rather than one independently justified social optimum.
The model choice matters to that search. The reported global random forest has lower ordinary test error than the global neural network, but the network is chosen for behavior on larger land-use changes beyond ordinary cases. The paper examines those changes against BLUE qualitatively; this is not a general extrapolation guarantee. The policies are still evaluated through the selected predictor. A prediction about that model's carbon response remains distinct from evidence of actual land conversion, ecological effects, food supply or realized emissions.
Expert material changes what search can readily find. No-change and forest-oriented behaviors are first fitted into compatible policy networks and used as seeds. Their ablation shows a benefit in the reported runs, particularly in the low-change region. Ancestry helps trace the returned policy's construction, but that policy's ecological or practical worth still depends on its consequences. Some simple heuristics remain competitive at very low change; the evolved approach's aggregate advantage does not mean it dominates every location.
Suppose the receiving use now must protect food production. Adding crop-area change as another objective exposes a trade-off, as in the paper; it does not impose a minimum food yield. The practitioner needs the production requirement, a model of the relevant yield and effects, and an admissible-action condition or an explicitly accepted trade-off. Reuse the already useful carbon approximation within its boundary, but compare whole feasible policies under the changed requirement. If no supported yield account is available, return that missing contribution instead of relabeling the old front as food-safe. This differs materially from the finite processing example: the reference model, spatial aggregation, extrapolation and unavailable field consequences determine what can be claimed.
A cyclic forecast chooses a trial, then an observation changes the result
A team varies three rules A, B and C for ordering a batch of inspection jobs. Every rule can be run in the permitted test environment. The receiving question concerns missed defects on a specified batch; fewer misses are better. Earlier compatible batch results supplied pair labels for a classifier. The following forecasts and later observations are invented to demonstrate the method, not measured industrial performance.
For this batch, the classifier gives p(A,B)=0.8, p(B,C)=0.7 and p(C,A)=0.9, with complementary reverse probabilities. Thresholding at one half creates A over B, B over C, C over A. Starting a winner-stays tournament with A against B and then C returns C; starting with B against C and then A returns A. Neither survivor is a justified overall winner.
Using the chosen score in :4.9 with S={A,B,C} gives t(A)=(0.8+0.1)/3=0.30, t(B)=(0.2+0.7)/3=0.30 and t(C)=(0.9+0.3)/3=0.40. The team provisionally selects C for a permitted comparative trial with A and B, because its apparent promise and the cyclic predictions leave a consequential result unsettled. The 0.40 is a modeled average win probability over this set, not a defect rate or confidence that C is usable. A's and B's equal scores do not establish equal quality.
Before that trial, a fourth rule D is proposed. No applicable comparisons involving D are available. Retaining the three-candidate pair predictions and adding the missing entries gives, with n=4, A and B each spanning [0.225,0.475], C spanning [0.30,0.55], and D spanning [0,0.75]. The former score order does not settle the four-candidate comparison. The team retains D as unresolved. Its current test allowance covers the already arranged A/B/C comparison; testing D would need a further worthwhile trial. Missingness supplies no loss by D, and the test's conclusion cannot cover all four candidates.
The comparative trial now produces A:12 misses, B:8 misses and C:18 misses under the same batch conditions. For these fixed results the supported order is B, then A, then C, contradicting two of the classifier's preferred directions. The team adds the measured results and pair labels to its construction data, examines the missed regime, and refits through the selected learner. It uses the observed order for this batch immediately; it does not wait for the new model to reproduce that fact. Reassessment on other relevant batches is needed before relying on its new predictions there. With only three inexpensive evaluations, direct comparison could have been the better initial choice; the constructed numbers expose the inference limits rather than demonstrate a saving from the model.
Finally, the receiving requirement becomes at most five misses on this batch. All three observed candidates fail it. B remains the best of those three observed results, yet choosing B does not satisfy the requirement. D remains unknown. An applicable bound or affordable trial could settle D. If no supported way is available, report what the search has established and which possibility remains open. If an adequate known rule already exists outside this search, using it can end the work. The relative search result and the absolute requirement now lead to different decisions.
An editable rule acquires an exception, then a proposed improvement fails
A workshop is constructing a rule that sends small jobs either to quick preparation or to manual setup. In this finite constructed case, job size n is one of 1, 2, 3 or 4, and a special-material flag is observable before the choice. The practitioner states that special-material jobs require manual setup. Ordinary jobs of sizes 1 and 2 are known to work with quick preparation; sizes 3 and 4 currently use manual setup. The task is to preserve those conditions while making an affordable rule, not to infer workshop performance from a model's confidence.
An existing program supplies four ordinary-job outputs. Searching the five threshold rules quick if n <= t, otherwise manual, for t in {0,1,2,3,4}, gives zero disagreement at t=2. That establishes agreement on the four supplied cases. The rule has no way to express the special-material exception. A high imitation score cannot recover it.
The practitioner supplies the missing condition. The representation is extended to inspect the flag, and the execution rule becomes: first, if special material, return manual; otherwise apply the size threshold. This first condition is protected from the threshold search. Enumerating the eight admitted size/flag combinations confirms the required routing: four special-material cases return manual; the four ordinary cases keep their earlier actions. The construction now implements the stated exception. Successful routing does not establish the quality of the resulting workshop work.
For readability, suppose the ordinary quick condition was written n <= 2 AND n <= 3. Removing its second clause preserves behavior for the stated integer domain because the first implies the second. Removing the special-material guard merely because no special jobs appeared in the original four examples would be a different change. Its first counterexample is a special-material job of size 1.
The practitioner next proposes increasing the threshold to 3 to save setup effort. It changes one of the eight outputs: an ordinary size-3 job now takes quick preparation. The workshop compares permitted actual attempts on that case, retaining the previous rule. In the stipulated trial, quick preparation yields an unacceptable finish; manual setup meets the existing requirement. The proposal is rejected and t=2 retained. This observation does not require changing the already-correct special-material guard or treating the human suggestion as a verified target. If a different material later permits quick preparation, that new observation can reopen its own case.
Finally, remove the flag from the later input while retaining the special-material requirement. The formerly successful rule is no longer executable as specified. Obtain the flag before routing, use an adequate permitted fallback such as manual setup when it is available, or return the missing input. Further fitting of thresholds cannot distinguish jobs whose inputs have been made identical. The affordable answer may be this small procedure; no neural network or population search is required.
Several participants continue a sound without inventing one shared preference
A sound workshop's first generator returns nearly identical single tones, and listeners cannot identify a useful continuation. An available sequencer can combine two sound settings with three pause lengths. The operator enumerates those six combinations as four-bar recordings, retaining their editable settings, and checks that each plays and fits that length. This inexpensive preparation produces distinguishable material, not an aesthetic ranking. In the next listening session a participant can select and explain a promising continuation, so the workshop proceeds with that contribution. If the six combinations remain uninformative, it must change the preparation or obtain a suitable starting example before asking for further choices.
The workshop now has a playable four-bar recording and the editable timing and sound settings that produced it. A listener selects a variant with a longer pause as promising material for a sparse composition. The workshop keeps that version and generates a small batch by varying the pause while retaining the other settings. The next comparison can therefore address the pause rather than a simultaneous unexplained change of tempo, instrumentation and timing. The selected version remains material for continuation, not an established best composition.
Another participant receives the permitted recording, editable material and needed settings and branches from that version for a denser composition. The participants can compare what their variations make possible while keeping their different questions. Sharing only a rendered recording could support listening but leave the second participant without the operation needed to continue this editable construction. More variants cannot repair that missing means.
Playback telemetry initially favors one version. Inspection shows that the interface automatically repeats the currently focused item; the count therefore does not establish the listener's preference. The workshop obtains an explicit comparison of the alternatives under matched presentation. When the listener cannot continue, it retains the versions and unresolved comparison rather than treating another participant's different preference as the first listener's answer.
A later request adds intelligibility of a spoken phrase over the sound. The earlier preference remains evidence about the earlier listening question. It does not establish intelligibility. A relevant listening comparison may reject the favored sound for this new use while leaving it available for the sparse composition. The new result returns to the affected branch, preserving both the editable material and the scope of each judgement.
Separate an inexpressible figure from an unreachable one
A small icon generator draws four horizontal bars in fixed rows. Their integer lengths must lie from 0 to 3, and bars 2 and 4 must remain equal. The current figure is (1,2,1,2); the designer asks for (1,2,3,2). These stipulated conditions make the full construction small enough to inspect directly.
With four independently stored lengths, the target is representable and admissible. Suppose, however, the only available change adds 1 modulo 4 to every length at once. Repeating it visits just four tuples and always preserves equality of bars 1 and 3. More repeats cannot reach the target. Keep the representation and add the needed local operation: change bar 3 to 3 while leaving the others fixed. For general variations, change bars 2 and 4 together and keep each length within its bound. The repair concerns the changes, not missing expressive capacity.
Now the editor instead stores two parameters and draws D(a,b)=(a,b,a,b), with a,b in {0,1,2,3}. Every produced figure has equal bars 1 and 3. No change of those two parameters can express the target. Extend the construction to D′(a,b,c)=(a,b,a+c,b), with integer parameters satisfying 0≤a≤3, 0≤b≤3 and 0≤a+c≤3. The old value (a,b) transfers to (a,b,0), preserving every old figure. The four-bar target remains (1,2,3,2). The new generator produces it from the parameter triple (1,2,2). Changing b still moves bars 2 and 4 together; changing c can alter bar 3 alone. Changes to a and c must respect their joint bound.
The old generator produces 16 figures; the extended one produces all 64 admissible tuples, since a selects the first length, b the equal pair and a+c the third. This finite coverage is a property of the stipulated construction. It does not show that arbitrary random changes sample those figures equally or find them cheaply. Because the desired tuple is known, direct construction already finishes the task.
Change the circumstances once more: c may change only by one unit per step, and selection discards every intermediate with c=1 before it can be changed again. The target c=2 remains expressible and reachable by the permitted operations, but this continuation policy never retains the necessary intermediate. Keep c=1 for one justified further step or provide a direct c=2 change; enlarging the representation is unnecessary. If the new requested third length is 4 while the hard upper bound stays 3, the request itself conflicts with admissibility. Return that conflict instead of treating it as another failure of search.
Preserve both a rendered notice and its next edit
A service desk makes notices from one editable template and a table of place names and opening times. Staff need to correct the time everywhere while keeping the name specific to each notice. The current generator substitutes each table row into the shared template. Its useful changes are a shared wording edit and an edit to one place's data; rendering then shows the receiving notice.
One place now needs an extra instruction. First construct a conditional slot in the template, supplied by an optional instruction in that place's row, and render both that row and an ordinary row. The empty slot preserves the earlier notices; the populated slot exposes the needed exception. Examine line wrapping and the instruction's meaning in the rendered notice. A correspondence between fields alone does not establish a usable result.
Exporting every notice as separately editable text can also supply the exception, but equal current renderings do not preserve the shared next edit. If a later general wording change remains required, retain the template and data or provide a working operation that updates the intended shared text without overwriting local exceptions. Compare that burden with the conditional slot before replacing the material. If the receiving editor supports only flattened text and no such update operation is available, report the lost shared continuation and retain the source separately; importing identical-looking notices does not restore it.
Now staff ask for a notice that changes when a desk closes early, but no timely closure signal is available. The template can express both messages and the generator can select between them when given a signal. The missing contribution is the observation needed to choose the true message. Another parameter, more generated notices or a learned selector cannot establish that missing fact. Use a sufficient static notice for the supported conditions or obtain a timely input before claiming the dynamic behavior.
Keep different tiles, then change what distinguishes them
Consider a stipulated design exercise with four adjacent cells, each black or white. Represent a tile by four bits, with 1 meaning black. The available variation flips one cell. For this exercise only, quality is the number of neighboring pairs with different colors. It ranges from 0 to 3; it makes no claim about people's aesthetic preferences. The designer initially wants a good sparse tile, with at most two black cells, and a good dense tile, with three or four.
The retained sparse tile is 0101. It has two black cells and three transitions, so its quality is 3. Flip its first cell to obtain 1101: three black cells and two transitions, quality 2. Global winner-only selection would discard it. Local comparison puts it in the previously empty dense group. The returned material now supports both intended kinds, although the second tile has a lower score. The bits and flip operation make both candidates available for continued editing.
From 1101, flipping its last cell gives 1100: two black cells and one transition. It belongs to the sparse group and fails to replace 0101 there. Its parentage does not keep it in the dense group. This failed local replacement supplies no reason to discard the retained dense parent. The designer can choose a tile now or continue from either retained tile if another useful difference warrants the work.
All sixteen tiles can be enumerated cheaply in this exercise. Doing so finds maximum quality 3 in the sparse group and 2 in the dense group. It also gives several tied candidates. Direct enumeration is therefore a sufficient alternative to an ongoing evolutionary search here. The example explains the local retention connection without claiming a need for expensive search.
Now change the receiving request: the two kinds are tiles whose first cell is white and tiles whose first cell is black. Reclassifying the retained 0101 and 1101 puts one in each new group; the observed tiles themselves have not changed. Their prior black-count groups no longer answer the new request. The two retained qualities are still 3 and 2, but the dense-group winner's previous qualification does not show that it is best among first-cell-black tiles. The available enumeration finds 1010, quality 3, in that group, so 1101 is replaced for this receiving use. Retaining 1101 for some future density-sensitive use requires that separate reason. A note saying only “two groups filled” would hide this changed comparison.
Two good encoders need one real output history
Suppose a stipulated stream writer accepts normal and bulky jobs. Two available encoders produce valid payloads: A uses four payload units for a normal job and twelve for a bulky job; B uses eight and five respectively. These exact sizes are given for this finite exercise. The receiving requirement is at most six payload units per job. Job kind is observable before encoding, so choosing A for normal jobs and B for bulky jobs supplies the required payload sizes; neither encoder alone meets both limits.
The complete stream protocol also requires headers X, Y, X, Y in order. Each original encoder maintains its own alternating header state, initialized at X. Each works correctly alone. A selector that sends the first normal job to A and the next bulky job to B produces X followed by X, violating the protocol. The two good individual performances did not supply a correct combined operation.
In this exercise the developer may separate each encoder's payload operation from its header generator. The available atomic emission operation reports either a successful complete output or failure with no output. Construct one stream writer that owns the next header, starts at X, selects the payload encoder using the observed job kind, emits the header and payload, and toggles the shared header only after an emission succeeds. On a normal job followed by a bulky job it produces X with four payload units and Y with five. Examine later alternation and a failed emission too: an unperformed output must not advance the stream history. The retained encoders become usable together through this additional construction and its checked conditions.
Suppose further development produces A-prime, using five payload units for a normal job and ten for a bulky one. Its standalone total over the two jobs is fifteen, smaller than A's sixteen. In the arrangement with B and the same selector, however, substituting A-prime raises the total from nine to ten. Both combinations satisfy the per-job limit; when smaller total output is preferred, retain A for this combination. That feedback concerns A's role with B and this selector. A different partner or selection rule can change the answer.
Now suppose job kind becomes available only after encoding. The previous selector can no longer use it for that action. A classification or trial-and-buffer arrangement could supply another candidate if its observation, delay and cost fit the receiving requirement. No such operation is stipulated here, so the selected combination remains unsupported under the changed condition. More retained encoders or more training examples alone supply no timely observation. Return that missing operation or a different sufficient whole way.
Obtain corrections where the learner actually arrives
A small generator emits opening and closing parentheses, then stops. The stipulated task accepts a nonempty balanced string of at most four symbols. An available training demonstration is (): its next outputs are ( at the empty prefix, ) after (, and stop after (). These examples contain no target after ((.
In a permitted simulation, an imperfect trained generator sometimes emits a second opening symbol and reaches ((, where it stops. Its result fails the task. Replaying the remaining suffix of the original demonstration would append one closing symbol and stop, leaving ((), which also fails.
The simulator can retain the exact prefix, and a checking construction can count unmatched opening symbols. At ((, that construction supplies the target ); after fitting this target, another simulated run can reach the new prefix ((). Obtain ) there and stop at the resulting (()). Fit those targets with the still-required earlier behavior, then examine complete generation again, including (), recovery through (()), premature stops and the length limit. The targets refer to the learner's reached prefixes, not to positions in a different demonstration. They establish corrections for this finite exercise, not a guarantee for every generator or longer language.
The teacher can itself need preparation. Suppose its response at (( is selected from the continuation strings ) and )), but its current table chooses ). In this finite exercise, the developer can evaluate both completions and change that table. Appending ) produces the invalid ((); appending )) produces (()), which passes the supplied check. Select )) for that teacher entry, use its first correction for the learner at ((, and examine the learner's subsequent completion. The check and editable table supply the teacher-improvement operation. If either is unavailable, the original demonstration alone does not obtain this prepared teacher.
A direct counter or a fixed generator already solves this small task. Use one when it is the receiving problem. The example exposes what an ongoing learner-development arrangement must obtain when it is being retained for a broader justified use.
Now remove the distinguishing input: at the decision point, both (( and (()) are reported to a stateless learner as the same present symbol, and neither the prefix nor earlier observations are recoverable. The first case needs a closing symbol next to finish within four symbols; the second must stop. Training on more copies of that same input with both targets supplies no reliable distinction. Restore an adequate observation or retained state, retain the checking construction as an operating supplier, or leave that proposed learner unsupported. If the correction supplier itself cannot handle ((, its competence on () does not fill the missing target either.
Compare an operation after its allowed refinement
A stipulated development exercise has two executable changing operations, T for local tuning and G for changing a grouping. Each trial starts from comparable material with quality 10 and receives the same two-unit refinement allowance. All intermediate constructions in this initial case remain admissible. Higher final quality is preferred, and the trials' gains use the same comparison scale.
Choosing only from immediate quality would discard G before its useful continuation. Complete the agreed trials, retain their operation and refinement conditions, and initialize the two mean gains at 2 and 6. Under :4.11's one-quarter uniform exploration rule, the next eligible proposal uses G with probability 3/4 + (1/4)(1/2) = 7/8 and T with probability 1/8. The choice still needs the random draw and execution; the probabilities are not produced candidates.
Suppose another comparable completed G trial returns gain -2. Its mean becomes (6 - 2)/2 = 2, tied with T. With the stipulated tie rule favoring T, the next probabilities reverse. Retain the individual outcomes: the mean alone conceals G's variable result and supplies no guarantee of a further gain. If the receiving requirement had already been satisfied by T at 12, the extra search might have been unnecessary.
If G instead denotes a group of two executable changes, the group choice still needs a member selection. With uniform choice inside G at the initial means, each member receives probability (7/8)(1/2) = 7/16; T receives 1/8. A completed member trial updates the group's mean while retaining which member ran. This shares sparse experience but can hide a useful difference between members. Split the comparison when that difference can change the next choice and enough examination is affordable; renaming two changes as a group did not make their effects equal.
Now a changed construction rule makes G's intermediate grouping inadmissible. Remove G from the eligible operations until a permitted version exists; a later score of 16 cannot waive that condition. If instead the available refinement time changes, the old final-gain comparison answers the old allowance. Obtain the affected comparison or retain its uncertainty before using it to allocate the next work. These are different returns from merely preferring the currently larger mean.
Reaching a target does not establish persistence
A stipulated three-cell process has state (left, centre, right), with each cell either 0 or 1. All cells update simultaneously from the previous state; positions outside the three-cell row are fixed at 0. Each cell reads itself and its immediate neighbours. The developer must obtain 111 from the seed 010, preserve it during continued execution, and recover after the centre is removed. In this exercise the three bits are the complete state and can be copied.
Compare two available rules. G outputs 1 when its neighbourhood contains at least one 1. P outputs 1 when the neighbourhood contains one or two 1s, but outputs 0 when it contains zero or three. Both reach 111 from 010 in one update.
A first-attainment test cannot choose between the rules. Return the reached 111 as another trial start while retaining the seed 010; continuing from those starts exposes P's alternating 111 and 101. Running longer from the seed with repeated checks would expose the same failure without a pool. Under G, 111 is unchanged by another update, which also supplies a direct persistence argument for this deterministic case.
Now remove the centre of 111, obtaining 101. Both rules restore the target in one update. Continue: P destroys it again, whereas G preserves it. A one-step damage test would therefore miss the consequential difference. A complete comparison of these two rules is enough to select G for the stipulated use; no neural training is needed. The example explains the trial construction and selection, not a claim that learning will find such a rule in any search space.
Finally remove all three cells. Both rules leave 000 unchanged. More generally, if an empty neighbourhood is forbidden to create a 1, simultaneous application cannot leave this all-zero state. Recovery now needs a new seed or another permitted operation. Repeating the same trials cannot supply either.
Choose a basis by the useful variants it can produce
A designer uses three on/off lamps to make indicator patterns. For the present family, the two outer lamps must agree. Two available representations have the same three editable input bits (a,b,c). I displays (a,b,c) directly; S displays (a,b,a), sharing the outer control and leaving c inactive. Both currently store 000 and display 000. The designer wants different acceptable patterns from one inexpensive change, rather than one already specified target.
Use :4.11 to compare the mappings at this starting input. The allowed operation flips exactly one input bit in a fresh copy. Enumerate all three choices; constructing and inspecting each display costs the same in this stipulated case. A trial output that violates outer equality is observable in this offline examination but cannot be returned as an acceptable indicator.
Each mapping produces three distinct trial outputs, so counting distinct trial outputs cannot choose between them. I supplies one new acceptable pattern; S supplies two. S's unchanged third output is acceptable but adds no new pattern. The joint comparison therefore favors retaining S as the mapping for these one-step variations, with its input and changing operation. The designer can also retain 101 and 010 as usable results; keeping those displays alone would not preserve the shared generating operation.
Now the next use permits asymmetric indicators and requires 100. Under I, flipping a alone produces it. Under S, the first and third output positions are equal for every input, so no sequence of its permitted input flips can produce 100. The earlier trial is still a correct observation of the old use, but it no longer supports preferring S for this one. Return to the mapping: select I for the direct change, or add and examine an independent outer control if preserving S's shared operation also matters. More attempts under unchanged S cannot repair this limitation.
This finite case compares three specified changes completely. It neither estimates untried tasks nor proves that a sample of good immediate variants supports a longer adaptive search.
Bias-Annotation
A strong incumbent can suppress useful intermediates, while novelty can attract effort without a receiving use. Examine what each variation permits next and include retention and opportunity costs. Keep the target's own conditions visible when enthusiasm for a source success would otherwise decide the transfer.
Conformance Checklist
-
The material, operations and grounds support the kind of work commissioned. Performing a search requires usable material and means of varying it. Describing a design or recovering past work requires grounds for that claim.
-
The result distinguishes a proposed variation, an available or derived consequence, and a change actually made and examined, as the receiving use requires.
-
Fixed-contribution and unsettled-contribution uses do not require one another.
-
Retained material and conditions make the stated continuation possible at a warranted burden.
-
Coupled use constructs meaningful problems and judges transfer on the target's own conditions.
-
A mixed or failed transfer changes the next problem or way question as its evidence permits.
-
The return supports the commissioned use. A known option can finish an order for one way. An order for a search design, a performed search or an account of what was tried is complete when its own result is supported.
-
Archive, front, selection, capability and permission claims keep their separate meanings.
-
When the candidate is an obtaining way, the comparison identifies its application and produced result, any approximation or inherited state, and the conditions of the receiving-use claim.
-
When learning changes the histories later encountered, targets correspond to those reached histories and the learner's available inputs; missing teaching or information has an explicit return.
-
For persistence or recovery, continuation states are realizable and compatible with the rule; original-start ability and functioning after recovery receive the trials required by their claims.
-
Adaptive operation choice supplies execution after selection, a result-update rule and the continuation conditions of its comparison.
-
Comparing future development returns a joint account of difference, quality and admissibility to the starting material or mapping; its trial supports only the continuation performed.
-
Repeated contraction specifies survival, feasible regrowth and the next event, with loss and total work included in comparison.
-
For model-assisted policy search, action realization, model predictions, actual outcomes and translated expert behavior have separate grounds; the returned policy retains the observations and conditions it needs.
-
For comparison-guided development, the pair evidence, learned prediction, derived order and actual receiving result retain their different grounds; missing pairs and changed conditions lead to an explicit next inquiry or bounded stop.
-
When a human contribution changes development, its input reaches an executable variation, learning signal, examination or branch choice; the resulting behavior is checked at its receiving use. Rule readability, copying fidelity, human approval and actual outcome remain distinct.
Common Anti-Patterns and How to Avoid Them
Consequences
Useful intermediate results can support later attempts without being promoted to current winners or final opportunities. The cost is generation, examination and recoverable retention. Coupled development can expose a better next problem, but trying every transfer can consume more than it returns. Small justified passes and sufficient-result exits limit that burden.
Architectural Rationale
The common operation is changing material, examining the difference and choosing a warranted continuation. These development steps serve both a selected contribution and exploration before a destination is settled; the examination question changes with the intended use.
Coupled development adds problem construction and target-tested transfer to that common work. C.40.CD explains their dependency: a changed problem can require another operation, while a partial result or new construction can change the next useful question. Its separate entry preserves the original receiving requirement through those changes.
C.39 remains separate because an obtaining explanation can finish before any variation is feasible. An archive or a continuation policy likewise has a useful independent result, but neither supplies the changing operation.
SoTA-Echoing
The Lila research synthesis proposes connecting generation, recognition of promising ideas and subsequent investigation. Sections 4.7 and 4.11 supply separate tests for those contributions. Creative Priming changes how a fixed responder is prompted. The Value of Not Knowing varies a generator's training and then adds a knowledgeable reviser. Unified Factored Organizers changes the training pressure on a network's internal representation. Lehman and Stanley study how a later change interacts with the existing representation and update. These experiments justify testing different components when current outputs agree. Evidence that the complete research practice sustains further development also requires an attainable next problem, means of investigation and a consequential use of the result.
These bounded source contributions connect the shared operations with conditional problem construction and target testing. Task generation for an expanding repertoire and curriculum generation for a fixed receiving use have different selection grounds. Later generator, representation and transfer-scheduling variants do not make one implementation universal. A broader claim of present algorithmic superiority would require its own current comparison.
For comparison-guided search, adapt DNESG (2023), §§4.1–4.2.5's learned pairwise comparator and tournament selection and CRSEA v2 (2025), §§3–3.3's population-relative aggregation and return to actual function evaluations in :4.9. The latter makes a usable derived order; it does not validate every comparison, recover distances or establish equality from a score tie. Direct comparisons and numerical surrogates remain alternatives when they answer the receiving question more cheaply. Duel-Evolve (2026), §2 and Appendix A.1 supplies a current sparse-preference alternative: fit a Bayesian Bradley–Terry model, direct comparisons toward plausible contenders and generate from selected parents. Adapt the explicit comparison acquisition and revision, while retaining the model and approximation assumptions. Its same-model judgements can amplify shared bias (§5); more self-comparisons do not independently validate a generated answer. Reopen when judge, context, reference set, missingness, tie meaning or the cost of actual evidence changes the useful construction.
For human contributions in :4.10, adapt the task-dependent choice among demonstrations, advice and environment shaping in Karpov, Valsalam and Miikkulainen (2011), §§4–7. This historical implementation shows different useful entry points; its sixteen-participant study does not rank them for every practice. Chapter 8 of Risi, Tang, Ha and Miikkulainen's Neuroevolution: Harnessing Creativity in AI Agent Design (2026) adds shared continuation and participation burden. Its §8.5 also develops preparation of initially uninformative material before useful human continuation. The study by Lehman and Miikkulainen (2013), §§2.3–4.2 supplies the historical image-generation and paid-selection construction; :4.10 retains the preparation-to-human-examination connection without adopting a universal aesthetic measure or a market implementation. Preserve those operations while testing the contribution in the receiving use.
For inspectable policies, adapt the executable grammar, variation and explicit action-selection alternatives of EVOTER, §§3 and 6. Direct construction and ordinary programming remain alternatives. The newer NeuroRule (2026), §§3–4 and 7 develops fitting rule sets to an existing network and balancing agreement with condition count. Adapt that obtainable copying method; qualify the query domain, lost behavior and distinct subject correctness. Both sources leave a convenient expert-intervention facility as future work. Their transparency and complexity claims do not establish human understanding, an effective intervention, fairness or safe deployment. A changed action-selection rule, consequential rare case or unavailable human contribution reopens the relevant comparison.
For representation and change in :4.11, adapt the connected representation, correspondence, developmental opportunity and incremental-construction argument of NEAT (2002), §§2.2–3.4. Its historical innovation markers align inherited structure; matching is not behavioral equivalence. HybrID (2009 manuscript), §§2–4 supplies an alternative: transfer generated structures to a direct encoding for local adjustment. Its difficulty can be finding an expressible exception, rather than an inability to represent one. Adapt that distinction and the explicit transfer; retain direct construction where it is sufficient. Neither historical benchmark comparison is a general superiority claim.
The newer line can develop the generator and its changes as well as its outputs. DDE (2020), §3 and Discussion connects a learned encoding, archived examples and a mixture of exploratory operations. Meta-evolved NCA (2024), Methods and Discussion evaluates a developing decoder through an inner search over encoded inputs; its learned regularities can also limit that search. Self-Referential Graph HyperNetworks (2025), method and Discussion lets inherited parameters control the variation of copied networks while policy outcomes guide selection. Its demonstrated parameter variation leaves architecture mutation as future work and incurs substantial generation cost. Adapt the supported feedback and change relations, while choosing implementation for the receiving use and examining any claimed improvement. Reopen when a changed generator alters retained meanings, loses a needed continuation, or costs more than the direct alternative.
For useful retained diversity in :4.12, adapt the production, local comparison and archive-reuse connection in MAP-Elites (2015), §3 and Cully and Demiris's modular QD account (2017), §III. The latter separates collection organization from parent selection and repairs an archive replacement that can erode the retained range. These historical algorithms supply constructions, not a universal requirement for grids or behavioral distance. AURORA, §IV, extends the line to learned descriptors and rebuilding the affected collection. Extract-QD (2025), §§2–3, adds budgeted reexamination, removal and reinsertion under uncertain quality and descriptors. Adapt those returns; the numerical parameters remain implementation choices. Lin et al. (2026), §§2–3 and 5 supplies a different set-optimization construction when differentiable quality and descriptors are available. A finite direct construction, or one conditioned way with its needed planner as in Batra et al. (2024), §§2 and 4, remains a cheaper alternative. The controller's storage bound does not establish the whole planner's cost. Reopen when changed descriptors, noise, lost intermediates or complete cost changes which collection and continuing operation are useful.
For use of a population, adapt Pardoe, Ryoo and Miikkulainen (2005), §§2–3 and 6's ensemble construction, its state/history limit and the proposed development of components by outcomes of their combinations. Adapt Tansey, Feasley and Miikkulainen (2012), §§2–3, 5.2 and 6's learning from locally useful examples with an explicit acceptability rule, learning groups and choice of whether acquired weight changes are inherited. Its foraging result does not rank inheritance arrangements for every changing environment. These historical neural implementations establish neither universal ensemble benefit nor transfer of a person's capability. Sections 4.7, 4.8 and 4.10 provide the relevant state, expert and human contribution conditions; :4.12 connects retained differences to the operation that uses them. Reopen when selector inputs arrive too late, component state no longer follows actual actions, teaching suppresses useful variation, or one sufficient way makes the combined arrangement unnecessary.
For restarting a converged search, adapt the saved-base, change-population and repeated-rebasing construction in Gomez and Miikkulainen, technical report AI96-248, §4.2, the earlier report behind their 1997 incremental-evolution treatment. Its ESP realization applies changes to neuron weights and permits occasional larger changes; the numerical distribution belongs to that implementation, and its experimental gains retain their tested conditions. Independent restarts, changed operators and a sufficient direct construction remain alternatives. Reopen when the retained base, allowed changes, comparison conditions or cost no longer support this continuation.
For learning on the histories produced by a developing controller in :4.7, adapt the iterative estimator-training construction of RMA (2021), §III. Its privileged simulation targets and history-conditioned deployment have different available inputs; its timing and disturbance limits remain implementation conditions. The newer SCOUT preprint (September 2026), §§3 and 5, Appendix A supports conditional teacher development on learner histories. Adapt that feedback connection within its access, outcome-checking and resource limits; long-horizon agent use remains unestablished.
For operation choice in :4.11, adapt SNAP-NEAT (2012), §§II–III and V's connection between initial examination, continued structural development and later operator credit. Its operator repertoire and numerical settings are historical realizations. Hanna, Blot and Petke (2025), §§4 and 11 supplies a newer contrasting use: improved intermediate repair outputs did not establish more bugs repaired than uniform selection. Retain that receiving-result comparison. Reopen the allocation when delayed effects, changed conditions, sparse examination or total cost defeat its useful continuation.
For developing persistence and recovery in :4.7, adapt the repeated-use and state-return construction of Growing Neural Cellular Automata (2020), Experiments 1–3. Its longer-trajectory alternative explains the memory trade-off behind pooled starts; the pool does not establish recovery from arbitrary damage. Horibe et al. (2022), §§2–5 connects discovery of a functioning body, learning from original and damaged starts, and simulated function after recovery. Adapt that connection and its observed separation between shape similarity and locomotion, without attributing the whole comparison to optimizer choice. A sufficient direct rule and long trials with repeated examination remain alternatives.
The newer Smart cellular bricks study (2026), Damage detection and recovery and Discussion adds an ambiguity: damage can make the intended shape indistinguishable from another legitimate shape; a supplied target embedding changes that information. Adapt this return to the missing target and means of repair. Its physical shape-recognition results and simulated directional detection/regrowth have different grounds. Moving and attaching physical material remain additional operations. Reopen the development arrangement when state compatibility, missing target information, tested disturbance range, actual function or complete cost changes the supported continuation.
For comparison through further variation in :4.11, adapt the joint quality/behavior account in Tarapore and Mouret, Evolvability signatures (2015), §3, the distribution update from sampled offspring in Evolvability ES (2019), §§3–6, and the explicit quality/variation comparison in Quality Evolvability ES (2021), method and main experiments. Their simulated tasks and descriptors do not establish universal adaptability. The meta-evolved NCA construction above supplies the distinct inner-search trial. Reject equating diverse retained answers, immediate offspring variety and useful further search. Direct construction and a sufficient fixed mapping remain alternatives. The Cost of Becoming (September 2026 preprint), §§3–7 adds a bounded trade-off among generativity, viable variation and inheritance in one compact controller simulation; equal genome dimensions and evaluation counts do not equalize mapping cost or the initial phenotype distribution. Its excluded lifetime-learning experiment supplies no adaptation evidence. Reopen when the task family, descriptors, permitted continuation or total cost changes.
For repeated contraction in :4.12, adapt the survivor–regrowth–next-event relation from Extinction Events Can Accelerate Evolution (2015), Methods and the divergent-search study (2015), §§3.3–6. The first keeps all members of five selected occupied niches; the second uses ten individual survivors and a temporary refill rule. These are different realizations, not interchangeable settings of one unstated algorithm. AURORA-XCon (2025), §§2–6 adds contraction with a protected elite and a learned behavior descriptor whose training distribution can change; it does not directly train the genotype-to-phenotype mapping. Adapt that consequential descriptor return, without claiming universal superiority or measured acquisition of future variability. Continued search, direct offspring trials and an adequate fresh start remain alternatives. Reopen when survival, grouping, regrowth, retained answers or rebuilding cost changes the useful effect.
Relations
-
C.40.CD constructs problems and obtaining ways together through examined target use; B.5.QD and C.39.RO supply question and operation construction.
-
C.39 develops a missing obtaining explanation for a local contribution; it is not a compulsory preliminary stage.
-
C.18 distinguishes archive, front, lineage and possibility-space claims about the developed material.
-
C.19 governs resource-bearing continuation of a live pool; C.11 governs a local choice.
-
C.38 constructs a finite same-result comparison from complete-enough ways.
-
C.17 characterizes generative novelty and value when those claims are needed; the applicable professional Method supplies the variation and examination.
-
E.23:4.1a and A.19.ECS:4.5 govern a justified change of evaluation; §4.5 here develops interacting challenges while preserving the comparison question.
-
Retain the sources and support needed for reliance on observations or transfer. Use A.10 when their source-to-use account remains implicit or must be replayable. Problem, capability, permission and actual Work claims retain their direct subject rules.
Opportunity and continuing development. C.40.OCN constructs a bounded opportunity before a useful contribution is settled; its inquiry, characterization, search, direction, reachability and requalification Methods can also be used independently. C.40.OD connects growth of knowledge, understanding or Methods with reproducible results and further attainable problems. C.40.JD addresses the narrower joint construction of a way, its holder and provision; C.40.NP begins from a qualified result and constructs its useful continuation. Ordinary work or acquisition of an existing Method needs no open-ended-development claim.
C.40:End
Develop Problems and Ways of Solving Them Together
Type: Method pattern Status: Stable Normativity: Normative
Problem frame
Use this pattern when developing a problem changes the way worth trying, and trying that way can change the next useful problem. A construction fails after a condition changes. A new operation makes a previously inaccessible question approachable. A partial result exposes a smaller answerable question while leaving the original need open.
The subject is the coupled development of problems and obtaining ways. Keep the sought result, conditions and available operations recoverable while changing them. The first useful result is an examined attempt that supports a receiving use or identifies the next worthwhile question and the operation it needs.
First useful move. Take one result, failed step or changed condition. Say which part of the problem it affects. Work out what the retained conditions already imply. Try a revealing case when its outcome could change the answer or the attempt itself is commissioned. If a conflict prevents a two-session assignment, its structure can support a question about the least number of sessions; the original two-session requirement remains visible.
This Method is for practitioners who can perform, obtain or construct the relevant subject operations. It connects those operations across problem changes. Use an adequate existing answer directly. C.40's ordinary variation is sufficient when only the material needs changing under an unchanged question.
Problem
Work can keep attempting a solution after the question's premises have failed. It can also make an easier question look like the solution of the original one. Conversely, a successful construction can remain confined to its first use because nobody develops the questions its new operation can now answer.
Problems and ways therefore need returns in both directions. The problem determines what a way must obtain; the attempted way reveals missing conditions, useful partial results, limits and further possible uses. Productive development preserves those connections through each change.
Forces
Solution
Recover the problem and an available way, construct a consequential change, establish what the new conditions imply, and use that result to choose the next question or use. Try a target case when its response can resolve the remaining question or the trial is itself owed. Keep the original receiving need connected to any replacement or subsidiary question.
Recover what is being sought and what can already be done
Identify the result being sought, the conditions it must satisfy and the material or situation to which it applies. Distinguish a result someone needs from a constructed challenge used to develop a method. Either can justify inquiry; they give different reasons for continuing it.
Recover an available way far enough to find its inputs, operations, output and supporting conditions. A worked case may be enough. B.5.RC recovers the construction; B.5.TU makes an available theory usable. Start at the point affected by the new question rather than reconstructing an unchanged whole.
Keep any protected condition explicit while it matters. A resource limit, physical condition or required output can be changed in an exploratory problem, but the resulting answer then has to be related back to the original requirement.
Construct a question that the available result makes worth asking
Locate a consequential change and carry it through the question:
- A premise or condition failed. Ask which result remains obtainable under the supported conditions, or which missing condition would permit the wanted result.
- A construction succeeded. Vary an input, condition or required output to find a useful broader operation or its limit.
- A new operation became available. Find a problem whose needed intermediate result that operation can supply; work the connection through to a possible use.
Use B.5.QD to form and examine the new question. Change a named condition, demanded result, available means or composition of operations. Explain how that change affects what an answer would need to establish.
If the opening is an incompatibility, B.5.QD.CF examines the premise that makes the demands conflict and constructs a question about another means or a legitimate distinction. Return its candidate to the retained requirements here; changing an assumption for exploration leaves the corresponding realization question open.
Check the formulation before investing in solving it. Its objects and operations must have meanings, its conditions must be interpretable together, and the proposed answer must be assessable. A question asking whether a configuration is possible can legitimately return an impossibility result. When a task generator promises a solvable challenge, support that promise with a feasible case. Otherwise keep feasibility open.
Where a further attempt can settle an open distinction, select a revealing attainable case. Its result may expose a counterexample, a useful intermediate construction or a missing operation. Leave difficulty relative to the available means and method where it has not been established more generally.
Construct the way for the changed problem
Use an existing result or derive what the receiving conditions already establish. Apply an operation when its result is still needed. Where the receiving problem breaks the old construction, locate the failed contribution: input meaning, intermediate result, operation, output condition or applicability premise.
Retain the usable contributions. Transform the input, replace a constituent or compose available operations so that each result supplies what the next step needs. C.39.RO develops that reusable operation and its support. A new physical or professional operation still needs the Method that constructs or performs it.
Trace the changed way on the target material. Distinguish a derived conditional result, an observation from a performed trial and a proposed operation whose effect remains to be obtained. The target question determines which of these can support the next use.
Examine what the target use establishes
Judge the result against the target's conditions. Recover the failed or satisfied condition and the contribution that produced it. An earlier source success is useful support for trying the operation; the target comparison decides this use.
For a mathematical construction, an argument may establish the result over the stated inputs. For a physical claim, use the model and observations required by that claim. A generated executable challenge may need both a solving construction and a check that an empty response cannot satisfy its result condition. Select the examination that can change the next action; a local inquiry need not become a separate assurance programme.
Keep distinguishable the useful result, the part still unresolved and any attempted transfer that failed. When no attempt was performed, retain the proposed way and its missing prerequisite as such.
Use the result to choose the next problem
Use a sufficient result and stop this development. Otherwise follow the part of the outcome that can change receiving work:
- A partial result can support a narrower use while another condition remains open.
- A failure witness can support a revised problem, a necessary-condition question or a justified change of operation.
- A successful new construction can support a further question about changed inputs, composition or use.
- A missing contribution can direct a bounded inquiry or specialist request when its attainable answer warrants the effort.
For each continuation, state what it would permit and which earlier requirement it answers, changes or leaves open. A problem constructed for exploration can be retained for a later possible use without claiming present demand. Its interest lies in the further reasoning or action it may enable.
Retain the material and explanation needed to continue. Compare the prospective gain with the cost of the attempt, adaptation and retention, including displaced work. C.11.DUA helps when the value of further information work is unsettled. Close an unwarranted branch without discarding an earlier qualified use of its source method.
Coordinate several attempts only when their results warrant it
Several problems and ways may develop together. Try a transfer because a usable constituent or intelligible adaptation could change the target answer. Characterize the difference needed for that choice: for example, which conditions defeat available methods or which new operations a result makes possible.
Use C.18 when archive, front, lineage or possibility-space claims are needed, and C.19 for a resource-bearing continuation policy. A small problem and one useful attempt need neither a population nor an archive.
People, AI agents and teams can contribute different operations to this work. Preserve the receiving participant's means and preparation when passing on a construction. A copyable program or policy can supply executable material; learning how to adapt a method may require separate acquisition or practice. C.36.RP coordinates that continuing availability when it becomes the problem.
Archetypal Grounding
The following constructed cases supply their task conditions and ordinary subject operations. They demonstrate the reasoning and its limits.
A mixed transfer leaves a narrower comparison available
An investigation requires each error/non-error comparison to share both setup and temperature regime. An existing grouping operation uses setup alone. The permitted target records are:
The source operation returns T1–T2 and T3–T4. Comparing those outputs with the target rule exposes the first pair's temperature mismatch.
The changed local problem is to return eligible pairs and expose where no comparison is supported. An available filtering operation adds temperature to the grouping key. It returns T3–T4 and no eligible S1 pair. This operation constructs the target comparison; repeating the old setup-only grouping would preserve the mismatch.
The result permits the S2/warm comparison if it serves the investigation. The S1 question remains open. A next inquiry can seek a comparable S1 record when obtaining it could change the needed explanation; alternatively, receiving work can use the narrower result or stop. The original source operation retains its uses where setup alone was the applicable rule.
The grouping has established eligibility for comparison. A causal explanation would need the separate reasoning and observations for that claim.
An obstruction changes the mathematical question
A scheduling model represents pairwise conflicts by the finite undirected path A-B-C-D. There are no capacity or precedence conditions in this model. Assigning A and C to session 1, B and D to session 2 satisfies its conflicts.
A new conflict A-C creates the triangle A-B-C. Applying two-coloring now exposes an odd cycle. Two sessions are impossible under the retained pairwise-conflict model.
That result supports the next mathematical question: what is the least number of sessions? The triangle requires at least three. The assignment A=1, B=2, C=3, D=1 satisfies every edge, so three suffice and are minimal. The obstruction and construction together answer the new question.
If the receiving project has only two sessions, the new mathematical answer identifies a resource mismatch. Receiving work must change a real conflict condition, obtain another session or accept that the stated assignment cannot be supplied. Merely asking the easier resource question has not removed that requirement.
The successful three-session construction opens a further question when more conflicts are expected: which additions remain compatible with three sessions? Adding A-D invalidates the old assignment because A and D both use session 1. Retain A=1, B=2 and C=3, and move D to session 2. This changed construction satisfies all five conflicts.
If B-D is then added as well, every pair among the four vertices conflicts. Four sessions are now necessary, and assigning one vertex to each session attains that bound. The successful construction has led to a new question, a local repair and then a limit on that repair. These results distinguish which anticipated conflicts can be accommodated with three sessions and which require changing the allocation conditions.
A physical invariant replaces a timed operation
A device should dispense 40 cm³ from a reservoir. An earlier constant-flow model used V = q t; with q = 20 cm³/s, it prescribed two seconds. The receiving situation has unknown time-varying outflow, so that duration no longer determines the volume.
The reservoir has constant cross-sectional area A = 10 cm², starts at height h₀ = 15 cm and has no inflow or leakage. The supplied model treats the level reading and shutoff as ideal. Conservation and geometry give the dispensed volume:
V = A(h₀ − h).
The changed construction asks which level corresponds to the required volume. It gives h = 15 − 40/10 = 11 cm. The new operating rule is to stop outflow when that level is reached. At that event the model gives 40 cm³, independently of the outflow's variation before it. If outflow stops earlier, the desired level is never reached; the volume relation alone supplies no completion guarantee.
This result connects physical conservation, a mathematical invariant and an event-based operating rule. Realizing it requires usable level observation and shutoff through C.16 and C.29.3. This supplies a design result under the model's premises.
Suppose receiving work also requires completion within eight seconds. The level rule has left that condition open. A supplied bound q(t) ≥ 5 cm³/s until the target level makes the accumulated outflow at least 5t, so 40 cm³ is reached within eight seconds under the same ideal model. If no such flow support is available, keep the deadline unresolved or change the physical arrangement.
Bias-Annotation
An attractive new question can make a partial result look like completion of the receiving need. Keep the old condition at the point where the new result is used. Conversely, fixation on the original formulation can hide a useful invariant, obstruction or intermediate operation. Examine what that result permits before discarding it.
Conformance Checklist
- The problem's sought result, conditions and available material can be recovered.
- A changed question identifies what changed and why its answer could matter.
- The formulation permits a meaningful construction, observation or impossibility answer.
- The changed way supplies the needed intermediate results under its retained premises.
- Target examination identifies the satisfied and failed conditions.
- A narrower result leaves the original unmet requirement visible.
- A next attempt has an attainable contribution and a justified burden, or the development stops.
- Retained material can support the stated continuation with the receiving participant's means.
Common Anti-Patterns and How to Avoid Them
Consequences
A failed transfer can produce a useful problem and a successful construction can open further work. The cost is constructing and examining the relevant changes and retaining enough material to continue. Local sufficient-result exits and the distinction between a receiving requirement and an exploratory problem keep that development usable.
Architectural Rationale
Problem and operation development constrain one another. A changed condition can invalidate a constituent of the way; a new intermediate result can make a different question answerable. Keeping both directions explicit permits local repair while preserving the receiving need.
B.5.QD constructs the question and C.39.RO constructs an operation for changed or combined use. This Method coordinates their results through target application and the next decision. Ordinary C.40 variation can finish under one fixed question. The coupled Method is useful when the question itself develops or a subsidiary problem must be constructed to progress.
The mathematical obstruction, restricted record comparison and physical invariant have different grounds. Their common contribution is the connection between a changed question, the operation that addresses it and the use its result enables. Subject Methods establish the respective proof, comparison eligibility and physical conditions.
An automated problem generator supplies one implementation of problem construction. Its representation, policy encoding and evaluation thresholds determine which transfers it can try. For people and teams, availability and acquisition of the needed operations enter the same coordination through different means and costs.
SoTA-Echoing
The selected line combines constructive problem development, operation reuse and target examination. Compare it with keeping the target fixed and improving a sufficient existing way: that cheaper alternative is preferred when it supplies the wanted result without developing a new question.
POET v3, §3, supplies a paired environment/agent construction with environment generation, optimization and cross-environment transfer. Adapt those connections in :4.2–4.5. Its fixed thresholds and copying of policy parameters belong to that computational implementation.
Enhanced POET, §3 and appendices A.1/A.4, distinguishes environment encoding from behavioral characterization and changes when costly transfer adaptation is attempted. Adapt characterization for the transfer choice and attention to adaptation cost in :4.6. A policy that skips adaptation after poor direct transfer is a bounded economy; an intelligible target-specific adaptation can warrant a different choice.
LLM-POET, §§2–4, changes the environment generator while retaining the need to make its generated representation usable in the environment. Adapt that validity condition in :4.2. Its reported performance comparison has substantial uncertainty and concerns the tested generators and environments.
Environment Evolution for Terminal Agents, §§3–4, develops tasks through changed situations, operations and sequence length. It realizes those edits as environments and checks solvability and rejection of an empty solution separately. Adapt the constructive connection between a proposed task and its usable examination. The preprint's difficulty account depends on a reference distribution and modeling assumptions; it supplies no observer-independent scale for all problems.
The worked comparisons show why the coupled result can be worth the extra operation. Setup-only reuse produces an invalid pair; adding temperature supplies one usable comparison and identifies the remaining need. Repeating two-coloring cannot supply an impossible assignment; its obstruction supports a minimal-resource question. Reusing a duration loses the volume condition under variable flow; a level construction recovers it while exposing the separate completion-time question.
Reopen the Method's use when the proposed problem cannot be interpreted or examined, an available operation already supplies the whole wanted result more affordably, or target use reveals a consequential condition that the construction omitted.
Relations
- C.40 supplies ordinary variation and warranted continuation of usable material.
- B.5.QD constructs the next question; B.5.RC, B.5.TU and C.39.RO recover and construct obtaining operations.
- B.5.MPC and B.5.MPC.R connect physical, mathematical and computational contributions and repair their changed relations.
- C.16 and C.29.3 supply measurement and realization work when a target operation needs them.
- C.18 and C.19 govern archive, front and continuation-policy claims when these are needed.
- C.11.DUA compares worthwhile information work; C.38 compares ways for a shared result.
- E.10.INT and C.36.RP connect action-relevant interest and continuing access to shared methods.
C.40.OCN develops a bounded opportunity when its useful contribution and attainability still need construction. If the resulting way, holder and provision require coupled change, C.40.JD constructs those joins. C.40.NP develops a further problem from an obtained result. C.40.OD connects these contributions into an open-ended developmental whole when that is the intended result; the co-development of this one problem and way can also finish on its own.
C.40.CD:End
Develop a Useful and Reproducible Use of a Construct
Type: Method pattern Status: Stable Normativity: Normative
Problem frame
Use this pattern when a material, device, living construct or supported way of working produces an unfamiliar response, or seems able to do more than its assigned function, and you want to turn that possibility into a useful repeatable application.
The working reader develops or investigates such an application and can perform or obtain the relevant domain operations. The subject is the connection between discovering a response, constructing a useful arrangement and making that arrangement reproducible. A successful probe can supply a useful clue while leaving the new application unfinished.
First useful move. Start with an observed response and its conditions, or with a proposed response and what could establish it for the intended use. Ask what receiving use it could change, and what uncertainty would change the next action. Use the response within its supported limits, or construct the smallest worthwhile contrast needed before relying on it.
The first result can be a bounded usable response, a candidate application, or a statement of the contribution still needed. A repeatable application additionally needs an obtainable method, capable participants and the conditions under which its result can be obtained. Use an already adequate application directly; ordinary material variation under C.40 is enough when no discovery-to-use connection is missing.
Problem
An unexpected success does not reveal which arrangement made it possible. The tested construct may have retained a capability, learned during the test, received essential assistance, or encountered a changed criterion. Conversely, a failed probe may be insensitive to a useful response.
Even after that uncertainty is bounded, the useful work remains. Someone must form an application, choose how to obtain its result, realize the relevant conditions and make the method available to later performers. A causal explanation alone supplies none of these automatically. Repeating the demonstration with its original expert can conceal an unavailable intermediate capability.
Forces
Solution
Connect an examined response to a useful application, construct an arrangement that can supply its result, and preserve or develop what a later enactment needs. Return from use to inquiry only when the unresolved distinction changes the application. Return from reproduction to the affected constituent or support condition instead of restarting every inquiry.
Recover the response and the configuration that produced it
Keep the relevant material, preparation, task, interaction, observation and time window. Include prior learning, stored state and external interventions when they can change the interpretation. For a living construct or a person, use a permissible preparation and observation; a desired comparison does not authorize resetting, training or disturbing the subject.
Distinguish these questions:
- What response was observed, or is only proposed?
- What did the interaction make observable?
- What did it make possible or change?
- What result might someone obtain from that response?
A new readout can expose an existing response. Training can create a new capability. A fixture can make an action unnecessary. Another participant can supply the decisive action. A changed metric can improve a reported score while the trajectory is unchanged.
Keep the tested configuration recoverable without attributing everything it did to one constituent. A capability claim under A.2.2 names its holder and conditions. The surroundings need not be one System or one Agent. A model or recorded trace describes something; its availability does not make it a performer.
If the present material and known conditions already support the receiving use, continue with them. The method does not require a universal inventory of hidden capabilities.
Construct a worthwhile receiving use
Describe a result the response could help obtain, the receiver, use and conditions of acceptability. Start from the effect rather than the construct's customary name or intended function. A weight can supply resistance, a surface can support alignment, and a response to a task can reveal something about the environment.
Try the connection between the available response and the needed intermediate result. If a gap appears, ask whether to change the problem, the construct, the method, the performer or the supporting conditions. C.40.CD develops that reciprocal change; C.39.RO makes a successful construction reusable.
Keep an exploratory question separate from an accepted application. A promising operation can justify a bounded further inquiry before anyone presently needs its result. Explain what the next attempt could make possible and what effort it displaces. C.11.DUA helps when the value of further inquiry is unsettled.
Use the receiving conditions to decide how much to establish. A fixed-route delivery use may need reliable delivery from a known configuration, not proof that the controller invented its route. A claim of adaptation to unfamiliar routes asks more. Preserve the narrower use while the stronger claim remains unresolved.
Build the contrast that can change the next move
When an uncertainty matters, identify the few materially different accounts or values still compatible with the response. Accounts can coexist: retained state and a live helper can both contribute. Seek a comparison that separates the distinction needed by the use, rather than an exhaustive taxonomy of mechanisms.
For a consequential causal question, C.28.CM constructs the candidate mechanisms and derives a differing implication. B.5.2 supplies hypothesis generation when a useful account is missing. Change the identified intervention while preserving, measuring or otherwise accounting for the other conditions on which the inference relies. State what each possible outcome would permit.
For a question about what an indication reveals, C.16.MR constructs the response relation, including preparation, interaction and registration. C.16.IR retains the different cases that fit it. C.16.RM chooses a change to the model, computation, subject conditions or measuring arrangement by what it would make distinguishable. Repeating a test on which the consequential alternatives agree adds no resolving power merely by being another run.
Keep two environmental questions separate. Changing the demand on a fixed arrangement can help infer what its environment supplies. Keeping the task fixed while changing the environment can compare two whole arrangements. The latter does not isolate one environmental property if other contributions also change.
The domain must supply any valid reset, intervention, calibration or observation used by the comparison. For an unavailable operation, obtain or develop it, use a sufficient existing bound, or retain the explicit limitation.
Return the outcome as a supported response, contradicted account, conditional relation or unresolved distinction, together with the material conditions. A negative result can reject the tested arrangement without establishing an intrinsic incapacity. Stop when the supported conclusion is sufficient for the receiving action.
Construct alternative ways of obtaining the useful result
Translate the qualified response into a candidate arrangement. State which contribution the construct supplies, what action obtains it, and which other participants and conditions make the whole possible. A function–bearer hypothesis and a claim that an action is available concern different relations; keep the one that the decision uses explicit.
Consider materially different ways, where they are feasible:
- Keep the construct and improve the performer's operation or acquisition.
- Change the construct, tooling or environment so the response becomes usable.
- Obtain a missing contribution from another performer or service.
- Change the arrangement so the previously required operation disappears.
For example, an orientation guide may remove the need to select a part's orientation during insertion. That is different from a robot performing the same selection. Guide design, setup, compatibility and wear remain contributions to the whole.
Choose a way of influencing the construct from evidence about how it responds. Direct physical constraint, specified commands, feedback, training and cooperation impose different requirements. An unfamiliar response is a reason to investigate an influence method, not evidence that the construct can understand a goal or consent. Actions affecting people, animals or controlled resources retain their applicable authorization and domain safeguards.
Use C.32 and C.32.P2S when architecture alternatives and their realization are the live question. Use C.38 to make the candidate ways comparable for the same result, use, horizon and protected conditions, then C.11 for choice. Include displaced work, setup, maintenance, recovery and capability consequences where they can reverse that choice. A low cost inside the performer can coexist with a high cost of preparing its surroundings.
Follow consequential reciprocal changes. Support can alter later task selection, learning, stored traces and other participants' behavior. A trace left for a later performer has an interpretation and a lifetime; a cue that is correct now can become misleading after the situation changes. Compare or refresh these conditions at the times that the intended use consumes them.
Select a candidate only within its evidence and authority. A proposed environment, trained performer or fallback is not already available merely because the comparison requires it.
Realize and examine the whole under the receiving conditions
Start at the first unsupported contribution of the chosen way. The relevant engineering, research, learning or organizational Method supplies that contribution. Carry its result into the next operation with its meaning and conditions intact.
Make the critical joins explicit. What does the receiver of a probe result treat as established? What does the designer change because of it? What must the operator recognize before using the resulting arrangement? Which result satisfies the original receiving need? B.1.5 governs composition; B.1.5.EW recovers how the constituent actions perform the encompassing work.
Examine a representative use against the same result conditions used to choose the arrangement. A derivation can establish a model-conditional result; a performed trial supplies observations within its conditions. Keep these forms of support distinct. If the result is partial, retain the qualified part and name the unmet condition.
When the whole fails, locate the first consequential mismatch: response qualification, architecture, realization, joining, performer capability, access or changed support. Repair that contribution and reuse the rest where its conditions still hold. A changed deadline can require a different operation; it need not reopen the origin of an already qualified response.
If reliance includes a transition between realizations, examine the transition as well as the endpoints. The running arrangement may need to detect the trigger, change realization and preserve the required result during the transient. A designer's possession of two alternatives does not establish that runtime ability. Use A.15.8 when configuration or recovery reliance is the question.
Make the use reproducible for another performer or occasion
Recover what made the successful enactment possible: preparation, inputs, operative steps, supporting artifacts or contributors, recognition conditions, acceptance basis and stops. Preserve the needed intermediate meanings, not just the final output.
Decide which contributions the receiver must perform and which can remain obtainable from elsewhere. Restore access or activation when that is sufficient; develop capability when the receiver cannot perform a necessary operation. E.23.CAE distinguishes those gaps under grounded capability claims, E.23.CDI develops the named capability, and C.36.RP arranges acquisition, provision and continuing access.
Check the operation the receiver is expected to perform with the help that will be available. An earlier performance by that receiver under comparable conditions may already answer the question. If adaptation to a changed condition is still in doubt, choose a task in which that change affects what the receiver must do. Identify which adjustment the receiver made and what a helper supplied. A successful joint performance can qualify the arrangement for the job. Credit the receiver with following an instruction when someone else selected the adjustment. To credit the receiver with adapting the method, establish which change in working conditions the receiver noticed, which adjustment the receiver chose and why, with the help that was actually available.
When a shared term matters, start with the decision the participants need to make together. For one handoff, check what the sender's classification tells the receiver to do with this item. People can understand the same term and take different actions because their jobs differ. Completing the handoff on the agreed terms can settle that joint task. If the item must also meet an acceptance criterion, check it against that criterion.
If later work depends on several participants using the same criterion, compare their classifications under the same facts and for the same question. If their answers diverge, locate the fact, instruction or criterion that led them to act differently. Where the work requires agreement, resolve the difference with the person responsible, repair the handoff and try it again on the disputed item. To distinguish criteria still consistent with their answers, seek a case on which those criteria would give different answers. If both participants give the answer that one criterion predicts, that observation counts against a rival that predicted a different answer. Other criteria may still fit. To rely on agreement across a class of inputs, establish why the criteria used give the same decision throughout that class, or state the range and limits of the evidence for agreement. An agreed instruction can settle which criterion participants should use. Evidence of their work is needed when the question is which criterion they actually use.
Retain an obtainable explanation, executable material or demonstration suited to that receiver. A complete account does not require every participant to internalize every constituent operation. It does require the arrangement to supply them when they are needed.
Retain the qualified use and renew it when its conditions change
Keep the useful construction, its supported scope and the means needed for another enactment. If a disturbance can reveal a different retained state or recovery behavior, qualify that behavior over the window the use needs. Normal appearance now is not evidence of indefinite retention or recovery after every disturbance.
A new input, environment, performer, criterion or time window can reopen a connection. Use B.1.5.RS when a constituent is replaced across several encompassing uses. Use C.40.CD when the result opens a new problem or requires a changed obtaining way. Keep the previous bounded application where it remains supported.
End the current development when the selected receiving use is supplied, when a sufficient conditional result serves the assignment, or when no attainable next distinction or construction warrants its burden. State which of these outcomes has been reached. A promising clue can be retained without claiming a realized application.
Separate recognition from assurance
Recognition needs a concrete response, a possible useful difference and one uncertainty or connection that changes the next move. A modest proposed application can begin there.
A reliance claim additionally needs support for the holder or configuration, obtaining conditions, interpretation, receiving result and material limitations. The receiving domain determines the necessary trials, safety, ethics, authorization and operating evidence. This pattern neither certifies a construct nor establishes intelligence, sentience, clinical efficacy or universal independence from assistance.
Archetypal Grounding
These constructed cases provide their domain facilities and observations as teaching assumptions. They illustrate the reasoning; they are not reports of experiments or evidence of this pattern's effectiveness.
From an unexpected detour to a reproducible delivery arrangement
A simulated mobile platform makes a detour in a run that also includes an operator route update and a changed map. The intended use is a fixed delivery task without live route guidance. The immediate question is whether the tested configuration can deliver without that update, not whether it invents new routes.
The stipulated sandbox can clone the complete relevant snapshot S: map, controller configuration, learned parameters, cached routes, task and prior interaction state. Both runs start from S with the same assigned route, detectable blockage, sensing rules, pre-update history and observation window T. Later sensor values follow each run's trajectory under the same deterministic environment rules. The map permits a detour. One run receives route R at the selected post-blockage instant; the other receives no route update. There is no other operator input. Documented replay and readout facilities expose channel deliveries, blockage detection and positions.
Suppose both deliver. The result supports the fixed-task question. It does not distinguish use of cached state from new route generation. If the receiving use requires unfamiliar-route generation, an additional domain-supported observation or intervention must distinguish those accounts. Deleting a cache without accounting for other stored state would not establish the intended contrast.
Now the developer considers two candidate ways for a corridor delivery application. Way A uses the retained controller state and operating checks. Way B changes the environment: a physical guide supplies the route constraint, removing runtime route selection on that path. Neither is established by the sandbox comparison. Engineering must still realize and qualify the physical arrangement.
For a stipulated local comparison, both ways meet the same load and protected separation conditions after their specified setup. A needs two minutes of route preparation for each delivery. B needs forty minutes to install the guide and 0.1 minute of preflight checks per delivery, covering the guide, route availability and permitted load. Other decision-bearing costs are stipulated equal in this teaching comparison. Preparation burdens are therefore 2n and 40 + 0.1n minutes for n deliveries: at n=10, A needs 20 minutes and B 41; at n=100, A needs 200 and B 50. B becomes lower on this burden for n greater than 40/1.9, approximately 21.1. A changing route, blocked shared corridor or unavailable installation authority can reverse or prevent that choice; the calculation cannot trade away those conditions.
Suppose the authorized application retains B for a stable hundred-delivery run. The candidate's receiving requirement is delivery within ninety seconds, timed from the start of preflight checks to completed unloading. Under the supplied model, preflight needs six seconds, the guided path is thirty metres, travel speed is 0.5 m/s and loading plus unloading needs twenty seconds. Completion takes 6 + 30/0.5 + 20 = 86 seconds. The conditional deadline result is supported; real operating reliability still requires the relevant engineering evidence.
The whole's operating Method includes preparing and inspecting the guide, confirming route availability and the permitted load, loading, initiating motion, recognizing arrival and unloading. A receiver who can press the start control but cannot recognize a displaced guide lacks a contribution the whole requires. The repair can be teaching that recognition, obtaining a qualified inspection or changing the guide so misplacement is prevented.
For a changed seventy-second requirement, the same construction fails. Merely repeating the no-update probe cannot repair it. A handling change to ten seconds would give 6 + 60 + 10 = 76 seconds and still fail. With the same path and preflight, handling would have to take at most four seconds; that is an engineering possibility to qualify, not a result supplied by the arithmetic. Alternatively, a permissible twenty-seven-metre path at the same speed with ten-second handling would give 6 + 54 + 10 = 70 seconds. Compare feasible changes or leave the new deadline unsatisfied. The earlier ninety-second use remains supported within its conditions.
A later route branching requirement reopens the removed route-choice operation. B is no longer an adequate way merely because its guide worked on the old path. The developer can retain the fixed-route use while constructing and qualifying the new branch. This is a change in the encompassing application, not proof that the platform lost a previously established general planning capability.
Use a worker's response to qualify environmental support
A stipulated deterministic worker reads a payload through an environmental service. For this constructed case, assume its response relation t = a + x/b holds for payloads from 10 through 30 MB under the named service conditions: payload x in MB, nonnegative setup time a in seconds, throughput b in MB/s and observed total time t. Applicability over that range is a supplied teaching premise, not a validation inferred from the two observations below. Fresh comparable requests preserve a and b; there is no caching, compression or retry, and timing resolves the needed difference.
If the receiving use only needs the same 10 MB request within four seconds under these conditions, the stipulated three-second response supplies that bounded answer; identifying throughput adds no necessary result. Now consider the stronger question of whether b is at least 8 MB/s. One response, x=10 and t=3, fits both (a=1, b=5) and (a=2, b=10). The same successful response supports opposite answers. Repeating the identical request under unchanged conditions does not distinguish them.
A 20 MB request predicts five seconds for the first account and four for the second. With shared a and b, subtraction gives b = 10/(t20 − t10). If t20=4, b=10 and a=2. If t20=5, b=5 and a=1. A result elsewhere must be interpreted through the relation rather than forced into these two example accounts. Timing uncertainty can yield an interval; changing setup, a nonpositive time difference or inferred negative setup defeats the stipulated inference.
Suppose the first outcome obtains in the teaching case. The environment supports the throughput threshold under the tested conditions. A proposed use must transfer 30 MB within six seconds. The relation predicts 2 + 30/10 = 5 seconds, so it supplies a conditional application result. It does not establish the same service during contention or in another environment.
Reproduction requires the receiver to know the payload and time meanings, use fresh requests under the shared-condition rule, and recognize when that rule no longer applies. The receiver can obtain the algebra from a tool; understanding why a cached response is unsuitable remains necessary somewhere in the arrangement. A changed-condition attempt with caching exposes that gap more directly than another identical demonstration.
If the new application has four-second total latency, a throughput threshold alone no longer selects a way. Reducing setup from two seconds to one would give four seconds at the same throughput; increasing throughput to fifteen would also give four seconds at the old setup. These are candidate ways requiring their own feasible realization, costs and evidence. The problem has moved from interpreting a response to constructing an adequate arrangement.
Keeping x=10 fixed and changing the environment could instead compare the two arrangements' total times. That comparison would not isolate b if setup also changed. The useful composition preserves the question being answered at each move.
Bias-Annotation
Novel behavior invites inflated attribution, while a familiar assigned function can suppress exploration. Preserve both the new possibility and its bounded support. A preferred theory must not decide which alternative explanations are allowed to remain.
Environmental redesign can hide burden transfers to installers, maintainers or other users. Reproduction can hide dependence on the original expert. Examine those contributions where they change the receiving application.
Conformance Checklist
- The response, preparation, intervention, observation and relevant window are recoverable.
- The receiving use states what result matters.
- Any selected contrast can distinguish a consequential uncertainty under supplied domain conditions.
- The conclusion separates what was revealed from what was changed or supplied.
- Candidate ways include the enabling and displaced contributions that can change their comparison.
- The chosen way is carried through its critical joins to the receiving result or to where a contribution is still missing.
- Reproduction supplies the necessary method, participants and support; independent capability is claimed only when supported.
- A changed whole condition returns to the affected constituent or connection.
- Retained use, proposed continuation and stop remain distinct.
Common Anti-Patterns and How to Avoid Them
Consequences
An unfamiliar response can become a useful bounded application without a premature universal capability claim. Inquiry, architecture, realization and reproduction inform each other while retaining their distinct results.
The method exposes the cost of making a demonstration repeatable. Some promising responses remain conditional constructions or clues because their necessary operations, evidence or supporting arrangements cannot yet be obtained.
Rationale
The same intervention can reveal a response and help produce it. Its interpretation therefore has to remain connected to the use that consumes it. The use, in turn, determines which conditions to realize, which alternatives are worth comparing and what to preserve for another enactment.
The contribution here is their conditional organization: qualify only what the application needs, construct the whole arrangement, test its result, and return from transfer failure to the affected contribution. Constituent Methods retain their own authority and detailed operations.
SoTA-Echoing
How should an unfamiliar response become a useful, repeatable arrangement? The selected line couples application-relative inquiry to construction of the whole arrangement and a receiver's reproduction. An alternative is to qualify a response through causal or measurement Methods and leave application and transfer as separate expert work. That alternative is sufficient and cheaper when the receiving question is only the bounded response or throughput value, as the first branch in 5.2 shows. It is insufficient when a practitioner must decide what to build and preserve: 4.4–4.6 add operation elimination, allocation of environmental support, constituent conditions and the receiver's missing contribution.
The two articles motivate these questions; their wider theoretical claims are not evidence for the effectiveness of this composition. The added cost is developing and trying the connections needed by the use. A directly usable existing whole can remove that cost; an unavailable domain operation or a failed receiving attempt reopens the affected construction. The changed deadline in 5.1 and changed service conditions in 5.2 show those returns without discarding the earlier bounded result.
Claims about living systems, learning and human cooperation require the applicable domain sources and safeguards.
Relations
- C.40 supplies material variation and retention; C.40.CD develops a changed problem and obtaining way together.
- C.39.RO constructs a reusable operation from an available case or composition.
- C.28.CM and B.5.2 develop consequential causal accounts; C.16.MR/IR/RM develop and improve interpretation of responses.
- B.1.5, B.1.5.EW and B.1.5.RS qualify composition, recover encompassing work and compare constituent replacement.
- C.32 and C.32.P2S develop architecture and its realization; C.32.MWA applies when practice structures require a joined synthesis.
- C.38 forms complete-enough comparable ways; C.11 selects among them.
- A.2.2 governs a capability claim; E.23.CAE/CDI separate supported performance conditions from needed capability development.
- A.15.8 governs configuration and recovery reliance; C.36.RP makes shared methods obtainable and renewable.
When realizing the application depends on co-adapting its way, the intended holder and provision, C.40.JD supplies that joint construction and keeps delivered work distinct from acquired capability. C.40.NP develops the continuation when a qualified result opens a further worthwhile problem. C.40.OD is useful when those connections form a wider open-ended development undertaking; an adequate application and its ordinary renewal need not invoke that whole.
C.40.CU:End
Construct a Bounded Opportunity
Type: Method pattern Status: Stable Normativity: Normative
Problem frame
Use this Method when an agent or arrangement of capable contributors is exploring what useful future Work could become possible for a named holder, but the problem, contribution or direction is not yet fixed. A new technology, an emerging use, a changed constraint or dissatisfaction with familiar options may prompt the inquiry. The result describes the proposed contribution and its conditions clearly enough to be examined outside the original inquiry. A list of fashionable subjects leaves that work undone.
The subject is a useful possible contribution and the conditions under which it could become obtainable. This Method governs how a possible direction is made meaningful and conditionally reachable. It does not govern an adviser’s recommendation, a development intervention, strategic commitment, procurement or the eventual choice.
A separate adviser is not required. A person exploring future engineering Work can begin here and stop with an opportunity or a question to pursue next. If another performer later gives advice to a named recipient, the applicable advising Method supplies the engagement and professional conditions; the decision rule and authority govern the later choice.
Do not use this Method to reopen an adequate option set merely because more ideas could be generated. Use the choice or comparison Method when the worthwhile result and live alternatives are already sufficient. Use a qualified specialist Method when it owns the whole question. An evolutionary account of a population without a recipient is not a development-direction choice; a researcher's intervention question needs its own science and authority basis.
A complete first return can be modest: “This contribution could address the named receiving-use problem, but only if access to these records and the required review support can be obtained. The next question is whether the data owner can provide that bounded access.” Missing feasibility does not have to erase the useful problem; it prevents a stronger reachability claim.
A useful result can be a candidate opportunity with a worthwhile problem or a proposed promise to a recipient, the conditions under which it is attainable and what is still uncertain. The inquiry can also establish that the proposal has no worthwhile use under the stated conditions. Give the recipient a further question when its answer could change a decision they face.
Problem
A promising means can be mistaken for a worthwhile result, and a plausible direction for an attainable one. Separate search, problem, support and qualification results then leave the practitioner to invent their connection. The practitioner needs to show the recipient what the proposed contribution could make possible and under what conditions it is attainable. The inquiry may show that the current arrangement is adequate. It can also identify a missing condition whose resolution would change the decision.
Forces
Solution
To construct a bounded opportunity, examine neighboring uses and technologies, distinguish the problem from what someone proposes to supply, develop alternatives where they matter and check which combinations can actually be obtained. Return what the recipient can rely on and what could defeat it. If an open question matters to the recipient's next decision, say how its answer would change that decision.
Its generic participants are the performer or performers conducting the inquiry, the developing holder or subject, possible receivers or users of the later contribution, and sources or suppliers of material premises. They are Method-side meanings, not actual assignments or Work participants. The inquiry performer may also be the subject of the proposed change; no distinct recommendation recipient is an applicability condition.
The minimum applicability is a recognizable inquiry question, later-Work concern or aspiration, a holder or bounded candidate holder, a useful horizon, and enough context to ask what would matter and what is protected. Some of these may be provisional. If they cannot be made specific enough to distinguish a useful possibility from a slogan, return the missing frame before claiming an opportunity.
The invariant result is an opportunity account or a gap, rejection or stop. It preserves the distinctions among possible value, supplied evidence, conditional reachability, actual access/support, recommendation, choice, plan, Work and effect. Holder, horizon, search seeds, technologies, suppliers and presentation can vary without changing that way of doing. Removing source qualification, treating a candidate as a commitment, or changing the result into advice to a distinct recipient changes the identification question.
When promise content itself becomes a relied-on claim, use A.2.3; commitment, provider, access and fulfilment remain separate. When a claim needs an actual Problem or problematic-for relation, use C.22.PFR — Problematic-For Relation to establish the actual condition, the applicable adverse criterion and the entity and use for which it is a problem. This Method can work with explicitly provisional problem statements without inventing those stronger relations.
State the holder, prospective later Work or concern, receiving situation, horizon, protected conditions and inquiry budget. Name what is known, what is merely suggested and which questions the inquiry may investigate. Include permissions needed for the inquiry itself; permission to read or ask is not permission to trial an intervention.
Do not demand a fully specified goal before discovery. “Explore a useful contribution to engineering handoffs in the next quarter” can be enough to begin if the possible receivers and stakes are recoverable. “Develop in AI” is not enough to end: it needs a useful problem, contribution and dependency account.
A small reversible inquiry may use a few source checks and conversations. A consequential commitment will need much stronger domain evidence, but that is not a reason to manufacture certainty during opportunity construction.
The following operations develop the opportunity. They express information dependencies and permit iteration; they do not impose a schedule. Return to the earliest changed claim, not automatically to the beginning.
The independently usable opportunity Methods supply the missing contributions: C.40.OI bounds the inquiry, C.40.OC characterizes its subject, C.40.NU searches, C.40.PC constructs the problem and contribution, C.40.DC constructs directions, C.40.JF qualifies joint reachability, C.40.OP retains useful continuations and C.40.RQ requalifies change. Enter at the missing result; an adequate existing contribution need not be reconstructed. The connections below explain how their results form one opportunity account.
Recover the later Work and the reason to explore
Describe what the holder might contribute, for whom, in what situation and over what horizon. Distinguish the receiver's useful result from the holder's desired learning, status, product or technology use. Ask what would remain unsatisfactory if nothing changed, and whether the receiver cares about that difference.
Keep alternative formulations when they materially change the opportunity. One inquiry may concern recovering an acceptance premise; another may concern supporting shift handoffs. Those are not merely two implementations of one already accepted result. Use C.40.PC to develop the worthwhile problem and proposed contribution; retain the material formulation difference and obtain any missing subject contribution before treating a premise as established.
Start from a qualified demand or problem account when one exists. If the reason to explore is only a supplier's enthusiasm, preserve it as a search lead, not as an established need. If the receiver says the proposed result is unnecessary, revise the contribution or reject that lead.
Search neighboring technologies, uses and characteristics
Begin in the direction the receiving question leaves open. Work from an affected use towards possible support, from a technology towards other applications, or from a desired characteristic towards ways of obtaining it. Seek another direction where it could change the answer or a range of alternatives was commissioned. For each lead, record its source and the difference that could matter.
Use explicit value questions to generate possibilities, rather than scoring only the first available products. Keeney's 2012 value-focused brainstorming is a historical Method anchor for that move and for individual generation before group anchoring when a group is involved. It is not a requirement for a workshop, proof of current prevalence or a reachability theory.
Name the generation operation used—for example, substitute the receiving use, vary the required characteristic, combine two support contributions or remove an assumed technical constraint. Keep the search scope, source dates and stopping reason sufficient for the next use. Search time, idea count, novelty score and geographic distance from an archive are not evidence of practical value.
Stop widening when the inquiry has enough materially distinct leads to expose the decision-bearing unknowns at its chosen effort, or when a missing domain premise prevents useful further construction. State unsearched areas if their omission can change the return. For a complete set, state its boundary and establish coverage by a rule, a complete source or justified exclusions. Otherwise report the searched extent and any consequential open areas.
Turn a lead into a problem and a proposed promise
For each lead worth further inquiry, connect four things in ordinary language: the receiving situation, the useful difference, the proposed contribution and the evidence or hypothesis linking them. Then ask whose facts or judgment can confirm whether that difference matters at the needed scale.
For example, “use a new model in engineering” may become “help a receiving engineer recover which acceptance conditions a supplied evidence package does not yet support.” The proposed promise concerns recoverable qualification information, not faster approval, higher quality or automatic acceptance. Those stronger benefits need separate evidence.
Keep the niche, problem and promise distinct. A niche names the possible fit. The problem explains why the missing or inadequate result matters. The proposed promise says what could be supplied and judged. A useful research opportunity need not be a commercial offer; it needs a worthwhile question and an inspectable result.
Reject or revise a lead that has no supported worthwhile problem after proportionate inquiry. If value remains uncertain, retain a problem hypothesis with the confirming question, not a positive value claim. A provider's existing promise is evidence about what it offers, not that this receiver needs it or that it will be fulfilled.
Construct different directions with their supports
Build the possibilities the inquiry needs for addressing the problem or realizing the contribution. Change the intervention, the work performed, the distribution of contributions, the support, the timing or the proposed result where those differences matter. Distinguish a different result premise from a different way of obtaining one result.
A direction should say more than “train,” “buy,” “automate” or “hire.” State the prospective contribution, the holder change if any, necessary access and support, protected conditions and the transition that would have to become possible. One option may develop a capability; another may make an existing capability usable through better access; a third may obtain a qualified external contribution.
Use C.40.DC to construct materially different directions and their support configurations. Obtain qualified holder, organization, AI and other subject results through A.15.9 for what the proposed changes would require. After one result/use/acceptance premise is stable, an applicable organizational Method can compare whole organizational obtaining arrangements. Before that point, forcing every lead into the same-result comparison would conceal the still-open opportunity question.
Retain the obtaining baseline, a smaller repair, a staged direction or no change when they answer the inquiry. Do not count a fragment as a complete direction merely to populate a table, and do not discard a potentially useful contribution only because its supporting arrangement is not yet designed.
Qualify dependencies and bound reachability
Identify the premises whose absence would block the proposed transition or change its value. Typical dependencies concern holder capability, usable access, available support, resources and time, supplier contribution, acceptance, authority, safety or continuing Work. State who owns each premise and what result is adequate for this use.
Use those results through the A.15.9 result-acquisition Method; do not perform an unqualified diagnosis or invent a supplier's answer. Inspect adequate existing results and retain the useful conditional opportunity. Select a new request only when its obtainable contribution warrants the complete acquisition burden for this inquiry.
For each material transition, state the starting condition, the proposed change, the enabling premises, the evidence and uncertainty, and what could prevent completion within the horizon. Keep the joint condition visible: two individually feasible steps may compete for the same person's time, data access or budget. A chain is not feasible merely because each box sounds feasible.
Use the strongest supported wording:
A person's capability envelope, an access/expression differential and a recovery test answer different questions. None alone selects an intervention or proves learning transfer. An AI benchmark, an organization allocation result and a provider promise have equally specific boundaries. Do not transfer one holder's mechanism or evidence to another by analogy.
Obtain the human, AI, engineering or other domain result establishing which changes are feasible for this holder and horizon; keep reachability conditional while that result is missing. A curriculum sequence, an archive entry or a “zone of proximal development” label can help identify a possibility to investigate.
Preserve useful stepping stones and return the next question
For a stepping-stone claim, name the option or region it could open, the enabling transition and the evidence that supports that connection. Distinguish a candidate stepping stone from one whose retention is already justified under an applicable policy. A data-access inquiry can be a candidate stepping stone toward representative evaluation; it does not itself demonstrate capability or authorize use of the data.
If a generation archive is useful, use C.18 with this named generator, its operators, descriptors, source/lineage and retention basis. A retained archive can preserve exploration value beyond the current front. Front membership needs its own comparator and admissibility; neither archive nor front is a recommendation or permission.
Claim a changed possibility space only when it matters and the earlier and candidate generation conditions are recoverable. A newly visited region inside the same admissible space is not an expansion. If a new access relation, operator or building block makes a formerly unavailable region reachable, state that change and its evidence. C.18 and the direct causal owner govern any stronger claim; this Method does not infer it from novelty.
When choosing which opportunity or question to consider next, declare the receiving policy: relevant value, burden, delay, protected conditions and uncertainty. A partial comparison can be enough. A C.19 pool policy is used only when that pool and policy exist; there is no default explore/exploit setting.
Return the useful opportunity and the unresolved distinction that limits it. Where the recipient needs a next inquiry question, state what its plausible answers could change. Before recommending a probe or request, compare its attainable contribution with its full burden and alternatives through C.11.DUA and A.15.9. A selected probe retains its feasibility, exposure, authority, and execution conditions; an inactive probe creates no request or omission record.
Return the opportunity and reopen the affected premise
Use a short account when it carries the needed distinctions. These are content positions, not a new universal record kind or a requirement to fill every field.
The account may return a qualified candidate, a conditional candidate with a missing premise, a rejected direction with its reason, or an unresolved frame. It can stop without a recommendation. A source bibliography, score or diagram alone is none of these results.
Reconsider an opportunity when a relied-on problem, promise, source, holder configuration, dependency, protection, horizon or comparison policy changes. Map the change to the candidate claim and any retained stepping stone or downstream advice that used it. Preserve independent claims.
For example, suppose the data owner in the constructed model case later qualifies only a narrower source set. Retain the receiving need and the manual-index direction. Narrow the model-assisted candidate to the permitted set; remove the unsupported representative-use reachability claim and request missing evidence only if its obtainable contribution to that broader use warrants the whole acquisition burden. No earlier actual Work is erased by this change of knowledge.
If the new source set also excludes the region that a candidate stepping stone was meant to open, revise or remove that retention rationale. Another retained candidate does not become invalid by association. If the proposed question can no longer change the recipient's decision, drop it. Choose a replacement only if another open question matters to work the recipient intends to pursue. An answer that settles the commissioned inquiry can close it.
A factual change in one application does not automatically change this reusable Method. Reopen the Method's claims when evidence or a stronger practice line shows that its search, problem/promise construction, dependency reasoning, reachability rule or stop omits a material action. G.11 supplies source refresh; C.40.RQ requalifies an affected opportunity. A recurring failure of this Method returns to its maintainer.
Recognize the contribution and qualify reliance
Recognition asks whether the useful future contribution and the conditions for achieving it still need to be worked out. One concrete later-Work concern plus a material unknown is enough to enter. Recognition is not an assurance judgment that a proposed development will succeed.
Assurance asks whether each value and reachability claim follows from sources qualified for this holder and use, whether important alternatives and protected conditions survive, and whether missing transitions remain visible. A hypothesis can be useful without being a feasible intervention. Before consequential enactment, obtain the capability, access, safety, professional, resource, authority and evaluation results required by that use.
Archetypal Grounding
A new model without a settled problem
This constructed example assumes the bounded facts below. It reports no actual model, organization or performed Work; missing facts remain unknown.
An engineering-services unit hears that a new model can process long technical records. No accepted development problem or target result has been selected. The inquiry concerns useful contributions to engineering handoffs within ninety days, under confidential-data, independent-acceptance and continuing-service conditions. Buying or deploying the model is not the inquiry's objective.
The inquiry searches in three directions. From the technology, it explores record retrieval and qualification support rather than assuming autonomous decision-making. From uses, it considers evidence acceptance and shift handoff. From characteristics, it considers traceability of limits and reduced recovery burden; a manual index and a clearer handoff structure remain non-model possibilities worth examining.
The case's receiving-use statements identify two different worthwhile concerns. An acceptance engineer needs to recover which claims an evidence package supports. Shift staff need a usable account of unresolved exceptions and who can take the next authorized action. These statements establish why the possible contributions matter in the example; they do not establish that the model can supply them.
The two directions address different result premises. They are not forced into one same-result comparison. If the unit later stabilizes Opportunity A's result, use and acceptance basis, an applicable organizational Method can compare complete manual, provider or mixed obtaining arrangements for that same result. That later comparison cannot supply the earlier choice of useful problem.
The return is two conditional opportunities with different missing premises. It is not “adopt the new model.” The original trend-only lead is not retained as a third opportunity because it identifies no useful receiving result.
For the next inquiry, suppose the data owner can provide a purpose-specific access disposition before the next commitment using an existing permissions account. Its answer would determine whether to keep the manual route alone or open the representative evaluation route, and that gain warrants the formulation, owner attention, delivery, and interpretation without protected-record processing or disruption to shift Work. Asking for this bounded disposition is then a useful selected acquisition. A possible later access probe is identified as a candidate stepping stone toward representative qualification work, not as an authorized probe or a proven development step.
An affirmative access disposition would open only the permitted evaluation possibility; it would not establish the model's usefulness, acceptance authority or the person's capability. A negative disposition would remove that processing route and reopen the manual or differently bounded contribution. An answer covering only public, non-representative samples would support only a narrow demonstration, not the original representative-use claim.
Both opportunities might later compete for the same review capacity. Their individual plausibility does not establish that they can be pursued together. The joint time/support question remains with its qualified owner before any combined reachability or plan claim.
A small archive is optional. If the unit retains these candidates for later inquiry, its record names this Method and the search operations used, the two candidate accounts and their source lineage, the missing premises, and the retention reason: preserving distinct useful questions until their dependencies can be resolved. It claims no non-dominated front, expanded possibility space, performed trial or effectiveness.
A person exploring future engineering work
This is another prospective, constructed use. An engineer wants to explore a useful next contribution within four months without asking an external adviser for a recommendation. The holder is the person. The example distinguishes two prospective Work families: technical evidence review and coordination of cross-discipline interface questions. No future assignment, promotion or learning effect is assumed.
The person searches neighboring applications of their technical experience, different characteristics of contribution and possible supports. Evidence review could emphasize identifying mismatches between claims and tests; interface coordination could emphasize recovering incompatible assumptions between specialists. The first is not renamed as the second merely because both involve reading documents.
A qualified demand statement in the case explains why each contribution could matter. For the evidence-review direction, access to a permitted example set is available, but the person's capability and the transfer from supported practice to representative Work remain unknown. For the coordination direction, access to the relevant participants and competent feedback support are not established. A job title and a course listing supply neither missing premise.
The first account retains two conditional opportunities. It names the evidence-review or coordination contribution, its prospective use, the four-month horizon, the relevant support and uncertainty, and the next question. For evidence review, that question is what bounded representative assessment can distinguish an absent capability from difficulty caused by the current support configuration. For coordination, the first question is whether an appropriate practice/feedback opportunity can be made available.
Current career information about these directions is qualified for accuracy, context and currency, following the human-only contribution of NCDG 2024, CD2/CD3. Results from the applicable human-capability and target-domain Methods determine capability, intervention and transfer claims. The framework of career information does not make a course effective.
The person may stop here. This inquiry produces no final advice episteme and needs neither a separate adviser nor an institutional conflict. If a separate adviser is later engaged, the professional decision-support engagement conditions open then. If the person instead already has adequate alternatives and is ready to choose, domain or C.11 guidance is enough.
A defeated or insufficient opportunity
A supplier proposes “AI transformation” to a small unit whose current receiving result is already adequate and whose users identify no worthwhile unmet contribution at the proposed scale. Novelty alone does not repair that missing problem. Reject the direction for this inquiry or change the question; an archive can retain a technical idea for another stated use without treating it as a present opportunity.
A qualified dependency result may also defeat an otherwise worthwhile direction. If the only proposed processing route violates the protected data condition, reject that route for this use. If access has merely not been assessed, retain the conditional question. Forbidden, unknown and available are different results. If the protection requirement itself is being questioned, use C.11.DUA §4.3 with the competent domain result; the competent subject practice supplies the value-account return when needed. A supported revision recommendation may finish that question while the current requirement continues to constrain the opportunity's feasible use.
A course sequence described as “from beginner to leader” has no bounded reachability here until representative later Work, the actual holder, required transitions, support and transfer evidence are recoverable. Likewise, evidence that one AI version improved on one benchmark does not establish which development changes are feasible for a person or organization.
An opportunity is not a recommendation, and a recommendation is not a choice. An authorized commitment, WorkPlan, intervention, provider agreement or observed effect must be established separately even when the opportunity account is persuasive.
Bias-Annotation
A supplier or preferred technology can anchor the problem and the option set. Recover an unlike receiving use and an adequate current way where they can change the answer. A neatly drawn path can also hide unavailable conditions; inspect the enabling result at each consequential join.
Conformance Checklist
- The subject, receiving use or exploratory question, horizon and protections are recoverable.
- The proposed contribution is distinct from the problem and its evidence.
- Materially different directions retain their operative support conditions.
- Joint reachability uses qualified premises for the same holder and horizon.
- The return distinguishes a qualified candidate, conditional possibility, rejected direction and gap.
- A useful stop or discriminating next question follows without inventing advice or commitment.
Common Anti-Patterns and How to Avoid Them
Consequences
The practitioner can preserve a useful possibility while refusing unsupported feasibility, advice or intervention claims. The construction costs search, interpretation and joining work; direct use of an adequate option set avoids that cost.
Architectural Rationale
The complete account connects changes in the question, receiving contribution and obtaining way. Its inquiry, characterization, search and qualification Methods remain independently usable. Their conditional connections produce the opportunity; the publication order imposes no work order. Subject practices provide the mechanisms and evidence for each contemplated change.
SoTA-Echoing
The working question is how to construct a useful possibility before the target result and its obtaining arrangement are settled. A short menu of available technologies is efficient when an adequate result premise already exists; it fails this use when nobody has established why the proposed contribution matters. This Method retains the useful menu as search input and adds problem/promise construction, dependencies and a bounded first return.
Phillips's 2025 practitioner account of decision analysis treats models as purpose-dependent support for exploring possible futures, using only the ingredients the situation needs. Its sections on requisite models, decision conferences and decision-analysis technology inform the non-compulsory inquiry and the separation between analytical exploration and the accountable person's decision. This is a contemporary practitioner synthesis, not empirical validation of this cross-holder Method.
Keeney's older generation anchor informs explicit value questions; the human career framework informs that branch's alternatives and information qualification. C.18 governs any archive/front and possibility-space claim. These contributions are not interchangeable and do not, separately or together, prove practical reachability for an actual holder.
The prospective construction connects neighboring-use search, a worthwhile problem and proposed contribution, support dependencies, bounded reachability and a discriminating return. An archive-only approach preserves candidates but needs a generator and domain facts. A same-result arrangement Method is powerful after its result premise is stable; it cannot substitute for this earlier question. Conversely, once the opportunity question is settled, this Method should stop and let the next owner work.
What changes in practice is concrete: “we should develop in this promising area” becomes a candidate contribution whose value, dependencies and next question can be inspected. The account can preserve a worthwhile possibility while refusing an unsupported reachability or advice claim.
Relations
C.40.OI/OC/NU/PC/DC/JF/OP/RQ supply the developed opportunity operations. C.40.CD develops a problem and obtaining way together; C.40.CU constructs a reproducible application from an examined response. C.40.OD connects a knowledge or Method gain to continuing development. The applicable decision-support Methods supply problem structuring, decision analysis and professional advice when requested. The applicable strategy Methods supply strategic choice and commitments. C.11 and the authority govern choice; A.15.9 governs obtaining a qualified result from another practice.
C.40.OCN:End
Bound an Opportunity Inquiry
Type: Method pattern
Status: Stable
Normativity: Normative unless marked informative.
Problem Frame
What inquiry is worth conducting when someone says “we should develop,” names a fashionable solution, or asks what to develop next before the receiving change is clear?
Use this pattern when a person, organization, community, AI arrangement, platform, facility, research programme, or another identified subject has a development concern or aspiration but the receiving use, horizon, protected conditions, reason to explore, or affordable inquiry effort is loose. The first useful result is a bounded inquiry that another agent can inspect, or one missing premise that prevents such an inquiry.
Begin here without appointing an adviser. A person can bound a self-inquiry and continue into opportunity construction. When a distinct performer is already being asked to advise a named recipient, this pattern can recover the subject, use, and horizon, but the applicable professional engagement Method governs whether the advising relationship itself is admissible.
Cheap exit. Reuse an existing inquiry when it already identifies the same subject, receiving use, horizon, protections, posture, and effort boundary and no material premise changed. If a qualified domain practice owns the whole question, use it. If adequate alternatives already exist and the remaining need is comparison, recommendation, or authorized choice, stop opportunity construction and use the decision-support owner.
Ordinary non-use boundary. A research account of change in a non-cultural population or lineage is not advice to that population. A separately authorized researcher may have an opportunity or decision question, but the researcher or sponsoring organization—not the population—then supplies the recipient and authority boundary.
A broad wish such as “develop in AI,” “become more mature,” “modernize the platform,” or “build a learning culture” can point to several different subjects and receiving uses. It may concern a person's capability for later Work, an organization's contribution and arrangement, an engineered system's service, an episteme or problem formulation, a research programme, or a proposed intervention. The same words do not make those subjects develop in the same way.
The practical danger appears before alternatives or advice. A course, tool, reorganization, model, supplier, or PatternID can become the answer before anyone names the receiver, useful difference, protected conditions, or evidence needed. The opposite failure is an unlimited discovery exercise whose growing list never returns a result that can be tested or used.
This pattern governs the inquiry boundary: the identified subject, receiving use, declared horizon, protected conditions, reason to explore, posture, inquiry budget, and first missing result. It does not establish a worthwhile problem, qualified opportunity, recommendation, choice, programme, Guide, performed Work, or observed development effect.
Problem
Without an explicit inquiry boundary:
- a solution label substitutes for the receiving contribution;
- “development” silently shifts among person, organization, technology, knowledge, and population meanings;
- a supplier's promise is treated as the receiver's problem;
- the horizon expands or contracts to make a favored direction look plausible;
- safety, workload, continuity, privacy, reversibility, or another protected condition appears only after a candidate has gained momentum;
- search effort grows without a proportional stopping reason; and
- missing authority or evidence is hidden by a polished opportunity statement.
The reader then cannot tell whether to characterize the subject, search neighboring possibilities, request a specialist result, bound an advising relationship, or stop.
Forces
Solution
Bound the inquiry so it can direct the selected useful action and identify a sufficient return or supported stop.
Identify the subject and claim posture
Work back from what the recipient needs to explain or change. Which subject could have that property? Would the same part have a different consequence in another permitted whole? Would the same readings for the parts conceal a different relation between them? Follow the connections on which the answer depends. The subject might be a person, organization, community arrangement, AI model-and-scaffold configuration, facility, platform, research programme, episteme, or another subject. If identification is provisional, state the candidate boundary and what would settle it.
Mark the claim posture. A proposed future contribution is not actual change. A plan is not performed Work. A model prediction is not an observed result. If development or evolution still hides the subject, continuity rule, or value claim, use E.10.DEV and return with the subject owner.
Identify a receiving use or an exploratory question
For a proposed contribution, state who or what could use it, in which situation, and what practical difference would matter. The receiving use may be the subject's later Work, a service used by another System, an organization's market or operating contribution, a research use, or a decision that consumes a qualified opportunity.
Keep the receiver's useful difference separate from the subject's desired learning, prestige, procurement, technology adoption, or internal activity. If no receiver or receiving situation can yet be named, keep that question open rather than inventing demand.
An inquiry can still start from material that someone can use or change: permitted records, a working procedure, a model or another construction. Name an available operation and a difference worth examining within the inquiry's resources and permissions. C.40 supplies that development Method. Its result can suggest a receiving use for C.40.NU or C.40.PC to investigate. A useful local result can also finish the inquiry; it need not become an opportunity.
Separate the return time from the development horizon
State when the inquiry's result must be usable and over what horizon the proposed development or its consequences matter. For example, an inquiry can return by Friday about a contribution intended to improve work over four months. The available effort within that window belongs in §4.5: six staff-hours is a resource ceiling, not the inquiry's elapsed duration. Add another change or consequence horizon only when it changes the question or protections.
Name conditions that the inquiry may not silently trade away. Depending on the subject and use, these can include safety, legal or professional limits, privacy, confidentiality, workload, service continuity, accessibility, resource ceilings, reversibility, distributional consequences, or preservation of future options. The practice governing those conditions owns their meaning and evidence.
State the reason to explore and separate its status
Write why exploration is warranted now. Distinguish:
- a qualified receiving problem or demand;
- an observed dissatisfaction or changed constraint;
- a proposed contribution or explicit problem hypothesis;
- a supplier promise, technology signal, or PatternID used only as a search lead; and
- curiosity about a question worth examining.
Preserve competing formulations when they would send the inquiry toward different contributions. Use the problem-structuring pattern when deciding among those formulations is itself the current Work.
Establish available means and permissions
Identify what can be used before the inquiry must return: the performers’ time and capabilities, source access, conversations, computation and other relevant means. Check who can obtain and interpret the proposed result. An available effort ceiling is not yet a reason to spend it; select the budget with the result in §4.6.
Count the work the inquiry would require from everyone affected, including preparation, obtaining the result, interpretation and use. Include delay, displaced work and downside where they can change the choice. A short interview can require substantial preparation or interrupt someone whose contribution is needed elsewhere.
Permission to inspect a source, interview a participant, or run a desk study does not authorize an intervention, access to restricted records, procurement, advice, or implementation. If the proposed inquiry cannot respect a protected condition, select a permitted alternative or return the gap. Keep any stronger evidence condition for a later consequential commitment visible.
Choose a worthwhile first result and enough inquiry
Identify a result that could improve the answer or next inquiry. Depending on what is already known, it may be:
- a change-relevant characterization from
C.40.OC; - materially different search leads from
C.40.NU; - a worthwhile problem and proposed contribution from
C.40.PC; - an admissible advising question or blocker from the applicable professional engagement Method; or
- a qualified result from the practice that owns a missing premise.
Say what plausible answers would change: the next question, a possible contribution, a supported claim, an admissible action or a justified stop. When the future receiver is unknown, use the local exploratory question from §4.2. Discriminating explanations or discovering a useful operation can justify inquiry before an intervention is selected.
Compare the attainable inquiry with what the present basis already supports. A narrower answer, smaller probe, different source, deferral or stop may serve the use better. Compare their whole burdens under §4.5 and the protections, not uncertainty alone. Use C.11.DUA:4.1–4.4 when the connection between the receiving question, evidence demand and worthwhile continuation needs development. Qualitative judgement is enough when it settles that choice.
Select the affordable inquiry whose attainable contribution is worth that burden, or finish with the supported answer now. State what would be enough for this use before spending the selected budget: for example, one evidenced difficulty and a question that distinguishes two plausible contributions. This condition need not settle every unknown. If the inquiry cannot reach it within the permitted means, return the narrower supported result or gap; exhausting a budget supplies no missing evidence.
For an outside-practice contribution, A.15.9:4.1, steps 2–3, connects reuse of an available answer with the decision to obtain more. When acquisition is worthwhile, its later steps bound the request and qualify the returned result for use. Asking for it does not supply it.
Start at that selected contribution. Publication order does not require the earlier PatternIDs to be used.
Return the bounded inquiry or exact gap
Use ordinary prose first. The following compact form can support a handoff when several parties must inspect the same boundary:
This is a local working form. An unsettled future receiver can remain unknown while the local question guides the inquiry. Keep the fields that change the next action or how another reader can use the result.
A complete return can be modest: “By Friday, use three of the six available staff-hours to locate one evidenced incident-handoff difficulty and the next question it warrants, using permitted examples and two brief clarifications. Preserve service workload and confidential records. This is enough to choose what to investigate next for Mira's work over four months; it does not select a course.” Section 5.1 shows why that inquiry earns its budget.
Stop, reuse, and reopen locally
Finish the boundary-setting work when the selected result, worthwhile effort, sufficient return and responsible practice are clear, or when the supported answer or gap already completes the present use. The inquiry itself stops at that sufficient return, a relevant blocker or its agreed limit; extending it requires a changed contribution comparison. Do not widen merely to increase idea count.
Reopen only the affected boundary when the subject, receiver, receiving use, horizon, protected condition, posture, authority, source premise, or effort budget changes. A changed characteristic need not erase an otherwise usable inquiry; a changed receiving use may.
What changes in practice
Instead of starting with a course, tool, provider, maturity label, or long opportunity list, the practitioner can now say what is being explored and why, over what horizon, under which protections, at what cost, and which result is needed next. A proposed contribution also needs its prospective receiving use. That is enough to begin; it does not establish a recommendation or choice.
Archetypal Grounding
Person entering construction without an adviser
Mira manages an engineering team and says, “I should develop in AI this quarter.” She names herself as the development subject, her incident-review and handoff work as the receiving use, and her team and service organization as its receivers. She wants to improve that contribution over four months. By Friday she needs a concrete question for the next improvement discussion. Current service work, confidential records and her ability to continue the role must stay protected. Six staff-hours, counting her time and colleagues' time, can be taken from optional reporting before Friday without shifting required work to others.
An existing summary establishes delayed handoffs but does not locate where they stall. Three permitted incident examples and two colleagues' fifteen-minute clarification slots are available. Mira compares what each feasible inquiry would add:
She selects the three-hour inquiry under C.40.OC. Its sufficient return is one evidenced handoff difficulty, the protected workload limit and the next question that could change the proposed contribution. A tool, a change in review support and personal development remain possible; none is selected from the word “AI.” If the examples cannot support that return, she brings the evidence gap on Friday instead of extending the study automatically.
Changed condition. If the existing summary already locates that difficulty and supports the next question under the same conditions, the half-hour interpretation is enough. She reuses it and ends this inquiry. The unused budget creates no obligation to interview anyone. Neither result requires an adviser or a course choice.
Organization with a fashionable solution
An engineering company announces that it should “become platform-led.” The inquiry identifies the organization and a product-development arrangement, the receiving use as faster and more reliable product-family delivery, a twelve-month horizon with two decision windows, and protected safety, customer continuity, regulatory obligations, and scarce specialist attention.
The provider proposal is recorded as a suggested premise. Strategy owns any commitment, organization-change practice owns changed arrangements, systems and platform engineering own enabling-system realization, and operations owns continuing service. The first missing result is not a platform purchase decision; it is the change-relevant characterization that can distinguish materially different contributions.
Advice can enter elsewhere
A committee already has three qualified directions for an identified facility, a complete characterization, and a current joint-reachability account. Its current need is an independent recommendation. Repeating C.40.OI adds no value. The adviser begins with the applicable professional engagement Method to establish recipient, mandate, competence, conflicts, disclosure, and reliance, then uses current decision-support results. The committee retains the choice.
Population anti-case
A researcher asks, “What should this non-cultural lineage develop next?” No population-local recipient or choice owner is identified. This pattern does not establish either role. Evolutionary description remains with the applicable domain owner. If the researcher is instead deciding whether to fund an experiment, the researcher or sponsor becomes the decision recipient and a separately bounded research opportunity or decision inquiry can begin.
Bias-Annotation
- Solution-first bias: familiar technologies and providers can appear more concrete than the receiving use. Keep them as leads until a contribution is qualified.
- Human-development default: do not transfer learning, career, or capability mechanisms to organizations, AI arrangements, engineered systems, epistemes, or populations.
- Sponsor-recipient collapse: the funder, subject, receiver, adviser, and later choice owner may coincide, but name distinctions only when their difference changes protection, disclosure, authority, or use.
- Open-search prestige: a larger idea list can look more rigorous while avoiding the first discriminating question.
- Benefit-by-word bias: development, progress, and maturity establish no beneficial direction without characteristics, polarity, and evidence.
- Form bias: a completed record cannot compensate for an unidentified subject or the absence of a useful question.
Conformance Checklist
A conforming use or companion account:
- MUST name the subject or provisional subject boundary and claim posture.
- MUST state the receiving use and receiver when known. Otherwise, identify the local question being investigated and keep the future receiving use open.
- MUST distinguish the inquiry's return time, available effort and relevant change or consequence horizon.
- MUST expose protected conditions before candidate selection.
- MUST distinguish qualified grounds, explicit hypotheses, supplier promises, and search leads.
- MUST select an inquiry budget from attainable contribution, whole burden, available means and permissions, or use a sufficient existing answer.
- MUST identify the selected result, its owner or next pattern, and what makes the return sufficient for the present use.
- MUST NOT present the inquiry as a recommendation, choice, programme, Guide, performed Work, observed effect, or intervention authority.
- MUST preserve independent Construction and Advising entry; no PatternID order may be presented as a mandatory lifecycle.
- MUST state a stop or local reopen condition.
- SHOULD use ordinary prose before any compact form.
- MUST return a specific gap or non-use case when the subject or a condition needed for the proposed inquiry cannot be established.
Common Anti-Patterns and How to Avoid Them
Consequences
The bounded inquiry identifies a useful next move and the effort it warrants. A person can explore without an adviser; an adviser can reuse an adequate boundary; and a missing premise remains an explicit question or condition. Request its answer only when the attainable contribution warrants acquisition.
The cost is explicit incompleteness. The result may be only a question, blocker, or narrow source request. Naming the subject, horizon, protections, posture and budget exposes which proposed directions still lack support.
Rationale
The earliest reusable contribution is not generic goal setting or decision-support engagement. Opportunity construction often begins before a decision, option set, or separate adviser exists. Yet a fully unbounded discovery activity cannot distinguish useful exploration from fashion following.
The selected boundary is therefore small: enough identity, use, horizon, protection, posture, and effort to choose one missing result. Characterization, opportunity search, problem construction, support configuration, reachability, portfolio retention, advising, and refresh remain separable because each changes the practitioner's action and can be entered or stopped independently.
SoTA-Echoing
Practice question. How can early opportunity inquiry remain revisable while earning its effort and returning enough to support the next useful move?
Selected best-known line. Phillips, Decision Analysis for Practitioners (2025), develops requisite modelling: revise the problem and model with participants, explore assumptions through sensitivity analysis, and seek enough understanding to move forward without settling every input. Its sections on decision conferences, transitional objects and decision-analysis technology already support exploration before commitment and the emergence of new options. This pattern adopts revisable inquiry and sufficiency from that line.
Useful adaptation here. C.40.OI joins those contributions to development-subject identification from E.10.DEV, inquiry appraisal from C.11.DUA and bounded outside-practice use from A.15.9. Sections 4.1–4.6 connect the subject, local question or receiving use, protections, available means and sufficient return. This supports a self-inquiry, an unsettled future receiver, or a question that an existing qualified result can already answer. None of those entries requires convening a decision conference or constructing a decision model.
Alternative and price. Requisite modelling is a strong choice when participants need to develop shared understanding, examine consequential trade-offs or test sensitivity to disputed judgements. Use the corresponding decision-support practice when it supplies the whole needed result. C.40.OI's smaller return can be useful before that need is established, but it supplies neither the model nor the facilitated shared understanding. Its explicit boundary costs attention too: repeating it around an adequate modelling brief adds work without a useful contribution.
Source roles and limits. ISO 56007:2023 provides the normative opportunity/idea-management boundary before solution development. Reliance here is on its public scope, not an inspected full standard or evidence of effectiveness. Phillips supplies a developed practitioner account and cases, not proof of this Method's comparative performance. The FPF contributions retain their respective semantic, appraisal and reliance roles; subject practices supply the actual development evidence.
Reopen condition. Reconsider this adaptation when a more economical entry supplies the same distinctions about the subject of development, protected conditions and sufficient return, when requisite modelling already closes the local use, or when a relied-on source changes the operation it supplies.
Relations
- Uses:
E.10.DEVto recover the subject, posture, and owner when development wording changes the next action. - Uses:
C.11.DUAto compare attainable inquiry with its receiving value and whole burden;A.15.9to reuse or obtain and qualify an outside-practice result. - Prepares:
C.40.OCwhen selected characteristics and current, intended, protected, or unknown readings are the first missing result. - Prepares:
C.40.NUorC.40.PCwhen the inquiry needs materially different leads or a worthwhile problem and proposed contribution. - Coordinates with: the applicable professional engagement Method when a distinct recommending performer and recipient make advising admissibility current; this pattern does not establish that relationship.
- Returns to: the practice that owns a missing premise, retaining its qualification and authority boundary.
- Hands later comparison and recommendation outward: the applicable decision-support Methods govern formed decision alternatives, comparison, recommendation, follow-up, and choice separation.
C.40.OI:End
Characterize a Subject for a Possible Change
Type: Method pattern
Status: Stable
Normativity: Normative unless marked informative.
Problem Frame
Which characteristics would change this opportunity inquiry?
Use this pattern when a person, organization, community arrangement, AI configuration, engineered system, facility, research programme, episteme, or another subject is described by a maturity label, Method list, role name, benchmark, dashboard, or one score, but the opportunity depends on more specific current, intended, protected, or unknown characteristics. The first useful result is a bounded change-relevant characterization with the unknown or limitation that matters to this inquiry. A measurement, comparison, or source request is a continuation only when its obtainable contribution warrants the whole burden.
Use it after C.40.OI when the subject, receiving use, and horizon were loose, or enter directly when those are already clear. It can serve either opportunity construction or bounded advising. It does not diagnose every property of the subject, choose a direction, prove capability, authorize intervention, or establish that development occurred.
Cheap exit. Reuse an existing characterization when it covers the same subject and configuration, receiving use, horizon, characteristic meanings and scales, applicable time and evidence conditions, protected conditions, and current question. If one domain practice already supplies the complete adequate characterization, use that result and retain its qualification.
Ordinary non-use boundary. Do not build a characterization merely because a dashboard exists. If no characteristic could change the next action, stop. Characterizations of populations, lineages, cultures, organizations, people, AI models, and engineered systems rely on different domain owners and evidence; similar metric labels do not justify transfer.
Development talk often compresses a subject into “junior,” “mature,” “ready,” “high potential,” “advanced,” “safe,” or a composite score. A list of methods or technologies can create the same illusion: possession or use is treated as proof of the needed capability or useful change.
An opportunity needs a more selective account. The practitioner must know which characteristics can reverse the worthwhile problem, proposed contribution, supported direction, reachability claim, portfolio position, advice, or protected condition. The account must preserve distinct result, resource, affected-System or distributional, and explicitly protected coordinates rather than hiding them inside a single number.
This pattern governs that inquiry-relative characterization. Measurement, diagnosis, capability, architecture, safety, financial, legal, professional, organizational, engineering, research, and other domain claims retain their owners.
Problem
Without a change-relevant characterization:
- a maturity word hides several unlike scales and value directions;
- a score is reused without its subject, characteristic, scale, method, uncertainty, or time stance;
- the intended condition is described without a current baseline;
- one desired gain suppresses resource use, affected-System consequences, distributional effects, or protected conditions;
- person evidence is transferred to an organization or AI configuration;
- architecture characteristics are applied when architecture is not the bearer;
- an unknown premise is silently converted to a favorable reading; and
- a later change cannot be traced to the smallest affected opportunity claim.
The resulting direction may sound specific while remaining unmeasurable, incomparable, or unsafe to rely on.
Forces
Solution
Build a characteristic account sufficient for this inquiry, exposing every material unknown and protected coordinate.
Bind the characterization to one subject, use, and horizon
Name the subject and configuration for each claim. If a conclusion needs several, distinguish the Method being designed, the Work that enacted it and the result it produced. Connect their accounts through the relevant use, configuration or occasion, or through a warranted transfer. For an AI case, distinguish model, data, scaffold, tools, oversight, access, and surrounding operating arrangement when their differences matter. For an organization, name the organization or bounded arrangement rather than treating its chart or provider as the subject. For a person, name the person and relevant later-Work context.
State the receiving use, horizon and scenarios. Include the time stance and evidence window needed to interpret or rely on the selected readings; a reported measurement retains its measurement-specific temporal basis. A characterization for today's incident-review Work is not automatically valid for a future management role. A model benchmark in one configuration does not characterize another configuration with the same product label.
If those boundaries are missing, use C.40.OI and return only after they are adequate for the current question.
Select characteristics by decision difference
For each candidate characteristic, ask: if its current reading, intended reading, polarity, or uncertainty were different, could the worthwhile problem, contribution, direction, reachability, protection, advice, or next question change?
Also try changing the relations among the parts while holding their individual readings fixed. If either the answer or its required grounds can change, retain that relation or obtain the information that distinguishes the arrangements. Keep an individual characteristic when it makes a consequential difference alone or in combination. Otherwise omit it from the working account, while naming any intentionally unsearched area whose omission could reverse the result.
Start from several directions when useful:
- the receiver's required result or acceptance expectation;
- resources and burdens needed to obtain it;
- consequences for affected Systems or groups;
- conditions that must not worsen;
- the subject's current limitations or enabling strengths;
- prerequisites, complements, constraints, interactions, and transition conditions; and
- uncertainty or information whose resolution could change the next action.
A generic capability, maturity, readiness, quality, or performance catalogue is a source of questions, not an automatic closed set.
Define each selected characteristic and reading
For every selected characteristic, state in ordinary language:
- what subject or relation bears it;
- what the characteristic means for this receiving use;
- its scale or admissible levels and any unit;
- polarity—higher-is-better, lower-is-better, target-is-best, or no simple order;
- time stance and evidence window when needed for interpretation or reliance, including the temporal basis of any reported measurement;
- current reading and uncertainty, or
unknown; - intended, threshold, exploratory, or comparison reading;
- explicitly protected reading or boundary when current;
- source, method, and qualification needed for reliance; and
- the next claim that this reading can change.
Use C.16 when a value, score, sensor indication, rating, or comparison must be made interpretable as a measurement. A display or source link can report or locate a result; reliance needs its measurement basis and supported relation to the subject. A MethodDescription explains a way of obtaining a result without establishing that the result was obtained.
Orient the complete comparison coordinate set
When a finite candidate change is current, use C.11.CRC to compare an actual or relied-on configuration S0, finite change Δ, and realizable candidate configuration S1 under the same boundary, horizon, scenarios, constraints, interactions, transition basis, evidence window, and receiving decision.
Keep four coordinate families visible:
- result coordinates: what useful results may improve, worsen, or remain unchanged;
- resource coordinates: time, attention, capital, computation, energy, materials and other consumed means;
- affected-System and distributional coordinates: consequences for other Systems, groups, services, or environments and how gains and burdens are distributed; and
- explicitly protected coordinates: values that may not be silently exchanged for a gain elsewhere.
State required authority separately as a condition for proceeding. Include the time, fees or other resources needed to obtain or maintain permission when they matter, without treating permission itself as a consumed quantity.
Do not hide one family inside another or average incomparable scales. A Pareto-improvement claim is warranted only when the declared complete set has no worsening coordinate and at least one improving coordinate under the same basis. Otherwise use only supported language such as beneficial, harmful, dominated, non-dominated, or indeterminate. None of these selects the direction.
Return characteristic meaning and evidence to owners
Use the practice that owns the substantive characteristic:
- human capability and representative later-Work evidence require the applicable capability assessment for that person, receiving Work and permitted support;
- organization contribution, positions, assignments, arrangements, culture, and change remain with current organization and strategy practices;
- AI/model evaluation, data, safety, security, oversight, and deployment claims remain with their engineering and governance owners;
- service, flow, reliability, safety, and operability remain with operations, engineering, safety, and domain practices as applicable;
- architecture characteristics use
C.32.ACSorC.32.ACEonly when architecture is the bearer; - measurement meaning and work use
C.16; and - evidence reliance, provenance, and currentness use their FPF owners.
Record an available MethodDescription as a way a result might be obtained, not as the result. Mark a source-qualified unknown when measurement, access, or currentness is insufficient.
When a reading was established for a different bearer or configuration, rely on it only when the practice that owns the receiving characteristic supplies a qualified transfer basis covering the receiving bearer and configuration, characteristic meaning, scale, intended use, evidence window, and uncertainty. Without that basis, keep the receiving reading unknown and limit the affected opportunity claim. This can finish a useful characterization. Use A.15.9 and C.11.DUA to select an evidence request only when its obtainable contribution warrants formulation, qualification, supplier Work, delivery, interpretation, delay, and displaced work.
Write the bounded characterization
Use ordinary prose first. This compact form can carry a characterization across a multi-practice handoff:
Include only fields that serve the receiving use; an inactive request needs no field or explanation. This local form is not a universal characteristic space, measurement result, diagnosis, criterion verdict, capability profile, decision, or authorization. Several selected characteristics may use different scales and result epistemes.
Choose the next result
Return one of these outcomes:
- an adequate characterization ready for opportunity search or construction;
- a worthwhile obtainable measurement or evidence request tied to one affected claim;
- a comparison question with the complete coordinate set;
- a narrower characterization that supports only part of the opportunity;
- a blocker because a protected or decision-reversing reading is unavailable;
- an explicit unknown that limits the answer while the supported characterization remains useful; or
- a stop because the existing characterization already serves the receiving use.
Continue to C.40.NU when material opportunity-generating dimensions are now visible, to C.40.DC when a direction needs a finite characteristic change and support configuration, to C.40.JF when joint reachability is current, or to the applicable professional engagement Method when a formed advice question needs an admissible relationship. These are conditional continuations, not a fixed order.
Reopen the smallest affected scope
Reopen when the subject or configuration, receiving use, horizon, scenario, characteristic meaning, scale, polarity, measurement method, evidence window, protected condition, or decision-relevant unknown changes. Trace that change to the smallest affected opportunity, comparison, or advice claim.
Do not discard unaffected readings merely because one characteristic changed. Do not retain a composite maturity score when its underlying meaning or weighting changed.
What changes in practice
The practitioner replaces “more mature,” “AI-ready,” “senior,” or “platform-capable” with the characteristics that alter this inquiry. Current, intended, protected, and unknown readings stay distinct. Missing evidence limits the affected reading; a worthwhile further contribution becomes a directed request. Later comparison retains benefits, burdens, affected Systems, distribution, and protections on their own scales.
Archetypal Grounding
Person and representative later Work
Mira's bounded inquiry concerns incident-review and handoff Work over four months. “Become better at AI” is discarded as a characteristic. The selected characteristics are: time to identify a material handoff gap, quality of the explanation supplied to the next reviewer, ability to detect unsupported model output, confidential-data exposure, review workload, and transfer to two representative incident classes.
The current account has qualified examples for explanation quality and workload, but no comparable evidence for transfer or unsupported-output detection. Those readings remain unknown. The intended contribution is faster detection without worsening explanation quality, confidentiality, or sustainable workload. The applicable human-capability practice supplies the human demand, capability, support, and transfer evidence. The characterization returns those useful readings and unknowns; it supplies no course recommendation or transfer claim. Representative-Work evidence becomes a request only when the attainable contribution warrants acquisition for the next use.
Organization without one maturity score
An engineering organization is called “low maturity” because releases are slow. The inquiry concerns reliable product-family delivery over twelve months. The characterization separates release lead time, rework, service interruption, platform reuse, security exposure, scarce-specialist attention, team autonomy, and customer-impact distribution.
Lead time has a lower-is-better quantity scale under a stated window; team autonomy has an ordinal scale and no ratio arithmetic; security exposure uses separately governed risk results. The intended configuration may improve lead time while worsening specialist bottlenecks or security. The account therefore remains multi-coordinate and indeterminate until organization, platform, operations, and security owners provide the missing premises.
AI arrangement and configuration identity
A team wants to “improve the model.” The subject is a model edition with a retrieval corpus, tool permissions, evaluator, human escalation path, and operating environment. Selected characteristics include task success under the declared evaluation, false-action rate, explanation adequacy, data exposure, escalation latency, compute use, operator workload, and robustness to changed inputs.
A benchmark from the bare model does not characterize the deployed arrangement. Human learning evidence does not establish model adaptation. The next question is whether the current evaluation covers the intended tool-enabled configuration and protected safety/security conditions.
Already adequate characterization
A facility team already has current, source-qualified service, safety, operability, resource, and affected-community characteristics for the same configuration and horizon. No meaning, scale, or evidence window changed. The team reuses the account and proceeds to joint reachability. Reformatting it into this pattern's compact form would add burden without changing action.
Bias-Annotation
- Score realism: a number can feel more objective than its method, scale, weighting, and uncertainty warrant.
- Maturity ladder bias: one ordered label invites a universal sequence and higher-is-better assumption.
- Human transfer bias: person learning and capability measures do not characterize organizations, AI arrangements, or engineered systems.
- Architecture projection: architecture characteristics are tempting general quality labels; use them only when architecture is the bearer.
- Available-data bias: easily measured coordinates can crowd out protected or distributional consequences that matter more.
- Benefit-only selection: desired result coordinates can hide resources, affected Systems, transition burdens, and option losses.
- Unknown-to-zero bias: missing evidence is neither absence nor a favorable baseline.
- Form completion bias: a full table does not supply qualified measurements or owner conclusions.
Conformance Checklist
A conforming characterization or companion account:
- MUST identify the subject and configuration, receiving use, horizon and scenarios, together with any time stance or evidence window required to interpret or rely on the selected readings.
- MUST select each characteristic by an explicit way it can change the current inquiry or a protected condition.
- MUST state the bearer, meaning, Scale or levels, Unit when applicable, and polarity for every relied-on characteristic. Include any time stance or evidence window required by its value, support or measurement claim.
- MUST distinguish current, intended or exploratory, protected, and unknown readings.
- MUST keep result, resource, affected-System or distributional, and explicitly protected coordinate families visible when finite comparison is current.
- MUST NOT force incomparable coordinates into one scalar or assert Pareto improvement without a complete oriented set under one basis.
- MUST identify the owner, source, qualification, and uncertainty for every decision-reversing reading.
- MUST NOT infer measurement Work or a qualified result merely from the presence of a MethodDescription, display, benchmark score, source link or tool output. MUST establish the measurement and evidence basis required for the receiving claim, reusing an adequate qualified result.
- MUST return each missing premise as a bounded request, unknown, narrower claim, blocker, or stop.
- MUST retain a next measurement, comparison, pattern, owner use, or local reopen condition only where it changes the recipient's continuation or interpretation. A sufficient characterization creates no further-work or omission-record obligation.
- MUST NOT present the characterization as a chosen direction, recommendation, programme, authorization, performed Work, or observed effect.
- SHOULD make the first useful result readable without requiring the compact form.
Common Anti-Patterns and How to Avoid Them
Consequences
Practitioners gain a characterization that is smaller than a full diagnosis yet strong enough to guide opportunity construction, source requests, finite comparison, or advice. Several scales and unknowns can coexist without being presented as one score. Protected conditions become visible before selection, and later premise changes can reopen only the claims that depended on them.
The cost is that a familiar word such as maturity may unfold into several claims with different owners and evidence windows. Some apparently precise numbers become conditional or unusable. Completing the account may require measurements or specialist results that the current inquiry cannot afford.
The characterization supports the inquiry. Obtaining a measurement, diagnosing a condition or changing the subject requires the applicable operation and evidence.
Rationale
The common contribution is not one cross-holder scale. It is the action of selecting and orienting characteristics for one subject, receiving use, horizon, and opportunity question, then returning their substantive meanings and evidence to owners.
In opportunity construction, use the characterization to vary opportunity-generating dimensions and specify finite candidate changes. In advising, use it to recover premises that can reverse a comparison or recommendation.
SoTA-Echoing
Practice question. How should practitioners replace a development maturity label or isolated score with a characterization that can guide a finite opportunity decision without inventing one universal scale?
Contributing Methods. FPF C.16 supplies the measurement chain: measurand or subject, Characteristic, Scale, method, model, calibration, dated Work, attributed result, uncertainty, and bounded later use. C.11.CRC supplies the finite-change orientation: actual S0, finite Δ, realizable S1, complete result/resource/affected-System/distributional/protected coordinates, interactions, transition basis, uncertainty, and option effects. This pattern adapts those lines at C.40.OC:4.2–4.7 by selecting only characteristics that can change one opportunity inquiry and by returning holder-specific meaning and evidence outward.
Alternative or default. A maturity model or composite readiness score can provide a compact summary when its construct, scale, weighting, evidence, and decision use are already qualified. When such a summary hides a distinction that can change the present inquiry—scale legality, trade-offs, protected coordinates or transfer—retain the separate coordinates. They keep those distinctions inspectable, at the cost of a less compact headline and a possible indeterminate result.
For organization and campaign cases, ISO 56008:2024 is a current opportunity/innovation measurement source and supports explicit measurement and portfolio review. It is an organization-scope comparator, not a universal person, AI, or engineered-system characterization Method and not effectiveness evidence.
Source roles and limits. C.16 and C.11.CRC supply measurement and finite contribution comparison. ISO 56008 is a bounded normative comparator. Holder, domain, safety, professional, legal, and evidence practices supply the characteristic meanings and results; this pattern does not outrank them.
Reopen condition. Reopen this stance if a current alternative demonstrates a lower-burden characterization that preserves subject/configuration identity, scale and uncertainty, the complete coordinate set, protected conditions, holder-specific ownership, and local reopenability, or if a relied-on measurement or finite-comparison pattern changes those obligations.
Relations
- Receives from:
C.40.OIthe subject, receiving use, horizon, protected conditions, posture, and effort boundary when those were not already adequate. - Uses:
C.16for relied-on measurement claims andC.11.CRCfor finite configuration-relative contribution comparison. - Prepares:
C.40.NUby exposing opportunity-generating characteristics and relevant unsearched dimensions. - Constrains:
C.40.PCandC.40.DCby making the receiving problem, contribution, finite change, and protections inspectable. - Supplies premises to:
C.40.JF,C.40.OP, the applicable professional engagement Method, andC.40.RQwithout establishing reachability, portfolio admission, advising admissibility, or refresh outcome. - Returns to: capability owners for human results; organization, strategy, engineering, operations, AI/model, safety, legal, professional, research, evidence, measurement, or other owners for their characteristics and conclusions.
C.40.OC:End
Search Neighboring Uses and Means
Type: Method pattern
Status: Stable
Normativity: Normative unless marked informative.
Problem Frame
Use this when an opportunity inquiry keeps returning to one course, provider, technology, role, or familiar future, while a different receiving use or support could reveal a more useful contribution. Also use it when idea generation is expanding but nobody can say what difference the next search should expose or when to stop.
Start with one bounded subject, a question worth investigating, a horizon, protected conditions, and an effort limit. The question may concern a known receiver or what a change to available material could reveal. Ask what could change if the receiving use, the way of contributing, a relevant characteristic, or the support were different. The first useful result can be a small set of materially different leads with their generating operations, sources and next question. An examined local result can also be enough to stop this branch.
This is opportunity-generating search: the worthwhile problem and target contribution may be provisional. A lead is a possibility worth examining, not yet a supported need, complete direction, or reachable change. A separate adviser is unnecessary.
Ordinary non-use boundary. Reuse an adequate lead or direction set when another search is unlikely to change the current inquiry. If a bounded decision-support question already needs a finite alternative set, use an applicable Method for generating decision alternatives. If one domain practice owns the whole discovery question, use it. If the only question is a requirement's justification, use C.11.DUA with its domain basis without reopening opportunity search. Do not open a generation archive merely to remember two possibilities.
Problem
A search can inherit its answer from the first suggestion. Searching for better courses still assumes that learning is the needed change; searching for better models still assumes that the model is the limiting part. Different brands are then mistaken for different opportunities.
Unbounded brainstorming fails differently. It accumulates possibilities without a recoverable receiving use, generating operation, source limit, or next discriminating question. Idea count and novelty replace usefulness; checking every lead consumes the time intended for development.
The practitioner needs enough contrast to expose consequential unknowns, without claiming an exhaustive possibility space or making evidence qualification a prerequisite for imagining anything.
Forces
Solution
Vary the few dimensions that could change the opportunity, make the resulting differences inspectable, and stop once the inquiry has a useful result or a specific limit.
Choose the search question and its limit
Recover the subject and configuration, the receiving concern when known, the local question, horizon, protections, and effort boundary. Use C.40.OI only if these are too loose to distinguish a useful lead from a slogan. Use C.40.OC when an unclear characteristic, scale, or current reading prevents a productive contrast.
Write a question such as: “What different contribution could help the receiving engineer recover unsupported acceptance claims within this quarter?” The question may explore what is worthwhile; it need not name one fixed target or preferred solution.
Name what may be investigated at the chosen effort and with the available permissions. A short source search and a permitted conversation may be enough. If protected records, experiments, or professional assessments are needed, return the corresponding access or specialist request before doing that work.
Search from more than one direction when it can change the answer
Use the following search prompts selectively.
Do not lift a binding safety, confidentiality, legal, or authority condition merely to generate a convenient lead. If the requirement's own justification is the live question, use C.11.DUA §4.3 with the competent domain contribution. Identify the protected person or System and outcome, governed quantity or activity, evidence and threshold basis, marginal protection, full burden and displaced protective work. Compare retaining, tightening, revising or replacing it only where those are meaningful alternatives.
Return the supported requirement recommendation and the reason or limit needed for this opportunity inquiry. Keep its present force, amendment authority and feasible window distinct. The appraisal can finish when timely amendment is unavailable; current leads still inherit the original binding limit. A possible later revision does not restore the opportunity lost under the present conditions.
Use explicit value questions: what would matter to a receiver, what burden should diminish, and what must remain protected? When group advocacy could anchor the search, let participants first propose possibilities independently, then use their differences to extend the set. Neither a workshop nor a separate group is mandatory.
Keep the operation, source, and claim status together
For each retained lead, say what was varied, where the idea came from, what concrete difference might matter, and any material limit or unsupported premise. Preserve a source date or edition when its currentness affects the next use. Distinguish the source's claim from the inference made for this subject.
For example: “The supplier describes long-record retrieval; changing the receiving use from automatic approval to claim-to-evidence recovery suggests a possible review aid. We have no result establishing its accuracy or permitted use with these records.” That is a search lead with a testable connection, not a supplier-validated opportunity.
Independent-looking pages that repeat one announcement supply one lineage of suggestion. A demonstration for another person, organization, model configuration, or environment is an analogy until the receiving practice qualifies its use. Use C.40.OC for a relied-on characteristic transfer and C.40.JF for a material reachability premise.
Retain meaningful contrasts without prematurely ranking them
Keep a lead when its difference could change the receiving problem, proposed contribution, support, burden, protection, horizon, or next evidence request. Merge duplicates whose differences change none of those. Do not discard a distinct useful question solely because another lead has a more impressive demonstration.
Keep the current way, a smaller repair, or no new development when each is a serious answer. “No new development” still has the continuing Work, costs, and exposure of the baseline.
Mark leads as ready for problem/contribution inquiry, conditional on a named premise, rejected for this use, or retained for a separately stated future question. These are working descriptions, not scores or a universal classification. The strongest available result may be that none of the searched leads yet identifies a worthwhile contribution.
Stop at a useful difference or a gap
Stop widening when the retained contrasts expose the unknowns that can change the inquiry at its chosen effort, when an adequate set already exists, or when a missing premise makes further generation uninformative. If the effort limit is reached first, report that limit and any consequential unsearched area; do not describe the search as sufficient merely because the time was spent.
When the receiving inquiry needs a further answer, state the discriminating question and how plausible answers would change the leads. A sufficient local result or adequate lead set can end the search. Use C.40.PC when the missing result is the receiver's worthwhile problem or proposed contribution, and C.40.DC when that result is already adequate and directions need support configurations. A return to one missing frame or characteristic is also a complete outcome.
A two-sentence return can suffice: “Changing the use and support exposed claim-to-evidence recovery and shift-exception handoff as different leads; three other products repeat the first mechanism. We stopped after the agreed source search because the next difference depends on whether receivers need either contribution; accessibility beyond the current office use remains unsearched.”
Use formal search machinery only for the questions it answers
Use C.17 when novelty, creativity, or value must be characterized for a named bearer, comparison basis, and objective. A novelty result alone does not establish worth for this opportunity inquiry. Use C.18 when a generation archive or front needs its generator, operators, descriptors, lineage, retention, or comparison basis. Use C.19 when the question is how to treat a live pool under its declared policy: widen it, keep its frontier, narrow it to a subset, or sunset a line. Selecting the next local candidate or probe under an unchanged policy is not that question.
Visiting a new region does not by itself expand the admissible possibility space. A stronger expansion claim needs the earlier and changed generation conditions and evidence of what a new operator, access relation, or building block makes available. Ordinary neighboring search need not make that claim.
Reopen only the affected search when a source, protected condition, receiving use, or generating assumption changes. C.40.RQ traces the effect into retained opportunities and downstream advice.
An available construction can therefore be an entry before a receiver or final contribution is known. Use C.40 when examining a feasible variation can disclose a useful continuation. Use C.40.CD when a changed question requires changing the way of answering it, and the attempted way can change the next question. Bring a suggested application back to C.40.PC for its receiving-use inquiry. Neither route requires performing a variation when an adequate lead already exists.
Archetypal Grounding
A model announcement becomes two different questions
This constructed case stipulates an engineering-services unit's ninety-day inquiry and a two-hour source-search allowance. Confidentiality, independent acceptance, and continuing service are protected. An announcement says a model can process long records; the unit has not established a worthwhile development problem.
The first search finds three model products with similar advertised retrieval functions. This changes a possible supplier, not the receiving contribution. A use contrast then separates two possibilities: helping an acceptance engineer recover unsupported claims, and helping shift staff recover unresolved exceptions. A support contrast adds a manual index to the first and a clearer human handoff to the second.
The return is two leads, each with a non-model possibility and a distinct missing premise. The acceptance lead needs a receiver's account of the qualification gap and purpose-specific record access. The shift lead needs the relevant exception problem, receiving-holder support, and attention limits. Neither demonstration establishes those premises.
The team stops widening within its allowance because receiver inquiry can now distinguish the two leads more usefully than another product search. It records that non-office accessibility uses were not explored. The next question—whether either receiver identifies a consequential unmet contribution—is the C.40.PC problem/contribution inquiry, unless an adequate existing account already answers it. Its answer can qualify, narrow, or reject the hypothesis. A positive answer alone establishes neither a feasible direction nor a commitment; rejection concerns this use, not every possible use of the technology.
A person searches contributions, not course names
Mira has four months to explore a next contribution in engineering Work. She contrasts evidence review, where the receiver needs unsupported claim-to-test links found, with interface coordination, where specialists need incompatible assumptions surfaced. Reading technical documents occurs in both; their receiving results differ.
She also varies support. For evidence review, permitted examples exist but representative capability and transfer evidence are missing. For coordination, access to participants and competent feedback are not established. The two leads retain different missing premises. Either conditional lead can finish this inquiry; request further evidence only when its obtainable contribution warrants acquisition through A.15.9 and C.11.DUA. A course catalogue may suggest a resource, but cannot merge the contributions or fill either gap. Mira can stop with these questions without engaging an adviser.
An adequate facility set is reused
A facility team already has current manual-repair, sensor-replacement, and changed-maintenance candidates for the same service concern. The remaining uncertainty is the safe outage window. Another technology search would not change that question, so the team requests the outage result and does not use this pattern to enlarge the set.
A useful question emerges from available material
A constructed workshop has permission to examine four records and an available sorting operation. The records are ordered by job: A starts at 09:00 and finishes at 09:08; B starts at 09:03 and finishes at 09:07. No future service or customer has been selected. A five-minute exploration can ask what a chronological view exposes.
Sorting by time produces A-start, B-start, B-finish, A-finish. The view makes a four-minute overlap visible. Keep the view and its sorting rule for the next question: could seeing overlapping jobs help someone coordinate the workshop? C.40.PC can investigate that possible receiving use. The overlap alone establishes neither a coordination problem nor competition for one resource.
If the workshop already has an adequate view and no worthwhile further question emerges, stop this branch. The examined result remains a local answer; no opportunity account or service proposal is needed to complete it.
Bias-Annotation
- Supplier anchoring: product variety hides the unchanged receiving-use assumption.
- Novelty preference: an unfamiliar idea displaces a useful small repair without a value reason.
- Source-count illusion: repeated announcements look like independent evidence.
- Available-support bias: an offered course or tool defines the future before the needed contribution is understood.
- Constraint wishfulness: an inconvenient protection is treated as removable without its owner's result.
- Search-completion bias: a spent budget or full table is mistaken for adequate coverage.
Conformance Checklist
A conforming search return:
- MUST name the subject, local question, horizon, relevant protections, and effort boundary, together with the receiving concern when known.
- MUST identify the generating operations and the differences that make retained leads consequential.
- MUST distinguish source claims, local inferences, analogy, and missing evidence.
- MUST NOT equate source count, idea count, novelty, or a supplier demonstration with local worth or reachability.
- MUST preserve binding constraints while separately identifying assumptions that may be investigated. A live merits question returns its supported judgement of protection and burden and the amendment limit, without treating the recommendation as a changed rule.
- MUST state the stopping reason and any unsearched area whose omission can change the receiving use.
- MUST return a useful local answer, a next question with action-changing answers, an adequate lead set, rejection, or a specific gap.
- MUST NOT require a workshop, exhaustive search, archive, live pool, recommendation, or selected direction for an ordinary search result.
- SHOULD reuse an adequate set and retain a serious baseline or smaller repair when relevant.
Common Anti-Patterns and How to Avoid Them
Consequences
The practitioner can explain how a search changed the possibilities under consideration and why it stopped. An old capability with better support can remain alongside a new technology. Rejected or unresolved leads retain reasons that make later reconsideration local.
The cost includes generating contrasts and retaining their source and difference basis. Search can end without a positive opportunity, and relevant unsearched areas may remain. The result supports the next inquiry; it supplies neither the receiver's value evidence nor the subject's feasible development change.
Rationale
Searching along several consequential dimensions loosens the first proposed solution without requiring limitless creativity. Naming the generating operation makes a lead criticizable: a receiver can question the use, a specialist can challenge the transferred characteristic, and a support owner can identify the unavailable relation.
Stopping at the next discriminating question converts breadth into a usable return. A missing premise then guides work instead of being hidden by more suggestions.
SoTA-Echoing
Practice question. How can an opportunity inquiry discover useful possibilities before its problem and target result are fixed, while keeping effort and reliance bounded?
Best-known line for this use. Value-guided contrastive search, coupled to source-qualified return, supplies the useful move. ISO 56007:2023 places opportunity and concept work before solution development and deployment and rejects one universal tool set. Its public scope supplies a normative boundary, not evidence that this search succeeds. Keeney's value-focused brainstorming supplies the historical generation anchor: use explicit values to generate alternatives and, when groups are involved, obtain individual ideas before group anchoring. This pattern adapts those contributions in :4.1–:4.5 to provisional receiving uses, explicit operations, protected constraints, and proportional stopping.
Comparison at comparable effort. A short supplier-menu search is efficient when the needed contribution is settled. In the constructed model case, the same bounded inquiry obtains more consequential differences by varying use and support than by retaining three brands. Unstructured brainstorming can reveal surprises but leaves the practitioner to reconstruct source limits and the next question. Adopt the contrastive move for that unresolved-use situation; retain a menu or domain Method when it already answers the question. The cost is explaining each retained difference instead of reporting only names.
Current neighboring contribution. An applicable decision-alternative generation Method supplies a finite alternative set for a bounded decision-support question; C.18 supplies archive/front stewardship. Neither result is required merely to perform this earlier search. Their boundaries prevent a search lead from being promoted by its position in a list or archive.
Source limits and reopen. The source combination is guidance, not comparative effectiveness evidence or a universal development mechanism. Reopen the affected move if a stronger current generation practice produces useful contrasts with less burden while preserving source limits, protected conditions, and a discriminating stop, or if a relied-on source changes those obligations.
Relations
- Receives from C.40.OI: inquiry scope, protections, horizon, and effort when missing.
- Uses C.40.OC: characteristic meanings and qualified readings that make a search contrast interpretable.
- Supplies C.40.PC: leads whose worthwhile problem and proposed contribution need inquiry.
- Supplies C.40.DC: leads with an adequate contribution premise that need distinct support configurations.
- Returns to decision-alternative generation: finite alternative generation when a bounded decision-support question is the receiving use; this is a conditional return, not the next numbered stage.
- Uses C.17, C.18, or C.19: only for the novelty/value characterization, archive/front stewardship, or live-pool treatment policy each governs; local candidate/probe selection is a different question.
- Uses C.40.JF and C.40.RQ: qualification of a material reachability premise and tracing of changed premises into retained uses.
- Returns to domain and holder practices: local need, feasibility, capability, access, safety, and other substantive results that search cannot supply.
C.40.NU:End
Construct a Worthwhile Problem and Proposed Contribution
Type: Method pattern
Status: Stable
Normativity: Normative unless marked informative.
Problem Frame
Use this when a development lead names an attractive niche, technology, course, service, or provider promise, but the receiving use and the reason to change it remain unclear. Someone may say “we need AI,” “become a leader,” or “offer faster analysis” without identifying whose result is unsatisfactory, what difference matters, or what could be supplied.
Recover from the receiver's account or adequate existing evidence what is missing or foregone in one bounded use. Relate that problem to a proposed contribution and an inspectable acceptance expectation. The first useful result is a supported worthwhile problem or an unconfirmed problem hypothesis, together with the proposed contribution, its material limits and, when needed, a question that could confirm, narrow or reject it.
The subject that might develop and the receiver of its later contribution can differ. A person may explore their own future Work without an adviser; an organization may investigate a contribution for a client; a research programme may seek an epistemic contribution rather than a commercial offer.
Ordinary non-use boundary. Reuse an adequate problem/contribution account when its receiving use, evidence, protections, and horizon remain current. Proceed to C.40.DC if only the way of obtaining that contribution is missing. Use an applicable problem-structuring Method when a bounded decision-support inquiry needs materially different problem formulations. This pattern does not establish a commitment, choose a direction, or justify intervention by itself.
Problem
A niche, a worthwhile problem, and a proposed contribution answer different questions. A niche is a possible fit between a contribution and a use. The problem explains why an inadequate or foregone result matters there. The contribution says what could be supplied and how the receiver would recognize it.
If these are collapsed, a supplier's enthusiasm is treated as demand, a proposed offer as a commitment, and a useful intermediate result as a promise of benefit. “Recover missing evidence” is read as “approve faster,” although neither the time saving nor the approval is established. Alternatively, a practitioner rejects a useful problem merely because no feasible way to address it is known yet.
The needed move is to make the value and contribution claims separately inspectable, preserving uncertainty without losing a worthwhile next question.
Forces
Solution
Write a small, source-qualified account of why a change could be worth pursuing and what contribution would answer it. Keep a supported problem, an open hypothesis, and a rejected lead visibly different.
Recover the receiver's situation
Name the subject of the proposed change, the prospective receiver, the later Work or use, the horizon, and the relevant conditions. Ask what result is currently inadequate or unavailable, what happens if nothing changes, and which evidence or participant account supports that description.
Use qualified evidence about actual Work or receiving demand when it exists. A receiver's report can establish their stated concern; its causal explanation, scale, representativeness, and effects on others may still need qualification. Do not infer organization-wide demand from one person's preference.
If the inquiry cannot yet identify a receiving situation, use C.40.OI or C.40.NU to recover a worthwhile next question. When usable material and an available operation permit a revealing local change, their C.40 entry lets that inquiry proceed before its future application is settled. A use suggested by the result is then a problem hypothesis to examine here. Keep an unknown receiver explicit; do not invent a customer to complete the account.
Separate the problem from the proposed answer
Describe the missing or inadequate result without building in one technology, provider, training programme, or staffing answer. Then state the proposed contribution separately.
For example:
- Problem: an acceptance engineer cannot recover which submitted claims lack a sufficient evidence basis before the handoff review.
- Proposed contribution: an inspectable claim-to-evidence account that identifies supported claims, missing premises, and qualification limits in the permitted source set.
- Acceptance expectation: the reviewer can account for the submitted claims covered by the order, trace support where it exists, and identify missing or inadequate support under the review criteria.
Faster approval is a different benefit claim; automatic approval is a different contribution and authority question. Preserve those differences if they become live rather than silently extending the proposal.
Retain rival problem formulations separately when they change the subject, boundary, evidence, intervention family or affected interests, and make those differences explicit. Obtain a bounded problem-structuring contribution through A.15.9 if the needed formulation cannot yet be constructed from the available grounds. A change of wording that leaves those consequences identical does not create another worthwhile problem.
Establish what makes the problem worthwhile
Ask who is affected, why the missing result matters, at what scale and horizon, and under what burdens and protections addressing it could be worth pursuing. Include material contrary evidence and consequences for affected people or Systems, not only the proposed beneficiary.
Use the domain practice that owns each relied-on claim. A human capability need requires representative later-Work evidence about the person and the permitted support; a service failure belongs with the relevant operations and domain evidence; safety, access, finance, legal, and other specialist claims retain their own qualification. Use C.40.OC when the value account relies on unclear or incompatible characteristics.
Return the strongest supported statement:
- a worthwhile problem for this receiving use under stated conditions;
- an attributed or contested value claim requiring a named resolution;
- an unconfirmed problem hypothesis with the confirming question; or
- rejection or revision of the lead because no worthwhile problem is supported at the chosen scale.
A receiver's desire alone does not defeat a binding protection. Conversely, missing feasibility does not disprove the problem: retain the worthwhile problem and name the unresolved way of addressing it. For a comparative claim such as “worth the transition cost,” obtain the comparison rather than treating this inquiry as its proof.
State the proposed contribution and its acceptance expectation
Say what could be supplied, for which receiving use, under what conditions, with what exclusions, and what the receiver could inspect to determine whether that contribution was supplied as proposed. Keep the proposed result, the acceptance criterion, and any acceptance separate.
Use A.2.3 when a promise's content must be identified or relied on. Read an existing provider promise from its source; mark a newly constructed promise as proposed. Neither description establishes a provider commitment, capability, access, provision, fulfilment, or achieved benefit.
A research contribution can be a criticism, a discriminating experiment, a reproducible comparison or a clarification. Specify the claim the inquiry should settle and what would count as evidence for it. A single admissible occurrence can establish possibility or disprove a claim with no exceptions. To establish how often something occurs, or that it holds throughout a class, use evidence that bears on that claim. The research result can be useful in its own right.
Link the contribution to the problem without assuming success
Explain why the proposed result could address the missing receiving result. Mark that connection as supported, conditional, disputed, or hypothetical according to its evidence. Identify what a plausible contrary answer would change.
When a consequential uncertainty remains, ask a discriminating question whose attainable answer warrants the inquiry. “Does the reviewer need a source-traceable account of unsupported claims, or is the delay caused by unavailable acceptance authority?” can redirect the contribution. “Would you like better AI?” cannot distinguish the problem from the preferred solution.
Use existing adequate results before requesting new Work. When another practice must supply an answer, use A.15.9 for the bounded result request and reliance boundary. The supplying practice answers its own substantive question. A proposed interview, measurement, or experiment needs the permissions and conditions applicable to that activity.
Return the account at the strength it supports
Ordinary prose is enough when the receiver can recover the situation, worthwhile difference or hypothesis, contribution, acceptance expectation, evidence limits and any next question needed by the use. For example:
The case's acceptance engineer reports that unsupported claims cannot be recovered reliably from the incoming package, and the stipulated review records support that concern. We propose a source-traceable claim-to-evidence account, not approval; the next question is whether the available records and review criteria cover a representative package without breaching confidentiality.
Use C.40.DC when the contribution premise is adequate and materially different directions need construction. Return to C.40.NU when the lead must change. When the assignment is to formulate a problem or decide whether this lead merits attention, a problem hypothesis or a reason to close the lead can complete it. If an adequate direction already exists, proceed to the missing reachability, comparison, or advising result; no mandatory cycle through earlier PatternIDs follows.
Reopen the affected problem or contribution under C.40.RQ when the receiving use, value evidence, acceptance expectation, source, horizon, or protected condition changes. A narrower permitted source set can narrow one proposed contribution while leaving the receiver's problem and another direction intact.
Archetypal Grounding
Qualification support is not faster approval
This constructed example stipulates an engineering-services unit's ninety-day inquiry. Its acceptance engineer reports repeated difficulty finding unsupported claims in incoming evidence packages; the example's review records corroborate that bounded concern. Confidentiality, traceability, independent acceptance, and tolerable review burden are protected.
The technology lead was “use a long-record model.” The problem/contribution account instead proposes a claim-to-evidence account for one class of package. The acceptance expectation is that the reviewer can recover the claims the order covers, their available support and any unresolved premises. A limited review can report its findings when it identifies the covered subset. A manual index and bounded retrieval assistance could both serve that proposal.
No result yet establishes faster approval or representative model accuracy. The first useful return preserves the worthwhile qualification problem while marking the processing route as conditional on purpose-specific access and evaluation. If the data owner permits only public, non-representative samples, a narrow demonstration remains possible; the original representative-use claim remains unsupported.
A rival lead concerns shift exceptions and the next authorized handoff action. That is a different receiving problem, not a second implementation of the same qualification account. The two remain distinct for later direction construction.
Receiver inquiry rejects a supplier lead
A supplier offers a dashboard to a small unit as an “AI transformation.” In this constructed case, the unit's current result meets its acceptance needs, its receivers identify no consequential unmet result at the proposed scale, and the proposal adds review burden. No independent protected or foregone-result concern is established.
The inquiry rejects this lead for the current use. It does not claim that the technology is universally useless. Retaining the idea for a different receiving question would require that question and a retention reason, not relabelling the rejected proposal as a worthwhile opportunity.
A research problem remains a hypothesis
A research group suspects that an evaluation loses important failure cases when several task types are pooled. The proposed contribution is a reproducible comparison between the pooled and separated analyses, with explicit limits on what any difference would establish.
First establish what the group wants to know. If it needs a conclusion for this dataset, return the two analyses and explain whether pooling changed that conclusion. If the separated analysis reproducibly reveals an important failure that pooling concealed on an admissible dataset, that dataset establishes that such a loss is possible. A change in the numerical summaries alone can leave every important failure visible. Agreement on that dataset leaves the possibility open elsewhere. A claim about how often the loss occurs needs evidence about the population of datasets. If the requested comparison answers the group's question, the work can end there. The group can choose a further inquiry when its answer would matter.
Bias-Annotation
- Solution-first framing: the supposed problem contains the technology or intervention it is meant to justify.
- Receiver-preference inflation: one person's interest is treated as established demand at another scale.
- Benefit inflation: an inspectable intermediate contribution is described as a guaranteed downstream improvement.
- Supplier authority: an offer is taken as evidence of need, capability, or fulfilment.
- Commercial projection: a research or community contribution is forced into a sales promise.
- Feasibility contamination: an unknown solution causes rejection of a worthwhile problem, or a feasible tool creates a supposed need.
Conformance Checklist
A conforming problem/contribution account:
- MUST identify the subject of the proposed change, receiving use, horizon, and relevant protected conditions.
- MUST distinguish the inadequate or foregone result, why it matters, and the proposed contribution.
- MUST distinguish a supported worthwhile problem, an attributed or disputed claim, an unconfirmed hypothesis, and rejection where those outcomes are live.
- MUST preserve material contrary evidence, affected-System consequences, and burdens that can reverse the worth claim.
- MUST state an inspectable acceptance expectation without reporting acceptance or achieved benefit.
- MUST distinguish a proposed promise from an existing promise and both from commitment, provision, and fulfilment.
- MUST expose the evidence or hypothesis linking the contribution to the problem, and a discriminating question when a material uncertainty remains.
- MUST NOT infer need from novelty, a supplier offer, or solution feasibility, or reject a supported receiving problem solely because solution evidence is missing.
- MUST return to the next missing result or stop; a direction, recommendation, choice, or intervention is not required to complete this account.
Common Anti-Patterns and How to Avoid Them
Consequences
The practitioner can distinguish what might be worth addressing from what could be supplied. A useful hypothesis can survive without invented evidence, and a supplier lead can be rejected before it consumes a development programme or procurement budget. Later direction construction has an inspectable premise rather than a slogan.
The cost is receiver and evidence inquiry. Different people may value the proposed change differently, and the available evidence may justify only a narrower problem or an unanswered question. This pattern makes those limits visible; it does not settle domain truth, comparative value, or authority by naming an opportunity.
Rationale
Separating problem, proposed contribution, and acceptance expectation lets each be challenged without discarding the others. A receiving difficulty can remain valid when one delivery route fails. A contribution can be precisely described while its worth remains hypothetical. The distinction prevents both premature commitment and premature rejection.
An explicit confirming question gives provisional value a practical use. It says what the next answer would change instead of treating uncertainty as either permission to promise or a reason for indefinite research.
SoTA-Echoing
Practice question. How can a promising lead become a worthwhile receiving problem and proposed contribution without importing the supplier's preferred solution or overstating benefit?
Best-known line for this use. Combine receiver-grounded problem inquiry with an explicit problem/contribution distinction. ISO 56007:2023 supplies the bounded opportunity/concept front-end before solution development. Problem-structuring Methods supply materially different formulations and their evidence consequences; A.2.3 supplies promise-content identity without conflating commitment and fulfilment. This pattern adapts those contributions in :4.1–:4.6 to one prospective development contribution and its confirming return. Domain evidence, rather than the source combination, supports the worth claim.
Serious alternative and trade-off. An offer-led brief is cheaper when a qualified receiving need and acceptance basis already exist. When they do not, it hides the problem inside the offered answer. At comparable bounded inquiry effort, asking what result is missing and what would count as a supplied contribution can distinguish qualification support from faster approval and from shift handoff in :5.1. A full decision-modeling engagement can handle richer competing values, but is unnecessary when a short receiver account or a negative answer resolves the lead. Use that fuller work when it changes the decision rather than making it an entry requirement.
Source roles and limits. ISO supplies normative scope, problem-structuring Methods supply formulation guidance, and A.2.3 supplies semantic distinctions. None establishes this receiver's need, a causal benefit, professional authority, or the empirical effectiveness of this cross-holder inquiry. A commercial promise is one possible use, not the template for every epistemic or community contribution.
Reopen. Reopen the affected move if current practice supplies a better low-burden way to expose false need, overpromised benefit, or consequential missing participants while preserving a usable first return; or if a relied-on source changes the problem, value, or promise boundary.
Relations
- Receives from C.40.OI and C.40.NU: the bounded inquiry and leads whose receiving contribution remains uncertain; adequate direct inputs may be reused.
- Uses C.40.OC: characteristics and scales needed to make a worth or contribution claim interpretable.
- Can supply a selected problem-structuring Method: the bounded problem/contribution account and unresolved formulation question for a separate decision-support inquiry.
- Uses A.2.3: identity and conditions of relied-on promise content, distinct from commitment and provision.
- Uses A.15.9: the needed bounded outside-practice result and its reliance boundary, without performing the supplier's Method.
- Supplies C.40.DC: worthwhile problem/proposed-contribution accounts, including explicit hypotheses and acceptance expectations.
- Supplies C.40.JF: the value and contribution claims whose necessary transitions need qualification; a worthwhile problem is not a reachability verdict.
- Uses C.40.RQ: local requalification when receiving need, source, acceptance, protection, or horizon changes.
C.40.PC:End
Construct Directions with Support Configurations
Type: Method pattern
Status: Stable
Normativity: Normative unless marked informative.
Problem Frame
Use this when a worthwhile problem or proposed contribution is clear enough to explore, but the proposed direction is still “train,” “buy,” “automate,” “hire,” or “learn Pattern X.” Those labels omit the change, later Work, access, support, transition burden, and evidence that would make the contribution possible.
Start by describing one finite change from the actual situation, together with the support needed to obtain the proposed contribution. Construct a materially different way when it could change the inquiry. The first useful result is an adequate contribution-bearing direction/support candidate, a set of materially different candidates, or a conditional candidate with a named missing premise.
A direction here is a prospective change in contribution, capability, use, or enabling conditions over a declared horizon. Its description is not an obtaining arrangement, selected programme, WorkPlan, performed Work, or choice. A person exploring their own development can use it without asking for advice.
Ordinary non-use boundary. Reuse adequate directions when only reachability or comparison remains. If the receiving problem or contribution is still missing, use C.40.PC; do not complete a training or tool configuration around an unexamined need. If a domain practice owns the whole configuration question, use it directly. Once one organization result, use, and acceptance premise are stable and whole obtaining arrangements are the decision, an applicable organizational Method supplies that comparison and its authorized return.
Problem
A direction can sound concrete because it names a product, course, Method, or person. Yet the proposed intervention may address only one fragment of what makes the later contribution possible. Training alone does not establish access; naming a model does not establish evaluation or receiving-holder support; a provider offer does not establish a usable result or a continuing arrangement.
Another failure compares fragments under different assumptions. One direction includes feedback and transition time while another is credited with its advertised benefit alone. A path compares every later step with today's baseline, hiding the configuration and costs that earlier steps create. “Learn this pattern” supplies no usable next development step when neither its complete contribution nor representative use is recoverable.
The practitioner needs possible configurations that expose what changes, what must support the change, what may worsen, and what remains to be established before reliance.
Forces
Solution
Construct whole-enough prospective directions around the receiving contribution. Preserve their finite changes, supports, consequences, and unresolved premises before asking which is reachable or preferable.
Recover the contribution and the actual starting condition
Name the subject of the proposed change and configuration, receiving Work or use, proposed result, acceptance expectation, horizon, scenarios, and protected conditions. Use the current problem/contribution account from C.40.PC or an adequate direct source. Preserve an unconfirmed problem as a hypothesis rather than strengthening it through detailed design.
Recover the actual starting configuration S0: relevant capability, access, support, Work, interfaces, resources, and current readings. Mark unknowns. A job title, organization chart, tool inventory or source publication may supply relevant information, but its presence alone does not establish the needed baseline.
Use C.40.OC for the characteristic meanings, scales, current/intended/protected/unknown readings, and evidence window needed by the inquiry. Characterize only what could change the direction or its qualification; do not commission a complete diagnosis by default.
Turn a label into a complete contribution
For every serious proposal, explain what the subject or supported arrangement could contribute in the receiving situation. State the recognizable situation, useful move, first result, and stop or non-use boundary. Include the support and substantive conditions that change how the contribution is obtained.
When a pattern is the proposed address, cite its complete current body and recover the action-bearing contribution from it. Do not substitute its title, PatternID, a summary card, or a list of topics for that contribution. Keep the body's relevant applicability, assurance, and stopping conditions; the direction need not copy the entire source.
If the receiving claim depends on a Method or MethodDescription, preserve its established relation under A.3.1 and A.3.2. A pattern is a MethodDescription only after A.3.2 establishes an already admitted Method as its EntityOfConcern and at least one substantive claim about that Method as a way of doing. Otherwise use its content as pattern guidance and leave that stronger relation unasserted. In neither case does reading or publishing the description establish the subject's capability.
For a person, connect the contribution to representative later Work. “Use A.15.9 to formulate and rely on a bounded outside-practice result in evidence review” can specify a contribution; “finish A.15.9” cannot specify capability or transfer. A capability or transfer claim needs qualified evidence about this person, the later Work and permitted support; obtain that result from the relevant capability and subject practice before relying on it.
Construct materially different directions
Vary the change that could make the contribution possible. Useful contrasts include developing a capability, making an existing capability usable through access or support, obtaining a qualified external contribution, changing the work or interface, or changing the proposed contribution itself. They are prompts, not mandatory candidate families.
Distinguish two questions:
- Different contribution premises: directions address different worthwhile problems, results, uses, or acceptance expectations. Preserve those differences for the receiving comparison.
- Different ways of obtaining one contribution: candidates hold the result premise stable but vary the supporting arrangement. For organization-change decisions, use an applicable organizational comparison Method when that whole same-result question is current.
For a separately selected decision-design inquiry, an applicable alternative-generation Method can develop a set of decision alternatives from the directions returned here. That Method can retain alternatives for different live formulations; same-result parity is not its universal entry condition.
Retain the actual baseline, a smaller repair, a staged direction, or no new development when each answers the inquiry. No change still carries continuing Work, resource use, and exposure. State when an alternative is missing; do not invent a weak candidate to make a preferred one win.
Specify the finite change and support configuration
For each retained direction, state the finite change Delta and candidate configuration S1 relative to S0. Give enough support detail to identify the material dependencies and contradictions:
- what contribution or Work would change, and who or what would perform or support it;
- required capability, access, feedback, tools, Methods, interfaces, provider contributions, resources, and authority where each is consequential;
- prerequisites, sequence or coexistence conditions, constraints, and interactions;
- transition and implementation Work, including migration, setup, evaluation, recovery, and continuing-service needs;
- what can be reversed, at what burden, and what remains irreversible; and
- later options that the change could open, close, delay, or preserve, with the basis and uncertainty of those claims.
Describe how the proposed performers, supports, and relations fit together under these conditions. A feedback provider needed by two candidates may have capacity for only one. A combined model-and-human arrangement may add review or recovery Work that neither component proposal named.
Use direct holder, engineering, operations, organization, provider, safety, and other practices for their mechanisms and quantities. Unknown support does not establish realizability. Keep the direction conditional, identifying the result needed to construct a realizable candidate; reject or revise it when a qualified condition rules out its necessary support. Do not discard the worthwhile contribution merely because this way fails.
Preserve the complete finite-comparison basis
When making a finite contribution comparison, use C.11.CRC. Compare actual or relied-on S0, finite Delta, and realizable candidate S1 under one system boundary, receiving decision, horizon, scenarios, constraints, interactions, transition basis, uncertainty, and evidence window.
Keep four coordinate families visible: useful results; consumed resources; consequences for affected Systems and the distribution of gains and burdens; and protected conditions. Retain each characteristic's scale and polarity. Do not hide one family in another or average incompatible scales.
For a Pareto-improvement claim, no coordinate in the declared complete set may worsen and at least one must improve under that same basis. Otherwise state only the supported relation, such as beneficial under a stated value judgement, harmful, dominated, non-dominated, or indeterminate. A direction with a possible gain and an unresolved protected condition is not thereby a Pareto improvement or a selected option.
If reachability or another necessary premise is missing, give the conditional configuration and comparison question to C.40.JF; do not manufacture a complete comparison. A numerical or analytical result from elsewhere remains limited to the configuration and use its supplying practice qualified.
For a later step, use its stage-specific S_i, finite Delta_i, and candidate S_i+1, not the original baseline by default. C.40.OP maintains the resulting finite path or branch account, including each step's burden, reversibility, option effects, and reopen condition. Neither the path nor its separate comparisons select a programme.
Return a usable candidate or identify what is missing
An ordinary direction account can say:
For this receiving Work and horizon, change this part of the current configuration to obtain this contribution, with these supports and protected conditions. The material transition and consequences are these; this premise remains unknown, and its positive, negative, or narrower answer would change the candidate in these ways.
Expand only what the next use needs. A complete enough direction may still be conditional; distinguish a qualified proposed configuration, a conditional direction with a named premise, an incomplete fragment, and a direction ruled out for the current use.
Use C.40.JF for jointly necessary transitions and competing support. Use C.40.OP for retention or a finite path when that is the question. If comparison or advice is current, provide the direction and its qualified premises to the appropriate decision-support or domain use; the applicable professional engagement Method supplies the development-advice relationship boundary only when advice is requested.
For human development, a direction choice may become an input to a programme-design Method, which composes and compares programme candidates without choosing. A programme-level advice request starts with those comparable candidates; it does not relabel an opportunity path as a programme. An applicable learning-material production Method supports realizing a Guide from an authorized selected programme and focal pattern contributions. A Guide is not produced by rearranging direction labels.
Reopen only the affected contribution, support, transition, comparison, or downstream use under C.40.RQ when its premises change.
Archetypal Grounding
A complete pattern contribution for a person's later Work
This constructed case concerns Mira's four-month exploration of useful engineering-review Work. The case assumes permitted example packages and a maximum of two inquiry hours per week. The stipulated review record identifies recurring delays from missing outside-practice premises. Her current ability to recognize those missing premises in representative Work is unknown. Independent acceptance, confidentiality, and sustainable workload remain protected.
One lead is “learn A.15.9.” The direction instead cites the complete A.15.9 body and recovers its contribution:
The pattern supplies guidance for the proposed contribution, not evidence of Mira's capability. The applicable human-capability and review-domain practices must establish representative performance, support, and transfer.
Two directions now differ materially:
The candidates do not yet share a qualified same-result comparison: the first proposes development of independent recognition; the second proposes supported result use. The next question is which contribution the receiving review needs and what human-capability or supplier result can qualify the corresponding transition. Combining them creates another candidate whose joint time and support need testing.
The result is two conditional directions, not a learning programme or recommendation. If a selected direction calls for human capability development, an applicable programme-design Method can supply programme candidates and comparison. Within that branch, any requested programme recommendation and authorized programme choice remain separate; only then can an applicable learning-material production Method realize the appropriate Guide. The person can also stop now with the conditional directions and the support limits that matter; a support request is a further selected contribution.
A faster retrieval candidate is not a cost-free improvement
This constructed example stipulates a service unit's manual qualification arrangement and an engineering estimate for a possible indexed arrangement over a ninety-day horizon. The result premise is fixed: a source-traceable claim-to-evidence account under the same acceptance criteria. The candidate does not replace independent acceptance. Review capacity and scarce index-maintainer capacity are separate resource coordinates.
The finite change is adding and maintaining the index, not merely subtracting twenty minutes. Discarding the index could reverse its future use but would not recover spent setup time. A later model-assisted retrieval candidate would start from whichever configuration is then assumed, including its index-maintenance burden; it cannot reuse today's S0 silently.
The return is a conditional configuration and two decisive questions: can the access owner qualify the proposed custody/use, and can the unit supply setup, verification, and maintenance without violating protected continuing Work? Even if latency improves, the resource increase prevents a no-worsening Pareto claim. C.11.CRC supplies the finite-comparison discipline; C.40.JF must qualify the joint premises before a stronger reachability claim. The illustrative numbers establish no measured effectiveness.
The AI subject is the arrangement being changed
A team says “improve the model,” but its receiving-use failure may arise from the retrieval corpus, tool permissions, escalation path, or operating support. The team constructs a parameter-change direction and a changed-retrieval-and-oversight direction, each naming its model edition, data, tools, evaluation, and environment.
A bare-model benchmark does not by itself qualify either tool-enabled arrangement. The receiving engineering and evaluation practice must establish any transfer basis for the intended configuration, characteristic, and use under C.40.OC:4.5. Otherwise the reading remains unknown and the direction retains that limit. A corresponding request follows only when its obtainable contribution warrants acquisition. Human curriculum or learning-transfer evidence does not qualify either AI change.
Direct same-result arrangement work is enough
An organization already has a stable contribution and acceptance basis, adequate baseline evidence, and whole manual, internal-specialist, and provider arrangements. It does not reopen development-direction construction merely because the arrangements could be given development labels. The applicable organizational Method supplies the remaining organization comparison and lawful disposition; the necessary authority and specialist results still have to be present.
Bias-Annotation
- Label concreteness: naming a course, tool, person, or PatternID feels more specific than the contribution it describes.
- Capability default: an access or support limitation is treated as a need to develop the holder.
- Fragment parity: a fully supported direction is compared with an advertised fragment.
- Benefit-only comparison: transition, maintenance, distribution, or protected conditions disappear behind a desired gain.
- Combination optimism: attractive components are presumed to work together despite shared capacity or incompatible premises.
- Baseline drift: later steps are compared with the original situation instead of their assumed starting configurations.
- Learning-route projection: a prospective contribution path is mistaken for a selected programme or a Guide.
Conformance Checklist
A conforming direction/support account:
- MUST name the subject/configuration, receiving contribution and use, horizon, acceptance expectation, protections, and actual or relied-on starting condition.
- MUST describe the useful move and first result rather than only a label; a pattern-based direction must preserve its complete cited contribution and applicable conditions.
- MUST preserve established Method/MethodDescription relations when relied on and must not infer them from a title or publication alone.
- MUST state what makes retained directions materially different and whether their contribution premises are the same or different.
- MUST identify the finite change, candidate configuration, material supports, prerequisites, constraints, interactions, and transition basis.
- MUST preserve result, resource, affected-System/distributional, and protected coordinate families for finite comparison, with each coordinate's declared scale and polarity under a common comparison basis.
- MUST state transition burden, reversibility, and supported or conditional option effects; a later compared step must use its own starting configuration.
- MUST keep unknown support and ruled-out support different and return the missing owner-qualified result before asserting realizability or joint reachability.
- MUST NOT claim Pareto improvement when any coordinate in the complete set worsens or its necessary qualification is absent.
- MUST distinguish a direction or path from a recommendation, choice, programme, Guide, performed Work, or observed effect.
- SHOULD retain a baseline, smaller repair, or no-change alternative when it answers the inquiry, and stop when the next missing result belongs elsewhere.
Common Anti-Patterns and How to Avoid Them
Consequences
Practitioners obtain directions that another agent can inspect, challenge, and qualify. A capability-development route can coexist with a support repair or external contribution. The next inquiry becomes precise: which access, feedback, transition, resource, or protection premise changes this candidate?
The cost is recovering the baseline and enough support detail to expose interactions. Some attractive labels become conditional fragments, and the comparison may remain indeterminate. A commitment may have to wait while the responsible practice supplies the necessary result.
A constructed direction remains a possibility. Realization, authority, programme design, evaluation, and later observed effects retain their own requirements.
Rationale
A contribution and its support jointly determine what a proposed change means. A course without representative practice, a model without permitted data, or a provider without a usable interface can be a relevant ingredient without being a way to obtain the contribution.
Finite changes and stage-specific baselines make hidden burden visible. They let a receiver retain a useful possibility while identifying the condition that prevents a stronger claim, and they prevent a gain at one stage from being counted again in every later step.
SoTA-Echoing
Practice question. How can a direction become a contribution-bearing possible change without treating a course, tool, or pattern label as a complete and reachable intervention?
Best-known line for this use. Construct the contribution and support together, then apply finite configuration-relative comparison. The applicable organizational Methods supply whole-arrangement construction and comparison when one organization result premise is stable. C.11.CRC supplies actual S0, finite change, realizable S1, complete coordinates, transition basis, and option effects. This pattern adapts their contributions in :4.3–:4.6 for development inquiries that may still retain different contribution premises. Human-capability and other holder practices retain the change mechanisms and evidence.
Alternative and comparison. A course or product catalogue is cheaper when the missing contribution and all material supports are already known. When they are not, it omits the difference between developing capability and making an existing capability usable. At the bounded effort of the Mira case, two contribution/support descriptions expose different evidence and provider needs; ranking two course brands would not do so. A direct organizational Method supplying the complete same-result arrangement comparison is the stronger alternative once that comparison is the whole question. Use it directly; no additional opportunity-construction pass is needed.
Front-end boundary. ISO 56007:2023 supplies normative guidance for opportunity/concept work before solution development and deployment. Here that boundary supports prospective configurations and gaps; it supplies no missing support, implementation, or effect evidence. A complete cited pattern contribution similarly supplies guidance, not holder capability.
Trade-off, limits, and reopen. Recovering the contribution and support takes work; compare that burden with using an adequate existing direction or obtaining the fuller implementation detail needed by the receiving use. The constructed cases demonstrate distinctions, not comparative effectiveness. Reopen the affected move if current holder or configuration practice preserves the contribution, complete consequences, and necessary support with less burden; if a relied-on comparison or source changes; or if a real use reveals a consequential omitted interaction or transition.
Relations
- Receives from C.40.OI, C.40.OC, and C.40.PC: the bounded inquiry, relevant characterization, and worthwhile problem/proposed-contribution premise; equivalent direct results suffice.
- Uses A.3.1 and A.3.2: identification of Method and MethodDescription when a direction relies on those relations; a PatternID remains a locator for the complete contribution.
- Can supply a selected decision-alternative generation Method: directions for a separate finite-alternative question, including alternatives addressing different formulations.
- Can return to an organizational comparison Method: a separate whole same-result organization arrangement comparison once its result/use/acceptance premise is stable.
- Uses C.11.CRC: finite contribution comparison under one basis with complete coordinates, transition burden, reversibility, and option effects.
- Supplies C.40.JF: direction/support candidates and the material premises whose joint qualification changes reachability.
- Supplies C.40.OP: contribution-bearing finite steps and their stage-specific baselines for retention, paths, or branches.
- Returns to human-capability, organization, strategy, AI, engineering, operations, and other direct practices: their interventions, mechanisms, capability, support, evidence, authority, and realization results.
- Returns human programme and Guide work: The applicable programme-design Method composes and compares programmes; authorized participants choose; the applicable learning-material production Method realizes the selected programme and pattern contributions as a Guide.
- Uses C.40.RQ: local requalification of changed directions or supports. A separately requested advising service retains its engagement conditions, established through the applicable professional Method.
C.40.DC:End
Qualify Dependencies and Joint Reachability
Type: Method pattern
Status: Stable
Normativity: Normative unless marked informative.
Problem Frame
Use this when a direction looks worthwhile, but it is not yet clear that its necessary changes, resources, support, and permissions can hold together for this subject and horizon. The warning sign is “each step is possible” without an answer to “can these steps work together here?”
Start with the proposed direction and name the condition most likely to defeat its reachability. Inspect the existing result for that condition, then check how it interacts with the other necessary conditions. The first useful result is a bounded reachability claim, a conditional direction with its missing premise stated, or rejection of the affected transition.
Development-opportunity construction and development-direction advising are applications of this joint-qualification Method. Reachability means that the proposed transition is supported as possible under stated conditions; it is not an observed effect, a recommendation, or permission to perform it. A person exploring their own direction can use this pattern without an adviser.
Ordinary non-use boundary. Reuse an adequate current reachability account when only retention, recommendation, choice, or execution remains. Enter from adequate direction descriptions; no earlier opportunity inquiry is compulsory. If the change itself is unspecified, return that gap to C.40.DC or the practice constructing it. If a domain practice already owns the complete feasibility question, request or reuse its result rather than building a second feasibility model here.
Problem
Necessary conditions can be credible separately and incompatible together. Two directions consume the same scarce specialist hours. Access exists for one environment but not the proposed receiving use. A later step requires a capability that an earlier step merely hopes to develop. A nominally reversible trial creates an irreversible disclosure or leaves continuing maintenance.
A checklist of obtained premises can therefore give false confidence. It can also cause needless exclusion: one unavailable route is mistaken for an impossible contribution, or one missing premise causes an otherwise independent direction to be discarded.
The practitioner needs to determine which combination is supported, which claim remains conditional, and where new evidence or a revised configuration would change the answer.
Forces
Solution
Qualify the necessary conditions together for one named transition or combination. Keep the work proportional to the claim that someone will rely on; do not turn every possible dependency into a new investigation.
Bind the claim before testing its dependencies
Name the subject, receiving use, actual or currently relied-on starting configuration, proposed finite change, candidate configuration, horizon, and protected conditions. Recover an adequate direction/support description when one exists. For a path, name the step and the starting configuration it inherits.
State the claim being qualified: one direction alone, several directions together, a sequence, or one branch under a stated contingency. For one possible outcome, establish the existence of a compatible route. To promise a response to any allowed choice, establish who decides, what they can know by each irreversible step and how their rule works across the allowed combinations. To exclude the entire opportunity, show that the obstacle or combination of obstacles covers all permitted routes. “A or B is available” does not establish “A and B can both be realized,” and two separately possible endpoints do not establish a possible transition between them.
Select dependencies by consequence. Ask what absence or failure would block the proposed transition, breach a protection, or materially change its contribution. Include dependencies on capability, access, evidence, acceptance, support, time, resources, authority, and continuing Work only where they affect this claim. An unstated material dependency is a gap, not a satisfied condition.
Use the supplying practice's qualified result
For each material dependency, state the required result and why it matters to this transition. Use A.15.9 — Request and Use a Bounded Result from Another Practice to inspect an existing result and use it at its supported strength for this receiving question. Retain a conditional direction where the needed premise is absent. Select a further request through A.15.9 and C.11.DUA only when its obtainable contribution warrants formulation, qualification, performance, delivery, interpretation, delay, and displaced work.
A provider's offer, a published MethodDescription, and a qualified result about this subject answer different questions. Recover the result's subject and configuration, characteristic or condition, receiving use, any evidence window, limitations, and uncertainty. If new Work is needed, the supplying practice retains its Method, competent performer, authority, and conclusion. Establish the source and support required for this premise use, reusing an adequate existing account. Use A.10 when the source-to-use account remains implicit or must be replayable.
Use ordinary wording that preserves the difference between supported, conditional, unknown, and ruled out. Keep these distinctions at the claim they affect. If authorized access is absent, an access-dependent route cannot be described as available. That fact need not defeat a contribution obtainable without that access.
An existing result for another holder or configuration is not automatically a result for this one. Use the practice's qualified transfer basis, covering the receiving bearer and configuration, characteristic meaning and scale, intended use, evidence window, and uncertainty as required by C.40.OC. Otherwise leave the receiving premise unknown.
Test the combination, not just its ingredients
Put jointly necessary conditions on one compatible basis. Recover the time intervals, resource quantities, operating and transition demands, availability restrictions, and protection limits. Before adding the demands on one provider, establish which operations must be performed separately and which one performance can supply to several recipients. Count shared work once and each separate demand in the time window in which it must be met. If sharing is undecided and would change feasibility, keep the two outcomes open.
Ask which conditions compete, complement, substitute for, or interfere with one another. Inspect common causes and shared suppliers, thresholds, congestion, delays, continuing obligations, and evidence dependencies where they can change the result. Two claims derived from the same unqualified assumption do not independently support the combination.
For a sequence, check whether the earlier transition can establish the next step's starting conditions within the declared horizon. A planned learning result remains a condition on the later step until learning evidence for the intended holder supports it. A later resource estimate must include continuing demands inherited from earlier steps. For alternatives, preserve their separate bases: a conflict between them may rule out simultaneous realization without ruling out either alone.
Use the domain analysis when the joint question needs scheduling, engineering, safety, organizational allocation, finance, or another specialized result. An ordinary arithmetic conflict can be stated when the underlying quantities and limits are qualified. Do not infer independence, additivity, or a probability of completion from the shape of a dependency diagram.
Include transition burden, recovery, and future options
Check what must happen to move from the starting configuration to the candidate, not just what could hold after arrival. Include material setup, learning, migration, evaluation, coordination, temporary loss of service, and continuing support. Distinguish the resources needed to find out whether a transition is possible from those needed to realize it.
When comparing the finite change's contribution, use C.11.CRC: actual or relied-on starting configuration, finite change, candidate configuration, common horizon and scenarios, and the complete relevant result, resource, affected-System/distributional, and protected coordinates. Each coordinate keeps its declared scale and polarity. A gain in the intended result does not cancel a protected loss or establish Pareto improvement.
Name what reversal would require and what it would not restore. Removing a tool may restore the earlier operating arrangement while leaving setup time spent, information disclosed, or a supplier relationship changed. State whether the transition opens, preserves, closes, or delays a later option, and which part of that claim is evidenced or conditional.
A proposed probe is itself a finite change when it consumes resources, exposes people or data, or changes later possibilities. Return its question and missing feasibility or authority result before treating it as an admissible recommendation or performed Work.
Return the strongest supported reachability claim
Write the result at the smallest affected scope.
“Conditional” is useful only when the condition changes the result. State what a positive and a negative answer would do. Do not retain a known-infeasible direction as merely awaiting more evidence, or call an unanswered question a negative result.
A short answer may suffice.
The two departments each need two hours of the same engineer's time every week. For this inquiry the engineer must review each department's confidential material separately. Two hours a week are available. Either review fits. Together they need four hours, so they cannot both be delivered in the eight weeks while urgent operating cover is protected.
This return need not become a large dependency register. Expand it when several recipients, delayed reuse, disagreement, or consequence makes the basis hard to recover.
Recognition and assurance. A reader can recognize a usable account from the named transition, joint limits, qualified or missing premises, and bounded return. Trust in the reachability claim requires the domain evidence and uncertainty assessment; high-consequence use retains its additional safety, authority, and assurance requirements. A clear account makes those requirements visible without satisfying them by description.
Return to the next use
Supply the qualified direction or gap to the current inquiry. C.40.OP may retain the candidate or reconstruct its path; the applicable professional engagement Method bounds an advising engagement, and decision-support Methods supply their results when recommendation is requested. Human-capability, organization, AI, engineering, operations, and other practices retain programme construction, allocation, realization, and evidence.
Reopen only the affected claim when its subject, starting configuration, use, horizon, resource, support, permission, protection, or evidence changes. Use C.40.RQ to trace that change through dependent opportunities and advice. Preserve an unchanged independent result.
Archetypal Grounding
Two feasible support directions cannot share the same hours
Consider a constructed engineering-organization case. Two departments want better acceptance preparation during an eight-week period. The current arrangement is manual review. Candidate A supports one department with a qualified reviewer; candidate B supports the other with the same specialist. The following are stipulated case premises, not reported field evidence.
The relevant practice has qualified each direction separately for its own use. Each requires the same specialist to review its own confidential work for two hours every week for all eight weeks. Each review occupies the specialist for the full two hours, and the departments' access conditions require separate reviews. The organization has confirmed only two hours per week for this inquiry, after existing obligations. The protected condition is no overtime or loss of urgent operating cover. Setup and the remaining dependencies are covered in each separate qualification.
The useful answer is not “development is impossible.” Both separate directions remain candidates. The organization can request the combined-work estimate for a shared-support configuration, examine another provider, or later consider a different horizon. Those possibilities require their own qualification. Choosing between A and B remains a separate decision.
Mira needs a transfer premise
In this constructed case, Mira, an engineer-manager, is exploring better review contributions over four months. A human-development result identifies what she can already do under specified support. A separate access inquiry finds that some representative examples are unavailable. A small recovery test shows improved performance with permitted examples and feedback.
Those results answer different questions. A capability envelope bounds current performance conditions; an access/expression differential concerns whether a capability can be expressed under changed conditions; a recovery test concerns that tested recovery. None alone establishes learning transfer to later independent review or selects an intervention.
If the direction requires independent later review, retain it conditionally while the applicable human-capability Method supplies the result about feasible development, practice, feedback, and transfer for Mira's intended use. If a different direction requires continuing specialist support, qualify that arrangement on its own basis. Do not silently replace independent capability with supported performance.
The return might be: “Supported review with the permitted examples is evidenced within the tested conditions. Independent transfer in the intended work remains unknown; this blocks the independent-review claim, not the supported route or the usefulness of the receiving problem.”
Data access does not complete an AI reachability claim
In this constructed case, a team has a model benchmark result and permission to inspect a small data sample. Its proposed direction is a retrieval-assisted review arrangement using a different configuration and representative operational data.
The benchmark does not establish the receiving configuration's performance. Sample inspection does not establish permission for operational processing, continuing data access, evaluation coverage, or adequate human oversight. Use the AI, data-governance, and operating practices' results through A.15.9. Until the required transfer and support results exist, retain the arrangement as conditional.
If the qualified data result excludes that operational use, reject the access-dependent route for that use. Preserve a permitted manual route when its evidence is independent. A population or lineage can likewise be the subject of a scientist's reachability hypothesis, but it is not thereby an advice recipient or authorized chooser; a proposed intervention remains the scientist's or organization's separately governed case.
All pairs fit but the whole combination does not
All pairs can fit while the whole set does not. In a separate arithmetic illustration, three non-shareable seven-hour activities each draw on one eighteen-hour budget. Every pair needs fourteen hours; all three need twenty-one. The deficit rules out that combination under those premises. It does not choose an optimal subset. Two separate six-hour reviews likewise exceed eight available hours if each occupies the same reviewer throughout.
Bias-Annotation
Optimism favors possible endpoints and hides the transition. Availability bias favors the one provider already known. Schedule and resource arithmetic can look precise while using unqualified quantities. Success in a supported test can be overread as independent capability or general transfer.
Ask which condition would defeat the combination, whose burden is missing, and whether two premises share the same unsupported source. Keep qualified contradiction, uncertainty, and absence of evidence different. These checks expose assumptions; domain evidence determines the reachability claim.
Conformance Checklist
- MUST bind the subject, receiving use, starting configuration, finite change, candidate configuration, horizon, and protected conditions.
- MUST state whether the claim concerns one direction, a joint combination, a sequence, or a conditional branch.
- MUST qualify material external premises through A.15.9 and preserve their supplying practice, use limits, currentness, and uncertainty.
- MUST test shared resources, timing, interactions, and inherited obligations where they can change joint reachability.
- MUST include material transition burden, reversibility, and supported or conditional option effects.
- MUST preserve C.11.CRC's complete coordinate basis when asserting finite contribution or dominance relations.
- MUST distinguish supported reachability, a conditional opportunity, a ruled-out transition, and an unresolved candidate.
- MUST localize a missing or defeated premise and retain unaffected independent alternatives.
- MUST NOT infer holder transfer, learning, capability, authority, observed effect, or choice merely from the presence of a label, source description, benchmark, or reachability account. Use a qualified result for the claim.
- SHOULD use a short result when the claim and its decisive limits are readily recoverable.
Common Anti-Patterns and How to Avoid Them
Consequences
Practitioners can stop making whole-direction claims from isolated premises. A conflict can narrow one combination while leaving useful alternatives intact. A selected worthwhile request becomes specific enough for the supplying practice to answer, while a useful conditional result can finish without one.
The cost is recovering compatible conditions and exposing burdens that a direction label omitted. Some candidates remain conditional, and a high-consequence claim may require specialist analysis. This pattern reduces neither that assurance obligation nor the uncertainty of an unperformed transition.
Rationale
Reachability depends on conditions holding together, not merely being individually conceivable. Binding the claim first makes both rejection and retention more precise: evidence can defeat one combination without defeating every use of the contribution.
Separating the qualification of each premise from the joint test also makes work reusable. A current access, capability, or support result can serve several candidates, while each candidate needs its own compatible combination and transition basis.
SoTA-Echoing
Practice question. How can a practitioner qualify a direction's joint reachability without mistaking individually credible ingredients for a realizable combination?
Best-known line for this use. Reuse holder-qualified results, then test their finite configuration and joint constraints. This pattern adopts A.15.9's single result-acquisition method in :4.2 and adapts C.11.CRC's finite-change, interaction, transition, and option-effect analysis in :4.3–:4.4. The supplying domain retains the feasibility model and evidence. This division makes the resource conflict visible without requiring a second domain investigation.
Serious alternatives and trade-off. A list of separately qualified premises is cheaper to maintain, but in the eight-week case it admits a combination that requires twice the available support. At comparable reading effort, the joint row reveals the decisive conflict. A complete domain schedule or feasibility analysis is the stronger alternative when the question is already an implementation problem and its conditions are stable; reuse that result. For an early opportunity inquiry, constructing such a model for every candidate can cost more than the bounded rejection or gap requires. This is a proportionality judgement, not a measured efficiency result.
AI-specific source return. The current NIST AI RMF Playbook, MAP 1.1 and MAP 4.1 provides guidance for mapping receiving context, non-AI alternatives, third-party components, and continuing support risks. This pattern adapts that line in :4.2 and :5.3: a model result and data access are ingredients whose receiving-use limits matter. Its AI risk guidance does not supply the missing client results or establish a common human/AI development mechanism.
Limits and reopen. The cases are constructed demonstrations, not evidence of intervention success. Reopen the affected move when domain practice supplies a less burdensome account preserving the same joint constraints and protections, a real use reveals a missed interaction, or a relied-on holder, configuration, evidence, or source meaning changes.
Relations
- Receives from C.40.OI, C.40.OC, and C.40.DC: the inquiry, relevant characterization, and direction/support candidates; adequate direct inputs suffice.
- Uses A.15.9: the single method for obtaining and qualifying a bounded external result. Use A.10 when the source-to-use account remains implicit or must be replayable.
- Uses C.11.CRC: finite configuration-relative contribution, complete coordinates, constraints, interactions, transition burden, reversibility, and option effects.
- Returns to holder and domain practices: capability, access, support, transfer, feasibility, safety, authority, allocation, and realization results.
- Supplies C.40.OP: bounded reachability, conditional or rejected combinations, and the gaps that change retained paths.
- Supplies bounded advising and decision support: qualified premises or localized blockers for a bounded advising engagement and its decision-support question.
- Uses C.40.RQ: affected requalification when a relied-on condition changes.
C.40.JF:End
Maintain Opportunities and Stepping Stones
Type: Method pattern
Status: Stable
Normativity: Normative unless marked informative.
Problem Frame
Use this when useful opportunities would otherwise disappear because they are not today's strongest candidate, depend on a later condition, or require a preliminary step whose immediate result is modest. The working question is “what is worth keeping available, for which later use, and at what cost?”
For an opportunity portfolio, name the later option that retention is meant to preserve, together with its contribution, conditions, lineage and next reconsideration. The first useful result can be two retained alternatives with one reason and reopen condition each; a large portfolio system is unnecessary. When no later opportunity can yet be named, the C.40 return in :4.2 handles retaining material for a worthwhile further inquiry.
This pattern governs opportunity retention and path construction, including development opportunities. A portfolio here is the maintained set of opportunity descriptions and their relevant relations. It is not a funding allocation, selected development programme, recommendation, or permission to realize every member. A stepping stone is a candidate step retained for what it may enable later, not necessarily for being best now.
Ordinary non-use boundary. Do not create a portfolio when one current question needs only a qualification or choice and there is no useful later retention. Reuse an adequate set with its current conditions. Enter from existing opportunities; no complete opportunity-construction sequence is required. When an authorized programme and its implementation are already the subject, use the owning programme or work practice. If a search archive, front, or live-pool treatment is the question, use C.18 or C.19 at the return in :4.5.
Problem
A current ranking can discard an opportunity that preserves a valuable later route. Conversely, retaining everything creates a costly collection with no reason to revisit or discard anything. A “roadmap” can make a conditional sequence look committed before its dependencies are qualified.
Path descriptions introduce another error: every later step is compared with the original situation. Earlier benefits are counted again, transition costs disappear, and the next step inherits capabilities, access, or evidence that have never been established. A preliminary inquiry can be called a stepping stone even when no supported link connects it to the later option.
The practitioner needs a small, revisable set in which retention has a reason, each step begins from its own configuration, and uncertainty does not become an implied choice.
Forces
Solution
Retain an opportunity when its expected use or learning value justifies the burden of keeping it. Describe paths as conditional finite transitions; qualify them before making a reachability claim.
State why and for how long something is retained
Name the subject, receiving use, protected conditions, and the horizon or horizons that affect this inquiry. Use the horizon needed by the opportunity; do not impose universal near-, medium-, and long-term buckets. When horizons differ, preserve that difference rather than merging their anticipated results into one comparison.
For each retained opportunity, keep its contribution and support configuration, present qualification, source or parent candidate, retention reason, material retention burden, and revisit or discard condition recoverable. Name who will reconsider it when someone has accepted that responsibility. A source pointer may suffice for an unchanged description.
Distinguish preserving an account from preserving an obtainable option. Saving a direction in a note does not reserve a provider, renew data rights, maintain a skill, or keep equipment ready. If retention needs resources or continuing Work, identify that burden and the authority or arrangement. Without it, retain a conditional possibility or description, not a claim that the option remains available.
An ordinary two-sentence account is enough:
Keep the permitted manual-review route as the current reference and the retrieval-assisted route as conditional on representative data use and evaluation. Revisit the second when those results arrive or at the end of the ninety-day inquiry; saving its description reserves no data, supplier time, or implementation budget.
Make each stepping-stone claim testable
Name the later option or reachable region, the proposed enabling transition, and the evidence connecting the step to that later use. Ask what becomes possible, less costly, or better qualified because of this step, and what could show that the connection fails.
When a later opportunity is unknown, a smaller retention claim may suffice: keep an examined construction for an affordable next question. C.40 supplies the development and continuation Method. Preserve the material, operation and conditions that let someone try that continuation. Judge its burden against what the next answer could change; do not invent a distant payoff or call this a qualified path to an unnamed opportunity.
Keep a candidate stepping stone distinct from one whose retention is justified under the stated policy. A plausible connection may justify a bounded inquiry, but does not by itself justify performing the step. State the value, burden, delay, uncertainty, and protected conditions that make retention worthwhile for this use.
A data-access question can be a candidate stepping stone toward representative evaluation. Its possible answers change which evaluation routes remain available. The question itself grants no access and establishes no model capability. A useful course, preliminary project, archive variant, or “zone of proximal development” label likewise needs the enabling relation and holder-owned evidence.
Do not require immediate dominance. Retention can be justified by later reachability, coverage, learning, or reversibility even when another candidate has a stronger current result. Conversely, novelty or diversity alone is not proof of future usefulness.
Construct paths from their stage-specific configurations
For every material step in a path or branch, name its own starting configuration, finite change, and resulting candidate configuration. Write S_i -> Δ_i -> S_i+1 only as compact notation for those three descriptions, not as evidence that the transition is feasible.
Recover what the starting configuration inherits: established or conditional capability, support, access, ongoing Work, resource demand, remaining uncertainty, and effects of earlier choices. If the earlier step has not occurred, the later starting configuration is conditional too. If the path can branch, name the condition that distinguishes the continuations and the stop when neither is supported.
State each step's transition burden, reversibility, and effects on later options. Do not repeatedly credit the full difference from the original configuration as each later step's contribution. Use C.11.CRC for finite comparisons with the complete relevant result, resource, affected-System/distributional, and protected coordinates. Each coordinate keeps its own scale and polarity under the common comparison basis. When step horizons or scenarios differ, obtain an adequate mapping or keep those comparison claims separate.
Use C.40.JF to qualify the jointly necessary transitions, shared resources, timing, and inherited obligations. A sequence of individually plausible steps remains a conditional path until the necessary joint basis exists. Return a missing holder, access, support, or implementation result through A.15.9. Do not substitute a path diagram for the supplying practice's conclusion.
A path can preserve alternatives without selecting a programme. Human capability-development programme composition and comparison remain with the applicable programme-design Method; authorized participants retain programme choice, and learning-material production Methods support realizing the selected programme and pattern contributions as a Guide.
Keep the set inspectable without forcing a winner
Preserve materially different candidates, including conditional directions and a useful baseline, when their retention reasons remain current. Keep a candidate ruled out for this use outside its live opportunity set, with enough explanation to avoid rediscovering the same defeated claim. Preserve other uses separately when their basis remains supported. A historical record can remain useful without being a live option.
Use a comparison for the question it answers. To establish non-domination, identify the eligible rivals and the declared comparator. For each rival, use supported values, bounds or another justified comparison to rule out its dominating the candidate. Name any rival whose unresolved comparison could change front membership. A useful stepping stone can be retained on the strength of what it enables. Different premises or horizons may support only a partial comparison.
Make resource interactions visible when retention itself competes for scarce attention, expertise, access, or maintenance. If the available retention allowance cannot support the set, return the needed allocation or reduction decision to its authorized owner. Meanwhile, describe only what the available maintenance supports. Do not silently make every member “available” by naming it in the portfolio.
Stop expanding the account when the next missing result belongs to a holder practice, a pool-treatment decision, or a local choice. The inquiry can return a useful partial set.
Use archive, front, and pool treatment only for the questions they answer
Use C.18 — Open-Ended Search Archive and Front Stewardship when generation or retention needs a search archive, front, descriptor space, lineage, or exploration policy. Its archive preserves exploration or stepping-stone value; its front reports current non-domination under a declared comparator and admissibility. Neither return is a recommendation, selected-set declaration, or work permission.
For generation, recover the generator, operators, descriptors, source/lineage, and retention basis that matter. Claim a changed possibility space only when that stronger claim changes the inquiry. A newly visited region under the same effective generation conditions is exploration, not evidence of expansion. Use C.18 for the earlier and candidate space declarations, changed component, supported mode, evidence, and blocked overread; causal claims require the causal result.
Use C.19 — Explore–Exploit Governor when the question is how to treat a live pool under its declared policy: widen it, keep its frontier, narrow it to a subset, or sunset a line. The policy may remain unchanged while that treatment question becomes current. Selecting the next local candidate or probe is a different question, as are generating a variant, declaring a selected set, and enacting Work.
Ordinary opportunity retention does not require all those practices. A few descriptions with recoverable reasons and conditions can be enough.
Return a discriminating question and reconsider locally
When further qualification or revision is needed, name the next question whose plausible answers would retain, narrow, reject, or open an opportunity or path. An adequate retained set can be used until its reconsideration condition occurs. Use the declared receiving policy to consider value, inquiry burden, delay, protections, and uncertainty. Reuse a sufficient partial comparison; no complete ranking is needed merely to state the question.
If the question becomes “which candidate or probe should we choose now?”, return to the local choice or decision-support question. A proposed probe needs its own feasibility, exposure, authority, and execution basis before it becomes recommended or performed. An information request is not that authorization.
Revisit at the declared trigger, including a material source, use, horizon, resource, support, permission, or evidence change. Use C.40.RQ to update the smallest affected opportunity and its dependent paths or advice. Preserve unaffected candidates and source lineage; retire a retention claim when its later use disappears, its cost no longer fits, or a qualified result defeats its enabling connection.
Recognition and assurance. A usable account shows what is retained, why, for how long, under which conditions, and what changes next. Whether a path is reachable, a step causes a later possibility, or a retained option will remain obtainable requires the appropriate domain evidence, C.40.JF qualification, and actual resource or authority basis. The clarity of the account supplies none of those assurances.
Archetypal Grounding
An index changes the starting point of the later model step
Consider a constructed review-support inquiry over ninety days. The case baseline is manual qualification, with sixty minutes per representative case and no separate index-maintenance allocation. A candidate index might reduce this to forty minutes, at two maintainer hours each week and six setup hours. A later retrieval-assisted configuration might reduce it to thirty minutes, with total maintenance of three hours weekly and four additional setup hours.
These are stipulated illustrative estimates under the same representative case type and compatible service assumptions, not observed effects. Data custody, representative verification, acceptance quality, and the feasibility of fitting the transitions within ninety days remain unqualified.
The second transition's marginal estimate is ten minutes per case, not thirty. A full path comparison against S0 can still be made separately, but it must include the whole path's transition burden and time-dependent operating costs without counting the first gain again.
The table is not a Pareto-improvement claim: maintenance demand increases and the protection/quality coordinates remain unqualified. Nor does it show that the index enables retrieval merely because it comes first. The engineering result must establish whether the index and its data representation are reusable for the proposed retrieval configuration.
If that connection is absent, the index can remain a manual-support candidate on its independent contribution, while its stepping-stone claim remains unknown or is rejected. The team returns that reuse question rather than treating the three labels as a committed roadmap.
A retained question can cease to be a stepping stone
In another constructed case, a team retains a representative-data access question because its answer could open a bounded model-evaluation route. The local retention policy permits keeping the question and its existing source notes until the data steward answers, at negligible additional burden. It authorizes no processing or trial.
An affirmative, use-specific result would allow the team to qualify that evaluation route's remaining conditions. A refusal for the proposed processing would close the data-dependent branch. A narrower permission would create a narrower candidate requiring its own receiving-use qualification.
The steward rules out the proposed use. The team retires the stepping-stone claim for that route; saving the historical answer can prevent the same unsupported request from recurring. It retains the independently permitted manual-support direction and the receiving problem. The changed premise does not erase the whole portfolio.
Mira retains directions before choosing a programme
In this constructed case, Mira is exploring review contributions over four months. She keeps two directions: develop more independent review capability, or improve the conditions for supported contribution. The first needs human-capability evidence about representative practice, feedback, and transfer; the second needs a continuing support arrangement. Both can remain described while those questions are open.
A preliminary exercise is only a candidate stepping stone toward independent review until its enabling relation and feasible development conditions are supported. Being the first item in a course sequence is not that evidence. Retaining its description does not reserve Mira's practice time or a specialist's attention.
When direction advice is requested, the bounded engagement and decision-support results remain separate from the authorized direction choice. The applicable programme-design Method then composes and compares human capability-development programmes when that is the selected question. An optional programme recommendation, authorized programme choice, and Guide realized through an applicable learning-material production Method remain further distinct results. If an adequate programme comparison already exists, the programme-level question can begin there; Mira need not reconstruct an opportunity portfolio to use it.
Bias-Annotation
Immediate-performance bias can discard a useful stepping stone; novelty bias can retain an attractive but unsupported one. Sunk-cost reasoning can keep a branch alive after its enabling condition disappears. A tidy stage sequence can conceal conditional starting states.
Ask what later use justifies retention now, what continuing burden it creates, and what answer would end it. Keep “worth remembering,” “worth maintaining,” and “chosen to perform” distinct. These distinctions make retention inspectable without guaranteeing future value.
Conformance Checklist
- For a retained opportunity or proposed path, MUST bind the subject, receiving use, relevant horizon or horizons, and protected conditions. Retaining material for further inquiry without a known later opportunity uses the C.40 return in :4.2.
- MUST preserve each retained opportunity's contribution, conditions, lineage, retention reason, material burden, and revisit or discard condition.
- MUST distinguish preserving a description from maintaining an obtainable option and identify resources or authority where the latter requires them.
- MUST state the later option or region, enabling transition, and evidence for a stepping-stone claim; a candidate connection may remain explicitly unqualified.
- MUST describe each material path step from its own starting configuration, finite change, and candidate result, including inherited uncertainty and continuing obligations.
- MUST preserve step-specific transition burden, reversibility, option effects, and the complete C.11.CRC basis when making finite comparison claims.
- MUST qualify joint reachability through C.40.JF or the adequate domain result before asserting it.
- MUST distinguish archive retention, current-front membership, live-pool treatment, local choice, and actual Work when those questions occur.
- MUST NOT force one scalar winner, universal horizon taxonomy, selected programme, or work permission from a retained set.
- MUST localize changed-premise effects, retire defeated retention claims, and preserve unaffected independent opportunities.
- SHOULD return a discriminating question when its answer could change a retained opportunity or path, and keep the account no larger than its receiving use requires.
Common Anti-Patterns and How to Avoid Them
Consequences
Practitioners can preserve a weaker current candidate for a specific later use without pretending it is best now. A path exposes what each step inherits and which premise would change the next continuation. A defeated branch can be retired without losing the receiving problem or an independent route.
Retention has a cost, even when only attention is scarce. Some later options remain conditional, and stronger claims need domain evidence. The portfolio can inform recommendation or programme work without deciding either.
Rationale
Immediate comparison and future retention answer different questions. Keeping their reasons separate lets a practitioner avoid both premature elimination and an indefinitely growing list.
Stage-specific configurations make the retained path intelligible. A later step can be valuable because of the conditions an earlier step creates, but that connection and its continuing burdens must be recoverable. Otherwise a path is only an ordered set of attractive labels.
SoTA-Echoing
Practice question. How can opportunity inquiry retain later possibilities and useful intermediate steps without forcing a current winner or disguising an unqualified sequence as a programme?
Best-known line for this use. Retain candidates for an explicit later use, while comparing each finite step from its own inherited configuration. This pattern adapts C.18's distinction between archive exploration/stepping-stone value and current-front non-domination in :4.1–:4.2 and :4.5. It adopts C.11.CRC's finite comparison and option-effect discipline in :4.3. C.19 supplies live-pool treatment, not a local next-probe choice.
Serious alternative and trade-off. Retaining only the current front is useful when immediate comparison is the whole question. It loses a presently weaker candidate whose supported enabling relation matters to a later use. Keeping an unrestricted backlog avoids that loss but hides retention costs and defeated premises. At the modest effort of the index case, adding a retention reason and a stage-specific increment exposes both the possible later use and the extra maintenance burden. The cost is reconsidering conditions that a static list ignores; comparative efficiency has not been measured.
Organization-specific source return. Proactive portfolio management in mission-oriented innovation policy, OECD, 2025, sections 2–4, supplies current guidance on complementary activities, development-stage dependencies, resource distribution, adaptation, and termination. This pattern adapts those interdependence and reconsideration concerns in :4.1, :4.3, and :4.6. The source's mission-management scope includes selection and funding; opportunity retention here supplies inputs to those separately authorized moves. It establishes neither a universal three-horizon rule nor a person- or AI-development mechanism.
Limits and reopen. The constructed cases show what must be distinguished; they do not demonstrate future option value or intervention effectiveness. This is not a complete comparison of investment-portfolio or real-options methods. Return such a specialist comparison when the receiving use requires valuation or allocation. Reopen the affected retention rule when a serious alternative preserves the needed possibilities at lower burden, the enabling evidence fails, or the governing use, source, horizon, or resource condition changes.
Relations
- Receives from C.40.NU, C.40.PC, C.40.DC or adequate sources: searched leads, proposed contributions, and direction/support candidates with source lineage.
- Uses C.40.JF and A.15.9: joint reachability and external-result qualification for retained paths and stepping stones.
- Uses C.11.CRC: finite step comparisons, inherited configurations, complete coordinates, transition burden, and future option effects.
- Uses C.18: generation, archive, front, descriptor, lineage, and possibility-space claims.
- Uses C.19: treatment of a live pool under its declared policy; local choice remains a different result.
- Supplies bounded advising and decision support: retained alternatives and conditional paths for a bounded advising and decision-support question.
- Returns to programme design and learning-material production: human programme comparison and later Guide realization; authorized choices and performed Work remain separate.
- Uses C.40.RQ: affected reconsideration, revision, or retirement while preserving independent candidates.
C.40.OP:End
Requalify Changed Opportunities
Type: Method pattern
Status: Stable
Normativity: Normative unless marked informative.
Problem Frame
Use this when something relied on by an opportunity, direction, path, or advising engagement changes, and a practitioner needs to know what can still be used. The change may concern the subject, receiving use, horizon, characteristic, contribution, support, resource, permission, evidence, or an enabling step.
Begin with the changed premise and the opportunity claim that used it. Compare the earlier and current conditions and return the useful result they support. Obtain new qualification only when the current use needs it and its attainable contribution warrants acquisition. The first useful result is an unchanged, narrowed, blocked, revised, or retired contribution for a stated use, with its affected downstream return.
This is application-specific requalification shared by Construction and Advising. A changed recommendation returns to its advice owner for the affected follow-up question; changed source-currentness or publication work retains its owner.
Ordinary non-use boundary. Reuse a current result when no material premise or receiving use changed. A new filename, link, date label, or layout alone does not require reconstructing opportunities. If the current question is only one source's currentness, use G.11; if it is one already-known evidence reliance, use the subject guidance and retain the support needed for that use. Apply A.10 when the source-to-use account remains implicit or must be replayable. Do not create an indefinite monitoring service to answer a bounded change.
Problem
Practitioners can respond to change too broadly or too narrowly. They repeat every assessment because a source version changed, or merely replace a link while the supported use has narrowed. They discard an independent manual route when one AI configuration loses permission, or carry an old whole-use estimate into a smaller data class.
Another error rewrites the past. A revised opportunity description makes a previous recommendation seem different from what was delivered. A later step is treated as implemented because it was retained in a path. A missed observation is interpreted as evidence that nothing changed.
The practitioner needs a local account of what changed, what depended on it, what is now supported, and who must act on the resulting question.
Forces
Solution
Trace the change to the smallest opportunity-specific result whose use can change. Qualify that result through its governing practice before reporting the stronger conclusion as restored.
Identify what changed and what the current use requires
Recover the earlier claim and its subject, configuration, receiving use, horizon, evidence, and conditions. State the new fact or uncertainty and how it became known. Compare meanings and applicability, not just version labels.
Distinguish the relevant kinds of change:
One event may contain several changes. A model replacement can change the actual configuration and invalidate an evaluation's applicability. A corrected report can change warranted belief even when the represented world stayed the same.
If a credible change indicates a possible protection breach, follow the existing applicable protection or escalation rule at its required speed. Identify that rule and its authorized owner. A routine requalification account neither suspends an applicable duty nor creates a new intervention authority.
Establish the affected use, not just a connection
For one known opportunity, inspect how the changed premise supports its contribution, direction, comparison, reachability, retention, or advising boundary. Ask which action or conclusion would change if that premise no longer held. A source listed in the same bibliography may do no work in this claim; an uncited assumption may do substantial work.
Use A.10.1 — Revalidate Affected Uses When a Relied-on Source Changes when a changed source claim may affect several receiving uses or some of the uses must still be found. It supplies the bounded search frame, complementary source- and receiver-oriented discovery, coverage limits, direct-reliance test, and a way to establish which receiving conclusions or actions could change. Do not substitute a link scan for that result.
Keep the returned discovery conclusions distinct from the opportunity judgement. A confirmed dependent use needs a supported subject judgment or scoped blocker; an adequate existing result may supply it. The direct return can finish the question. Add a common account only when a receiver needs that overview, and group equivalent irrelevant mentions without separate exclusion records. An unresolved use or inaccessible source remains a gap. Do not call an uninspected branch unaffected or extend a local negative search result beyond its covered frame.
Follow an established dependency only while a receiving action or claim can change. Include a combined recommendation when its ranking used the affected comparison, but stop before a separate result whose inspected basis is independent.
Obtain the result that changes the opportunity claim
Use A.15.9 for the missing external result: inspect an adequate existing result first and qualify its supported use for the present subject, configuration, use, and window. Retain a useful narrowed or conditional answer. Select a request through A.15.9 and C.11.DUA only when its obtainable contribution warrants the full acquisition burden. The supplier retains its Method and conclusion.
Return to the opportunity contribution that the change affects. A changed receiving need may reopen C.40.PC's problem/contribution account. A changed support or finite configuration may require C.40.DC and C.40.JF. A defeated enabling relation may retire a C.40.OP stepping-stone claim while leaving the candidate's independent contribution useful. A changed recipient, conflict, consent, or disclosure condition reopens the advising engagement and its service boundary. Preserve an independently qualified domain result within its established scope, and reconsider a domain claim whose qualification itself depended on the changed condition.
Do not transplant the earlier qualification to the new subject or configuration. A model name, similar job title, same provider, or narrower dataset is insufficient. Use the practice's qualified transfer basis where it applies; otherwise preserve the receiving reading as unknown under C.40.OC.
Narrowing a candidate also reopens the quantities and comparisons that depended on the broader use. A guarantee for every member carries to a subset under the same conditions. A combined average alone can leave each class’s average open. If an earlier path step has become actual, use that configuration and inherited obligations as the next starting point. If it remains prospective, keep its resulting state conditional.
Use C.11.CRC for the changed finite comparison: complete result, resource, affected-System/distributional, and protected coordinates, each with its scale and polarity on a common basis. Use C.40.JF for the changed joint conditions. A request for either result is not the completed result.
State the smallest warranted disposition
Write what can now be used and which earlier scope no longer has support.
These are scoped returns, not a universal state machine. One change can leave a baseline unchanged, retire one route, and open a revised conditional alternative. “Revised” does not mean “validated,” and “blocked” does not mean the subject lacks every possible direction.
An ordinary return can be short:
The indexed manual route remains supported for its inspected use, but the earlier retrieval direction no longer has permission for data class B. Retire that broad route and retain the class-A direction only as a new conditional candidate; use an applicable prior result, a justified bound or a new evaluation to establish what the permitted narrower use can deliver at its incremental burden.
Return affected advice and action questions to their owners
Make the changed contribution and its scope available to those who use it, under the disclosure and service conditions. Preserve the earlier opportunity or recommendation as an account of what was returned under the earlier basis. Record a correction where needed without silently rewriting that history.
When a recommendation or follow-up question is affected, return the qualified change to the responsible advice owner. Keep the observation, interpretation, earlier advice, established decision and implementation distinct, and obtain only the affected professional return. A later recommendation is a separate result when needed. An applicable follow-up Method and a recommendation Method can supply those distinct results for a separately selected decision-design engagement.
A recommendation, its delivery, and its acknowledgement establish neither the recipient's choice nor an implemented change. If an operational action must change, recover the applicable authorized decision or protection rule. If the receiving choice is unknown, say so rather than treating the old advice as a standing instruction.
A change in professional scope or disclosure can restrict the service return while leaving a qualified technical result intact. Contact, referral, confidentiality, and closure depend on the relationship. Do not expose protected material merely to make the update easier to explain.
Close the bounded update and keep a feasible reopen condition
Close this update when the inspected use has a supported disposition or an explicit blocker and its material coverage limits are visible. A compatible source-link repair can finish. Name the next receiver only when further action needs one; neither branch requires a duplicate common account. Keep an independent unresolved branch open as such; do not hold every resolved branch unfinished.
For a continuing use, state the observation or condition that would reopen its result and identify any responsibility on which that use relies. A completed reference repair needs no new observation arrangement. If later advice depends on future evidence, the observation, interpretation, access, timing, burden, and receiving response must be feasible under the advice and observation arrangement. A missed or late observation is a present reliance limit when that evidence was necessary.
Use G.11 for source currentness, decay, scoped refresh planning, or reporting. It does not decide the new domain result or replace the opportunity judgement. If the guidance itself fails across applications, return the bounded failure to its maintainer; a changed fact about one subject normally changes only that application.
Recognition and assurance. A usable update shows the changed premise, the inspected dependence, the current scope, and the next use or gap. Assurance requires support for the observation and subject result relied on, and for the authority needed by the receiving use. A recent date or a complete-looking trace does not establish any of them.
Archetypal Grounding
A narrower data permission does not preserve the old whole-use estimate
This is a constructed continuation of a review-support inquiry. The indexed manual arrangement S1 has now been realized. A separate retrieval-assisted direction remains prospective and concerns data classes A and B. Its estimated contribution was described for that combined workload; no later retrieval arrangement has been implemented.
The data steward supplies a qualified change: external processing of class B is no longer permitted, while the existing local manual use of both classes and the stated external use of class A remain permitted. The update must distinguish the receiving claims.
This result can finish with the manual arrangement preserved, the prohibited broad route retired, and the class-A direction conditional on its unqualified performance and transition. A class-A evaluation or finite transition result becomes a request to the AI or engineering practice only when its attainable contribution warrants the whole burden relative to continuing from S1. C.40.JF qualifies the revised joint conditions when that question is pursued.
If a probe was already authorized or begun, the responsible operational owner applies the applicable permission and protection rules. The adviser does not infer continued authority from the earlier recommendation. If no decision or implementation evidence exists, those remain unknown.
The update can close with the short return in :4.4. The narrowed branch is retained as a possibility while its qualification is missing; the realized manual arrangement is not silently reverted.
A corrected source can change use even when the world did not change
In a constructed case, a human-development report's corrected edition states that a performance result concerned supported review, not independent review. The person's performance history has not changed. The correction changes which claims the report can support.
The known opportunity account used the independent-performance premise to qualify a later step. That step now lacks its transfer basis; the supported-performance result can remain useful within its proper conditions. A new course is not inferred from the correction.
Several other possible receivers are not fully known. Under A.10.1, the responsible practitioner bounds discovery to the current opportunity inquiry's opportunity notes and issued advice, excluding unrelated historical projects. Source-oriented search finds the cited report; receiver-oriented inspection also finds an uncited independent-performance assumption.
One note merely lists the report as background. Another uses its corrected claim in a path. A private recipient-held copy cannot be inspected. The outcomes differ: the background mention supplies no affected use; the path needs its human-capability and opportunity qualification; the private copy remains a named coverage gap, not an unaffected result.
The opportunity return can finish with the locally blocked path, supported-review result and unresolved recipient-copy question. Subject qualification is still required before asserting a restored independent-performance claim. If an adviser needs a common overview, it cites these supported results and groups the irrelevant background mentions; no separate exclusion note is needed. Neither return claims complete no-impact coverage.
Changed service terms leave technical evidence distinguishable
In a constructed organization engagement, the consultant discovers a material provider interest that was absent from the agreed service basis. The existing technical comparison was independently produced and remains qualified for its stated configuration and horizon.
The applicable professional engagement Method now reopens whether this adviser can provide the requested recommendation and on what terms. The professional result may require narrowed scope, an independent recommender, or referral. That service outcome does not by itself refute the technical comparison. The applicable advice-follow-up Method receives the affected advice-use question if advice was already returned.
At the planned three-month follow-up, no capable observer or permitted source is available. That gap supplies no evidence that conditions stayed favourable. If the advice depended on that observation, return the missing observation and its reliance effect now. Existing professional contact or protection duties still follow their basis.
Take the cheap stop for unchanged meaning
A source is moved to a new repository path with the same relevant claims, applicability, and accessible content. The user verifies that the relied-on source and receiving use remain the same. Updating the access reference answers this change; no opportunity reconstruction, multi-use discovery, separate subject-result record or common summary is needed.
If access is lost, a relied-on window expires, or one changed word narrows the claim, the premise is different. Return that reliance or claim-comparison question instead of using “only a document change” as a blanket exemption.
Bias-Annotation
Recency bias can make a new document look more trustworthy without relevant qualification. Consistency pressure can preserve a previous recommendation after its premise fails. Conversely, a visible change can encourage excessive rechecking and discard independent value.
Ask what action-changing claim differs and what inspected evidence supports each retained use. Distinguish unresolved coverage, missing observations, and negative findings. Preserve history without treating it as present warrant.
Conformance Checklist
- MUST identify the earlier claim, present change, subject/configuration/use/horizon, and the basis on which the change is known.
- MUST distinguish world-side, source-claim, receiving-use, reliance-window, and representation changes where they imply different returns.
- MUST establish dependence before propagating an effect; inspection and coverage limits must remain visible.
- MUST use A.10.1 for its several-use source-change discovery question, not as a mandatory pass for every changed fact.
- MUST obtain the independently governed subject result through the relevant practice before asserting restored qualification.
- MUST requalify altered scope, quantities, transfer, and finite comparisons; a narrower candidate does not inherit every whole-use result.
- MUST preserve actual versus conditional stage configurations, incurred burdens, and unaffected independent contributions.
- MUST state the supported unchanged, narrowed, blocked, revised, or retired disposition and its affected scope. Name a gap or next receiver only when one is needed.
- MUST preserve earlier returned advice and distinguish renewed recommendation, authorized choice, protection response, and implementation.
- MUST retain professional and observation duties without inventing a continuing service.
- SHOULD close a bounded update when its result and coverage limits are usable, while leaving independent unresolved questions explicit.
Common Anti-Patterns and How to Avoid Them
Consequences
A change can now produce a small useful correction rather than a global restart or a cosmetic source update. Practitioners can retain independently supported routes, revise a conditional branch, and explain which premise would restore a stronger claim.
The cost is recovering the earlier use and enough current evidence to distinguish dependence from proximity. Some coverage may remain unresolved, and an urgent operational or professional duty can continue while the opportunity update is incomplete. A local closure claim has support only within the inspected conditions.
Rationale
An opportunity's use depends on contributions and premises, not on everything stored beside it. Local requalification preserves this useful structure while allowing a changed need or configuration to have wider effects when the dependence warrants them.
Keeping the earlier return, current qualification, and later authorized action separate also prevents history from being rewritten to fit a new recommendation. The practitioner can learn from the change without pretending it had already been known or acted upon.
SoTA-Echoing
Practice question. How can an opportunity inquiry respond to changed premises without either replaying every result or preserving unsupported opportunities and advice?
Best-known line for this use. Qualify the claim-sized change, establish reliance, and return the affected question to its owner. This pattern adopts A.10.1's bounded source-to-use discovery only for that source-change question in :4.2, and adapts the professional follow-up distinction among observation, interpretation, advice, decision, and implementation in :4.5–:4.6. C.40.PC, C.40.DC, C.40.JF, C.40.OP and the applicable professional engagement Method supply the opportunity-specific result being reconsidered; G.11 retains currentness and refresh work.
Serious alternatives and trade-off. Rechecking everything can be justified when the affected basis cannot be separated. It is unnecessary for the independently permitted manual route in :5.1. Updating only a source pointer is cheaper but misses the corrected support claim in :5.2. At comparable inspection effort, recovering the receiving premise distinguishes the two cases and identifies the smallest needed result. Bounded discovery costs more than a citation scan and may end with unresolved coverage; that is preferable to unsupported complete-preservation claims. No measured efficiency advantage is asserted.
AI-specific source return. The current NIST AI RMF Playbook, MAP 1.1 emphasizes receiving context, assumptions, operational requirements, and downstream effects of system changes. This pattern adapts that line in :4.1, :4.3, and :5.1: a changed configuration or use requires qualification of the receiving contribution, not merely a fresh product label. The source does not supply the client's evaluation, permission, or causal result.
Limits and reopen. The cases are constructed probes of the distinctions, not evidence of field effectiveness. Reopen the relevant requalification rule when actual uses reveal missed dependence, an inadequate discovery frame, unaffordable burden, or a stronger direct-domain method. Changes to the guidance itself go to its maintainer.
Relations
- Receives from C.40.OI, C.40.NU, C.40.PC, C.40.DC, C.40.JF, C.40.OP and the applicable professional engagement Method: the inquiry, characterization, contribution, direction, reachability, retained path, or advising boundary whose premise changed.
- Uses A.10 and A.15.9: A.10 for a source-to-use account that remains implicit or must be replayable; A.15.9 for bounded external-result qualification.
- Uses A.10.1: changed-source discovery and revalidation across receiving uses, with coverage and independently governed subject results.
- Returns to C.40.PC, C.40.DC, C.40.JF, C.40.OP, the applicable professional engagement Method and holder practices: the smallest opportunity or relationship question affected by the new conditions.
- Uses C.11.CRC: changed finite comparisons with complete coordinates and actual or conditional stage-specific baselines.
- Can return to the applicable follow-up and recommendation Methods in a selected decision-design engagement: follow-up and any later recommendation, separate from the choice and implementation. Other advice retains its professional owner.
- Uses G.11: currentness, decay, scoped refresh planning, and reporting when those are the questions.
- Returns to professional and operational owners: contact, disclosure, protection, authority, and realization duties.
C.40.RQ:End
Construct and Continue Open-Ended Development
Type: Method pattern Status: Stable Normativity: Normative
Problem frame
Use this Method when worthwhile development depends on several kinds of work and their connections are not yet workable: a direction has been formulated but its means are missing; a provider supplies a result but the intended recipient acquires no new capability; research, product work and learning compete for the same performers or equipment; or a successful intervention leaves no way to recognise and organise the next worthwhile transition.
The intended result is a feasible, revisable construction of development for named subjects, with the Methods, performers, means and receiving uses that make it possible. It includes a description of the connected transformation flows and an arrangement that can perform the needed work. The description is a means of construction and coordination. Its production is not the development of the described subjects.
Open-ended development here concerns growth of knowledge, understanding or Methods that expands attainable reproducible results and enables further worthwhile problems. The growing result can be public knowledge, a better way of obtaining results, a holder's capacity to construct such ways, or their connected change. Identify which growth is intended: a newly delivered service or a person acquiring an already established Method does not by itself establish this continuation.
A difficult episode can begin before anyone has named a problem. Its continuation can require a different question, a different Method, a different arrangement or a different understanding of value. Completing current work, learning an established Method, creating a Method and implementing an agreed change can contribute to development; none establishes the whole by itself.
The Method can guide self-development with assistance, work by organizational leaders and participants, a provider’s contribution to another subject, or support for a practice among independent carriers. Establish which decisions a provider can make and which require the recipient’s agreement. Independent carriers can take up a practice without belonging to a common organization.
For one adequate bounded improvement with available means and no consequential connection problem, use its direct Method. Do not build a network description merely to redescribe that work. Conversely, a bounded contract may be one contribution to continuing development without promising to provide its whole future.
To direct this development from the conditions it faces, use C.43:4.8 to connect observation, advice, choice and intervention across the working Method, developing capability and provision. Supply it with the present possibilities, proposed directions and the conditions under which contributions can be obtained. C.40.DC supplies alternatives from successive attained configurations; C.40.JD establishes what the receiving activity obtained, what the holder demonstrated with the available help and what the work revealed about the Method and its provision. A claim that the holder gained a capability also needs evidence of what it could do earlier. When continuing development is wanted, C.40.NP helps those responsible choose a worthwhile problem from the new starting point. The control account returns which dependency, timing, allocation or observation must change for this construction to proceed. This relationship preserves new problem formation and epistemic results rather than treating development as execution toward one permanently fixed target.
Problem
Useful local results can fail to become development. A research finding does not reach an engineering choice; a course supplies knowledge without the conditions for its use; a new arrangement performs well only while its designer supplies every judgement; or support remains organized around an earlier problem after its recipient has changed.
A single lifecycle diagram hides these failures. It can put an epistemic result, a person's capability, an organization and a provider's work into successive boxes as though one object passed through them. An organization chart can hide the same problem from the opposite side: every box has an owner, but the required work and its conditions do not connect.
The practitioner must construct a connected whole without turning it into a compulsory universal sequence. Different subjects, questions and time scales need different Methods and different evidence. Some work can proceed independently. Some results need one another, some merely share scarce means, and some reveal a need to revise a previous decision.
Forces
Solution
Recover what the undertaking owes and what is already established. For a design, trace how each needed contribution will be obtained, joined and used on the receiving occasion. For a commission to realize the design, secure the means, carry out the work and establish what the recipient obtained. Return a warranted new possibility, a useful continuation or a reason to stop according to the commission. A difficulty may first show up as a workaround, dependence on a helper or a cost borne elsewhere. A success can reveal a possibility that was previously out of reach.
Name what is being developed and who undertakes the work. For a system, show what it can now do under stated conditions (A.1). For an episteme, identify the subject matter, what can now be claimed or inferred about it and on what grounds (C.2.1). For a Method, show how a change to the reusable way of doing the work makes a result attainable across qualifying performances (A.3.1). For a discipline, show how its organized body of knowledge and its repertoire of practice sustain the change through later work (C.20). One undertaking can change several of these. An author can extend what a proof establishes without changing the research team’s capability. The team can also learn an existing proof technique while the proof stays the same.
For each connection identify the work, what performs it, what changes or is returned, and what the next performer needs in order to use the result. Assign formulation, search, comparison, choice, action and checking according to the capability, access and authority each operation requires (C.32.MWA.PP). An engineering unit can commission a supplier to construct alternatives and use a robot to test them. Either contribution can be obtained for one trial or retained in ordinary work. The authority to select an alternative or commit resources follows the actual assignment. Where cultural continuation matters, other carriers must be able to reconstruct and use the Method.
When the intended change is a capability, name the system expected to gain it and what it should be able to do on a later occasion with specified support (C.40.JD). That system can differ from the performer of the intervention. A service can produce a result for a client today while the client still needs the service for the same operation tomorrow.
The construction below is prospective. It describes transformations to arrange and the contributions on which they depend. It does not assert an obtaining E.18.NET network or make the following descriptions an exhaustive taxonomy of flows. Use the E.18 identity and change rules when selecting actual flow members. One described strand can require several members; two distinctions useful for the explanation can belong to one member.
Identify the growth and its receiving continuation
Name the subject of development, the situation that warrants attention, whose purposes it affects and the basis for characterizing its current condition. The performer, beneficiary, provider and payer may coincide. In research the subject of development can be the account of a phenomenon. The investigators choose and perform the inquiry.
Use C.40.OI and C.40.OC to bound and characterize the development inquiry. B.5.PI supplies the entry from an unnoticed or weakly understood difficulty in ordinary work. C.40.PC and C.40.CD help construct the problem and possible contribution when neither is settled. A discrepancy from a target is one entry; a newly accessible possibility, a dependence that has become costly, a changed environment or an unexamined limit after success can be another.
For each proposed change, say what becomes attainable, for whom, at what cost and with which uncertainty. “Become a learning organization” leaves the choice open. A team might obtain a qualified answer from an adviser, make that judgement itself or construct and improve its own way of making such judgements. The three results have different costs and call for different evidence.
Keep the present work that must continue visible. Development need not suspend useful operation, nor must every limitation be removed before the next contribution. Rest, consolidation, deliberate termination of one direction and protection from overload can preserve the conditions for subsequent development.
An epistemic question can supply the consequential situation. State what knowledge, understanding or way of obtaining results should grow and what new results or questions that growth could make attainable. A scientific explanation can be a worthwhile result before a practical application is found.
Distinguish three claims: the current result became better, the available way or understanding grew, and that growth changes which further problems can be handled. Explain each relied-on connection. A routine service improvement can end with the first claim; it need not be redescribed as open-ended development.
Formulate the problem and the inquiries needed to choose a direction
Keep ordinary work running where it remains useful. Preserve enough of a consequential episode to recover what was being attempted, the way used, its premises and what happened. The episode may be a failure, an unexpectedly successful response or a newly available means. Use a sufficient established account. When different live explanations would change the continuation, distinguish them before selecting the repair. For example, inability to supply a result can concern an inadequate Method, unavailable authority or an unlearned operation. Sending all three into a training programme would give them the wrong continuation.
Problematisation produces a formulation that can guide the search for a solution. Characterize what happened or what has become possible, for whom it matters, and under which conditions. Specify the change or gain in understanding sought and the evidence by which a proposed solution could be judged. If the present way breaks down, locate the failed premise or the observation needed to find it. A promising opportunity can also warrant a new problem. B.5.PI helps notice a difficulty and open an inquiry. C.40.OI bounds the inquiry. C.40.PC and C.40.CD support constructing and revising the problem and a possible contribution.
Use the thin ProblemCard in C.22.2:2.1 to check the formulation. The signal, affected situation and criterion should let an evaluator judge a proposed result without access to the formulator’s preferred solution. The formulator can also apply the criterion. Assign a separate evaluator when the assessment requires independence. Record a missing criterion or observation and obtain it before using the card to judge a solution. The formulation can be ready for a search while the method is still unknown.
A question asks for an answer. “Which field marks a revised record?” requests information that may help choose a processing Method. A research problem can itself be expressed as a question. “What is the smallest number of sessions compatible with these conflicts?” names a result to establish. A schedule attaining the number and a reason no smaller number suffices settle it. For an inquiry whose answer form needs construction, use B.5.QD:4.3 to name the objects, premises, allowed changes and what would answer it. A known, directly obtainable fact can be requested in ordinary language.
The formulation and available answers supply two different receiving operations:
- Direction construction compares worthwhile attainable changes for the subject, including their means, losses and dependence on others' contributions. It uses the understood difficulty and known possibilities, and identifies what further answer could change the choice.
- Method or knowledge inquiry searches, adapts, constructs or investigates the missing operation or relation. State what result would settle its present question and how that result could be used. The inquiry need not wait for a fully settled strategy when what can be done affects what is worth choosing. Once an adequate method is known, plan the executable work under A.15.2. Specify who will use the method, the result to deliver, its conditions of execution and how to check it. Once the performer has carried out the task, compare the actual result with the problem's acceptance criterion.
The returns matter. A technical inquiry that establishes a cheaper usable way can change the direction. A participant's objection can expose an unwanted result and change the problem formulation or proposed direction. An unavailable decision power can require a different undertaking. Do not freeze the formulation simply because one branch has begun work.
A direction is ready for the next construction when it specifies enough of the subject's changed possibilities and receiving conditions to distinguish ways of obtaining them. It need not settle every future question. “Use a consultant's answer”, “make this judgement with obtainable help”, and “organize and revise our own way of making this judgement” specify different results even when all three begin at the same failed project.
Construct the joins that make a developmental whole
Connect the formulated problem or selected inquiry to an obtainable epistemic or practical contribution. Then connect that contribution to the conditions of its use and to a further possibility opened by the changed position. A research result can change a problem formulation before a new application is built. A newly developed Method can make another result reproducible before anyone needs to acquire a new personal skill. A holder-development branch is selected only where the intended continuation requires it.
When changes to the obtaining way, the holder's capability and the conditions for performance depend on one another, use C.40.JD to construct those joins. It distinguishes the delivered work, the holder's performance with the retained help and any supported change in capability or in the Method. When a qualified result is already obtained but the next worthwhile problem is unclear, use C.40.NP. Its return identifies what the changed position enables, including a justified local stop and preserved future possibility. These Methods can be used independently of this whole: one joins a present construction, the other starts from a result.
Do not require all branches. A theory-building inquiry may continue through a new question and a discriminating explanation without a training programme. A practitioner may use a developed way with an unchanged provision arrangement. A better AI configuration may remove a human operation while a different undertaking deliberately retains human mastery as an independent result. Keep the chosen end visible when changing its means.
The useful whole is an explained construction of transformations and their dependencies. A list of research, teaching, operation and management activities leaves the joins to be invented. For each consequential join, show what the supplying work returns, what the receiving work does with it and what condition makes that use possible. Adequate existing contributions are reused at their qualified scope. Obtain or develop only what is missing.
Connect the needed contributions to means, access and authority
For each proposed use of a Method, identify what the performer must receive from other work, which means it must be able to use, when it needs them, and who may authorize that use. Keep the supplying work and its result distinct from the receiving work: an expert examines a proposed experiment and returns a judgement; the researcher uses that judgement to choose or revise the experiment. The researcher's access to the apparatus and permission to conduct the experiment are additional relations. None can be inferred merely from the expert's availability.
Reuse conditions that already hold. For a missing condition, identify the work that could establish it and the participant able and entitled to undertake that work. Obtaining a usable input, reserving equipment, preparing a performer and securing agreement can require different actions and can return different results. Explain which receiving action each result makes possible. This construction belongs at every consequential connection, including problematisation, inquiry, practice, use and reconstruction by another carrier; it is not a final supply stage after the direction has been settled.
When existing contributors cannot perform the required work, identify which System, capability or Method must change. The development of a provider's knowledge, Methods or capability can be a dependency of another subject's development or an independently worthwhile undertaking for several later uses. Specify that growth and its further possibilities separately from routine work obtaining an input or arranging access. The same apparatus can remain in use while its measurement Method develops; replacing an apparatus or reserving its time alone establishes a different result.
The same performer can deliver a service, develop its Method and help another performer use it. These are distinct Work contributions without necessarily being distinct positions or organizational units. Compare their joint demand against capacity. Likewise, a research result can have its own scientific purpose while contributing to a product direction; the product does not become its only justification.
Embed the development undertaking in the surrounding activity. Preserve the ordinary results that cannot be displaced, identify what each development alternative would displace, and settle that choice with the resource owner. A project boundary, a Method boundary and an organizational boundary answer different questions. A residence coordinator can develop a way of running programmes and also run current cohorts; drawing two project branches has not doubled that person's available time.
Use and vary the connected description
The following drawing compresses the explained construction. Boxes identify described work, results and conditions; labelled arrows summarize prospective receiving connections. They do not assert generic world-side relation occurrences or one object passing through a lifecycle.
The drawing omits internal professional transformations, work that establishes access and authority, the multiplicity and timing of occurrences, participant signatures and most failure returns. Read those from the explanation and the Methods governing those details, not from arrow direction. For an E.18.NET claim, identify member flows and qualifying cross-flow relation occurrences under their governors. The proposed drawing does not supply that assurance.
Enter where a sufficient result is already available. An existing adequate Method can enter at the way and its receiving use; an established capable holder need not repeat learning. A missing source claim can open an inquiry without creating a whole training programme. Remove a branch only when its contribution is unnecessary or supplied elsewhere under compatible conditions. Keep an unresolved contribution visible instead of calling it done.
Compare alternatives that remain whole
Construct alternatives far enough to compare their attainable effects, costs and receiving uses. C.40.DC constructs possible directions, C.40.JF tests their joint reachability, and C.40.CU develops a reproducible way. Take the characteristics that matter to the affected roles and specify the constraints for acceptance and the improvements sought. For alternatives compared under common conditions, one dominates another when it does at least as well on every selected criterion and better on at least one. The remaining trade-offs call for a choice. A change of criteria can reveal a different problem and calls for a fresh comparison.
Retain an alternative in an accessible archive when its later use warrants the cost. The portfolio contains the undertakings actually selected within the available capacity, time, authority and resources. A trial can earn its place by opening a direction that current performance alone would miss. Strategizing supplies a justified choice and commitments when the direction or means are unsettled (STR.18). Work on the chosen way and evidence from its use establish what changed.
Several routes can be live together:
- A usable Method exists. Obtain its result, use it, adapt it or develop the capability to perform it, according to the sought result. A course is not required when a qualified provider is the sufficient alternative.
- The Method exists but is not feasible locally. Obtain missing inputs or assistance, change access or allocation, or adapt the Method's contextual realization. If that exposes a poor direction, return to direction construction.
- The proposed way is underdeveloped. Use C.39 and C.40.CD/CU to construct and try it. Do not sell established provision of a result that the proposed way cannot yet supply.
- A consequential knowledge claim is unresolved. Use the applicable research practice. Independent research can also be worthwhile without an immediate product customer; give it its own result, conditions and resources.
- The participant can perform the work only with continuing help. Compare maintained support with a learning transition and a different division of work. Neither dependence nor independence is universally superior.
- The current arrangement produces the result but blocks further development. Compare changes to problem formulation, authority, learning opportunity and support, not merely faster execution of the same routine.
These branches can rejoin or reopen one another. They are not stages that every undertaking must complete in order. A fully formed local question can enter near its receiving use; a failed trial can return to the underlying purpose rather than repeat an unchanged solution.
Obtain the contributions that realize the construction
Use the first real receiving occasion to define a bounded realization. Which contribution will be made, by whom, with which means, on what terms, and what will its recipient then be able to do? Keep any uncertainty that the first attempt is meant to resolve visible. A charter, budget and meeting schedule can enable this work without being its intended result.
Identify the contributions needed for the first useful handover. If an existing service agreement and interface already cover a contribution, check its operating conditions and obtain it. Where a decision or a new commitment is needed, prepare the proposal with the participants who can make it. Voluntary participants must be able to decline or qualify their contribution. For a subject uncertainty, use B.5.SD; for incompatible contributions, identify the conflict in result, burden, timing or authority and obtain a workable agreement, revised proposal or stop before relying on the combination.
When the proposal calls for agreement, discuss it in the participants’ own working terms and allow them to consult trusted advisers. Distinguish a misunderstanding from an objection to an accurately understood proposal. Revise the work, its conditions or the distribution of gains and losses when an objection warrants a change. Make the accepted contributions mutually knowable before relying on them. This work can change formulation and selection, not only implement a completed strategy.
An employer's proper authority, a professional obligation, a voluntary commitment and agreement to a service are different bases. A sponsor cannot settle all of them through one signature. When a needed contribution is unaccepted, outside the contributor’s authority or beyond available capacity, return to the relevant construction rather than hide it in the implementation plan.
When organizational realization is needed, obtain the joined contribution under its ordinary and exception conditions while preserving continuing service and concurrent commitments. Where a human capability is part of the promise, reuse sufficient current evidence about that person, later Work and permitted support. Obtain learning, practice, feedback and transfer results only for the needed acquisition, change or new use; an adequate supplied service can instead answer a promise that requires the result rather than the person's own capability. Engineered means need their own construction and qualification. The relevant subject practices retain the mechanisms and authority for these results. Observation and evaluation use the evidence required by the claim, not the mere completion of activities.
When the first attempt fails, locate what the result changes. Inadequate purpose, a wrong Method, a missing capability, unavailable means, a broken cross-flow connection and an unsupported success claim need different repairs. Preserve useful performed work and warranted results while changing the failed connection.
Make the way reconstructible in further use
Keep the decisions needed to operate the arrangement available at their point of use. Explain the purpose and strategy, which work supplies which result to whom, how the receiver uses it, and which means, access and authority the work requires. Keep this architecture explanation distinct from the effective commitments on which performance relies. Research questions, where present, specify what remains to be learned; Methods explain how recurring work obtains its results. These are different products. They need not become one document per named pattern, and their hierarchy need not reproduce the Library's framework hierarchy.
A local framework is justified when recurrent local practice needs a coherent language of Methods and their relations. Use E.4.DPF to construct it from the local problems and supplier results; identify variation, authority and the conditions that reopen an adaptation. Do not copy whole DPFs into every project. A contract or resource allocation is not replaced by an LPF, and a single project may need only a few adequately explained local instructions.
Carry a proposed Method into an actual work occasion with sufficient explanation, capability and means. Observe both conformity where it matters and the receiving result. When local adaptation works, distinguish what can be reused from what depends on local facts. When it fails, return to the responsible Method or architecture decision rather than treating failure as disobedience by default.
The typical network thus remains an aid to constructing and revising concrete practice. Its useful invariant is the recoverable connection among developing subjects, consequential problems, viable transformations, means and further development. Its layout and allocation can change.
Qualify continuation without promising infinity
Compare what the construction promised with what it made possible. An improved explanation, a new Method, a prepared performer and an accessible offer are different results. Use each result's own qualification and preserve the support on which it depends. The same observation can support one claim and leave another unresolved.
Show what the changed position makes newly attainable. Formulate a worthwhile next problem and name the means and support for a first attempt. For an inquiry, state what observation, argument or construction would establish its result. For a proposed practical change, identify who can choose it and what that choice depends on. A feasible continuation establishes an opportunity. To show self-development in action, identify the developing agent's own choice of problem, the work it initiated and the help it used. To establish growth beyond greater output on a familiar task, identify the new class of problems, the change in how candidates are generated or evaluated, or the improved means of making further improvements.
Compare the proposed continuation with what was attainable before the change. State the new problem, the capability or understanding it calls for, and which retained method, representation, evidence or provision now makes an attempt feasible. When two ways yield the same present result, the useful difference may be the next change one can make while retaining a required capability. C.40:4.7 and :4.11 help locate the constraint. A novel proposal can justify an inquiry into feasibility. An attainable problem requires a warranted path to an attempt. An initiated investigation additionally requires a performer, means and work actually begun.
Finite observations qualify particular transitions and the means available for continuation. Stop an exhausted or harmful branch, preserve useful knowledge and take up another worthwhile problem when conditions warrant it. A provider's bounded contract can end while the recipient or independent cultural carriers continue.
Archetypal Grounding
From repeated replanning to a different way of committing work
This is a constructed case. Its observations and resource figures are stipulated to expose the Method; they are not a report of an intervention or estimates for other organizations.
A small engineering company integrates business-data feeds for customers. Two project leads use a reliable planning routine for familiar integrations: inspect an agreed input format, divide the work into known tasks, estimate them and reserve people. In a new integration, the first sample passes the existing tests, yet later exports contain revised records that must replace earlier versions. The plan treated every later record as a new item. Adding more execution capacity would produce the wrong report sooner.
The immediate customer result is an adequately processed report. The proposed development result is different: project leads should be able to notice when the basis of a commitment is unestablished, construct a useful inquiry and revise the obtaining way before making that commitment. The company also wants its leads to identify their next development problem and initiate work on it after the adviser leaves. A consultant could repair the current integration without producing those latter results.
The first lead initially frames the problem as inaccurate estimation. With a development adviser, she reconstructs the episode: which input premise was used, why the first sample supported it only weakly, and where the later output ceased to mean the intended business quantity. They contrast the episode with a familiar feed whose replacement semantics are documented and covered by suitable tests. This shows that the consequential difference is not simply project size or programmer speed. In the new case, the necessary processing operation depends on a source behavior that has not yet been established.
The lead can now state the problem for this report. When an export corrects an earlier record, the customer needs the corrected amount to count once. Replaying a known correction must produce that total and retain the other records. This formulation gives the lead a test for any proposed processing method. It also changes the question from “How many more days do we need?” to “Which source behavior must we establish to choose a valid processing way, and what can we responsibly commit before that is known?” The adviser asks for the competing implications rather than supplying a schedule. The subject specialist helps formulate two feasible possibilities: a feed with stable record identities and explicit revisions, or one without sufficient information for reliable automatic replacement. The first permits a bounded processing construction; the second requires changed source provision or a narrower agreed result. A small obtained source history, its interpretation with the data owner and a replay against a known correction can distinguish those possibilities for this decision.
This inquiry is not automatically a scientific research programme. Existing professional knowledge may be enough to construct and interpret it. The lead arranges access to a permitted history and obtains the owner's explanation; the specialist examines the relevant semantics; the lead works through what each possible result permits. They preserve the existing task-estimation routine for the part whose premises remain valid.
The resulting way contains a conditional commitment, not an optimistic average of incompatible plans. Obtain and interpret the decisive source behavior; choose the compatible processing operation; qualify it for the receiving meaning; then commit the dependent work with its remaining assumptions. If source provision cannot support either acceptable route, return that restriction to the customer and direction choice. Protect unaffected current service while this work proceeds.
The supplied contributions now have receiving uses. The episode gives problematisation a failed premise and its consequence. The revised question gives the specialist an interpretable inquiry. The qualified source result gives the lead a basis for selecting a processing way and commitment. The explanation of that selection supplies the learning work with the judgement to practise. The eventual project result is consumed by the customer; the lead's changed capability is a separate claim.
Preserve the construction under materially different conditions
For independent practitioners, no employer can allocate every carrier's time or require disclosure. The same explanation and specialist return may be supplied by a publisher, association or service on acceptable terms; cultural continuation can occur without a communicating group.
Consider a software agent acting as the test engineer on a weighing line. It can inspect calibration records, vary belt speed on a test rig and read a reference load. It separates a constant offset from variation during motion, constructs a test that distinguishes the two, and tries it on previously unused loads. To establish a gain in its diagnostic capability, compare its choice and conduct of the test with its earlier performance under comparable conditions. Record which measurements and assistance it used. If the supplier chose the test and the agent merely ran it, the run establishes execution. Releasing the changed production line is a further decision for whoever holds that authority.
For MIM's development, research can have an independently worthwhile scientific result as well as possible product uses. Developing a residence Method, preparing coordinators, running current cohorts and helping a participant develop are not one flow. A coordinator can contribute to all of them; their joint demand remains real. The typical construction supplies the joins to expose and develop, while strategic inquiry, research, human-capability development, comparison of competing undertakings under shared provision and the professional practice supply their substantive work. The MIM case neither converts all research into immediate client service nor establishes the Method's general boundary.
An episteme, a Method and a discipline in development
Episteme. In the constructed scheduling case of C.40.CD:5.2, events sharing a conflict must occupy different sessions. The conflicts are A–B, B–C, C–D and A–C. A schedule with A and D in session 1, B in 2 and C in 3 establishes that three suffice. The open problem is whether fewer could suffice for this same graph. A, B and C conflict pairwise and therefore require three distinct sessions. The existing schedule and this lower bound together establish that the minimum is three. The subject matter and conflict rule stayed fixed. The successor account entails a claim that the earlier upper bound could not establish. The researchers supplied the argument. A scheduler who must also respect room capacities needs further grounds.
Method. An editor restores an uncertainty condition lost when an instruction was shortened. To help with later edits, the editor adds a reusable operation. For each operative condition stated in the source, compare what the reader may do before and after shortening. On a second instruction this detects a lost restriction on who may approve an action. C.40.NP:5.2 follows the different problem that arises when the source itself leaves that restriction implicit. The repaired instruction, transfer of the existing operation and extension of the Method each have their own test.
Discipline. Consider an established practice of event scheduling. Several planning groups share conflict models, scheduling Methods and a review practice. Their previous standard accepted a feasible schedule as evidence for a published claim of minimality. When two groups submit different feasible counts for the same event, that standard admits incompatible minima. The groups develop the knowledge and practice together. They distinguish feasibility from minimality in their published accounts and require a feasible schedule attaining the claimed count and a lower-bound argument at that same count under the same model. Independent reviewers apply the distinction to later schedules and return counterexamples to the model builders. With room capacities added, for instance, a conflict coloring alone cannot show that everyone will fit. The new case requires a capacity-feasible construction. Following what the groups now publish, check and correct shows a change in the same organized practice-and-knowledge whole (C.20). It also makes a previously unsupported class of minimum claims open to investigation. A committee vote alone could establish the decision to adopt the standard.
Bias-Annotation
The main distortion is to begin with a familiar institution, course or flowchart and make every problem fit it. Recover the developing subject, work and consequential variation before selecting those forms. Another distortion is to equate autonomy with absence of support. Compare the degree and kind of independence worth developing.
A further bias treats success on a prepared case as evidence for the whole proposed culture. Ask which subject changed, which further use was shown and what remains a supported expectation.
Conformance Checklist
- Can the reader distinguish the intended development from current work, Method acquisition, Method construction and delivery of help?
- Does each consequential connection name the contribution and receiving use, with a workable return when its conditions fail?
- Are the proposed flow description, actual transformations, performed Work and organization claims kept at their supported status?
- Can the selected arrangement supply the needed capability, resources and authority together with continuing work?
- Does the next transition arise from the changed subject's position, with a bounded attainable contribution rather than a compulsory endless workload?
- Can another carrier obtain and adapt the relevant Methods without the unpublished invention of the first organizer?
- Does the claimed growth concern knowledge, understanding, Methods or the means of obtaining further development, with ordinary output and acquisition of an existing Method separately described?
- Which problem has become attainable, for whom, and what supports the first attempt? If an independent continuation is already under way, who chose the problem and started the work?
Use the questions needed by the claim; they are not a demand to complete every branch.
Common Anti-Patterns and How to Avoid Them
The office is the solution. Recover the connected work, its performers, results and conditions before and while comparing organization concepts.
Everything involving Methods is development. Identify the subject, worthwhile change and continuation being claimed.
The arrow establishes the relation. Qualify the direct relation and its local conditions; retain a proposed description where facts are missing.
One successful service proves recipient mastery. Keep delivery, assisted performance, transferred capability and self-development distinct.
All research feeds today's product. Preserve independent research ends and their relation to the rest of the portfolio.
The new AI removes every human purpose. Recompare the whole allocation against the retained result and duties.
The example defines the Method. Keep case assumptions local and test consequential changes that could defeat the promised use.
Consequences
The undertaking becomes constructible and revisable across disciplines. The practitioner can locate an unavailable contribution, compare provision and change a failed connection without restarting everything. Development can continue after a bounded success without requiring the same office, provider or project forever.
The Method costs effort in recovering relations and comparing whole configurations. It does not remove professional knowledge, supply missing authority or guarantee cultural uptake. Use an account sufficient for the consequential choices and deepen it where missing structure would change action.
Rationale
The recurring difficulty is a failure of the connected whole, not necessarily the absence of an elementary operation. Existing FPF and DPF Methods already provide problem formation, construction, realization, capability development, organization design and cultural continuation. This Method makes their receiving relations and returns explicit for continuing development.
A connected description and a worked case can support understanding and prospective construction. Qualification of concrete outcomes remains with the subject, work and evidence. This distinction lets a typical network guide a future arrangement without asserting that its work has already occurred.
SoTA-Echoing
The whole question is how to construct an attainable development undertaking, realize it through available or obtainable capabilities and means, and continue from the changed position without confusing that continuation with indefinite operation of the same project. For this question, an existing Library answer is direct use of C.32.MWA.PP, C.40.CU and the qualified subject Methods. C.32.MWA.PP already compares Method, Work, allocation, provision and other decision-changing relations, including joint burden and realization conditions. C.40.CU already connects an examined response to a useful application, alternative obtaining arrangements, realization, reproduction and condition-sensitive renewal. Subject Methods supply the research, learning, engineering or organizational operations. This is a sufficient alternative, not merely a choice of office diagram. When the practitioner can obtain the required result and its next continuation from that composition, use it; the present pattern requires no additional modelling exercise.
The remaining difficulty arises when individually adequate contributions leave the developmental connections to private invention: how a newly noticed limitation changes the problem and the obtaining way together; how the agent's present capability enables its next capability development; how immediate delivery, development of a performer and provision rejoin without one standing in for another; and how success opens further development and cultural continuation. C.40.OD:4, C.40.JD:4.1–4.3 and C.40.NP:4.1 develop those joins. C.40.JD:4.4 derives the changes when a supplier changes, and C.40.NP:4.2–4.3 explains continuation after success and reconstruction by another carrier. The worked case follows a failed commitment premise into a changed decision capability and then a harder carrier-development question. These moves reuse A.12, B.5.PI, C.39, C.40.CD/CU, E.23.CDI and C.36/C.36.RP; E.18/E.18.NET, C.30.TFS-REL and C.32.MWA retain the distinctions needed to describe and realize the construction.
For organizational provision, Rebecca Peters's CIPD analysis, 9 July 2024, compares central, local and hybrid HR arrangements in relation to purpose, continuing operations and provider capability. This pattern adapts that comparison to the support-remit choice; it does not transfer an HR configuration as a generally optimal development office.
Ina M. Sebastian, Peter Weill, Thomas Haskamp and Jan vom Brocke's MIT CISR study, “Designing Decision Rights for AI”, 18 June 2026, distinguishes framing, action and learning decisions under risk and ambiguity. Its interview-based examples include local business ownership combined with common AI expertise and infrastructure. Here that is a source for examining which decisions and contributions can be distributed or shared; its human organizational roles are not mandatory roles for every agent arrangement.
Neither this study nor the CIPD cases establish comparative effectiveness for every community. The applicable organizational Methods and C.32.MWA.PP retain the operative organizational construction; implementation-support practice supplies needed help in an enactment, not the prior definition of every development question.
The 2024 essays on role communities inside a company and around a company motivate attention to overlapping organization/community relations, participants' own journeys and competing formal and informal influence. They are conceptual contributions, not evidence for universal predictions about human replacement or office disappearance.
C.18 distinguishes exploration within a possibility space, expansion of operations or regions, and a changed generation or evaluation regime; C.19 governs which live branches receive further resources. E.23 distinguishes improvement from improvement of its obtaining means. C.36/C.36.RP supply cultural reconstruction, use, selection and renewal. Their differences prevent one score, one new version or one copied exercise from standing in for open-ended development. Subject practices supply learning, research, computational and organizational mechanisms.
Relations
C.40.OI, C.40.OC, C.40.NU, C.40.PC, C.40.DC, C.40.JF, C.40.OP, C.40.RQ and C.40.OCN supply opportunity and direction construction; this Method uses those results and returns changed premises. C.32.MWA.PP compares how supplied Methods can be performed together, including their required contributions, means and authority; C.39 and C.40.CD develop an obtaining way. Organizational realization, human learning and transfer, engineered means, research and strategic commitments remain separately qualified subject contributions when the undertaking needs them. A.15.9 obtains the missing result and its reliance limits without prescribing one framework as its supplier.
C.36/C.36.RP and C.36.MC supply the general cultural-continuation questions. For a separately selected leadership or community engagement, the applicable leadership Method supports proposal preparation, decision and first use, and the applicable community Method supports voluntary participation. E.4.DPF supplies a local language when recurrent local practice warrants one.
These use and support relations do not by themselves establish constructive Method parthood, a universal work sequence or a containing organization.
C.40.JD develops the joint construction of a way, its holder and provision. C.40.NP develops continuation from a qualified result. Each can be used for its own result without constructing a larger undertaking. C.40.CD constructs a problem and way together; C.40.CU constructs a reproducible application. C.40.OCN can supply a bounded opportunity without a continuing-development commitment.
C.40.OD:End
Develop a Way, Its Holder and Its Provision Together
Type: Method pattern Status: Stable Normativity: Normative
Problem frame
Use this Method when obtaining a worthwhile result requires coupled changes in a way of doing, its performer or holder, and the conditions that make performance possible. The difficulty can be a good Method that cannot be enacted locally, capable help that bypasses the intended learning, or learning that produces an operation current work cannot use. The result is a jointly workable construction with separate claims for delivery, acquired capability and provision.
An already adequate way, capable holder and sufficient provision can be used directly. A simple request for help belongs with the applicable implementation-support practice or the supplier. This Method does not make every supported episode open-ended development; C.40.OD governs that larger claim.
Problem
Developing each component independently leaves their joins unconstructed. A course can teach an operation missing from actual work, a provider can deliver while the recipient remains dependent, and an organization can allocate time without supplying the missing competence. Altering one branch can also make another unnecessary. The practitioner needs to construct and revise the combination without protecting obsolete steps.
Forces
Solution
Join development of the way with development of its performer
Construct one adequate way to obtain the chosen change. Its explanation connects the material or information available in the work, the operative distinctions and actions, the intermediate result needed by the next action, the receiving result and the conditions on which those connections depend. Use C.39 and C.40.CD to develop this construction; obtain a qualified research contribution through A.15.9 for an unresolved research claim that matters to the intended use. A method description that names only its final result does not yet supply the next performer.
This way then separates into three possible receiving contributions:
- A capable performer can use it to supply the immediate result.
- A capability-development Method appropriate to the holder can use the explained operations and revealing contrasts to develop that holder, under its own mechanism and evidence conditions.
- Provision and organizational design can use its requirements to construct the access, time, tools, authority and cooperation needed for performance.
These branches are joined by the same intended use and its operative conditions, not by a shared document title. In a human-learning branch, the learning task must require the consequential operation that later work needs. In a provision branch, capacity must be available when that task and later work consume it. When immediate delivery and learning use one occasion, distinguish the helper's contribution from the learner's. A helper can complete the project while accidentally removing every opportunity for the learner to make the judgement being developed.
The branches also return results to one another. If a learner's attempt exposes an unexplained distinction, return it to the Method construction. If a sound operation cannot be executed under local information access, return the missing access to provision or alter the way. If changing the arrangement removes the operation entirely, reconsider whether learning it still serves a retained development aim. A machine that eliminates an unnecessary decision can be a better whole solution than training everyone to make it.
Use present capabilities to develop the same agent
An agent can use a capability it already has to perform work through which it develops another of its capabilities. This can concern a person, an AI system, a team or an organization. The available capability supplies a contribution to that work; it need not supply the whole undertaking. For example, being able to construct a revealing comparison can help someone develop a new professional judgement, but does not supply missing subject knowledge, information access or authority. Begin with the intended changed possibilities, not with a presumption that every exercise or Method change is development.
Work backward from the wanted capability through E.23.CDI:4.2–4.2.1, then follow the proposed development forward with the agent's means. Which operation must become possible? What would let the agent construct or acquire its way of performing that operation? What can it already do to obtain that contribution? An agent may already be able to recover an explanation, construct a trial, distinguish an error, obtain expert help or revise an arrangement. Explain how the selected action produces the input needed by the next action. If the agent can reconstruct a demonstrated solution but cannot judge an unfamiliar one, a trusted worked comparison can supply the missing judgement while it learns the relevant distinction. Do not require the target capability in order to acquire it. If the development operation itself is unavailable, obtain that contribution or develop it under an attainable smaller question.
Use A.12 to expose who acts on what. Where acting and changed constituents are distinct parts of the same agent, Reflexive Split supplies the internal account. An organization's method-development group can change the preparation of another group; an AI's existing orchestration component can construct tests and install an authorized change in its retrieval component. Their part relations and the interaction belong to the chosen account. A capability, role or later edition alone does not identify either constituent. Select enough physical, software or organizational detail to make the intervention intelligible and decide it; an ordinary development choice does not require an anatomical account of a person. Preserve an adequate existing Reflexive Split and supply the missing development operations through the relevant Methods. A.12 identifies the participants and relation; it does not by itself teach the subject operation or establish its outcome.
The working connection has to survive enactment. In the proposed AI arrangement, the orchestrator's existing test-construction ability produces contrasting queries and expected distinctions. A qualified source or evaluator supplies any judgement that the orchestrator cannot make. The permitted change modifies the retrieval component, and representative receiving queries test the capability of the resulting agent with its tools and remaining help. A changed file is not that capability result. For a person, a worked explanation, a chosen practice task and feedback can play corresponding roles, but the learning Method must explain how the person's attempts and corrections develop the operation. Software installation is not a model of human learning. In an organization, changing an instruction still requires the relevant groups to acquire or obtain the operations, coordinate their work and have authority to perform it. The agent-specific Methods supply these mechanisms.
Keep external contributions visible without treating useful assistance as a failure. An outside specialist remains an outside contributor unless the chosen whole includes that contributor; its successful work is not automatically the recipient's acquired capability. State which contributions the agent can subsequently perform or organize and which remain supplied. When a further worthwhile question is needed, use the changed position through C.40.NP to form it. When the acquired capability changes how the agent obtains subsequent development, trace that additional connection: for example, a newly learned way of examining errors can change the construction of the next practice task. Reusing an adequate way is legitimate. Improving the way of developing is needed when it changes an attainable result or burden, not as a compulsory extra cycle after every success.
Rejoin in consequential work and distinguish what changed
Bring the available way, performer and provision together in one consequential receiving occasion. Ordinary work supplies the material and stakes; the developed way supplies the operation; the performer enacts it; the arrangement supplies the necessary conditions. The work changes its own subject, such as a service, engineered system or decision. This performance can show what the holder can do or organize with the help intended for later use. To judge a change in capability, compare it with grounded evidence of what the same holder could do earlier. Account for differences in the task, access and help that could explain the change in performance. If the earlier capability is unknown, qualify the present performance and leave the change open.
Use the work and its appropriate examination to recover three distinct results where they matter:
- what the receiving activity obtained;
- what the developing subject can now do, under which retained assistance and conditions;
- what the attempt revealed about the way and its provision.
The receiving activity can use its result. Those choosing further work can use what they know about the holder's present capability, including a capability it already had. What the attempt reveals about the Method or its provision can change the next decision. When a specialist made the critical choice, the delivered service and the holder's own contribution can each be credited at the scope the work supports.
Use the development or provision Method selected for the holder to make the named operation available. Check the receiving performance under the promised assistance and state what capability the evidence supports for the holder and arrangement. C.40.CU, E.23.CDI and the subject evaluation Methods provide the corresponding construction and qualification.
Change a connection and derive the changed arrangement
Local adaptation changes the construction, not just its names. Begin with the operative condition that differs; follow only the joins whose receiving result depended on it; reconstruct those joins; and compare the changed whole with the original intended result.
Consider a team developing its ability to frame and revise consequential operating decisions. Its first construction requires development of a technical calculation, learning how to interpret the result, use in a real decision and cultural continuation to another team. Two alternatives differ substantively:
The changed supply removes one internal branch and adds a provider dependence. It does not automatically remove the subject's development or transfer responsibility for the operating decision. Conversely, if the only retained aim was obtaining a qualified decision and no independent judgement is wanted or needed, revise the direction openly and compare a service solution. Do not retain compulsory training under a discarded aim.
A different change gives a different answer. If the external service fails the local confidentiality condition, the technical result can remain adequate while its provision is unusable. Consider an authorized local execution, a less sensitive input, another provider or renewed internal construction. “The Method exists” does not settle that arrangement. If an AI arrangement can perform the whole retained contribution with the required powers and supported reliability, compare it as a whole alternative rather than preserving human steps by habit.
For ordinary leadership without a separate office, allocate these same contributions to managers, practitioners and obtainable specialists. For a dedicated development office, keep it responsible for the development help it can provide and make its technical suppliers explicit. For a common service covering several masteries, share only contributions that are reusable: for example, access, practice design or maintained infrastructure. Preserve specialized problem formulation and judgement where substituting one subject's standard would defeat another's use. Federation combines these where common provision and local judgement both matter.
Use the same adaptation on connections omitted from the drawing: a provider's authority may not extend to a voluntary participant; a learning occasion may compete with a production deadline; two branches may require the same scarce expert simultaneously. Follow the changed condition through the receiving work and compare the moved burden. These changes can alter the direction, the Method, its allocation or the service remit; not every failure calls for another organization chart.
When the resulting performance and provision arrangements still need comparison, use C.32.MWA.PP. Use it to obtain complete alternatives and their combined demands, then return that comparison to the choice in C.40.JD:4.5. It can use adequate existing Methods while changing their Work order, allocation, provision or provider relations. Constructing or changing an organization requires a Method for the organizational change being proposed; comparing arrangements does not enact that change. An existing arrangement that meets the present requirements needs no new architectural exercise.
Before relying on a consequential connection, distinguish a proposed requirement from an established relation:
A small local construction can remain in ordinary language when it supports the decision. Stronger formal reliance requires the corresponding direct basis. The purpose of that qualification is to preserve meaning while making the undertaking usable, not to defer every practical decision until a complete formal model exists.
Synthesize the organization from the work and conditions
For each alternative, recover the Work contributions, capability holders, means, authority and mutual dependencies needed to enact it. Use C.32.MWA to distinguish the relevant Method and Work structures, and C.32.MWA.PP:4.3–4.8 to construct and compare the arrangements and their combined demands. Return the comparison as a basis for the authorized decision about whether and how to proceed with the development aim and its arrangement. Keep comparison distinct from that choice: a limited test, retained alternatives, a revised remit or an adequately grounded stop can be a complete return. Preserve who supplies each operation, who receives its result, the effective assignments and authority, enabling conditions and exception return. C.38 can compare human, AI, robotic and provider ways for one result.
Developing the organization concept, creating positions or establishing enabling relations requires the corresponding organizational design and change work; comparing alternatives does not establish these results.
Start with the contributions that must remain with the developing subject or at the place of work. Where self-development is intended, someone must notice and help formulate their own next difficulty, judge what matters to them and use the changed possibilities. A helper may assist strongly; outsourcing all those judgements changes the self-development claim. Where the purpose is obtaining a competent result rather than personal mastery, more work can properly be delegated.
Then locate professional help, common means and decisions that need a wider view. They may be performed by the same person, ordinary managers, peers, a temporary team, a specialist office, a common service, an external provider, AI systems or combinations. A name such as “development office” establishes neither a capability nor an authority relation.
Compare at least the serious feasible alternatives exposed by the case, not every possible organization chart. The usual differences are:
Provider remit and realization arrangement are coupled choices. “All needed masteries are obtainable” does not entail “one office owns and teaches them all.” A tunnelling company can use specialist technical providers for tunnelling mastery and a narrower development practice for constructing its next worthwhile transitions. MIM can coherently specialize in development while referring a client's geological problem to another provider. Both must connect the obtained result to their receiving work.
In an organization with autonomous AI agents, distinguish development of the agents, development of the human participants, and improvement of the combined arrangement. A cheaper AI answer can remove an unnecessary human processing job; it does not demonstrate human mastery. Conversely, preserving human mastery as an explicit objective can justify practice that would not be needed to obtain the immediate answer.
Compare complete configurations through C.11.CRC and C.11.DUA over the same horizon and sufficient result. Include construction, acquisition, learning, adaptation, operation, maintenance, coordination, checking, transition and displaced work. Count a shared contribution once only where it can serve both uses. A benefit that depends on unavailable authority or capability is not available merely because its direct monetary cost is low.
Close the joint construction at the claim it supports
Return the obtaining way, the operations the intended holder must perform or organize, the provision that makes those operations attainable, and the consequential joins between them. State what was realized and what remains a prospective requirement. Reuse an adequate existing arrangement without changing its organization merely to make the description symmetrical.
Test a change that could defeat a join: the competent helper is unavailable, protected material cannot be shared, an AI can now supply the whole service, or a recipient still wants independent mastery. Follow the changed premise into the relevant operation and receiving result. Retain unaffected contributions. If the result is a service solution with no retained holder-development aim, name that result and stop its unnecessary learning branch.
The construction succeeds for its declared use when the needed operations and provision can work together and the relied-on results have appropriate grounds. A proposed feasible arrangement, a delivered result and acquired capability remain distinct conclusions. Use C.40.OD only when this joint construction belongs to a larger open-ended development undertaking; use C.40.NP when the changed result opens a further problem.
Archetypal Grounding
Derive a feasible arrangement without making the office the premise
In this constructed case, two integration leads repeatedly promise delivery before checking whether revised source records can change an earlier total. The intended change is their capacity to identify the consequential premise, obtain a qualified subject answer and form a warranted commitment. A data specialist supplies source semantics; an adviser helps the leads construct and vary the inquiry. Current delivery must continue. The case distinguishes the improved integration Method, the leads' capability and the support needed for both.
The company compares three whole arrangements. Ordinary managers can provide the help if they have the relevant development skill and protected time, obtaining the data specialist where necessary. A development specialist can supply problematisation and learning support while obtaining technical judgement from elsewhere. A general competence office can supply both through a broader maintained service. The office gains no presumption merely because several kinds of help occur.
For this initial horizon, suppose the ordinary managers lack time to supply the development help, an available adviser has that capability, and the subject specialist can provide the bounded technical contribution. The proposed general office would add fixed coordination work without removing another current constraint. Choose the specialist arrangement for the first undertaking. Keep the conditions under which recurrent demand and reusable common support would change that choice.
The adviser has six hours available each week. Existing support consumes two. Joint work on the failed premise, question and conditional commitment requires two; each lead's own contrasting case requires one; preparing another helper would require two more. The initial request is therefore eight hours. Defer helper preparation for this first attempt, leaving six. This does not establish the wider cultural aim or make preparing a successor unnecessary forever.
The joint work is shareable only because both leads need the same operative distinction and can lawfully use the shared material. Each individual case remains separate because its local premise and decision must be reconstructed. If either confidentiality or the problem difference defeats sharing, the two-hour contribution cannot be counted once; change the timing, supply or promise.
The leads' work time, subject-specialist access, permitted data and the manager's authority must also be available. In this constructed case they are stipulated as obtainable for the bounded attempt; the arithmetic establishes only the adviser constraint. A real organization must replace those premises with its actual availability. Six adviser hours cannot compensate for a lead assigned entirely to current delivery.
Proposal discussion changes the arrangement. A manager wants uncertainty reports included unqualified in performance ratings. A lead explains that this would make early discovery of an unsupported commitment personally costly. The concern is understood correctly, so further persuasion is not the answer. The manager agrees a different use of the material and makes the corresponding work-time allocation. If that authority or commitment were unavailable, the design would retain a dependency, not claim a supportive culture.
The same adviser and manager continue ordinary work during this change. No separate project box supplies extra people. Development of the integration Method, present integration delivery, the learning work and provision of help remain distinct contributions even when one person performs several.
Show the changed capability in consequential work
In the first attempt, the specialist establishes the source semantics and the lead explains why that result changes the permissible commitment. The adviser then varies the next case. In another feed, stable semantics are already adequately documented; in a third, source latency rather than revision semantics defeats the promised reporting time. Applying the new inquiry unchanged to all three would be another fixed routine.
The lead has to locate which condition is consequential. For the well-supported feed she can use the familiar way without adding a discovery exercise. For the latency case she distinguishes a calculation defect from an unavailable input at the promised time and constructs alternatives involving the receiving window or provision. She obtains the relevant professional answer where needed and makes the commitment within her authority. Help remains available on the agreed terms.
Those enactments would support only what the performance evidence establishes. If the lead independently notices the premise difference and formulates the consequential question but still needs a specialist's technical result, retain that bounded capability with its support dependence. Do not call it complete subject expertise. If the adviser still chooses the difference and every consequential operation, retain the delivered project result and assisted performance, and continue the missing development.
The later cases show what the lead can do with the agreed help. If comparable earlier work shows that she could not identify the premise or frame the question under those conditions, her later choices support a gain in those operations. If earlier records show that she made those choices whenever the source information was available, the skill was already present. If missing access had blocked an earlier commitment, restoring that access removes the documented obstacle. Protected time or specialist help may also have made the present commitment possible. If those conditions changed together, the contribution of access alone may remain unknown. Either finding can guide the next assignment. Attributing a gain to the adviser's intervention also requires grounds for separating its effect from other changes.
Support several masteries in a company without centralising every judgement
This constructed application stipulates its working conditions. It illustrates the linked Methods, not measured superiority of an office form or evidence about a named real organisation.
A company wants project leads to notice and question consequential premises in engineering estimates, resource accounts and customer commitments. Its first proposal is one mastery office. The intended result is more specific: each lead should recognise when a premise needs examination, obtain the appropriate answer and revise the commitment within their authority. The leads are also to become better able to formulate their next development question. They need not each become an engineer, accountant and market researcher.
Use the difficulty inquiry in C.40.OD:4.2 to recover three contrasting episodes with the people doing the work, then identify the missing contributions through C.40.JD:4.1. Stipulate that two failures concerned unexplained specialist limits and one concerned a schedule that left no time to examine a known uncertainty. An adequate technical Method and capable specialists already exist. Further research is unnecessary for these gaps; the company needs usable specialist returns, preparation for the leads and changed work conditions.
Compare the complete alternatives through C.40.JD:4.5, including authority and commitments through the applicable Method for agreeing feasible joint contributions and the receiving work and learning contributions through C.40.JD:4.3. In this case the line managers can prepare and debrief the relevant work, have protected time within revised current obligations, and can obtain the specialists before the decision windows. An existing service maintains permitted source material. These premises make a distributed arrangement viable for the bounded start. Creating a separate office would add no missing contribution to that start. Keep the comparison open if the premises change.
At each workplace, the lead brings the local episode and identifies the decision whose premise matters. The subject specialist returns the answer, applicability and any unresolved question. A qualified learning contributor supplies a contrasting practice task and feedback on the lead's own judgement. The work owner provides the receiving occasion and any permitted change to current obligations. The lead then performs the receiving task under those conditions. This construction uses C.40.JD's receiving-work join; it does not infer mastery from the specialist's successful answer.
The shared contribution is a usable way of asking for and interpreting a qualified answer, together with permitted access and maintenance. Its application still requires different engineering, accounting and customer evidence. A common AI assistant may retrieve the approved material and offer contrasting cases within its qualified scope. It does not acquire a right to make the company's commitment, and its fluent answer cannot silently replace the lead's intended practice.
Suppose later work shows that leads can now question an estimate, but commitments from different teams use incompatible meanings of “available capacity”. The completed learning result remains useful. They now need to make commitments across teams using comparable capacity claims, a more demanding use of the acquired judgement. Use these cases to develop the shared repertoire; connect the practices through a useful joint contribution. The relevant accounting and operating Methods must supply the common comparison's meaning. If that construction is unavailable, obtain it before making a commitment that depends on the comparison.
Now stipulate that the preparation needed for recurring cross-team work exceeds the time local managers can provide within their retained commitments. A common service becomes an alternative if it can supply that shared preparation at the required times while local specialists retain their judgements. A specialised development provider is a different alternative if the missing contribution is sustained problematisation and improvement of how the company develops its ways. Its technical suppliers remain explicit. A federation can combine that common work with local conditions. Compare these arrangements against the changed demand, available competence, total work and powers.
If the proposed service cannot receive the necessary confidential material, return to a permitted local execution, a narrower common contribution or another qualified arrangement. Do not carry the old sharing estimate into this changed configuration. For example, suppose one shared two-hour session, four half-hour case reviews and two hours of existing work use all six available specialist hours. If confidentiality requires four separate two-hour sessions, demand becomes twelve hours against the same six-hour capacity. The earlier fit to that capacity limit fails even though the technical distinction remains supported.
Bias-Annotation
An educator can overvalue retained learning, a service provider can overvalue delegation, and an organizational designer can overvalue a new unit. Keep the intended result and an adequate current arrangement in the comparison. A visible human operation need not be retained when another arrangement supplies the whole required contribution.
Conformance Checklist
- The wanted result and any independent capability aim are explicit.
- The way explains the operations and conditions needed by the holder.
- The learning or change mechanism belongs to the subject.
- Performance, holder development and provision return relevant results to one another.
- Joint costs, authority, access and continuing work are feasible at the relied-on scope.
- A changed connection has been followed through the affected whole.
- Delivery, assisted performance and acquired capability receive distinct conclusions.
Common Anti-Patterns and How to Avoid Them
Consequences
The construction can retain development that a cheap service would bypass, or remove training that no longer serves the selected end. It exposes the cost of coupling and the support still needed. This can justify a smaller arrangement, a different remit or a stop; no organization form is selected in advance.
Rationale
The same intended use connects the branches while their results remain different. Cross-return makes learning, construction and provision mutually correctable. This relation is more informative than placing training before performance or naming an office responsible for both. Its realization still depends on the subject Methods.
SoTA-Echoing
The question is how a reusable way, an intended holder and its provision become jointly usable without treating delivered work as acquired capability. C.40.CU already constructs alternative obtaining arrangements, realization and reproduction. E.23.CDI supplies capability development for a work family. Their adequate direct use is an alternative. The present Method develops the cross-return: learning can reveal an incomplete explanation, provision can defeat an otherwise adequate operation, and a changed way can remove the need for that operation.
For organizational provision, Rebecca Peters's CIPD analysis, 9 July 2024, compares central, local and hybrid HR arrangements in relation to purpose, continuing operations and provider capability. This pattern adapts that comparison to the support-remit choice; it does not transfer an HR configuration as a generally optimal development office.
Ina M. Sebastian, Peter Weill, Thomas Haskamp and Jan vom Brocke's MIT CISR study, “Designing Decision Rights for AI”, 18 June 2026, distinguishes framing, action and learning decisions under risk and ambiguity. Its interview-based examples include local business ownership combined with common AI expertise and infrastructure. Here that is a source for examining which decisions and contributions can be distributed or shared; its human organizational roles are not mandatory roles for every agent arrangement.
Neither this study nor the CIPD cases establish comparative effectiveness for every community. The applicable organizational Methods and the Core Method constructions retain the operative organizational construction; implementation-support practice supplies needed help in an enactment, not the prior definition of every development question.
Relations
C.39 and C.40.CD develop the obtaining way. C.40.CU supplies application, realization and reproduction. E.23.CDI supplies the general capability-development construction; the holder needs its applicable development mechanism and evidence. A.12 qualifies a Reflexive Split. C.32.MWA and the applicable Method-and-support comparison supply coupled structure and burden. Organization construction and practical implementation help remain separate subject contributions when needed. C.40.OD composes a developmental whole, and C.40.NP continues from its result. A separately selected strategic comparison may use an applicable strategy Method to compare competing undertakings under shared provision.
C.40.JD:End
Continue from a Result to New Attainable Problems
Type: Method pattern Status: Stable Normativity: Normative
Problem frame
Use this Method after a qualified result changes what can be understood or obtained, but the next worthwhile problem and means of continuation are unclear. A successful project can reveal a deeper limit, make another question answerable, or provide a construction another carrier can develop. The result is a justified attainable continuation, a useful retained possibility or a local stop.
Routine repetition of the same service, more exercises under an unchanged aim and a mandatory next project are not the intended result. Use ordinary operations when the current way is adequate. Use C.40.JD when a present way, holder and provision still need to be made jointly workable.
Problem
Success can conceal a new limit or supply the means to overcome an old one. Treating it as final closure, or repeating the successful activity on schedule, can leave a more demanding problem undiscovered. Demanding novelty at every occasion can consume useful practice. The practitioner needs to decide what the changed position now warrants.
Forces
Solution
Recover the changed starting position
Start from the result and the conditions under which it holds. Recover what a performer can now notice, attempt or combine, or what an improved account now enables someone to infer. A revised Method can open a new class of performances. A change in a discipline's repertoire or standard of evidence can make previously unsupported investigations possible. Compare the old and new position and formulate the worthwhile problem that has come within reach. A completed project can satisfy all its acceptance conditions. Repeating it on schedule can also remain useful routine work.
For example, a lead who can now revise a planning assumption may become able to notice that the organization has no way for another lead to dispute an inherited assumption. The next problem concerns the conditions under which a judgement can be reconstructed, challenged and renewed across people. Repeating the planning course does not supply that result. The new direction can require a different way of discussing evidence, a different information-use rule and learning support for another carrier. It uses the first result without making the first project endless.
Cultural continuation is a connected but distinct strand. Use the operative explanation, decisive contrasts and conditions from the successful work to make the practice obtainable by another carrier. Through C.36.RP, arrange reconstruction, first consequential use and renewal in that carrier's conditions. A faithful copy can fail when those conditions differ. Return a useful new variant or a revealing failure to the common repertoire; preserve local differences that explain why the variants work. A shared employer or an organized community can support this work, but is not required for it.
When several occasions reveal the same costly limitation in noticing, framing, constructing or continuing, those responsible for that work can develop the way they conduct development. The returned result must make a subsequent worthwhile transition more attainable or less burdensome. Producing a new framework, helper or training programme does not establish that gain by itself. Reuse the same construction at that subject and purpose; do not create a new universal level of development merely because its object is another development Method.
Construct the next worthwhile attainable problem
Recover the starting position established by the qualified result: attainable work, support dependence, new constraints, retained possibilities and newly exposed limits. C.40.OP and C.40.RQ preserve opportunities and requalify changed premises. B.5.PI and C.40.CD support the next inquiry even if the completed project meets all its declared criteria.
Ask what became worth attempting that was previously inaccessible or invisible. A Goldilocks problem stretches what can currently be achieved and is within reach of the agents undertaking it with the support, time and resources available for the attempt. Judge the value of the advance and the cost of an unsuccessful attempt. A reversible trial can make a demanding problem worth taking up. Record the fit for this use under C.22.2:9. A change in support can change the difficulty. Ending one undertaking can free the means for a more important one.
Separate the end of a project, a provider's contract, one chosen direction and the subject's continuing development. A provider may complete a good bounded service while another provider or the subject takes the next work. If nobody can yet continue without the original helper choosing the problem, retain the supported-development result and identify what remains before claiming the intended self-development. For a claim that self-development has begun, identify the developing agent, the problem it chose, the first step it took and the support it used.
Improvement of development itself is another possible direction. Identify the limitation in noticing, framing, choosing, obtaining means, checking or continuing; construct a better way; use it in a subsequent development occasion; and compare the result and its total burden. Preparing a replacement helper sustains supply. Changing how future helpers are developed can improve the means of development. Neither recursion nor a new generation is valuable by label alone.
A finite observation can support a repeatable next transition, transfer to another carrier or reduced dependence. It cannot demonstrate that development will continue indefinitely.
When the changed position exposes several costly or previously unattainable ways of working, C.39.WI develops their discovery and selection. Compare the receiving gain and complete improvement burden, including effects beyond throughput and work displaced at a constraint. The selected improvement returns to this Method through what its use makes newly attainable; neither the largest reported cost nor a fixed quota determines the next problem.
Enable further carriers without requiring compliance
When development is intended to become ordinary practice, identify what must be reconstructed and used by further carriers. The object is not necessarily a fixed procedure. It can include ways of noticing difficulties, disputing formulations, obtaining professional help, constructing a transition and reopening the question after success.
Use C.36/C.36.RP to distinguish generation, transmission, reconstruction, use, selection and retention. Use C.36.MC for the bounded cultural claim and its receiving construction. Where the continuation relies on distributed participation, obtain the contributions, usable occasions, access, support and agreements that make another carrier able and willing to act and return a useful result. Retain their voluntary limits and shared demands. The applicable organizing and subject practices supply those mechanisms; a communication campaign does not establish their results.
A workplace, association, customer community and population of independent practitioners expose different constraints. Organizational membership may support some work assignments; participation in a community or receipt of a public text does not confer the same authority. Independent carriers can use and develop the practice without a shared meeting, employer or provider.
Create conditions for another capable carrier to recognise and organize a materially different next development problem. Successful performance of the founder's prepared exercise demonstrates less. Examine what the next carrier can reconstruct from available material and help, what changes under local conditions, and which result is returned to the common repertoire. Qualified disagreement or an independently improved variant can sustain development better than uniform repetition.
Distinguish reuse, recursion and growth
An adequate way can be reused without changing it. Select improvement of the developing means when a new limitation or opportunity makes that change worthwhile. To claim recursion, trace how an improved means contributes to a subsequent improvement: a new error-analysis Method changes construction of the next learning task, or a prepared practitioner changes preparation of the next practitioner. Merely applying a corrected instruction to its usual task shows use, not this further relation.
Characterize what has become attainable: a new problem class, a previously unavailable combination of quality and cost, a stronger way of constructing questions, or an improved capacity to obtain further changes. Keep the original result and the investment in future change visible. Preparing another contributor can temporarily reduce current output while making a valuable continuation possible.
Use the mechanisms appropriate to the subject. Software may permit restoration of a previous version; human experience, changed authority and damaged work relationships cannot be rolled back by replacing a file. Limited trials, parallel work, revised commitments and recovery of necessary functions need their own subject Methods. A cross-subject analogy identifies a question, not its mechanism.
Where criteria change, retain the earlier result and explain the new question. A better score under changed criteria is not evidence that the subject improved. If quantities are not comparable, state the acquired possibility and its conditions. Qualify the observed transitions and retained means; do not extrapolate finite success into guaranteed endless development.
Archetypal Grounding
Continue from success into a harder, attainable problem
In this constructed case, two integration leads repeatedly promise delivery before checking whether revised source records can change an earlier total. The intended change is their capacity to identify the consequential premise, obtain a qualified subject answer and form a warranted commitment. A data specialist supplies source semantics; an adviser helps the leads construct and vary the inquiry. Current delivery must continue. The case distinguishes the improved integration Method, the leads' capability and the support needed for both.
The first undertaking has succeeded within its declared scope: the leads can distinguish adequate existing premises from a consequential unknown. Its adviser account allocated two hours to incumbent support, two to a joint construction and one to each lead’s individual case, totaling six hours per week.
After the first projects work, the lead reviews a newcomer's planned integration. The newcomer has copied the visible lesson as “always run a discovery trial”. He proposes a week of replay on a feed whose operative premises are already adequately supported. Asked what result would change his action, he can name no such distinction. The previous successful project concealed a new limitation: the organization transmits the exercise more readily than the judgement that made the exercise worthwhile.
The lead now identifies a further problem herself. How can another person reconstruct and dispute the reason for an inquiry, without either copying a ritual or requiring the original adviser to authorize every case? This is harder than making her own commitment: another carrier must recognize the distinction under different preparation and conditions, and the working arrangement must permit a justified disagreement. It is attainable because the first development supplied contrasting cases, an explained construction and a lead able to examine its limits.
She does not solve it by adding a rule saying “do not run unnecessary trials”. She obtains help for the missing learning operation from the applicable capability-development practice and uses C.36.RP for reproducible access and renewal. Together they construct a comparison in which the newcomer must relate the input premise, proposed inquiry result and changed action. For one adequately supported case he explains why no new trial is needed. For a differently unsupported case he constructs the result he must obtain and the consequence of each possible answer. A disagreement with the lead is examined against the material and receiving decision, rather than treated as a failure to remember the house style.
The lead also changes provision. The usable account retains the decisive contrasting material and the reasons for the selected action, with access to a subject specialist for technical questions. It is not a catalogue of approved answers. At a later real occasion, the newcomer must formulate the issue and obtain the needed contribution under the support conditions the company intends to maintain. An adviser silently repairing the case would preserve delivery but defeat this transfer claim.
The earlier two-hour joint construction has ended. In the next explicitly agreed horizon, its released adviser time can support this carrier-development work while the two hours of incumbent support and the two individual hours remain. That gives six hours again, subject to the allocation being confirmed. If first-stage support has not ended, the transfer undertaking needs another arrangement. A programme's successful milestone does not free capacity by wording alone.
The new result can establish a usable way for another carrier and a better local means of development within its observed scope. It does not prove endless continuation. The practice keeps returning from the changed position to worthwhile attainable problems; a later one might concern several teams' incompatible commitments, another subject's learning needs or a better way to expose hidden assumptions. Select it for its contribution and difficulty relative to the subject. Do not manufacture increasingly elaborate exercises simply to keep the programme running.
An editorial result opens a different condition-loss problem
An editing Method shortens a cost comparison. The source permits selection only when the complete uncertainty interval is below the limit. The shortened text leaves only a point estimate. Restoring the interval condition repairs this instruction. The editor adds a reusable operation. For every condition stated in the source that could change the reader's action, compare what the reader may do before and after shortening.
A second editor tries the operation on a publication instruction. The source says that only the volume editor can authorize release, but the shortened draft lets any reviewer do so. Comparing the two permitted actions detects the loss. The operation already works for this stated authority condition. Its use here establishes transfer.
A later instruction says simply “approve the order”. The purchasing rule gives that power only to the budget holder and the source omits the rule. Comparing source wording with shortened wording preserves the same misleading reading. The next problem is to catch an operative condition supplied outside the source. The editor extends the Method to consult the rule governing the action when authority is implicit. In a further case about approving a refund, another editor locates the applicable rule and keeps its approval limit in the public instruction. Compare the resulting decisions and the cost of finding the governing rule to judge where the added operation earns its use.
A proved bound makes a new mathematical problem attainable
In C.40.CD:5.2, three sessions are necessary and sufficient for the conflict graph with edges A–B, B–C, C–D and A–C. The proof supplies an obstruction and a working schedule. Researchers can now ask which additional conflicts the three-session construction can absorb. Adding A–D invalidates the old schedule, but assigning A=1, B=2, C=3 and D=2 satisfies all five conflicts. Adding B–D as well makes all four events pairwise incompatible. Four sessions are then necessary and sufficient. The earlier claim still holds for its graph. What develops is the account of when the bound survives and why it fails. A decision to reserve a fourth session concerns a particular event and whoever has authority over it.
Bias-Annotation
A successful exercise can anchor the next undertaking, while a programme owner can invent novelty to retain work. Compare the acquired possibility, a useful ordinary continuation and a stop. An unfamiliar case is not automatically a new problem class; show what operation or understanding must grow.
Conformance Checklist
- The prior result and its qualification are stated.
- An attainable continuation is grounded in the changed starting position and a supported next contribution; a retained possibility keeps any unresolved condition explicit.
- Any selected continuation is worthwhile relative to its complete burden and retained obligations.
- Reuse, transfer and improvement of the developing means remain distinct.
- Independent carriers retain their own ends and consent.
- Changed criteria do not silently rewrite the earlier comparison.
- The return permits a justified stop and makes no claim of infinite future performance.
Common Anti-Patterns and How to Avoid Them
Consequences
The result of one undertaking becomes material for a more capable next construction while preserving adequate ordinary practice. This can expand knowledge, repertoire or the means of further development. It also exposes occasions where continuation is not presently worthwhile or attainable.
Rationale
A qualified result can put a previously inaccessible problem within reach. Those undertaking or authorizing the next work can choose it when the expected gain justifies the cost of an attempt. Tracing how a result changes the next obtaining way makes recursion examinable. Cultural continuation adds independently acting carriers without making a population one managed agent.
SoTA-Echoing
C.40.CD already permits a useful construction to change the next problem. C.18/C.19 preserve exploration and selection beyond a fixed objective; E.23 distinguishes improvement from improvement of its means. This Method develops the practical return from a qualified result to a newly attainable problem, including an inquiry whose result is knowledge. It considers the means and cost of pursuing it. If a current practice already supplies that return, use it without another cycle.
C.36/C.36.RP supply cultural reconstruction and renewal. The 2024 essays on role communities inside a company and around a company motivate attention to overlapping organization/community relations, participants' own journeys and competing formal and informal influence. They are conceptual contributions, not evidence for universal predictions about human replacement or office disappearance.
The integration-lead and editorial cases are constructed explanations of these connections, not measured effectiveness results. Growth, transfer, retained independence and comparative burden require evidence from their use. Scientific or computational open-endedness mechanisms remain with their research and subject Methods; this general construction does not supply an executable learning algorithm.
Relations
C.40.CD develops problem and way together. C.40.OD constructs the wider open-ended undertaking; C.40.JD joins way, holder and provision for a present change. C.40.OCN and C.40.OP develop and retain opportunities. E.23 addresses improvement of improvement means; C.18/C.19 govern relevant search and selection; C.36/C.36.RP support cultural reconstruction. Human-capability development, organizational change, community practice and the technical practice supply the subject-specific realization.
C.40.NP:End
Select Checklist Questions for a Work Result
Type: Method pattern Status: Stable
Use and combine the checklist Methods
Choose the missing contribution
Skilled work can fail through a small omission, a forgotten conversation or an answer that outlives its grounds. More documentation can make that worse when it hides the question, interrupts useful work or rewards reassuring entries. Checklist practice concerns the design and use of a limited attention aid within that wider work. It does not supply all professional knowledge, prove the usefulness of its described characteristics, or establish a whole result from constituent passes.
This family offers six Methods. Selection establishes which questions are worth carrying. Construction preserves those questions in a reusable template. Fitting arranges a usable occasion and access to truthful answers. Use obtains answers and informs action. Reconciliation connects local answers with the shared result. Qualification tests whether proposed criteria and checking means distinguish what matters.
Choose the Method whose inputs are available and whose result you need. Selection starts from a working problem and descriptions of the result; construction needs selected questions; fitting needs a template and a work arrangement. To use the checklist, interpret its questions for the subject and occasion. Reconcile local answers when people rely on them together; qualify a criterion or checking means when its grounds are uncertain. An existing adequate result can supply any of these inputs.
A direct skilled question, conversation or observation can finish a simple use without a maintained list. The two connected cases in C.41:5.1 and C.41.SR:5.1 show conditional combinations, not six compulsory phases.
Subjects, descriptions and useful characteristics
A team can describe a seminar set as twenty readable cards with wording v2, assembled for delivery. The cards have characteristics. The description says something about those characteristics. The team also assumes that a set with those characteristics will serve its exercise. Comparing cards with the description, testing that usefulness assumption, and deciding what to change are different activities.
A mismatch can reveal a defect in the cards, a mistaken description or an unsuitable intended use. It supplies information for that decision; it does not impose a duty to change the cards. Legislative requirements, when applicable, retain their source and applicability. They are not the general meaning of a checklist criterion.
A criterion here is the condition or distinction used to interpret an answer. “All twenty described cards are present” and “The wording matches v2” are examples. Their grounds and useful coverage remain open to examination. Some questions use direct observation; others consume an already competent result or a reasoned judgement. Neither a blank field nor a plausible answer supplies missing evidence.
This follows the distinction between a subject, an episteme about it and a selected use of that episteme in FPF E.10.D2. A template describes questions for reuse; a filled record describes answers from a use. People and technical systems can perform work. For descriptions, files and card sets, identify the relation: for example, a file describes the cards, the cards enter an exercise, and an observation supplies grounds for an answer.
Change, simultaneous work and the whole result
Checklist questions select conditions within structures of transformations and structures of states of their subjects. The questions need not form a ladder. Wording can change while a set remains assembled; handling a replacement can defeat an earlier count; two preparations can proceed at once. A table, diagram or graph may represent useful relations, but its layout does not determine the states or transformations.
During a workshop preparation, testing a projected example can contribute both to checking the display and to preparing the session. The whole matters at that moment: a technically sharp image of an obsolete example does not serve the exercise. The FPF B.1.5 Method relations and A.15.10 changed-premise relations help identify what the combined work consumes. A supplied projector enables this work; its availability alone does not establish a performed part of the session preparation.
A missing intermediate ability can stop the combination. Someone able to count cards may lack the judgement needed to select instructional examples. The next contribution is then help with the examples, not another count. Three people may also need the same room at incompatible times. Keeping a shared operating account helps make that conflict recoverable across participants. Selecting a feasible resource arrangement is additional coordination work; filling three lists cannot resolve it.
For a fuller account of change and subject states, return to FPF A.3.3, A.3.4 and A.22. These foundations help describe the structures used here; a constructive composition of transformations needs its own justified account. Context-state descriptions in E.18 concern the bindings and locality they describe, rather than every state of a checklist subject.
Choosing and combining the Methods
For a recurring omission, select a useful question set with C.41, express it with C.41.TF and arrange first use with C.41.OF. C.41.UA then produces answers when the checklist is used. Compare the contribution and burden, reusing an adequate comparison or using C.11.MV when that result is still needed. Use A.3.1.MI when introduction and later use must inform a change. The result can be a clearer question, a different occasion, new support, or retirement of the aid.
For shared work with uncertain criteria, use C.41.QC to challenge those criteria and checking means. C.41 retains the useful coverage; C.41.TF and C.41.OF give participants workable local views. C.41.UA supplies local answers. C.41.SR tests their combined use, locates gaps and reopens affected reliance when premises change. Several Methods can be active within the same larger performance; an earlier result is needed only where another action consumes it.
One centralized checklist can be sufficient when everyone has the relevant access and it adds little burden. Separate local lists can be sufficient when their uses are independent. Where people rely on one another, this family keeps the useful local arrangements and connects only the shared claims that affect work. The cost is maintaining those connections. Use the simpler arrangement when it gives the same practical result.
Judging use and learning from it
Across this family, ask whether the important work questions are covered, their meanings survive the chosen forms, people can answer at the right occasion, the answers have adequate grounds, shared conclusions follow from the contributions supplied, and uncertain criteria have been challenged. Inherit adequate answers from prior use; revisit the part changed by new evidence or conditions. Each pattern's Conformance Checklist makes its own contribution inspectable.
Recognition begins with the ordinary situations in §0.1 and the Problem frames. Assurance comes from examining the questions, observations, mappings, counterexamples and consequences in the later sections.
What does a filled checklist establish? Read the existing answers and their grounds before asking for another check. If they meet the acceptance conditions for this delivery, the recipient can act on them. A completed checklist can express a model of one delivery (L0). Its template may embody the team's model of a repeatable local check (L1). When the commission is to develop a disciplinary model of what to check (L2) using the transdisciplinary model of question selection here (L3), show which conditions of the discipline's work matter to the decision and how the chosen questions reveal them. A filled form can also carry that reasoning if its author records the design choice. When the assignment is to use an approved question set, the performer can take it as given. To credit the performer with applying the checklist, trace the answers to the checks they made or the prior results they chose to reuse. To credit them with selecting the questions, identify which choices the brief, form or helper supplied and which the performer made and justified. An earlier selection or a change the performer made to the form can supply that evidence.
The constructed examples here illustrate reasoning and possible use. They do not report measured effectiveness. Human-use studies and AI evaluation studies inform the Methods under the limits in C.41:11 and the supplying patterns' SoTA-Echoing sections. Practical adoption depends on capabilities, access, working conditions and the result people value.
The arrangement favors visible, discussable omissions. It can underrepresent tacit expertise, unusual events and concerns that people cannot safely voice. Keep open inquiry alongside the selected questions. Watch for time displaced from the work, repeated uninterpretable answers, prefilled records and harmful reliance on a narrow pass. Use those observations to change or remove the aid.
What the checklist account preserves and where to return
Checklist tables compress questions, answer meanings and their relation to the work. To interpret or adapt a table, recover the domain knowledge, subject states and work descriptions on which its questions depend. To explain its adoption or effects, examine the wider working conditions. Return to the criterion's domain source when a judgement exceeds the question set; to the named FPF patterns for stronger ontology or structure claims; to the supplying Core Method-development patterns or DPF for capability development, operations or administration.
The source line combines the supplied Methodology manuscript's checklist practice (R7, chapter 5 and its Method, qualification and change context) and Systems Management manuscript's coordination practice (R10, operational and administration passages), Gawande's historical account of professional checklist use, later human outcome and documentation studies, and current primary work on AI criteria and checking arrangements. Their useful contributions converge on situated questions with accessible grounds. Their differences survive: a human conversation, an executable test and a model judgement supply different observations and burdens. The shared language connects these contributions without treating one as a universal substitute for the others.
Problem frame
Use this when a recurring omission matters to a work result and you are choosing, shortening or revising the questions that could expose it. An organizer receives twenty cards, yet the exercise fails because one card uses an obsolete phrase. A developer demonstrates an import, yet no one asks what happens to existing data when parsing fails.
The Method selects a bounded question set for a stated use. Start with the result people are trying to obtain and one consequential omission. The first useful result is a small set of answerable questions, each connected to its subject, interpretation basis and practical contribution.
Use a direct question when that settles the immediate difficulty. Use domain inquiry when nobody yet knows what characteristics would serve the work. Selection can reveal that missing contribution, but cannot replace it.
Problem
A list copied from another setting can be both complete on paper and irrelevant in use. A list generated from a task description can omit an important consequence or inherit a mistaken assumption. Adding every imaginable question makes the aid harder to use and can obscure the few answers that change action.
The selection problem is to direct limited attention toward consequential differences while keeping the selected coverage and its basis visible. The object being changed is the question set; the subject of each question remains separately identifiable.
Forces
Solution
The criterion comes from the work and its subject, not from the checklist form. C.39.RQ can supply criteria when the unresolved question concerns a Method contribution; use that result where applicable without requiring every checklist question to become a Method requirement. The construction below selects the consequential questions, their answer basis and the coverage worth maintaining.
Recover the work and the consequential difference
Name the work result and the occasion at which an answer will matter. Identify the actual or prospective subjects: a physical set, a document edition, an import operation or the catalogue after it. Ask what was omitted, could plausibly be omitted, or remains difficult to notice. Include the encompassing use: why would this difference matter to the exercise, delivery or operation?
Use your own and others' past omissions to propose questions, keeping the conditions that made each omission consequential. Look for patterns across incidents or records where a single account hides the connection. Ask what observation could challenge the favored explanation or desired answer. Prior mistakes suggest questions; they do not enumerate every possible failure. Place a revealing question early enough to avoid further work that its answer could make unnecessary.
Follow the actual production and use relations far enough to expose a consequential omission. Who or what produces this result, what enables that producer, and who uses the result? At a relevant link, distinguish the producer or user, descriptions of it, its way of working and its actual work. A missing intermediate service, tool or capable performer may deserve attention even when the final subject looks satisfactory. Stop following links when further detail cannot change the selected questions. If the prospective subject does not yet exist, ask about the available materials, descriptions or enabling results without treating a description as the existing subject.
Recover the descriptions used to judge the result and the assumptions connecting their described characteristics to that use. Use C.39.RQ when the missing criterion concerns a required Method contribution, retaining the subject/description distinction in §0.2. When the criterion's basis is unavailable, return that question to competent domain work.
Propose questions and test their contribution
Write each candidate as a question with a recoverable subject and distinction. “Is the set ready?” may hide count, wording and assembly. Replace it with questions that can receive different answers where that difference changes the next move. Conversely, combine questions when splitting them creates effort without changing interpretation or action.
Choose how to obtain the questions from the uncertainty that remains. Direct practitioner selection can suffice for a familiar omission. Extract candidate questions from a usable description when the relevant properties are explicit, then ask whether those properties serve this use. Use generated proposals or comparisons between candidate results when they can expose a missing distinction. Compare the alternatives for the same needed coverage, including the effort of examining proposed criteria and removing duplicates. If the missing input is domain knowledge, another round of generation is useful only if it helps obtain that knowledge.
Comparing candidate results can reveal a missing question. Put a useful result beside a plausible but inadequate one and identify the consequential difference. Recover why it matters from the work and available domain knowledge; turn that difference into a candidate criterion. A generated contrast helps propose the question but does not establish that its preferred answer or coverage is sound.
For each candidate, establish:
- what it asks about, including an edition or time when relevant;
- which description, observation or judgement gives the answer meaning;
- why a different answer could change the work or a decision;
- who or what could supply sufficient grounds at the intended occasion.
These can be spoken answers in a small case. A maintained list benefits from nearby explanations where readers would otherwise guess. A criterion author can consult a specialist; the eventual user need not reproduce the author's entire reasoning.
A question can support prevention, preparation for a possible failure or recovery after it. For the workshop, confirming usable backup materials can improve the response to a display failure without making that failure less likely. Ask what makes the action suitable for these participants and circumstances: printed copies may be ordinary access for one participant and backup for another. Keep the intended contribution explicit. Available backup materials do not establish a successful response, and preparation does not remove the possible value of improving the underlying technical Method.
A canvas can prompt inquiry as well as confirm conditions. “Who will use this result, and why?” may initiate a conversation that develops an account; “Have we identified the intended users?” asks about that account. Select the prompt that serves this occasion, and distinguish obtaining an answer from judging its adequacy. Filled sections can expose missing connections without establishing that the described proposal will work. For a seminar proposal, naming participants, activities and available support does not establish that those participants can attend or that the activities help them learn. Keep the substantive assumptions and grounds available for discussion; use the appropriate domain practice to test them.
When the questions also determine rejection, ask whether any allowed satisfactory result could be rejected. A derivation may settle this for the whole intended use. Examine a distinguishing case where the criterion’s effect remains uncertain. Recover allowed alternative ways of meeting the terms and the effect of a failed necessary condition. A known failure within the proposed use that would pass every question already refutes the claim that passing suffices for acceptance. If the question is which difference the set can detect, compare a satisfactory result with a plausible failure. Would their answers lead the user to act differently where the work calls for it? Distinguishing one pair shows which difference the questions can expose. For reliance beyond that pair, trace the consequential conditions of the work to the questions and their grounds, or limit the proposed use to the conditions supported so far. Reuse an adequate qualification. An existing observation can answer a question for this delivery. Before dropping that question from the maintained list, check how later users will obtain the answer or why they can make their decision without it. Use C.41.QC where the criterion or checking means still needs examination.
Select coverage and a useful stopping point
Retain questions whose expected contribution warrants their interruption and interpretation cost. Remove decorative, redundant or routinely unanswerable items, unless naming the unresolved contribution is itself useful. Keep the wider Method description or domain instruction reachable rather than copying it into a long list.
State the useful coverage in ordinary language: “These questions concern the physical set's contents, wording and assembly.” State an uncovered question when it changes use: “They do not establish that the exercise teaches the intended distinction.” FPF B.1.5 helps distinguish the constituent result from the encompassing result; adding a whole-result question is worthwhile only when someone can supply its grounds.
When a reusable template is needed, pass the selected questions and their meaning to C.41.TF. Stop when further additions do not improve the intended use at worthwhile cost. Reopen selection after a missed consequential difference, a changed purpose or description, or repeated burden without benefit. Include a question about continued usefulness or feasibility when that premise can disappear during the work: an earlier decision to begin does not settle it for every later occasion. Obtain that judgement from the practice competent to supply it.
A discovered omission can suggest checking analogous subjects or implementations. Carry over the question and relevant source passage, then establish its applicability, observation route and answer for each subject. Keep assumptions about use, interfaces and dependencies visible. Agreement among several implementations can preserve the same mistaken interpretation; return to the substantive description and its grounds where that possibility matters. Neither an earlier failure nor an earlier pass transfers merely because the subjects look alike.
Archetypal Grounding
A facilitator describes set S as twenty readable cards with wording v2, assembled for delivery. She assumes those cards will support a comparison exercise. The organizer's draft asks only, “Are the cards ready?”
The facilitator retains a separate exercise trial because a complete, readable and accurately worded set can still confuse participants. That trial uses domain judgement about the exercise. Selection has produced three useful physical-set questions and a return for the encompassing question.
In a catalogue importer, “Did the file import?” does not establish identifier preservation, duplicate reporting or preservation of the previous catalogue on parse failure. The team recovers those three described properties and selects questions about each. Input A17 becoming stored identifier 1 demonstrates why generic success is inadequate. Whether the three properties are enough for the team's use remains a further selection question.
Suppose the cards are produced on a shared printer. A complete file does not establish that the printing service can produce readable cards at the needed occasion. The selection follows one extra link: the printer's configuration and the person able to use it. If the print setup itself is being changed, questions about that enabling work become useful too. The design file describes the proposed cards; materials enter the print run, which produces the physical set. Questions about the setup, its description, the printing Method and the actual run concern different subjects. Preserve those relations when connecting their answers to the card preparation. If those enabling results are already competently supplied, the organizer can use them instead of designing a second printer-maintenance checklist.
For a community running the seminar, distinguish the organizer's recruitment and support work, members' learning and practice, and their joint use of the exercise. Sending invitations does not establish attendance; attendance does not establish a member's ability to use the distinction; individual ability does not establish that members coordinate their contributions. A useful selection might therefore ask what support was provided, what a member can do, and what happens in the joint exercise. Choose only questions that matter to this community's work. One person may both organize and participate, but an answer about her organizing work does not answer the question about her learning. Return the capability question to the applicable human-capability diagnosis Method and an organizational participation difficulty to the applicable organizational response Method.
These cases are constructed. The gain illustrated is a more discriminating question set, not an empirical claim about fewer errors.
From a recurring omission to a useful aid
An organizer repeatedly receives physically complete seminar cards containing wording from an earlier exercise. The facilitator can explain the exercise, and the packer can count and inspect cards; neither currently sees the whole question at the useful moment.
Begin with C.41 if useful questions are missing. The first result is a question set tied to the work, or a clear statement of the domain judgement still needed. With adequate questions already available, start at the contribution that is missing. The mantra below follows the aid through use and appraisal; a sufficient direct question can end the search immediately.
Mantra — from omission to a worthwhile aid: Name the omission and the work result it threatens. Reuse adequate questions, forms and checking means; obtain only what is missing. Tie each question to its subject, description and intended use. Make an answer possible at the useful moment, challenge uncertain criteria or means, and answer from adequate grounds. Keep the answer, the decision and the work distinct. Connect answers that affect the whole. After change or interruption, recover the question and its grounds; reconsider only affected reliance. Compare the aid's contribution with its full burden. Improve what obstructs useful work, or retire the aid when a simpler arrangement suffices.
The worked case below explains those conditional moves. Resume at the unresolved contribution after interruption; if the questions have been answered but the aid's worth remains unknown, continue with Worth and improvement and C.11.MV.
Selection. Here the facilitator supplied the description, and the three questions concern contents and readability, wording and assembly. With grounded answers the packer can settle whether this set meets that description for delivery. To commission the packer to design checks for future sets, give the packer the work and its acceptance conditions and examine which omissions the packer identified, which questions the packer selected and why. An earlier design with those choices can serve. The facilitator's separate trial addresses what the set enables participants to do in the exercise.
Construction. C.41.TF puts the set and description reference once above the questions. The packing view carries count/readability and assembly; the facilitator's view carries wording comparison. Both retain S and v2. The answer forms allow a supported negative or an unresolved question. The transposed example in C.41.TF shows the same answers without changing their subject.
Fitting. C.41.OF puts the questions in the preparation workspace, with the description accessible before the box closes. The packer reports physical conditions; the facilitator compares wording. An assisted first use reveals an ambiguous readability condition, which returns to criterion construction and help.
Use and change. C.41.UA supplies answers from observations at 10:00, 10:05 and 10:10, entered at 10:12. At 10:15 v3 changes one phrase used on the cards. Known unchanged v2 cards already mismatch it. If current cards may differ, current wording is instead unknown. The facilitator decides to keep v3 and replace the affected card, then obtains the changed comparison and any physical checks affected by handling. The old v2 answer remains historical.
Worth and improvement. A representative trial through A.3.1.WT observes whether the aid catches the wording problem, whether users can interpret the questions, and how much time or duplicated work it adds. C.11.MV compares that contribution with burden across the preparation work. A timely wording check can avoid reopening, replacing and repacking the set; completing the checklist faster is no gain if it misses that problem or merely moves the delay elsewhere. If a conversation at packing already prevents the omission with less cost, retain the conversation and retire the redundant aid. If the aid is useful but its description is hard to find, A.3.1.MI selects that focused improvement rather than enlarging the list.
Keep use, detected problems and work outcomes distinct. A caught wording mismatch is an observed contribution; it does not establish that the exercise would otherwise have failed. Another participant might have caught it later. Conversely, a session that went well does not establish that every check was sound. When appraising the change, account for the training, support, other changes and case differences accompanying the aid. Staff enthusiasm or a rising count of ticks does not isolate its effect.
The worked route uses C.41, C.41.TF, C.41.OF and C.41.UA with the trial, worth appraisal and focused improvement described above; uncertain criteria or shared reliance bring in C.41.QC or C.41.SR when needed. It can return at several points: unsupported usefulness calls for domain inquiry; lost references call for template repair; inaccessible information calls for fitting; unchanged adequate grounds support economical reuse. The three physical-set answers never settle the exercise's usefulness by themselves.
Bias-Annotation
Selection reflects the experience of those asked. Frequent, visible failures can crowd out rare consequential ones, and the person creating the list may never see the costs borne by its user. Include relevant practitioners and people consuming the result. Ask what their work reveals that the proposed questions miss.
A generated list reflects its prompt, examples and model as well as the task. Treat its fluency as help in proposing questions, not grounds for coverage.
Conformance Checklist
To examine this Method's use, inspect the selected questions:
- Can a user identify the work result, each subject and the intended occasion?
- Can the criterion's basis and the assumption about usefulness be distinguished?
- Does each question contribute a consequential distinction, with adequate grounds obtainable or the missing contribution explicit?
- What do the answers establish for the work entrusted to this checklist? If they are to decide acceptance or rejection of the whole, do the grounds cover the required connections and allowed alternatives?
- Is the bounded coverage usable at its cost, with a concrete reason to reopen selection?
An answer can be inherited from a still-applicable prior selection.
Common Anti-Patterns and How to Avoid Them
Consequences
The selected set gives template authors and users a shared account of what answers mean. It also makes unsupported criteria and uncovered work visible early. Selection costs domain attention and can expose a need for inquiry before the aid is usable. Its result remains revisable; it does not close open-ended diagnosis.
Architectural Rationale
Selection is independently useful because a well-presented, faithfully completed list can ask the wrong questions. Construction cannot repair absent coverage merely by adding fields, and checking cannot validate a criterion merely by applying it.
Use direct skilled attention without a maintained list for an isolated simple question or where the list's cost exceeds its contribution. A recurring, consequential omission justifies a retained question only when the aid improves that use.
SoTA-Echoing
How should questions be selected when expert attention, description-based extraction and generated refinement are all available? Adapt situated selection from the supplied Methodology manuscript (R7, chapter 5 and its Method/qualification context) and Systems Management manuscript (R10, operational and coordination passages) together with the proposal-and-challenge line in AutoChecklist: use the least costly way to expose the consequential difference, then judge the proposed question against the work and its grounds. The selected answer is conditional. An explicit, usable description favors extraction; a familiar local omission can favor direct expert selection; uncertain coverage can justify generated contrasts and refinement.
These are serious alternatives with different costs. Extraction can preserve a description's blind spot; direct expertise can miss an unfamiliar case; generation can add duplicates or plausible but unsupported criteria. SPECA supplies a specification-derived audit alternative and counterevidence about mistaken scope assumptions. Its single case and correct-specification premise do not establish general coverage. AutoChecklist supplies evaluated generation/refinement methods in its stated tasks, not evidence that generation wins for this workshop or every domain. The supplied Methodology manuscript (R7, chapter 5 and its Method/qualification context) and Systems Management manuscript (R10, operational and coordination passages) supply the practice connection to subjects and work; they do not compare model performance.
Section 4.2 therefore selects the proposal method by the missing distinction and includes adjudication effort. For the three explicit importer properties in §5, extraction plus a practitioner's usefulness challenge supplies the stated questions without a generation loop. If a plausible imported result satisfies all three yet defeats the intended use, comparing that result can expose an additional question. Reject automatic iteration merely because more criteria can be produced. Accept the added examination cost when it can reveal such an unresolved consequential difference; retain direct selection when it already provides adequate coverage.
The sources do not rank these alternatives at equal effort across domains. Where that ranking changes a substantial investment, use a representative local comparison through A.3.1.WT and C.11.MV. Reopen the local choice when a missed difference defeats coverage, the description's usefulness is in doubt, or another proposal method produces needed distinctions at lower total effort.
Relations
C.41.TF consumes the selected questions and their interpretation basis. C.41.QC can return a counterexample or qualification that changes selection. C.41.UA can reveal an unanswerable or irrelevant question during use; that finding returns here. C.41.SR can reveal a whole-result gap invisible in local selections.
C.39.RQ supplies construction of criteria from subject and situation; domain practice supplies their substantive grounds. FPF E.10.D2 governs the subject/description distinction and B.1.5 the Method relations behind constituent and encompassing questions.
C.41:End
Construct a Checklist Template That Preserves Its Questions
Type: Method pattern Status: Stable
Problem frame
Use this when selected questions are becoming a reusable form, or an existing checklist is being transposed, split, translated or moved into another tool. A row of green cells is easy to scan, but its user can no longer tell which set, description edition or observation each cell concerns.
This Method constructs a checklist template: a reusable description of the questions, their interpretation and how answers can be expressed. A filled instance records answers for an identified occasion. Start by carrying one selected question from its wording to a plausible filled answer.
The result is a form that preserves the selected distinctions in its intended use. If the difficulty is missing or unjustified questions, return to C.41 or C.41.QC. If the existing representation already works, retain it.
Problem
A heading can survive a redesign while its meaning disappears. A copied checkmark can lose the subject it refers to; a date can become ambiguous between observation and entry; a binary checkbox can force an unresolved question into a false answer.
The construction problem is to make truthful interpretation easy without turning the aid into a burdensome evidence repository. The template's layout serves the question set. It does not define the kinds or states of the things being checked.
Forces
Solution
C.37 selects a representation for one receiving use. Here the construction binds each selected question, subject, answer meaning and ground so that the template can be filled, transposed or shared without changing the claim. Use an adequate existing representation; a different layout is useful only when it repairs the receiving difficulty.
Carry the question and its interpretation together
Take the selected questions and coverage from C.41 or another adequate source. Identify information shared across the list, such as the physical set and comparison description. Place it where users can see its scope. Keep question-specific conditions with the question.
Phrase the questions in the language practitioners use for their work: the card set, print run or catalogue import. If adaptation changes what counts as satisfactory, reconsider the criterion before relying on equivalent answers from the forms. Adapt the selected subjects, conditions and questions as well as their names. Replacing “system-of-interest” with “cards” is insufficient if “ready” still hides an unfamiliar or unsuitable account of readiness. Compare the wording with a real case and the recipient's way of working. An applicable explanation-design Method helps recover the explanation needed for that use. Framework distinctions can guide the author's reasoning without requiring every recipient to learn their terminology.
Develop ambiguous headings with someone who understands the work. For a printing checklist, distinguish the person who wrote the comparison description, the system assigned to checking, and the system that performed this check. One person may occupy all three positions, but a name under “Responsible” does not tell the recipient which relation is asserted. A system's role kind and assignment, its capability and its performed work remain distinct; an applicable organizational Method helps when the assignment or enabling arrangement itself needs repair. For a file or description, name the relation, such as input to printing or basis of comparison. Use the resulting practitioner wording in the form and retain further explanation only where it helps interpretation.
For each answer, make recoverable what is being asked, what counts as an answer and what grounds could support it. An instruction may name the observation; a more complex judgement may point to a domain procedure. A.3.2.DE supplies the broader description of how to carry out the Method. A short checklist can refer to that description without trying to replace it.
Distinguish a reusable instruction from a filled claim. “Count the cards against the contents description” is part of the template's help. “Twenty cards present in S, observed at 10:00” is an instance answer and its basis. FPF A.3.2 governs a MethodDescription claim when the template has an already admitted Method as its EntityOfConcern and makes at least one substantive claim about that Method as a way of doing.
If a form opens with an answer already entered, make its source recoverable where the user needs to act on it. A suggested value still needs grounds. An observation from an earlier occasion can supply the answer when its subject, criterion and conditions still apply. A record of a new check identifies what was actually examined. These distinctions let the recipient use adequate prior work and identify any check that remains to be made.
Give answers enough expressive range
Use answer forms that preserve consequential distinctions. In the examples here, Yes asserts the question's condition on the stated basis; No denies it on such a basis; unknown says the available grounds do not settle it; not checked says no check was performed for this occasion. These are local answer meanings, not a universal state taxonomy. An old adequate result may be reused even when no fresh check occurred.
Use not applicable only with a reason why the question does not concern this subject or use. Missing access is a reason for an unresolved answer, not for inapplicability. Some questions call for a quantity, identified alternative or short account rather than Yes/No. Choose that form when it carries the distinction more faithfully.
When a label such as “prepared” abbreviates several conditions, recover which conditions hold together, which are alternatives, and which depend on another subject. A later label in a displayed list does not establish every earlier condition. A supplied plan can answer a question about planning while leaving resource availability or actual performance unanswered. Keep those conditions explicit enough for the intended action; an ordinary explanation can suffice.
Where the aid supports planning, distinguish an expected occasion for a condition from when it holds, is observed, and is recorded. Preserve the scope of a question that continues across several editions or attempts, as well as questions confined to one edition. Repeated attempts and alternative paths do not form a universal ladder in which every answer stays positive.
Add observation time, description edition, performer or evidence reference when the receiving use depends on it. Keep entry time distinct from observation time. Avoid fields with no practical consumer. A three-question aid used immediately can be spoken; a record reused tomorrow usually needs enough context to recover its grounds.
Transform the form and compare the meaning
Choose paper, editable text, a table or a working application according to the occasion. Transpose or split the presentation only while preserving the associations needed to interpret an answer: subject, question, comparison basis, answer and grounds. Keep shared reference information visible or directly reachable in each local view.
Retain one form when it serves the receiving uses. Make another view when it will resolve an access or interpretation difficulty. To hand over one record, compare the answers and their grounds as they reach its recipient. To approve a view for future records, establish how the form keeps each answer and its grounds with the right question, subject and comparison basis across the entries it must support. If two fields could be confused, try values that would call for different actions if exchanged. Following the field definitions through to the displayed questions may already establish how entries are carried over. Include preparation, looking up references and updating copies in the cost of maintaining a second view. Generating the views from shared content can be worth the effort when it prevents enough copying and update errors.
For a maintained aid, make clear which question set and template edition are in use. A corrected wording question in the working application can coexist unnoticed with an obsolete printed form or presentation. Derive needed views from the same maintained content where practical; otherwise identify how affected copies are updated or withdrawn. Preserve the earlier question and comparison basis when historical answers still refer to them. The useful result is a recoverable current form and interpretable earlier records; a separate modeling application is worthwhile only if it helps sustain that result.
Inspect what entering an answer does in the chosen tool. A mark may record a comparison, but the application may also use it to release a file or notify another participant. Make that additional effect apparent and preserve the distinct decision behind it. A truthful “matches v3” answer does not, by its meaning alone, select distribution of v3. Use separate controls when the two moves need separate decisions.
An outline that puts each item under one heading can hide another relevant relationship. One comparison of the projected example with the cards may contribute both to display preparation and to exercise preparation. Preserve both uses when they affect the next action; use a cross-reference or a small relation table if one parent heading cannot express them. Do not infer that the subject has only one such relation from the form's single indentation.
Where interpretation of the answer forms is in doubt, replay the different answers that matter to the next action, including a supported negative or an unresolved answer where users need them. If the intended recipient's interpretation remains uncertain, ask someone who will use the form in those conditions to recover the question and decide what to do with the answer. Use that response to locate the missing explanation. Compare the repair's benefit with preparation, answering and maintenance effort. Return a missing answer distinction to the template, a question that asks about the wrong condition to selection, and an inaccessible observation to C.41.OF.
When transfer changes the speaker or recipient, recover relative references through FPF A.6.P.RI: “my cards” or “our current description” can acquire a different referent in a copied view. Check the question's associations after the change even when its words are unchanged.
Archetypal Grounding
The selected card questions become this reusable template:
A constructed instance supplies the references and grounds:
Suppose the wide table on a small screen separates the answer from its S/v2 reference. Transpose the same information so the recipient can recover both together:
If entries are recorded at 10:12, that time describes the recording. It neither changes the observation times nor shows when each condition first became true. A split view for counting and a view for wording comparison retain the S/v2 references and the relevant observation grounds.
The explanations and times help the recipient check the three associations in this record. All three verdicts happen to be Yes. A display that copies the wording verdict into the assembly row would give the same verdicts here. Keep the described contents and wording, but leave the set unpacked. The assembly row must now show No with the observation of the unpacked set. The wording row can still show Yes from the comparison with v2. The two answers show that the wording matches v2 and the set is unpacked. If an assembled set is still due, the packer assembles it. If the recipient asked only for its present state, the packer can return the two answers with their grounds. It tests this particular copying error. Other fields and answer meanings need grounds of their own when approving the view for wider use.
For an AI-assisted importer, a template can ask for an observed catalogue before and after a malformed input, with a reference to the input and the comparison performed. A field headed “Tests pass” loses the question if it is detached from those cases. When an observation tool fails, the form supports “unknown: post-import catalogue was not obtained”; it does not coerce the agent into a positive completion entry.
Bias-Annotation
A form convenient for its author may be difficult on a phone, inaccessible to a participant or opaque to an AI tool. Colour can supplement but cannot carry the only answer distinction. Specialist abbreviations and apparently obvious dates deserve testing with intended users.
Recorded evidence privileges what can be captured. Allow a relevant conversation or observation to contribute. An empty attachment field does not invalidate that contribution.
Conformance Checklist
For a particular handover, inspect what the recipient can recover from that record. For a reusable form, follow how its fields and transfers carry the necessary distinctions over the allowed inputs. A filled case can resolve an uncertainty about the mapping or about how a recipient reads it.
- Can the intended recipient recover the subject, question, comparison basis, answer and sufficient grounds?
- Can the form represent a negative answer and an unresolved one without conflating them?
- Is any inapplicability reason about the question's scope rather than missing evidence?
- Do shared references and answer meanings survive the actual transpose, split or tool transfer?
- Does every retained field help the intended use, and is nearby help enough for the expected user?
Common Anti-Patterns and How to Avoid Them
Consequences
Users can adapt the form without losing the differences that matter. Shared information can be stated once instead of repeated in every cell. The cost is maintaining its scope and links through change. More elaborate evidence fields can improve later reuse while slowing immediate work; retain them only where that trade-off is worthwhile.
Architectural Rationale
Template construction is separate from question selection because a sound question set can be damaged by representation. It is separate from fitting because a clear form can still appear at the wrong time or in an unused tool.
Editable text and tables are useful when people adapt, transpose or split the aid. A diagram can be useful when it preserves a conditional relation better. The choice follows the receiving action and demonstrated preservation, not a universal preference for one medium.
SoTA-Echoing
When is a local form worth maintaining alongside a shared checklist? Adapt the supplied Methodology manuscript's template/instance and transposed-view practice together with the supplied Systems Management manuscript's local interpretation: retain shared question meanings and provide a different view only where a receiving action needs it. The alternatives are one fixed master form and independently adapted local copies. The master is preferable when it already permits the needed interpretation; independent copies can be cheaper for a small one-off use but introduce a continuing update problem when reused.
Section 4.3 compares these choices with the same filled cases and includes preparation, reference recovery and later updates. In §5, the small-screen condition justifies transposition because the subject/basis and answer can be read together while the three questions, answers and grounds remain unchanged. It does not justify another checklist meaning. If the wide form already shows those associations adequately, keep it. If several views change frequently, accept the setup cost of deriving them from shared content when it avoids more costly inconsistent updates; a single occasional copy need not acquire that machinery.
The supplied Methodology and Systems Management manuscripts supply the competing arrangements and preservation question, not comparative reading-time measurements. The study by Facey and colleagues supplies failure evidence about documentary completion and actual use. It supports preserving unknown, observation and record meanings in §§4.1–4.2; it does not show that a layout is superior. Reject uniform appearance as the selection criterion when it hides a needed distinction, and reject richer fields whose later usefulness does not warrant their entry cost.
The conditional choice improves access at the accepted cost of maintaining another expression of the same questions. Where neither interpretation nor total effort can be judged from the filled cases, obtain a bounded comparison of the recovered question, next move and total effort in representative use. Reopen the choice when a recipient loses a reference, an answer meaning diverges after an update, or the extra view costs more than the difficulty it removes.
Relations
C.41 supplies questions and bounded coverage. C.41.OF consumes the usable template to arrange an occasion. C.41.UA supplies real answers, and C.41.SR consumes local views without assuming that identical layouts imply identical claims.
A.3.2.DE supplies the broader Method description; the applicable explanation-design Methods can recover what the recipient needs explained and compare explanations for that use when those separate questions are current. FPF A.3.2 governs a MethodDescription claim; A.6.P.RI supplies recovery of references whose meaning depends on the participant or occasion.
C.41.TF:End
Fit a Checklist to Its Occasion of Use
Type: Method pattern Status: Stable
Problem frame
Use this when a meaningful checklist is skipped, completed too early, used without the right participants, or stranded in software that the work does not otherwise use. People and AI agents can know the questions yet lack a workable occasion to answer them.
This Method arranges the interaction through which the aid contributes to work. Start by following one real or representative occasion: who encounters the question, at what moment, with which observation and means to respond? The first result is a usable arrangement or an account of the missing capability, resource or decision.
If the person can answer the question with existing support, retain that arrangement. A new training programme or software workflow is not the default response to an unused list.
Problem
Possession of a checklist does not supply time, attention, understanding, observation access or a reason to use it. A form can be visibly complete while the conversation or check it was intended to prompt never occurred. Requiring another completion report can strengthen that substitution.
The practical problem is to connect a worthwhile aid to a moment when participants can use it truthfully and act on its results.
Forces
Solution
A.15.8.MS supplies methodological material and support across tasks; A.3.1.MI supplies a complete bounded introduction when that is the question. This Method constructs the checking occasion: when the question can change action, who can answer it, which observations and means are available, and how ordinary use will expose the remaining obstruction.
Locate an occasion that can change the work
Trace the work around a consequential question. Before which action is the answer useful? When is the subject observable? Is there a natural pause, handover or decision point? Who can supply the answer, and who can use it? FPF A.15.11 helps make an applicable Method noticeable at a work occasion; ask which cue can lead to a useful attainable action.
Choose an interaction mode for the question. In read-do, the aid prompts an action and its check during performance. In do-confirm, a capable performer first does familiar work and then confirms selected conditions. A mixed arrangement can fit different questions.
Identify work that must remain viable while the aid is used. Arrange attention, timing and capable contributions so reading or recording does not defeat that work. During a live exercise, one facilitator might sustain participants' activity while another person investigates a display fault; otherwise pausing the exercise may be the useful choice. Include interruption and a short time window in the trial when those conditions matter.
Name the initiating cue and the response to a missing participant or inaccessible observation. An aid that appears before its subject exists can invite speculation; one that appears after the only useful decision can become documentation alone.
Make a truthful answer possible
Put the question, help and response mechanism in the working environment. Agree who asks, who answers and how a negative or unresolved answer reaches someone able to decide. Where everyone already has access, a short spoken exchange can be enough.
Recover the capability needed to answer. Counting supplied items may be easy; interpreting a specialist defect can call for another person. The applicable capability-diagnosis Method helps identify the capability demanded by the work. Nearby explanation can help a capable user apply a distinction; it does not create all missing competence.
For human participants, inspect whether time pressure, hierarchy, mistrust or incentives discourage a truthful response. Explain the practical purpose, involve people who bear the burden, and change the arrangement that prevents use. Where introduction changes established organizational practice, the applicable organizational Methods supply assignment, enabling-relation and participation-condition work beyond this fitting Method.
When the proposed change needs examination and agreement, recover the proposal with those who supply and receive the contributions, let them obtain relevant advice, and establish the decision or voluntary agreement needed for this occasion. Return the accepted contributions and their conditions to the work: who supplies the observation, who has time to check, what help is available and who receives an unresolved answer. Confirming those arrangements can support first use. When their contribution to a useful handover remains uncertain, use matching observations or the trial below to settle that question. A new consequential objection or an unavailable resource returns to the affected decision before dependent use.
For an AI agent, provide the applicable description and criteria, usable tools, relevant retained observations and a way to express uncertainty. Mixed human/AI work needs a clear handover of unresolved judgement.
Try the arrangement and repair the obstruction
Start from the fitting question and reuse observations that already answer it under the relevant user and work conditions. Use A.15.8.MS for an unresolved support arrangement.
When a new observation of representative work is needed, use A.3.1.WT. Observe whether the user encounters the aid, interprets the question, obtains the grounds and chooses a useful continuation. In shared use, observe who contributes, whether another participant understands and responds, and whether a raised concern changes the work where it matters. Examine the intended interaction and the aid's contribution separately: opening the form, reading its items and obtaining a useful conversation are different observations. Include a negative or unresolved case when the response to it can change the intended reliance.
Choose observations from the unresolved question. An existing reliable record may settle whether the aid appeared at the intended occasion. That appearance alone does not settle whether a participant understood or acted on a concern; use evidence of that interaction when it matters. If the value of continuing oversight remains open, seek an existing or attainable observation that could change the decision. Where supervision is required, establish its availability at the occasion the work needs it. Include the observer's effort and any change their presence makes to the work.
During an assisted use, establish who interpreted the question, obtained the observation and chose the response. The result may be sufficient for this handoff with the available help. If someone supplied the interpretation, credit the user with the work they actually did. When deciding what the user can do on other occasions, identify the support that will be available then and the decisions the user will have to make. Examine the effect on the work result and the effort moved to other participants or tools.
If ordinary use is still in doubt, look for an existing performance by the intended user with the applicable form, support and work conditions. Observe another occasion when the available evidence leaves a consequential question unanswered. Compare when the aid influenced the work with the observations and entries made. A late record can faithfully report an earlier observation. For entries made without the required check, repair the obstacle, such as missing access or incentives that discourage a truthful answer. Choose continuing oversight only where it serves the work.
Return a meaning defect to C.41.TF, a coverage defect to C.41, an inadequate check to C.41.QC, and a missing capability or resource to its supplying practice. When the fitting decision still needs a worth comparison, use C.11.MV. Use A.3.1.MI when a decision to change or retire the aid depends on its introduction and later use. Stop when the arrangement works at worthwhile cost; do not keep adding ceremony merely because some use remains irregular.
Archetypal Grounding
The seminar organizer originally emails a spreadsheet a day before delivery. The person packing the cards sees it only after the box is closed. She marks “ready” from memory. The current description is in a separate document that she cannot open.
The repaired arrangement puts the three questions and the v2 description in the preparation workspace. The packer counts and inspects during assembly; the facilitator compares wording where instructional interpretation is needed. Before closing the box, the organizer confirms the set and unresolved questions.
Another reminder email would leave the missing access and late occasion unchanged. The organizer joins one initial use to inspect access and interpretation. To decide whether the new arrangement can be used routinely, the organizer can draw on a comparable ordinary use under the revised conditions. If the available evidence covers only the assisted use and the independent decision still matters, observe that decision at an ordinary packing occasion. Continued help or a second observer is worth arranging when it addresses a difficulty still present.
On first use, “readable” turns out to be interpreted differently by two people. They inspect the cards under the intended display conditions and ask the facilitator what the exercise needs. Repair that interpretation in the criterion and nearby help before asking the packer to answer again.
In the importer case, an agent is given criteria but no access to the stored catalogue after an error. Adding “verify preservation” to its prompt cannot produce that observation. The team supplies a bounded test environment and a tool that returns the before/after catalogue. If that tool fails, the agent records the unresolved question and returns it. A session restart restores the criteria and relevant evidence references in the agent's working context.
These are constructed arrangements. In practice, judge their expected benefit and full burden on available grounds; use a representative trial when its observations can change that decision.
Bias-Annotation
The person introducing the list can mistake resistance for indifference. Non-use may be a rational response to poor timing, inaccessible information or repeated low-value work. Conversely, convenient local use can conceal burdens moved to a coordinator.
When a trial is needed, include the people and tools that will bear those costs. A successful assisted demonstration does not establish routine independent use or broad cultural adoption.
Conformance Checklist
Inspect the arrangement at an actual or representative occasion:
- Can the user encounter the current questions when their answers still matter?
- Are the needed observation, interpretation, participants and response means available?
- Can a negative or unresolved answer be expressed and used without substitution by a reassuring entry?
- Does the fitting judgement use adequate grounds for interpretation, interruption and moved burdens, with any assistance explicit?
- Is the chosen repair directed at the observed obstruction, with a simpler sufficient arrangement retained where possible?
Common Anti-Patterns and How to Avoid Them
Consequences
The aid becomes part of a usable work arrangement. It can prompt a consequential exchange and preserve honest uncertainty. The costs include observation access, help, coordination and tool maintenance. A successful local introduction establishes neither universal effectiveness nor cultural spread; broader change needs its own evidence and Methods.
Architectural Rationale
Fitting remains distinct from constructing a good form because timing and access can defeat clear questions. It consumes the existing aid and changes its occasion, participants and support.
The alternative of uniform mandatory completion may simplify reporting, but it can leave the underlying obstruction unchanged. The selected Method diagnoses that obstruction and tests a proportionate arrangement. Skilled interaction remains preferable when a retained aid adds no useful support.
SoTA-Echoing
Which support and observation arrangement makes truthful checking attainable at worthwhile cost? Adapt situated design, assisted trial and subsequent ordinary use, drawing on Atul Gawande's historical pause-and-redesign cases in The Checklist Manifesto (2009) and Anthropic's long-running-agent arrangements. The alternatives are an additional reminder using existing support, integration into the working environment, and continued assisted or directly observed use. Choose by the missing input: a cue for an unnoticed but attainable check, access or tooling for an unobtainable observation, and assistance for an interpretation or capability difficulty.
Sections 4.2–4.3 and the packer case illustrate that choice. Another email leaves the inaccessible v2 description and late occasion unchanged; making the description available during assembly removes those obstacles. A bounded joined first use can expose the remaining interpretation problem. Continued observation is worth its greater cost only when the unresolved use question or reliance warrants it. Reject using an attendance or completion record to answer a question about understanding, speaking up or acting on a concern.
The study by Facey and colleagues supplies counterevidence about documentation versus observed use. CheckPOINT supplies an alternative for observing team behavior; Ridgeway and colleagues distinguish conversation-aid implementation from the practice it supports. Their instruments help select an observation that answers the question; they do not warrant observing every use or transferring a generic engagement score. Urbach and colleagues found no significant mortality or complication reduction in their population-level before/after study, so introduction and completed forms cannot carry the outcome claim alone. Anthropic's accounts are bounded engineering demonstrations, not comparative human/AI effectiveness evidence.
This choice accepts the cost of integration or assistance only for an obstruction a cheaper cue leaves in place. The sources do not rank all arrangements at equal effort. Section 4.3 therefore uses matching observations or needed comparisons of representative and later ordinary use, including observation cost and burdens moved elsewhere. Retain the existing arrangement when it already supports the needed action. Reopen the choice when access, capability, tools or incentives change, or when ordinary use fails despite successful assisted use.
Relations
C.41.TF supplies a usable template; C.41.UA enacts it at the arranged occasion. C.41.SR identifies shared contributions whose coordination affects fitting. A.15.8.MS, A.3.1.WT, C.11.MV and A.3.1.MI supply enactment preparation, representative trial, worth appraisal and improvement.
A human-capability diagnosis Method is available when that separate result is needed. Separately selected organizational introduction can use the applicable Methods for assignments, enabling relations and participation conditions. FPF A.15.11 supplies noticing an applicable Method in the working situation.
C.41.OF:End
Use a Checklist to Answer Questions and Act on the Results
Type: Method pattern Status: Stable
Problem frame
Use this when a checklist is being used on an occasion, or an earlier answer is being considered for reuse. The list may be familiar, but the cards, description, room or software state can have changed.
This Method obtains answers with sufficient grounds for their intended use and informs a separate next action. Start by identifying the subject and the question being answered. The useful result can be Yes, No, an unresolved answer, an applicable existing result, or a missing contribution. A positive completion record is only one possible outcome.
Use C.41 or C.41.QC when the questions or checking means lack a sufficient basis. Use the work's ordinary decision or authorization arrangement when a consequential next action calls for it.
Problem
People can confuse doing the work, checking its result, recording an answer and authorizing a continuation. An agent can report that it implemented error handling while the catalogue was never inspected. An organizer can reuse yesterday's green summary against today's changed description.
The problem is to answer the present question on sufficient grounds, preserve what remains useful and avoid inventing certainty to finish the form.
Forces
Solution
A.15.10 supplies changed-premise recovery. The operation here uses that boundary to answer the selected checklist questions from adequate present or retained grounds, then distinguishes the answer from the decision and any corrective work. Reuse a supplied current answer; reopen only the reliance whose relevant premise changed.
Bind the question to this subject and occasion
Identify what is being checked, which description or criterion applies, and the relevant time or condition. If every admissible reading warrants the same action on sufficient grounds, give that answer. Resolve any difference that still affects the work, including an order to identify the exact subject. A file name can refer to a different edition; “the cards” can refer to another box.
Recover prior answers or competent results that might already settle the question. Administrative claim-checking Methods supply useful distinctions; apply the competent examination for the present domain. Record-maintenance Methods supply record-use distinctions when the answer is retained for later reliance. Decide whether the subject, criterion, observation and conditions supporting those results still apply. Current reliance can survive a new occasion without a new observation when nothing relevant has changed and the grounds are sufficient.
After an interruption, recover what the checks were to support and whether that work is still due. Resume the first unfinished operation that can change the remaining work, with the content already established. A marker at card seven helps only if the next person knows which cards were compared, with which description, and what remains unresolved. Reconsider any changed subject or basis before resuming. Retain only the context needed for that continuation; a position marker is neither a completed answer nor proof that earlier grounds still apply.
Use a position cue alone when the performer can already recover the question, basis and unfinished work unambiguously. Otherwise retain the missing content alongside it. Compare the effort of recovering adequate prior grounds with obtaining a fresh observation. Observe again when the earlier basis no longer applies, or when recovering it would cost more than obtaining adequate current grounds. Do not repeat an observation merely to confirm a result already settled for this use.
A changed description can settle a new comparison immediately. It does not automatically make all prior facts unknown. A changed subject can instead remove the basis for a current answer while leaving the historical answer intact.
Keep evidence of effort and evidence of the intended result separate. A packer's report of two hours spent preparing cards does not settle their count or wording; inspect or reuse grounds for those conditions. The result described in a report is also distinct from the report's own completion.
Obtain and express the answer
Perform or obtain the check needed. Use a direct observation, executable comparison or competent judgement according to the question. Where the mode is read-do, carry out the prompted work and then establish the result; where it is do-confirm, check the selected conditions of work already performed.
If conditions change during use, reconsider what can usefully be checked now while sustaining essential work. A limited time window may call for a competent choice of the most consequential feasible contributions, help from another participant or a pause. Keep omitted or interrupted checks distinguishable from positive answers and communicate their consequences for continued reliance. Completing the visible list is not sufficient reason to sacrifice the result it was meant to support.
Express what the grounds support. A supported negative answer is No. Unknown means the available grounds do not settle the question. Not checked describes the absence of a check for this occasion, not the subject's condition. If a competent earlier result settles it, record the answer as reused, not as newly observed. Give a reason for not applicable that concerns scope.
Keep the answer associated with the subject, criterion and grounds needed by its recipient. A tool failure can be evidence about the checking attempt while leaving the subject question unresolved. Another agent's agreement may contribute judgement; it does not supply an unperformed observation.
When the grounds are in a document or model, examine the relevant content using an appropriate way of interpreting it. A polished report can omit the decision being asked about; unfinished meeting notes can already identify that decision and its grounds. The document's existence, formatting or completion does not settle the condition of what it describes. Obtain enough detail for the recipient's reliance, and retain a usable route to it when the answer will be reused.
Decide what the answer changes
Ask what follows for the intended work. A mismatch can support changing the subject, revising a mistaken description, changing the intended use, obtaining more information, or choosing a bounded continuation. Recover the decision grounds and any separately applicable authority. The answer itself imposes no duty and grants no permission. If recording it also triggers an operational action, identify that coupling and establish the grounds for the action separately.
If a change is selected, recover how it could produce the intended result: the suitable Method or alternative Methods, a capable performer, needed means and a useful occasion. Return missing Method knowledge to the domain practice; writing another question does not supply it. For a noisy workshop, closing a door, changing rooms and moving the discussion online are different possible interventions, with different conditions of use. An expected time for quiet is a planning judgement, not an observation that quiet has been achieved.
Distinguish the selected work from checking and recording it. A decision to replace a card does not establish that the card was replaced. After the change, check the affected question on adequate grounds. An operating-case continuation Method separates continuation, performance, records and resulting state when an operating case needs another action; an applicable operating-control Method addresses the quality and reliability question in continuing operation.
Keep historical answers and unaffected grounds. Reopen other questions only where the change can defeat their basis. Stop when the selected questions are sufficiently answered for the intended use, or report which observation, interpretation, capability or decision is missing. Use C.41.SR when a decision depends on how several answers fit together or a changed premise can affect a conclusion drawn from them. One person can need that reconciliation. Several recipients can also use a single sufficient answer directly.
Archetypal Grounding
Set S has twenty readable cards matching wording v2 and is assembled for delivery. Count/readability was observed at 10:00, wording at 10:05 and assembly at 10:10. The answers were entered at 10:12.
At 10:15, the facilitator changes the comparison description to v3, replacing one phrase.
In this constructed case, the facilitator decides to keep v3 and replace the affected card. Replacement is the selected work. A subsequent comparison establishes the new answer. If opening the set and changing a card could leave it incomplete, recounting is useful. If the remaining cards were protected and their wording unchanged, repeating every wording comparison adds no corresponding gain.
The v2 answer remains a historical answer to its question. Recording the v3 answer later does not alter when the earlier observation occurred, and neither observation alone establishes when a condition first became true.
A spill can make already checked cards unreadable while their wording and the completed printing work remain unchanged. Observed damage supports a current negative answer to readability; it does not erase the earlier work. In another branch, the workshop is cancelled while every physical-set answer remains true. The decision about further preparation then uses the changed occasion and value of the work, rather than trying to turn the physical answers negative.
For an importer, the agent's message “error handling implemented” reports work. To decide what happened in a particular failed import, use an adequate comparison of the catalogue before and after that run. If the only observation tool failed and the later state is unknown, the question stays open. A confirmed change to the state that should have been preserved settles it in the negative. One such failure within the promised range refutes the claim that the importer preserves the catalogue on every failed import. If preservation is required, this run violated that requirement. The same requirement applies to the next import. An unchanged catalogue settles the question for that run. Extending that success to other inputs calls for grounds that cover them. Another run adds nothing to a negative answer already established for the decision at hand.
Bias-Annotation
Completion pressure can turn uncertain or negative answers into apparent passes. An expert may also rely on tacit judgement that others cannot recover. Record the grounds proportionately to how the result will be used, and preserve an unresolved answer.
Familiarity can encourage both stale reuse and unnecessary repetition. Examine relevance of the premises rather than preferring either “always recheck” or “already green.”
Conformance Checklist
For the answers on this occasion, ask:
- Is each answer bound to the subject, criterion and relevant conditions?
- Do its grounds settle that question, including when an earlier result is reused?
- Are No, unknown, not checked and inapplicability interpreted without changing the subject claim?
- Are observation, entry, performed change and any decision kept distinguishable?
- Did a changed premise reopen exactly the affected reliance while preserving historical and unaffected information?
Common Anti-Patterns and How to Avoid Them
Consequences
The Method produces usable information even when the answer is negative or incomplete. It supports economical reuse and targeted recovery after change. Its cost is attention to the premises of reliance. It cannot supply a missing domain judgement or make an unauthorized action acceptable merely by recording a pass.
Architectural Rationale
Question selection and checker qualification happen at different scopes from an answer. Repeating them in every use wastes effort; omitting them when their premises fail produces unsupported certainty. This Method uses adequate supplied criteria and checking means; when either proves inadequate, return the affected question to C.41 or C.41.QC.
Keeping decision and corrective work separate preserves the user's choice to revise a mistaken description.
SoTA-Echoing
How much recovery or fresh checking is needed after an interruption or changed premise? Adopt sufficient existing grounds for the question, and adapt the resumption cue to the missing context. The alternatives are a position marker, a marker with question-specific content, and obtaining a fresh observation. Section 4.1 chooses among them by whether the question and prior grounds are recoverable and still apply, including the cost of that recovery.
A bare “card seven” is enough only when the performer already knows what was compared and against which description. When that context is missing, retaining it avoids either guessing or repeating all comparisons. The trade-off is a little more preparation and reading for a more dependable continuation. If recovering a trustworthy record costs more than an adequate inspection, inspect again. In §5, however, known unchanged S and the changed v3 phrase support No: another observation adds no needed information. If S may have changed, current inspection supplies a missing basis. These branches reject both automatic repetition and automatic reuse.
Chen and colleagues' laboratory experiments supply a contextual-cue candidate and comparisons with other task support, with effects that depend on the experimental condition and measure. They motivate testing which content matters; they do not rank every workplace recovery arrangement or establish a universal interface. The supplied Methodology and Systems Management manuscripts distinguish useful work from reporting. Anthropic's 2025 account supplies a concrete testing-before-pass arrangement; its 2026 account motivates revisiting the means as capabilities change. Those engineering cases do not establish that every resumed occasion needs a new test.
The selected line spends recovery effort only on information needed for a sound current answer, while preserving historical and unaffected answers in §§4.2–4.3. Reopen the choice when a cue fails to recover the next operation, a subject or basis changes, or a cheaper adequate observation becomes available. If cue alternatives remain indistinguishable on that use question, keep the simpler one.
Relations
C.41 supplies selected questions, C.41.TF their usable expression and C.41.OF an occasion with access to answers. C.41.QC supplies a qualification when criterion or checker uncertainty is material. C.41.SR connects answers across shared work.
Administrative practice supplies claim checking and fit-for-use records. Operating practice supplies case continuation, exception handling, and quality and reliability control. FPF E.10.D2 preserves subject and description, and A.15.10 governs changed-premise use across work.
C.41.UA:End
Reconcile Checklist Results Across Shared Work
Type: Method pattern Status: Stable
Problem frame
Use this when several people or agents rely on local checklist answers to conduct shared work, one person combines several such contributions, or a change may have invalidated those answers. A technician's equipment list, a facilitator's material list and an organizer's summary can all be green while the session lacks a usable exercise.
This Method establishes which local answers support a stated shared conclusion, what remains unresolved, and which reliance changes after a new premise. Start by naming the shared result and the local claim someone is about to use.
A conversation can supply the whole useful result. If local answers have independent uses and no shared reliance, leave them local.
Problem
Local summaries compress away the subject, time, description or evidence that another participant needs. A missing assignment is reported as if it were an unknown subject state. A whole-result pass is inferred from constituent passes without examining their connection.
The reconciliation problem is to recover those connections at useful granularity. Recover only the claims and relations needed for that shared use.
Forces
Solution
B.1.5 governs a Method-composition claim, and A.15.10 governs changed-premise recovery. Their results do not themselves reconcile local answers about different subjects and occasions. The construction below makes those claims usable together, tests the additional whole-result conditions and returns a missing arrangement, observation or interpretation to its proper contributor.
Recover the shared question and contributing claims
Identify the result or next action that consumes answers from the contributing work, whether performed by different participants or by the same person. Recover for each relevant claim its subject, description or criterion, observation conditions and grounds. Translate local labels by their meaning: one person's “ready” can mean powered equipment; another's can mean content suitable for participants.
Keep the operating subject, its condition, an observation and a record of it distinct. Recover the small shared account needed for this reliance: which result and occasion the claims concern, their sources and conditions, what changed, what remains unknown and who needs the answer in time. Retain a record only when later use needs it, with enough meaning and currentness to recover the relied-on claim. Keep people or technical systems performing work distinct from descriptions, materials and observations used in it.
A work assignment does not establish capability or completed work. A request sent, an accepted assignment, performed work, reported completion and a result examined by its recipient answer different questions. Recover which event a local status reports; where waiting matters, include the needed coordination as well as performance.
If the technician declines the requested check, the request remains unfulfilled without establishing that the equipment failed. Recover another capable contribution, revise the arrangement or choose a bounded continuation. Similarly, unreceived completion news can leave another participant uninformed even though the check was performed. Make the consequential handover and its recovery clear without requiring a formal coordination protocol for every exchange.
Preserve a small connection rather than merging all local lists. It may be enough to identify the session, room and materials edition once, with the local claims that bear on them. A richer shared account is useful only where readers need it. Put the current account where those participants can find and interpret it in time. Where their work arrangement allows it, they can choose a next contribution from that information: the technician sees that the materials are available and starts the comparison when the room is free. A dashboard helps only if its subjects, grounds and freshness support that decision; visible colours alone cannot replace them.
Choose the communication arrangement for its consumers. An exchange can suffice when the participants can settle the shared question now. Use an existing operating view when later participants need to recover the subjects, grounds and changes. Add a separate retained account only for a consequential reliance that neither the exchange nor that view can support. Include updates and reconciliation between copies in its cost.
In a progress conversation or report, examine what is absent as well as what is stated. If materials are discussed but participant preparation has disappeared from view, ask what is known, who can supply grounds, and who can carry out any chosen follow-up. An omission does not itself establish failure or concealment. A person accepting a contribution does not thereby demonstrate the capability to make it. Arrange a worthwhile return to the other consequential questions while attention is concentrated on an urgent problem; choose the occasion from the work rather than imposing a universal reporting interval.
Test how the contributions support the whole
Start with the decision about the whole and its acceptance conditions. Which conditions must all hold, which alternatives are allowed, and what has to work at the connection between contributions? Establish which conditions the local answers cover for the same configuration and occasion, including the availability of intermediate results and the joint demands on time, people and resources. If those grounds settle the shared question, give the answer and its grounds to whoever makes the decision. An established failure of a required condition can settle the question in the negative. If one admissible alternative fails, judge the others on their own grounds. If the question remains open, identify what would settle it. Participants may need to interpret an observation, obtain a missing one, agree on a handover or allocate a resource. Tell the decision maker what the available answers already establish.
Name the relevant relation instead of inferring it from a checklist's nesting. A supplier can enable work; a schedule allowance can affect its feasibility; a model can describe the subject; one observation can support several conclusions. Identify the facts it supplies to each. Two records derived from the same observation count as one source when independent confirmation matters. For a physical part–whole claim, identify the part and the whole in the work itself. Recover what a changed local condition changes for each consuming claim.
The construction of an encompassing Method determines which local Methods it admits. FPF B.1.5 explains the composition distinction; its consequence here is to examine the whole's conditions when interpreting constituent work. To explain an actual shared job, establish which performance was part of that Work and which Method it enacted. An outside service or an earlier conclusion can support the job through its result.
Separate three kinds of gap. A missing performer or resource returns to enabling and coordination work. An unknown subject condition returns to an observation or judgement. A conflict between claims first calls for comparing their subjects, meanings, times and grounds; if those match, investigate the substantive disagreement. Do not resolve a conflict by averaging labels or accepting the newest entry.
Where the case can reveal something the fixed questions miss, prompt a short exchange among the people whose contributions interact. Before the workshop, ask what has changed, what each participant is concerned about, and what the others need to know. A technician may report an unfamiliar connection; a facilitator may reveal that one participant cannot use the planned materials. Recover the consequence, the needed judgement and any chosen follow-up. Holding the conversation does not establish that the concern was resolved. Leave competent local choices with those able to make them while communicating changes that affect another contribution.
Recover after change or interruption
When a room, description, tool, subject or intended use changes, identify which earlier claim consumed that premise. Follow the reliance to the shared conclusion. FPF A.15.10 supplies changed-premise recovery; C.41.UA supplies the affected answers.
Retain the historical account. Reuse unaffected observations where their basis remains sufficient. A continuation after interruption consumes this account of current questions and grounds, not only a remembered set of colours.
Record or communicate the resulting shared conclusion, any unresolved contribution and any needed next move. Stop at enough detail for the current dependence. If maintaining the connections costs more than the shared reliance warrants, simplify to a conversation or an existing coordination arrangement.
Change, simultaneous work and the whole result
Checklist questions select conditions within structures of transformations and structures of states of their subjects. The questions need not form a ladder. Wording can change while a set remains assembled; handling a replacement can defeat an earlier count; two preparations can proceed at once. A table, diagram or graph may represent useful relations, but its layout does not determine the states or transformations.
During a workshop preparation, testing a projected example can contribute both to checking the display and to preparing the session. The whole matters at that moment: a technically sharp image of an obsolete example does not serve the exercise. The relevant FPF B.1.5 Method relations and A.15.10 changed-premise relations help identify what the combined work consumes. A supplied projector enables this work; its availability alone does not establish a performed part of the session preparation.
A missing intermediate ability can stop the combination. Someone able to count cards may lack the judgement needed to select instructional examples. The next contribution is then help with the examples, not another count. The three people may also need the same room at incompatible times. Keeping a shared operating account helps make that conflict recoverable across participants. Selecting a feasible resource arrangement is additional coordination work; filling three lists cannot resolve it.
For a fuller account of change and subject states, return to FPF A.3.3, A.3.4 and A.22. These foundations help describe the structures used here; a constructive composition of transformations needs its own justified account. Context-state descriptions in E.18 concern the bindings and locality they describe, rather than every state of a checklist subject.
Archetypal Grounding
For a workshop, the venue technician checks projection and sound; the facilitator checks materials against the exercise; the organizer coordinates the session. Their local results include:
The shared account binds these claims to session W, room A and materials v2. The facilitator projects the example and compares it with the cards. That action contributes to both checking the presentation and preparing the exercise. A sharp image containing an obsolete example yields a useful equipment observation and an unresolved or negative content comparison. The person who notices the content problem brings that finding into the session decision instead of passing along an unqualified overall “ready.” The local equipment answer can remain true while the joint result needs another action.
If the three participants are together, can inspect the same materials and will act immediately, this reconciliation can remain a spoken exchange. If another technician will take over after a room change, the shared operating view needs to preserve W, the new room, retained material grounds and the room questions still open. An existing view that supplies those distinctions is sufficient; a second reconciliation report would add another copy to update.
The same difficulty can occur when one facilitator also packs the cards and operates the projector. After revising the slides, she can forget that the printed set still carries the earlier wording. One performer does not make those contributions consistent; compare their subjects and editions just as in the multi-person arrangement.
If the venue moves W to room B, prior room-A observations remain historical. Room-B projection, audibility and relevant arrangement questions reopen; unchanged card wording need not be compared again merely because the room changed.
If the facilitator instead revises one example, the content comparison and its display can reopen while room-specific equipment facts remain useful. If handling the cards changes the set, affected physical questions reopen too. The structure has several distinct dependencies, not one inevitable progression from “unprepared” to “ready.”
No technician has been assigned in a third branch. That is an enabling gap. In a fourth branch, the technician is assigned but no one knows whether projection works. That is a missing observation. A shared “in progress” entry hides the difference and cannot select the right next action.
For an importer, parser checks and storage checks can both pass locally while their integration mishandles an error. Reconciliation connects the supplied parser result, importer's response and resulting catalogue state. Whether the catalogue entries are suitable for publication remains a different question. A session restart preserves those distinctions so a new agent can reuse grounds or perform only the missing check.
Shared work with uncertain criteria and changed premises
A rented-room workshop combines projection, card exercises and participant coordination. The technician uses a venue tool, the facilitator a materials workspace, and the organizer a session calendar. The team is uncertain whether “readable on screen” can be judged reliably at the technician's desk.
When local answers exist but their support for the session is unclear, C.41.SR first connects them to the shared result and exposes missing or conflicting contributions. Follow the return that answers the difficulty. In this case, uncertain projection checking calls for qualification before reliance on its verdict.
Mantra — from local answers to a shared result: Name the shared result. For each contribution, recover the subject, question, conditions and grounds behind its answer. Reuse what is adequate; qualify uncertain checks and obtain what is missing. Ask what the whole still needs and whether the contributions can combine. Return a missing performer or resource to coordination, an unanswered subject question to the needed observation or interpretation, and incompatible claims to those who can resolve them. After change, retain unaffected grounds and reconsider dependent questions; after interruption, recover these distinct next moves. Use a conversation or an existing view when enough; record more only for a concrete later reliance.
The worked case below connects these moves to the room, materials and workshop. Stop reconciling when the shared question is sufficiently answered; the participants still make the continuation decision and perform the selected work.
Qualification and selection. C.41.QC challenges the criterion under the intended viewing conditions. A representative participant examines the projected example from relevant seats. If desk readability fails to distinguish a difficulty, the team changes that criterion or its checking means. C.41 then retains questions about the room setup, content and exercise use; it leaves unrelated venue questions in the venue's own practice.
Local representation and occasion. C.41.TF preserves the session, room and materials references in the relevant local views. C.41.OF arranges the observation while the room and content are available. A missing capable interpreter returns to capability support; an unavailable room returns to scheduling.
Local answers and simultaneous contribution. Through C.41.UA the technician establishes equipment conditions and the facilitator compares the projected example with the physical card. One well-grounded comparison can settle whether those two versions agree. Both participants can use it, even if they record it in separate lists. If projection is essential to the commissioned exercise, a confirmed projector failure rules out running it as planned. If the exercise also permits printed copies, that failure rules out the screen option. The organizer can use the printed option when the copies, participants' access and other conditions of that exercise are established.
Reconciliation. C.41.SR connects the local answers to workshop W. A whole-session account includes the materials and arrangement, not a count of green local lists. If the facilitator and technician need the room at incompatible times, a shared operating account makes the conflicting claims visible to the people authorized to select a feasible arrangement or report that the allowed arrangements cannot meet the terms before those contributions combine.
Changed premise and continuation. Moving W from room A to B reopens the room-dependent observations. Revising the example reopens its content and display questions. Known unaffected card facts remain usable. An absent technician assignment returns to coordination; an assigned technician with an unknown projection result returns to observation. After an interruption, the shared account lets participants recover these different next moves.
The route uses all six Checklist Methods because this constructed case has uncertain criteria, local forms, an occasion, answers and shared reliance. A simpler workshop may need only a conversation and a check. The example explains the connections; it does not prescribe six phases for every session.
Bias-Annotation
A shared view often reflects what the coordinator can see. Local work or a participant's uncertainty can disappear during aggregation. Allow the person supplying an observation to correct its shared interpretation.
A central record can also shift maintenance work to whoever has the least authority to refuse it. Include that burden when deciding how much reconciliation is useful.
Conformance Checklist
For the shared conclusion being used, inspect:
- Which local claims support it, and do their subject, meaning and occasion match its use?
- Which conditions and connections does the shared decision require, which do the available answers establish, and what, if anything, remains open?
- Are missing arrangement, missing observation and substantive conflict distinguished?
- Does each changed premise reopen the reliance it affects?
- Can participants recover the next contribution in their working tools without unnecessary duplication?
Common Anti-Patterns and How to Avoid Them
Consequences
The team gains a shared result with interpretable limits and a targeted recovery route. Reconciliation can reveal integration failures before they become a false overall pass. It costs attention to connections and changes; excessive centralization can make that cost dominate the benefit. Its output does not remove the need for domain judgement about the whole.
Architectural Rationale
Local checking and shared reconciliation solve different problems. Sound local answers can still concern incompatible occasions or omit a combining condition.
The selected arrangement preserves local usefulness and adds only the shared claims with consumers. A single shared checklist remains a valid simpler alternative when all participants can interpret and maintain it at low cost.
SoTA-Echoing
When should local checks be reconciled in conversation, an operating view or a separate retained account? Adopt a shared question with recoverable contributing claims, using the supplied Systems Management manuscript's operational/coordination practice. Adapt the communication arrangement to who will rely on those claims and when. Exchange and an already adequate operating view are alternatives to another maintained record.
Sections 4.1–4.3 choose the arrangement for the same needed session conclusion. In §5, participants who can inspect the same materials and act together can resolve the question in conversation. A later technician cannot depend on having heard that exchange; the existing operating view can preserve W, the changed room, still-valid material grounds and missing room observations. Adding a second report provides no further needed distinction when that view is sufficient. Accept the extra recording and update effort only when a later consequential use still lacks its subjects, grounds or changes.
The supplied Methodology manuscript's dependency cases and the supplied Systems Management manuscript's views supply the practice comparison, while Atul Gawande's open-team-exchange cases in The Checklist Manifesto (2009) supply a coordination candidate for trouble outside fixed questions. These accounts explain why a shared conversation can be the useful action; they do not measure the cost of every arrangement. Anthropic's account supplies bounded engineering evidence for retaining criteria and progress across agent sessions. It motivates recoverable continuation, not a universal reporting format or cadence.
The defect overcome is a lost local-to-whole dependency or an unrecoverable handover, not a low completion percentage. Reject aggregation of labels as the substitute for that relation. The selected arrangement trades communication and maintenance effort for the later reliance it enables; where an existing exchange or view suffices, it adds no new record. Reopen the choice when a new participant, delay, changed work or tool prevents recovery of a consequential connection, or when duplicate maintenance outweighs its remaining use.
Relations
C.41.UA supplies local answers and their grounds. C.41 supplies questions about the shared result, C.41.QC tests uncertain criteria, and C.41.OF arranges participants' access and interaction.
Operating practice illustrates distinctions among cases, queues, resources and records and a recoverable shared operating account; administrative practice illustrates records fit for their uses. These are domain applications of the distinctions used in :4.1. FPF B.1.5 supplies Method composition and A.15.10 changed-premise recovery. This Method constructs the connection among checklist answers used across shared work; its subject and record distinctions do not require one of those domain applications.
C.41.SR:End
Qualify Checklist Criteria and Checking Means
Type: Method pattern Status: Stable
Problem frame
Use this when a proposed criterion or checking means may fail to distinguish the difference that matters. This is common with inherited lists, generated criteria, model judgements, new tools and changed operating conditions. A checker can say “pass” while an importer replaces A17 with 1.
This Method establishes a bounded basis for using a criterion and its checking means, or returns a counterexample or unresolved qualification question. Start with one claim the proposed check is supposed to support and one plausible case where it could be wrong.
Reuse an applicable qualification when available. A second agent, a formal experiment or a broad benchmark is useful only when it answers an uncertainty at worthwhile cost.
Problem
A criterion can be irrelevant, vague or based on a mistaken description. A checker can implement a sensible criterion badly or observe a convenient proxy. Agreement among judges and high aggregate scores can hide the individual failure that matters to the work.
The problem is to examine both the criterion and the means of checking it before placing more reliance on their answers than the evidence supports. This differs from selecting which questions belong in the aid and from checking the subject on one occasion.
Forces
Solution
A.10 governs evidence reliance; A.19.ULSAM governs a numerical aggregation. This qualification constructs the cases and expected distinctions needed to judge a criterion and its checking means. Separate answers need no aggregate, and an applicable earlier qualification needs no repetition merely because the questions are now carried in a checklist.
Recover the claim, basis and intended reliance
Name the criterion, the question it interprets and the subject of that question. Recover the description and the assumption connecting it to useful work. When the criterion needs construction, use C.41 §§4.1–4.2 to obtain a consequential work question and its basis. Use C.39.RQ when the missing criterion concerns a Method contribution.
Then name the checking means: a person applying a judgement, a comparison script, an observation tool or a model-based evaluator. What does it observe or consume? What answer does it produce? What will another person or system infer from that answer? Administrative claim-checking Methods supply the distinctions needed for that domain.
For a model judge, include the supplied evidence and instructions, the criteria grouped in each call, and any rule combining repeated verdicts when these can change the relied-on answer.
Separate two uncertainties. A criterion can faithfully express the description while the description poorly serves the work. A checking means can also fail to apply that otherwise useful criterion. Investigate the uncertainty that changes the intended use.
A cheap indirect signal can justify further examination without certifying the larger result when it passes. An obsolete edition label on the card box can prompt inspection of its contents; a current label alone does not establish that every card was replaced.
Name this intended inference and next move. Challenge both false reassurance and unnecessary alarms. If people can satisfy the visible signal while bypassing the substantive work, examine whether the signal remains informative for the intended use.
Construct a discriminating challenge
Choose a challenge whose answer could change the proposed reliance. Use cases with known or competently established differences relevant to the criterion. When qualification depends on detecting both a satisfactory result and a plausible failure, include both and vary the property of interest while keeping other differences small enough to interpret the verdict. One established false pass can already rule out a checker as a guarantee that every accepted item meets the criterion. For a subjective criterion, recover the judgement basis and a meaningful disagreement.
Predict which answers those differences support before inspecting the checker output. Then compare the outputs with adequate grounds for those expected distinctions. A false pass and a false failure can have different costs; inspect both where they affect the work.
Where a rubric may drive the verdict independently of the subject, also hold the subject facts fixed and change the question. Establish what answer the new predicate supports before asking the checker. A reversed condition can call for the opposite answer; if only wording changes and the question stays the same, the expected answer stays the same. This distinction tests responsiveness to meaning without requiring every wording change to reverse a result.
When a decision about the judging arrangement depends on whether an alternative would give adequate answers with less effort or interference, compare the alternatives on the grounded cases. A question about whether the existing arrangement can be relied on may already be settled by its observations or by a known failure.
Grouping criteria can save calls; separating them can reveal interference. Repetition and a rule for combining verdicts may reduce sampling instability, but repeated agreement cannot correct a consistent error or supply independent truth.
Select the arrangement from consequential errors and total effort; repeat only when that comparison can change reliance.
If the receiving use consumes a total or percentage, examine what that number means and how the questions affect it. Splitting one easy condition into ten items can increase its influence without improving the subject. Ten successful formatting checks may conceal one failed catalogue-preservation check. Challenge the implied priorities, duplicate contributions and treatment of missing answers. Keep consequential answers visible.
Where a numerical aggregate is needed, FPF A.19.ULSAM governs the admitted measures and lawful aggregation; a convenient average does not supply its justification.
Start cheaply. One counterexample can defeat an overbroad claim. Use a direct executable comparison when the question and observable property admit it.
A selective rubric or pairwise judgement can help with an interpretive or relative question, but choosing the better of two results does not establish that either meets an absolute condition.
Qualify that judging arrangement on cases whose consequential answers have adequate independent grounds. Compare false passes, false failures and obtaining those grounds as well as the cost per judgement.
Broader reliance may call for representative cases, independent judgement, executable observation or a field trial through A.3.1.WT. Choose another reader or agent when it supplies a materially different basis, expertise or observation. Mere duplication of the same unsupported inference does not establish independence.
When work is repeatedly optimized against a score, compare earlier and later work results on consequential characteristics and suitable new cases. A rising score can reward surface changes while the intended result deteriorates. For the importer, a more persuasive success report leaves identifier loss unchanged. Inspect the proposed shortcut and obtain evidence beyond the optimized signal.
A stronger second model can help investigate disagreement, but its score needs grounds and does not become ground truth by rank.
State the bounded result and a return condition
State which decisions the qualification supports. A grounded check on one item can settle the disposition of that item. When relying on a checker across a class of items, establish why its observations and decision rule detect the differences that matter throughout that class, or bound the claim by the evidence available. Testing one checker on one kind of item and another checker on a different kind establishes what happened in those pairings. To use them interchangeably, examine the untried combinations that matter or establish why the change cannot affect the answer the recipient needs. If the intended decision permits a rate of error, evidence about that rate must cover the population and operating conditions to which it will be applied. Report detected defects and remaining uncertainty where they change what the recipient can accept or needs to inspect.
If the criterion misses a work question, return to C.41. If the form corrupts its meaning, return to C.41.TF. If access or capability is missing, return to C.41.OF or its supplier. A false pass, a false rejection or a failure to return a needed answer can justify withdrawing the checker from that use or limiting it to conditions for which adequate evidence already exists. Before relying on a repaired or replacement checker, establish that it gives the answer the criterion calls for under the affected conditions. Repeat the affected challenge if that evidence is still missing. Compare the gain with the qualification and operating cost through C.11.MV.
Keep the qualification separate from subject answers. C.41.UA establishes what is true of the subject on its occasion.
Reopen qualification when a material criterion, model, tool, observation route or operating condition changes, or when a new counterexample defeats its scope. Changed evidence, instructions, criterion grouping or verdict combination can trigger this even with the same model and criterion.
Retain qualification for the unaffected observations and inferences. Stop when further investigation cannot improve the intended reliance at worthwhile cost, and preserve the limit rather than extending the claim.
Archetypal Grounding
A team describes an importer as preserving item identifiers, reporting duplicate identifiers and retaining the previous catalogue when parsing fails. The description embodies assumptions about the team's use. A proposed generated check asks only whether an input file was accepted.
That check establishes none of the three described properties. A17 can become 1 during an otherwise successful import. A successful run also says nothing about duplicate reporting or the state after a parse failure.
The team constructs these qualification cases:
These are three separate questions. A candidate checker that observes only an exit code cannot settle the third unless an independently established relation makes that code sufficient for the stated use.
The team tests its proposed before/after comparison against an altered catalogue and an unchanged one. If it passes both, it has failed the intended discrimination; if it distinguishes them, that supports the comparison for those cases.
Complement that changed-catalogue challenge with a changed-question challenge. Use a constructed fixture whose before and after catalogues are both C. For these fixed facts, “Are the before and after catalogue contents identical?” has the expected answer Yes, while “Do the after contents differ from the before contents?” has the expected answer No. “Do both snapshots contain the same catalogue content?” preserves the first question and therefore its expected Yes. These answers follow from the fixture before any checker is run.
A checker that returns the same verdict for the opposite conditions misses the question; one that changes the first answer merely because it is reworded is unstable on that case. Passing these contrasts supports only this bounded discrimination.
To examine grouped judging, give a model the same before and after snapshots C, plus a report headed “Import summary”. Before any output, content identity has expected Yes, and the presence of that heading has its separate expected Yes. Compare the identity verdict when asked alone and when asked alongside the heading question, with the same evidence supplied. The extra question changes neither snapshot nor the identity predicate.
Repeat the contrast with a fixture whose after snapshot changes A17 to 1: identity is then No in both arrangements, while the heading answer stays Yes.
If grouping changes either identity verdict, investigate or restrict that arrangement; do not average the two distinct questions into a pass. Combining repeated identity verdicts has to answer to these fixture facts. Where grouped judging preserves the needed distinctions at lower cost, keep it.
For stored data, an adequate comparison of its values remains the cheaper starting choice.
A reliable record or applicable rollback guarantee can settle the post-state even when it was unobserved. Where neither the observation nor another adequate ground settles this run, report the state as unresolved. Tool failure is not an importer failure or success. Local parser correctness does not settle the encompassing response and catalogue state; C.41.SR connects those contributions.
For the workshop display, “readable” may be judged from a desktop preview while participants will read the projected content from the back of a room. Qualification varies the viewing condition relevant to the exercise and asks a representative participant to interpret the content. A font-size threshold can be a useful local proxy only if its basis supports that use. Reconsider that proxy if glare or contrast makes text unreadable despite the chosen size.
Both examples are constructed probes. They illustrate how to challenge a claim, not measured checker performance.
Bias-Annotation
People tend to invent tests that confirm their implementation. Generated criteria and generated answers can share the same blind spot. Recover a failure from the receiving work, seek an unlike case where it matters, and distinguish the data used to tune a checker from the evidence used to judge its broader use.
A benchmark's accessible cases can underrepresent the domain or participants who bear false failures. Make that limit part of the reliance decision.
Conformance Checklist
Inspect the qualification's claim and probes:
- Are criterion, checking means, subject and receiving inference separately recoverable?
- What grounded comparison, subject rule or counterexample warrants reliance over the intended cases?
- Have false passes, false failures and unresolved observation been interpreted at their costs?
- Does any independent contribution supply a different basis rather than repeated agreement?
- Is the resulting reliance bounded, with material change and counterexample triggers for reconsideration?
Common Anti-Patterns and How to Avoid Them
Consequences
The Method can replace an attractive but uninformative check with a defensible, limited one. A single counterexample may prevent large misplaced reliance. Qualification also costs representative data, judgement and maintenance. Some questions remain uncertain or too expensive to settle; the useful outcome can be a smaller reliance claim or a different work arrangement.
Architectural Rationale
Qualification is triggered when the adequacy of criteria or checking means becomes uncertain. Folding it into every ordinary use wastes effort; omitting it from design lets repeated unsupported answers accumulate.
An applicable prior qualification can settle the uncertainty. Otherwise begin with a concrete difference and scale the evidence to the intended reliance, using direct executable tests of the property or competent human judgement as appropriate.
SoTA-Echoing
Which checking means and qualification effort fit the intended reliance? Adopt a direct comparison of the observable property; adapt selective rubric or pairwise judging for an interpretive or relative question, with discriminating cases when its adequacy is uncertain.
These are alternatives, not successive compulsory stages. In §§4.2 and 5, inspecting identifier or catalogue equality supplies the relevant grounds. Asking a model for a persuasive overall score adds cost without settling those properties. Pairwise judging can instead help choose between two explanations, but the preferred one needs a separate answer to an absolute suitability question.
Selective Checklist Evaluation supplies a current rival to full-rubric/direct scoring, with effects that differ by setting and comparison mode. It warrants choosing and qualifying the mode for the receiving question, not declaring one universally superior. The study by Hong and colleagues supplies counterevidence about detail and uneven emphasis in generated criteria. Section 4.2 therefore examines consequential errors, missing answers and induced weighting alongside judgement cost. The extra qualification is worthwhile when it can defeat misplaced reliance or select better means; an applicable prior qualification remains sufficient for its unchanged scope.
The study by Bagaria and colleagues supplies bounded counterevidence about judges responding to changed answers and reversed criteria. Its generated and qualitatively inspected contrasts leave uncertainty about some expected verdicts.
The adaptation in §§4.2 and 5 fixes expected answers from a fixture, then varies subject facts or question meaning separately.
RIPD supplies a further benchmark-versus-target counterexample motivating examination after rubric edits; it does not qualify a deployed judge.
For single, grouped or repeated model judgements, adapt protocol comparison to the consequential uncertainty in §§4.1–4.3 and the fixed-fixture case in §5. RuVerBench v2 supplies counterevidence about prompting, grouping and voting under long research/coding contexts; its selected binary cases exclude substantial ambiguity. PReMISE v1 separates stability from preference fit and robustness under model-mediated evaluation. Neither determines this deployment's best arrangement. Compare saved calls with consequential errors and examination effort; retain adequate prior qualification, and reopen its affected scope when a changed arrangement can alter a relied-on answer.
AutoChecklist and RLCF supply criterion refinement and checklist-feedback training alternatives. Reject their benchmark or training gains as substitutes for qualifying this checker.
CHERRL and Rubric Dropout expose proxy exploitation and divergence in bounded training settings; the latter's single-run configurations and fallible external judge limit the result. Its training intervention gives no general reason to omit operational checks. Section 4.2 instead examines work results and new cases beyond the optimized signal.
The trade-off is the cost of obtaining discriminating grounds against the cost of a consequential false answer. None of these studies supplies a common error/cost ranking for every domain.
Anthropic's harness-design account supplies a worked calibration candidate as models and tools change; A.3.1.WT and C.11.MV guide the local trial and worth comparison when the choice remains consequential.
Reopen qualification for a material criterion, tool or use change, an unexplained verdict on a known contrast, or a cheaper adequate observation.
Relations
C.41 supplies proposed work questions and receives coverage corrections. C.41.TF preserves qualified meanings, C.41.OF makes observations accessible, C.41.UA performs checks, and C.41.SR interprets their combined use. FPF A.19.ULSAM is a conditional return when the receiving use needs numerical aggregation; ordinary separate answers require no such calculation.
C.41 supplies construction of work questions and their criteria in §§4.1–4.2; C.39.RQ supplies criteria for a required Method contribution. A.3.1.WT supplies representative trial, and C.11.MV worth appraisal. The subject examination supplies the observation or judgement whose adequacy is being tested; administrative claim checking is one domain example. Qualification uses the required results for the uncertainty in checklist criteria or means; it does not replace those Methods.
C.41.QC:End
Reduce Work While Preserving the Needed Result
Type: Method pattern Status: Stable Normativity: Normative unless marked informative.
Problem frame
Use this when obtaining a needed result consumes work or time that might be avoided, but a local cut, faster operation or new tool could lose something the whole undertaking needs. A team repeatedly rebuilds reports after source changes. A service spends more time waiting for decisions than processing requests. A learner can obtain an answer instantly but needs to acquire the ability to produce and challenge it.
The working reader is a practitioner responsible for improving a way of obtaining a result, with access to the people and knowledge needed to judge that result. The subject is the change to the whole way of obtaining it: what can cease, what must remain, what changes the troublesome condition, and how the change reaches subsequent work.
First useful move. Take one costly occurrence or proposed saving. Follow it to the result and receiving use it serves. Distinguish a contribution that is no longer needed from a needed contribution obtained with avoidable effort. This can already justify retaining an activity, removing an unnecessary duplicate, or identifying the condition a change must address.
Practical gain. The method helps the reader construct a worthwhile reduction instead of accelerating unwanted output, moving the burden to somebody else, or removing a condition of later success. Its result can also be a warranted decision to keep the present way.
Ordinary non-use. When an adequate known correction or domain Method already supplies the whole answer at acceptable burden, use it. If the only question is whether one proposed requirement contributes anything, A.11.OP supplies that test. If the main problem is a missing way to obtain the needed result, C.39 develops that construction. Return here when the question is which avoidable work to remove and how to make that reduction hold in the receiving work.
An immediate protective action follows the applicable safety and authority rules. A broad improvement inquiry is not a prerequisite for protection whose grounds are already sufficient.
Problem
A proposed saving can name a change at one place: fewer fields, fewer checks, fewer people involved, a quicker response or less time at a machine. The receiving result can depend on something outside that view. Removing a comparison can create rework; shortening a local queue can enlarge another; outsourcing can reduce the requester's work while increasing verification and support.
The reverse mistake is to preserve every familiar action because it might have value. An obsolete output continues, a recurring error is corrected forever, and a capable person keeps performing an operation that could be removed or supplied more simply. Merely listing possible losses does not identify a change worth making.
The needed answer joins diagnosis and construction. It explains which work is avoidable under the purpose and conditions, how a proposed intervention changes those conditions, what the complete alternative costs and preserves, and what subsequent performance establishes.
Forces
Solution
Use a result-preserving comparison of whole obtaining ways. Recover what the work is for, locate a consequential avoidable demand, construct changes that address it, and compare those changes with continuing the present way. Introduce the selected change under adequate conditions and judge what happens in later work. Retain the present way or stop the improvement effort when a proposed reduction is not worthwhile.
A recall cue is:
Keep the needed result; remove unneeded work; prevent avoidable rework; distinguish waiting from processing; change a cause worth changing; carry the change into the next performance.
This is an order of explanation, not six compulsory gates. Enter at the question whose answer is missing. A new technology can remove rework, waiting and processing together. An independently supported change can proceed while another cause remains unknown. The dependency is between the grounds and the decision that uses them, not between positions in the mnemonic.
Establish what the lighter way must obtain
Name the result in its receiving use: who or what will use it, for which action or judgement, under which conditions and over what relevant horizon. Include a later use when the proposed saving can affect it. A result can be usable equipment, a decision, knowledge, an explanation, a capability or a maintained condition. These are not interchangeable successes.
Recover the conditions the recipient relies on. For a dataset, values may be insufficient without units and source identity. For a maintained service, ordinary output may be insufficient without recovery from a foreseeable failure. For learning, a correct answer from a helper may be insufficient when future independent performance is the purpose.
Obtain the professional judgement that establishes those conditions; cost reduction does not establish them.
Follow a proposed deletion through its actual uses. Ask what would become unavailable, unreliable or more burdensome if this contribution ceased. Reuse a supplied contribution where it remains sufficient.
When a local action belongs to several larger undertakings, B.1.5.EW helps recover those encompassing uses; B.1.5.RS evaluates a replacement against what each affected whole relies on.
Decide which work can cease. To omit an operation on one job, establish what each affected recipient can obtain on that job without it. To remove the operation from a recurring method, say which jobs would proceed without it. One eligible job in which a recipient would lose a required result rules out that proposal as written. A narrower change may still work if the performer can recognize the exceptions in time and keep an adequate way to serve them. Follow an indirect use when its loss could change this decision.
Separate three conclusions:
- The result or contribution is no longer needed: ending its production may be the useful change.
- The contribution is needed, but this action does not supply anything further: remove or replace the redundant action while preserving the contribution.
- The contribution is needed and this action currently supplies it: construct another adequate way before withdrawing it.
A.11.OP supplies the contribution test for one proposed mandatory requirement. A positive contribution does not establish that its burden is worthwhile; C.11.DUA helps resolve that further question.
An existing obligation remains binding until changed by its competent authority or agreement. A practitioner may propose its removal without being entitled to stop complying.
Recover where the effort and elapsed time arise
Use a recent representative occurrence, an already available account, or another source sufficient for the decision. Follow the work from its initiating need to the recipient's usable result. Recover consequential actions, handoffs, corrections and support.
Include unsuccessful or exceptional occurrences when excluding them would make the proposed change look better than its actual use.
Distinguish active work from elapsed time. A request can spend two hours in processing and three days awaiting an input or decision. Do not add those days to labour hours or treat elapsed-time reduction as a reduction in labour without a further argument. Also distinguish resource occupancy from attention: a device can remain occupied while its operator does other work.
Locate the condition that makes an occurrence wait. It might lack usable input, permission, a decision, an available resource or the result of another operation. Trace a consequential dependency until there is a condition someone can change or a justified limit. If the condition is shared, examine the combined demands: a proposed earlier slot is not available merely because each request separately fits it.
Collect only the detail needed to distinguish the changes. An existing timestamp and a conversation with the responsible participant may settle the question. If measurement or inquiry would be substantial work, C.11.DUA compares its attainable decision contribution with its whole burden.
Diagnose the avoidable demand without calling every difficulty waste
Use the distinction that changes the proposed intervention. The following cases are common entries, not an exhaustive partition:
Examine configuration-related rework. If an accepted instruction requires one edition while the performer uses another, recover which edition was required, how the performer obtained the used one, and where the mismatch could have been prevented.
An authoritative source, usable retrieval, recognition of the required edition and a proportionate check can form a preventive change. Its added effort belongs in the comparison.
A failed hypothesis is not automatically failed work. If the intended research result was a distinction between explanations, a well-grounded negative answer can supply it.
Likewise, repeated practice can produce a capability even when it produces many similar outputs.
Recover the intended result and the actual contribution before classifying that effort as avoidable. Repeating an invalid measurement may be wasteful; learning why it was invalid may supply valuable knowledge.
For waiting, distinguish an unavailable prerequisite from an ordering problem among already feasible actions. Making an earlier decision, preparing an input in advance, changing batching or allocating an available resource can shorten the path.
Each candidate needs its receiving conditions. Starting more items can increase contention; moving a compulsory decision to the end can create extensive unusable work. A shorter first completion and greater sustained throughput are different improvements.
For processing, construct a different obtaining operation. Explain what the tool or performer receives, what it does, how the recipient obtains an adequate result and what happens outside its supported cases. Reuse, delegation and automation are possible means.
Construct an intervention through the condition it changes
Work backwards from the desired difference: what would have to be different for this effort to cease or decrease without losing the required contribution? Then work forwards from a feasible action: what changes, how does that affect the operation or dependency, and why does the needed result remain obtainable? Reconcile the two explanations.
For example, repeated copying between incompatible records may cause mismatches. Removing one copy, using a common interpreted source, checking correspondence, or redesigning the receiving operation are different interventions. Compare how each would preserve the required contribution and what support it would need. If the proposed action merely moves the copying to the recipient, it has transferred work rather than removed it.
Where a causal claim matters, use C.28.CM to state the intervention and outcome, develop the relevant causal explanation, challenge plausible rivals and preserve its limits. Several causes can coexist or interact. A useful intervention can affect several at once; a single root-cause label is not a required result.
Keep causal explanation and protection distinct. Suppose a wrong edition could arise either from a stale local copy or from an incorrect retrieval index. A supported comparison with the authoritative edition may protect the next use under both explanations.
Use that protection when its conditions and burden warrant it. Decide separately whether distinguishing the causes would change a worthwhile preventive action. The protection's success alone would not prove which explanation was true.
Construct a complete candidate only as far as the decision requires. Include an unchanged adequate way when it remains an option, a small correction when available, and removal or replacement when those could win. C.39 develops a missing operation or join; C.38 constructs comparable whole ways when the proposed alternatives are disconnected ingredients. An existing adequate alternative needs no new construction.
A changed purpose requires a new comparison basis. Ending an unwanted report and producing the same report faster do not obtain the same result.
First establish the remaining needed use and the authority to change the commitment; then compare ways that serve that use. Do not claim preservation by weakening the result.
Compare complete alternatives over the relevant horizon
Carry each candidate through construction, transition and subsequent use. Include what makes it obtainable: competent performance, data, access, equipment, permissions and support. Explain how exceptions are detected and handled, how the new means are maintained, and what happens if they are unavailable. Use existing adequate arrangements rather than inventing a new process for each concern.
Compare the whole burdens that can change the choice. They can include preparation and integration, learning and verification, ordinary execution, correction, support, maintenance, recovery and withdrawal.
Determine who bears each burden and when. Separate a transfer accepted by its bearer from an unprovided condition. A recipient's capacity limit can make a locally cheap alternative infeasible.
Use compatible quantities where a calculation is useful. Compare the ways under the same relevant external demand and commitments. Where the Method itself changes the number or mix of jobs, recover the jobs, repeats and exceptions that would arise under each arrangement and compare their totals. A common set of jobs answers a question about effort per comparable job. A sufficient bound may settle either choice. For a fixed number of repeated uses, a simple account can compare initial effort plus recurring effort, with separately stated exceptions and uncertainty.
Do not add labour time, elapsed delay, money, failure severity and future capability into an unexplained total. C.11.CRC develops a comparison when several contributions and shared constraints make those interactions consequential.
Compare the joint intervention, not the sum of imagined independent gains. If a common source both removes copying and prevents the resulting correction, distinguish the effort categories so the same minutes are not saved twice. Include remaining failures; complete elimination is a claim needing grounds, not a consequence of installing the change.
Check the horizon and the cost of reversal. An expensive preparation may be justified by many future uses and unjustified by a single remaining occurrence. A simpler manual way can win when volume is low, conditions change rapidly or exception handling dominates.
When the demand or horizon is uncertain, establish whether variation within the supported range could reverse the choice and what would do so. Obtain more information only when it can change a worthwhile decision.
At a later decision, compare what each feasible continuation would require from the current situation. Already incurred effort is not charged again and does not by itself justify continuing.
It remains in any account of the undertaking's total effect. Include new costs of withdrawal, restoration and lost capability; an available tool and one needing construction are different starting conditions.
Preserve required margins and later ability. Unused capacity can support recovery or demand variation. Human practice can preserve a needed capability. Deleting either may reduce present activity while increasing future vulnerability.
Establish the needed reserve or capability from its use; do not preserve every idle resource or repetition merely by naming possible future value.
When one candidate preserves the same needed conditions and is no more burdensome on any relevant coordinate, with at least one strict reduction, it is preferable for this comparison.
When gains conflict, make the trade-off explicit, such as accepting more setup effort for earlier availability, or retain alternatives until the person responsible for that choice can settle it. A binding safeguard is not converted into an optional cost by this comparison.
Choose a supported change, a bounded trial, a narrower use, continued operation, or no further improvement work. A new safeguard can be part of the lighter whole. Worth is not equivalent to the fewest operations.
Introduce the change without substituting a document for practice
Make the selected change usable where the next affected work occurs. Identify the performers and support providers; obtain the necessary permission or agreement, access and available resources. Resolve simultaneous demands rather than assigning the same scarce support twice.
If the work remains beyond current capability, obtain an adequate performer or develop the missing capability. If capability is already sufficient, an additional course is not required.
Use A.3.1.MI to distinguish the proposed way, introduction strategy, preparation, introduction and later work. Keep an adequate existing instruction where it suffices.
Where an error recurs, change the conditions that allow it at the point of use: withdrawing an obsolete copy, supplying the correct source, making an exception route usable or providing the missing decision can matter more than another announcement.
Select a trial when its attainable result can alter adoption or scope. Choose cases that expose the disputed condition, including consequential exceptions, and preserve the necessary result during the attempt. Establish when to continue, restrict, revise or stop. A known adequate correction may be introduced under the applicable rules.
Plan recovery appropriate to the subject. A previous software configuration may be restorable; lost trust, interrupted service or a person's acquired experience cannot be undone by replacing a file.
Use limited exposure, retained alternatives, parallel provision or another suitable protection where reversibility is limited. Include these costs in the choice.
Judge later performance and return to the consequential question
Compare the intended difference with what occurs in the receiving work. Was the needed result obtained, with its safeguards and support? Did whole effort or relevant delay decrease, for whom, and under which conditions? Include transition work and exceptional cases at the scope of the claim.
The decision about work still to come uses the cost comparison in §4.5. Keep the grounds for two conclusions distinct. To retire a step across a class of work, establish what happens to the need it served. The output may have no further use. A change at the source may prevent the difficulty that required correction. If a recipient still needs what the step supplied, establish how they will obtain it throughout that class. A checked property of a common source may settle that question. Where exceptions are possible, establish how the performer recognizes them before reliance and what the recipient can then use. A hold is adequate only for a use that permits it. To claim a saving in ordinary use, compare the work that actually recurs, including exceptions and work transferred to others, over the period claimed. The replacement can preserve every required result and still cost more to run.
Distinguish a useful result from its explanation. Observed improvement can support keeping a local arrangement while leaving causal attribution uncertain. If the claimed advantage depends on attributing the effect to one intervention, C.28.CM supplies the further causal question. A.3.1.MI keeps introduction, contextual support and the later performance claim separate.
Return to the part whose grounds changed. A newly required output reopens the purpose comparison. Recurrent mismatch reopens the preventive construction. A longer support queue reopens the whole-burden comparison. An unhandled input reopens the permitted scope or fallback. Preserve earlier answers that hold; one exception does not require rebuilding every unchanged part.
Keep the change when the supported gain remains worthwhile. Restrict or replace it when only some uses remain adequate. Restore a suitable earlier way when the new one loses a needed condition. End this improvement undertaking when its useful result has been obtained or further work is not worthwhile. Review again on an action-changing event, such as changed demand, lost support, a new receiving use or evidence that the preserved condition no longer holds.
Continue development without turning economy into its only purpose
A worthwhile reduction can release attention, capacity or resources. Establish what has become available before promising it elsewhere. Then ask whether that changed position makes another useful problem, investigation or capability attainable. C.40.NP develops this continuation from an obtained result.
The next undertaking may change the problem, create knowledge, expand a capability or improve the way future changes are obtained. It can require more current work. Judge that contribution in its own right. This method governs economy of obtaining a result; it does not define all development as obtaining the same result more cheaply.
Archetypal Grounding
These are constructed cases. Their numbers and observations are premises for inspecting the reasoning, not evidence that the method has produced a measured benefit in practice.
A repeated research packet: removal, prevention and automation
A research group must supply a weekly packet of ten matched observations for a comparison. The recipient requires correct source identity and units, a usable comparison, and an explicit hold for unresolved rows.
A separate presentation duplicates the accepted table. The owner reviews the agreed deliveries for the eight-week period. This is the only recipient entitled to the presentation. The recipient and the work owner have authority to change its form and agree to the table in fulfilment of that delivery. The owner can then stop preparing slides.
The group traces recurring correction to mismatched source editions. It can obtain the authoritative edition and check the row identities before preparing the comparison. This adds a small action but prevents the specified mismatch. The validity of the subject measurements remains a separate required condition.
Three complete alternatives fit these stipulated requirements. All figures are staff minutes, include the recipient's changed work where any, and assume the stated weekly demand:
The present 100 minutes comprise 60 minutes of necessary preparation and verification, 20 of duplicate presentation work and 20 of correction of the specified mismatch. These categories do not overlap. The revised manual way keeps the 60, adds 10 for the identity check and 5 for support of the changed arrangement, and removes the duplicate and specified correction.
The tool uses the qualified source identifiers to prepare the comparison and hold unsupported rows. Its 45 minutes include 25 for preparation and verification, 10 for exceptions and 10 for maintenance.
Its 300 initial minutes include configuration, checking, instruction and transition. The stipulated comparison grants neither candidate new authority to approve unsupported observations. The same substantive result, controls and access conditions remain.
For eight remaining weeks, the revised manual way uses less total effort than the tool, despite more effort in each ordinary week. At twelve weeks the totals become 1,200, 940 and 840: the tool then wins under the same assumptions.
Its extra initial burden over the revised manual way is 260 minutes and its weekly saving is 30, so their order changes after 8⅔ weeks. Uncertain duration can therefore matter more than a small refinement of weekly estimates.
The tool was a legitimate candidate from the start: one integration can remove duplicate preparation and a mismatch mechanism together. It did not have to wait for a calendar sequence of earlier improvements. Conversely, the tool's attractive weekly figure did not make it the eight-week choice.
The group chooses the manual change for the eight-week undertaking. The work owner makes the authoritative source accessible, replaces the old preparation instruction and provides a responsible contact for unresolved rows.
In a constructed first-use variation, the source is temporarily unavailable. The agreed response is to hold the affected packet or use an independently adequate fallback, not reuse an unqualified copy. That exception's burden would enter the subsequent comparison.
Suppose the next four packets meet the agreed conditions, including delivery within the permitted window after the temporary hold. Their recurring effort totals 300 minutes; the exceptional source recovery adds 30, and initial preparation took the stated 40.
The resulting 370 minutes include the recovery instead of hiding it. The record supports four acceptable packets with that effort. A claim of 30 minutes saved additionally relies on the comparability of the 400-minute baseline; the record alone does not establish that counterfactual or the remaining weeks.
The group can keep the change within these supported conditions while continuing the ordinary comparison.
Consider a group with the same four weekly records whose month-end reviewer also needs the reasons that observations were excluded. Only the presentation gives those reasons. The four packets still support acceptance of the weekly work. The month-end use rules out dropping every presentation. If preparation staff know which packets go to month-end review, they can retain the presentation for those packets. To retire it there too, the group can put the reasons in the authoritative table and establish that the reviewer can retrieve them for the edition under review. The work of recording and using those reasons belongs in the revised comparison.
If correction persists because some supposedly matching identifiers have different meanings, the original prevention premise fails for those rows. The group reopens that interpretation. It can still omit the presentation for packets where the presentation serves no further use.
If the recipient later needs an annotated presentation for a new audience, the deletion decision is reconsidered for that use; the source-identity improvement can remain.
In a separate constructed variation, the integrated tool has already been built, is available and preserves the same required conditions. Four weeks remain. Continuing with it requires 180 minutes; changing to the not-yet-introduced manual way requires 40 minutes of transition plus 300 minutes of recurring work, or 340 in total.
The original 300 minutes of tool construction remain part of the undertaking's historical effort, but the current choice is not 480 against 340. Any new withdrawal, recovery or capability loss would enter the future comparison separately. The initial eight-week comparison and this later choice differ because the available means and remaining work differ.
A service delay is not its processing time
A translation service spends two hours preparing and checking a document, but completed requests wait three days for an authorized release decision. A proposed tool saves twenty minutes of preparation. For the request at issue, the next available authorized release slot is three days away. Twenty minutes less preparation does not create an earlier slot, so this change alone cannot provide next-day delivery.
The service follows the delay to its condition. A qualified release decision is required; only one authorized reviewer currently has a weekly slot. The alternatives include keeping that arrangement, obtaining an earlier decision slot with the same reviewer, or qualifying and authorizing another reviewer. Omitting the decision does not preserve the result.
An earlier slot is useful only if the document and necessary evidence are ready then and the reviewer can serve the combined requests. Moving a slot on one request's plan supplies no extra capacity.
If these conditions can be met, it can improve elapsed delivery time with little change in labour. If they cannot, the preparation tool may be worthwhile for effort, but it does not solve the promised delivery-time problem. The receiving commitment needs another complete arrangement or a truthful revision.
The service can introduce a supported scheduling change while studying why the queue formed. It can also compare a combined scheduling and preparation change. The common question is what obtains the recipient's adequate document; no universal order makes one intervention wait for the other.
An answer, a capability and a useful negative result
A learner practices interpreting unfamiliar statistical results. A tool can produce a correct explanation of today's example.
If the needed result is only that explanation, suitable tool use and checking can be an adequate obtaining way. If the need is later independent interpretation, the supplied explanation does not establish the learner's capability.
The practitioner separates practice that develops the missing operation from clerical copying that serves neither learning nor later use. A competent teacher can replace the copying with supplied material while retaining the learner's interpretation, feedback and a consequential variation.
Whether this changed practice develops the intended capability is a learning question requiring its method and evidence. The elegance comparison keeps that result visible and includes the teacher's preparation and feedback effort; it does not infer a learning mechanism from repetition alone.
Similarly, a research group may perform a careful test that rejects its preferred hypothesis. Removing the test because it produced no usable product would discard the knowledge it was meant to obtain. Repeating the same test without a new discriminating purpose can have a different value. The decision follows the receiving knowledge or capability, not the visible count of outputs or the emotional experience of success.
Bias-Annotation
The Gov, Arch, Onto/Epist, Prag and Did lenses apply. Scope is transdisciplinary for the comparison and change construction, with subject-specific mechanisms, authority and evidence.
The method foregrounds avoidable burden. That can bias attention against reserves, inquiry and capability development whose contribution arrives later. Sections 4.1, 4.5 and 5.3 keep those results and their grounds in view. The opposite bias is speculative preservation of every activity; a merely imaginable future use does not establish its worth.
Conformance Checklist
Inspect an explanation or application claiming this Method:
Common Anti-Patterns and How to Avoid Them
Automate the unwanted output. A quicker duplicate consumes work. Recover the receiving use through 4.1 before treating production as necessary.
Call all repetition waste. Similar practice attempts and negative tests can obtain capability or knowledge. Use the result distinction in 4.3 and the contrasting case in 5.3.
Turn the mnemonic into stage gates. Waiting for every earlier diagnosis can postpone a well-supported combined intervention. Follow the grounds and dependencies in 4.4.
Claim the local saving as the whole gain. A requester stops entering data while support staff reconstruct it. Follow the transferred contribution and its bearer through 4.5.
Install a rule and declare the work changed. A correct instruction cannot provide missing access, a performer or an available decision. Carry the change into practice through 4.6 and examine the receiving work through 4.7.
Consequences
A practitioner can stop an unnecessary contribution, prevent recurring avoidable work, distinguish an elapsed-time opportunity from a labour saving, or select a worthwhile replacement without assuming automation is preferable. A warranted decision to keep a manual way is as legitimate as introducing a tool.
The price is attention to interactions and later use. An elaborate accounting exercise can exceed the attainable saving. Reuse sufficient existing knowledge and compare only distinctions that can change the decision. More difficult uncertainty can justify a limited trial or a narrower conclusion.
A retained protection or learning activity can make a candidate look less minimal while making the whole undertaking less burdensome or preserving a necessary later result. Savings remain conditional on demand, support and use; they can disappear when those conditions change.
Architectural Rationale
The governing invariant is the needed contribution in its receiving use, not the visible activity that currently produces it. This makes removal and replacement comparable without treating all outputs as desirable or every familiar action as indispensable.
The method joins questions whose answers constrain one another. Purpose determines which losses matter. The observed operation and its causal explanation constrain interventions. Whole-burden comparison determines whether an intervention is worth obtaining. Introduction and subsequent performance determine whether the proposed gain is available in practice. A change at any of these points can justify a bounded return without imposing a universal sequence.
A.11.OP tests one proposed requirement's contribution. C.11.DUA judges the worth of advice and inquiry. B.1.5.RS determines which encompassing uses tolerate a replacement. Those results can be sufficient for a local question. The present method adds the connected construction from an avoidable demand to a worthwhile change and its continuation in work; it does not replace those narrower judgements.
The commonality across people, organizations and technical systems lies in this reasoning about obtaining ways. Their causal mechanisms, permissible interventions and means of recovery remain different. A shared comparison does not establish a shared mechanism.
SoTA-Echoing
Practice question and selected line. How can a practitioner reduce avoidable work while preserving the result and conditions over the relevant horizon? The selected answer is the conditional synthesis in section 4: purpose and use, diagnosis, intervention construction, complete comparison, introduction and later reconsideration. It is a method proposal grounded in the comparisons below, not an empirical claim of universal superiority.
Toil reduction is an alternative. Google's 2018 SRE treatment already compares elimination, simplification and automation, including maintenance and indirect costs. Adapt its whole-burden and cause-removal reasoning in 4.3–4.6 to the broader receiving-use question. Use the domain practice when it already supplies that answer.
For the broader question, sections 4.1 and 4.3 also compare changes when knowledge or capability is the needed result and purposeful repetitions contribute to it. The trade-off is extra diagnosis where those distinctions matter; this synthesis offers no established gain from repeating them when they are already settled. Eliminating Toil.
An improvement cycle can already suffice. IHI's PDSA guidance connects prediction, a bounded test, observation, learning and adaptation; its implementation guidance distinguishes testing from changes in ordinary provision. Adopt that separation in 4.6–4.7. Adapt the earlier construction in 4.1–4.5 when the uncertainty is which demand to remove or which complete alternative to test.
PDSA can handle combined changes and unintended effects. If an existing improvement practice already carries the needed construction and comparison, use it without another pass. Reopen the selected synthesis if an equally usable simpler treatment preserves these distinctions at lower burden. Testing Changes, Implementing Changes.
Subtraction deserves consideration, not automatic preference. Adams and colleagues' experiments support including subtractive possibilities explicitly when people generate changes; they do not show that removal generally outperforms addition. Later work reports task and population variation and further bounded replication.
Adapt the candidate search in 4.4 to include ending or removing work; reject subtraction as the selection rule. An added preventive check can win, as 5.1 shows. Reopen the search cue if evidence for the reader and task makes it ineffective or unnecessarily burdensome. Adams et al., 2021, Juvrud et al., 2024, Fillon et al., 2025.
The worked cases establish consequences of their stipulated premises. They do not validate expected savings, a universal human or AI bias, or the comparative effectiveness of the whole method. A stronger practical claim needs evidence from its conditions.
Relations
- A.3.1 supplies Method identity and the distinction from descriptions and performances.
- A.11.OP tests one proposed requirement's contribution; C.11.DUA resolves consequential questions of advice or inquiry worth.
- B.1.5.EW recovers encompassing work; B.1.5.RS evaluates the compatibility of a constituent replacement with its affected uses.
- C.28.CM constructs and challenges a causal explanation when that explanation changes the intervention or attribution.
- C.39 develops a missing obtaining operation or connection; C.38 constructs whole alternatives when the available ingredients are not yet comparable.
- C.11.CRC compares contributions with shared constraints and consequential interactions.
- A.3.1.MI supplies introduction into practice and judgement of what to keep or change.
- C.40.NP develops a further worthwhile attainable problem from the position reached after a result.
C.42:End
Connect Observation, Choice and Action in Control
Type: Method pattern Status: Stable Normativity: Guidance within the stated applicability
Problem frame
Use this pattern when observations, rules for influencing a system, performed actions and their consequences have to be connected to answer a question about control, but the separate accounts do not yet provide that answer. You may need to explain what an existing arrangement does, including why it works badly, or decide how to continue and what to change. The connections matter in both uses; the required results differ.
A researcher can reproduce a recorded response with two different mechanisms and not know which observation would distinguish them. In practical work, a forecast can be sound while its proposed response arrives too late. A profitable change can consume the capacity needed for another commitment. A useful measurement can arrive when the action it would have changed is no longer available.
First useful move. Identify the consequential change or maintained condition and recover how the available arrangement influences it. Ask what information reaches which rule or participant, what action follows, how it affects the system and what can be observed afterward. Keep an unsupported connection as a question to resolve.
For explanation, work the proposed relations through to a consequence that could distinguish competing accounts. For a practical continuation, take the next consequential choice: explain what is currently known, what can be done, what each continuation could change and what would change the choice. Follow the unresolved necessary connection far enough to supply the receiving answer or identify the condition that prevents it.
What this buys. An explanatory use returns supported control relations, their consequences under stated conditions and any unresolved alternative. That can explain an unsuitable response without prescribing a redesign. A practical use returns what to do, why it is a suitable continuation, how it can take effect and how later information will change it. It can justify continuing an existing practice, taking a bounded action under uncertainty, obtaining information, or changing the goal, model, method or control architecture. Neither use requires a new document when the working explanation is already available.
The intended reader is a practitioner or researcher who can obtain the subject knowledge required by the question. The common Method organizes that knowledge; a claim about a machine, organization or organism needs the relevant subject explanation and evidence.
Use an adequate controller, plan or established response when it already supplies the needed connections. Apply A.15.7 to select the next action within an understood working Method. Apply C.32.LCA when the unresolved result is the construction or alteration of the control arrangement. Here the arrangement may remain unchanged: the missing result is a coherent continuation through its observations, choices and actions.
When explaining observed control, an attributed goal or internal model is a hypothesis to test, not an instruction addressed to the observed system. Explaining how a person, animal or organization acts and deciding what it ought to do are different receiving questions.
Problem
Useful answers are often developed separately. One account estimates the situation, another predicts consequences, another compares benefits, and another assigns work. Their assumptions about time, available actions, resources and purpose can differ. The reader then has to invent the connections that determine whether the answers can be used together.
The difficulty is not resolved by arranging their titles in a cycle. The practitioner needs to know why one answer is usable by the next, what to do when they disagree, and which earlier conclusion a new observation changes. Otherwise a repeated cycle can reproduce an unsuitable action, and an accurate local calculation can support an infeasible whole.
The same missing connections obstruct explanation. A report of a response does not reveal which information determined it; a stated rule does not establish that the corresponding action occurred; and different mechanisms can produce the same observed trajectory. The investigator needs to derive how a proposed observation-to-action relation produces its consequence and which changed condition would distinguish it from an alternative. The resulting conclusion concerns what the arrangement does. Whether to retain or change it additionally depends on the intended outcome and available means.
Forces
Solution
Connect the current situation, available information, the rule or choice that produces an intervention, its realization and its consequences. For an existing arrangement, recover the relations that obtain and derive what they do under the relevant conditions; compare alternative explanations where the evidence leaves them open. For a practical continuation, construct and compare what could be done through that arrangement or a changed one, and establish how the selected action can take effect. An explanatory result can supply a premise for the practical choice without selecting that choice.
Use the following operations where their results are missing. Return from a consequence to the observation, rule, execution condition or purpose on which it depends. Preserve a sufficient contribution when the returned question leaves its grounds intact. The order below is a logical unfolding for learning and attention, not a required temporal sequence.
Establish what is being controlled and what continuation is needed
Begin with the subject of the intended change and the systems that can realize it. A controlled system can be a machine, a person, an organization or another physically realized system. A controller can be embodied in one participant or distributed across several. These descriptions designate what the systems do in this relationship; they do not introduce essential kinds of system.
The change being organized can also concern work, capability, an applied Method or the model used in decisions. Establish separately what content is revised and which systems interpret it, act and produce the intended consequences. A model, goal, command and report are content used in that work, not additional physical plants.
Recover the present conditions from which action must continue. Include completed work, unfolding effects, resources already occupied, and commitments that constrain what can happen next. Maintaining the current practice continues its effects; it does not freeze the system.
State the result being sought, the conditions to preserve, the affected participants and the span over which the choice matters. These can be qualitative. An emergency response may need only a justified action until assistance arrives; an investment can require consequences over years. Establish whose decision can change a goal or commitment. Conflicting interests may require agreement, a legitimate governing decision or separate comparisons rather than one invented common preference.
A useful continuation is not always movement toward a fixed target. It can preserve an acceptable range, avoid a loss, keep several futures open, or investigate which outcome is attainable. When the goal is unresolved, obtain the missing purpose decision through the governing practice while identifying any bounded action that remains appropriate across the alternatives.
Recover control relations before judging their quality
Relate the change being considered to the means of influencing it. Recover both why an input or rule is set in relation to the relevant outcome and how its execution can affect that outcome. The rule can be embodied in a mechanical arrangement, physiological mechanism, program or agreement. It need not represent alternatives explicitly or employ a conscious chooser.
In an explanatory inquiry, the proposed regulated characteristic and its relation to the mechanism are claims to investigate. Compare what would happen under a changed reference, observation, available intervention or disturbance. This gives the account explanatory content beyond describing a trajectory after it happened.
Keep three questions available. Description recovers the relations that obtain, their participants and the consequential gaps. Explanation derives what those relations do under the stated conditions, including an unwanted result. Design compares changes that could make the continuation more suitable. A descriptive or explanatory inquiry can finish without prescribing a new arrangement. In design, an intelligible description of the incumbent supplies a comparison basis, not a reason to retain it.
Control can be open over the interval being examined. A controller may follow a fixed schedule or adjust its input to a measured incoming disturbance before the disturbance affects the process. Trace how it sets the input and how the input reaches the controlled system. If the controller also uses an observation of the resulting state to select a later intervention, that return supplies a feedback path. Both paths can operate in one arrangement.
Feedback operates when an observation returned from the controlled system can affect a subsequent intervention on that system. Identify the decision and the interval. Which different reading could change the action, its timing or the decision to continue it? Several readings can call for the same action. To establish what happened on a particular occasion, recover the rule or judgement actually used and the information that reached it. A fixed schedule and an observation-dependent rule can produce the same recorded command. Evidence of how the decision was made can distinguish them even when the command stayed the same.
For a claim about the whole arrangement, state the property at issue and how it follows from what the parts do. Work back from the property to the contributing parts and the connections it depends on. A rule may operate within every unit while their joint response depends on how their actions affect one another. For example, if two loads account for the entire circuit load, their instantaneous powers at one instant add to the total at that instant. Mean powers measured over the same interval add to its mean total. Two displays read at the same time may report different averaging intervals. Limits that apply to both loads in the same operating mode give an upper bound on the total. Whether both loads can reach their limits together is a further question. Where regulators share capacity or change one another’s inputs, follow the connections that affect the claimed response. A unit can correct its own backlog while its allocation follows a fixed schedule. Several units can also produce a common response through their interactions. To extend a conclusion to other operating conditions, establish which limits, decision rules and connections still apply. Use C.30.LCA to describe the control relations and C.32.LCA when their organization needs to change.
The subject of the decision, the information and interventions available to each participant, and the physical realization can have different boundaries. One device can estimate, choose, act and modify its means. Several participants can jointly supply one influence, and each can be influenced by the others. Recover the contributions and how they are connected before allocating them to parts.
A combined mathematical description can retain joint dynamics when a clean separation of states is unsuitable. Its coordinates do not themselves establish physical parthood.
For distributed control, preserve what each participant can know and do. Writing local rules as u_i(t) = pi_i(I_i(t)) makes their information dependence explicit; collecting those rules into one notation creates no shared observation, communication channel or authority. Compare their joint effects with shared resources, timing and constraints. The constructive work is obtaining the required cooperation.
Explain controlled self-change through its acting and changed sides
A continuation can change the means of later control: a model used for interpretation, a policy, a performer's capability, an allocation of work or the controlled system. Identify what is changed, what makes the change, why this intervention is selected or triggered, and how it alters later action. Then connect the resulting observations to the decision to retain, revise or undo that change where those alternatives are available. This extends the same account to changing control arrangements without assuming that their mechanism remains fixed.
Use A.12 to recover the acting participant and changed subject. Its Reflexive Split applies when both are distinct entity parts or subsystems of one containing holon, with established parthood. Shared hardware does not require that their functional contributions collapse, while different roles or successive versions alone do not establish different entity parts. An ordinary account can retain the demonstrated action and change when that stronger internal-parts claim is not available.
Reflexive Split answers who acts on what inside the whole. The control explanation additionally connects the change to its regulating rule or intended result and to the consequences for subsequent work. An identifiable updater can install an unsuitable setting. Temperature drift can alter the same device without being the intervention selected to obtain that result. Random generation of candidates can be part of controlled search when the generation, examination and retention of variants form the arrangement; determinism alone does not establish such an arrangement.
For adaptation, distinguish updating the estimate of the current situation, changing an estimated model, adjusting the controller and changing the subject dynamics. Direct adaptation can adjust controller parameters without first constructing a new plant model.
When a property depends on those parameters, examine the combined operation of acting and updating under the conditions of the claimed property. A result for a fixed setting answers that fixed-setting question; the technical Method must establish which updates preserve it. Development adds the further question of changed attainable problems and means, developed in §4.8.
Describe incomplete and unsuitable arrangements without inventing missing parts
Start from the incumbent's relations. An unexecuted command, a delayed observation, a response based on a stale assumption and competing local allocations can each explain a different failure. Retain the working contributions and recover the connection whose change could improve the receiving result. A globally unsuitable arrangement can contain useful local control.
A mathematically degenerate case is useful when the mapping is justified: a constant policy removes dependence on new observations; a zero coupling removes a specified influence. A missing observation in the evidence is not such a zero.
Likewise, a coupled nonlinear arrangement need not be a limiting case of one linear model to be describable here. Choose or reuse an adequate subject model and establish its relevant behavior.
Apply C.29 when a mathematical representation's choice, transfer or supported use remains unresolved. This preserves both a description of imperfect control and the stronger question of what should replace it.
Recover the control claim behind the field wording
Read control, management, regulation, governance or steering in the claim where it is used. First identify what is being discussed: a field of practice, its methods, the people doing the work, an arrangement, an intervention, a power to decide, or information about the result.
Then recover the subject being influenced, the condition or change at issue, the available information and means, and how they determine what happens. Preserve the established field name while making this relationship explicit.
In operations management, practitioners control how an undertaking produces and delivers its results: they connect demand, capacity, work, commitments and observed performance to changes in what is done. The word management does not exclude this practice from the common control construction. The discipline also studies and designs those operations. A calculation, forecast or inspection is a contribution to control when its result is used in that connection. The same distinction applies to strategic direction and to control at other horizons.
The organizational account must retain the conditions through which a decision can influence action: commitments, participation, rights, incentives and available resources. A right to issue an instruction and the means to have it carried out are different relations that can both be necessary.
Governance may concern how such rights, obligations and accountability are constituted or exercised; recover those acts and their effects rather than translating the word into a physical feedback mechanism. Management may also designate the people exercising a responsibility. Name those people when they are the claim's subject.
Likewise, an inspection called a control can establish a test result without selecting or realizing a corrective action. Identify the response that receives the result when the claim concerns continuing control. In translation, English control and Russian контроль are not interchangeable by default: the latter can mean checking or monitoring, while управление can name the wider arrangement. Preserve a source's intended meaning through the recovered relationship; do not replace every occurrence with one preferred English word.
The result is an ordinary, usable account such as: “The operating team changes job release using queue information, due dates, capacity and existing commitments.” Use C.30.LCA when the resulting observation, command or supervision relations need structural qualification; use the pattern that governs the remaining authority, commitment, measurement or evidence question. If the text already makes its intended meaning clear, retain it.
Build the account of the situation needed for the choice
Separate the situation from what the decision maker can know about it. Identify observations, their times, relevant past actions and the assumptions used to interpret them. An observation of an earlier state becomes useful for a present choice through an account of what could have changed since then.
Choose the representation by the decisions it must support. A point estimate can suffice when its remaining uncertainty cannot change the response. A set of compatible situations is useful when one action must remain acceptable across them. Probabilities are useful when the evidence and decision justify comparisons of expected consequences. A causal explanation, a production schedule and a physiological model retain different distinctions because different actions depend on them.
Recover what the retained information allows you to infer. If two possible situations require incompatible actions, a representation that treats them as identical is insufficient for that choice. The response can be to obtain a distinguishing observation, retain uncertainty explicitly, select an action acceptable in both, or change the requested outcome. More detailed modeling is only one possibility.
Check the information needed at the moment of action. Prediction can permit a justified action before a delayed measurement arrives, or compensate for some consequences of that delay. Retain the delay and the model assumptions that make this response possible. Preserve the effects of actual inputs and relevant history when they affect what the current observation means. Use C.30.LCA to recover an unclear control structure and C.29 to choose a suitable mathematical representation. Neither a mathematical vocabulary nor a detailed model supplies missing observations.
Construct possible continuations, including conditional future actions
When you cannot yet formulate relevant alternatives or the conditions that would distinguish them, B.5:4.1.2 develops that basis from available material, a subject relation and a revealing consequence or trial. Use that construction here. An unknown condition of a described continuation differs from not yet having a useful continuation to describe; a classification of conditions alone cannot supply the latter.
Construct alternatives from the available situation and from what must be obtained. Reason backward from the required result to the contributions it needs, and forward from available means to the results they could produce. Where the accounts fail to meet, obtain the missing contribution through C.39, revise the proposal, or retain the incompatibility as an explicit limit.
A recommendation or evidence demand can leave that alternative partly unspecified. Recover what following it would require, who could perform the work and which receiving result it could improve. "Validate the model" may hide a study that finishes after the response window, while a narrower check could support a useful action.
Use C.11.DUA:4.1–4.4 when this connection is missing: recover the demanded work, compare its attainable contribution and burden, and return the warranted advice. Bring the resulting proposal, sufficient present answer or impasse into this control account. A binding requirement continues to constrain the available continuations until its authority or agreement changes it.
A continuation can contain an immediate action, a planned sequence, a rule that selects later actions from later information, or a change that makes new actions available. These have different uses.
A fixed sequence is adequate when relevant future differences do not alter it. A conditional policy is needed when later choices should depend on information not yet available. Replanning after each observation can implement such responsiveness, but a present comparison must also anticipate that future responsiveness when it changes the value of the action chosen now.
Include action that obtains information. Its consequences can be both physical and informational: a trial can produce an item, expose a failure mode, consume capacity and change subsequent choices. Explain which possible observation would change which later action, and whether that action will still be available. This is the constructive connection between exploration and control.
Avoid demanding a complete tree of all futures. Begin with contingencies capable of changing the decision, the constraints that protect continuation, and an adequate response to a remaining range of uncertainty. A bounded robust action can be preferable to a costly attempt to identify the exact situation. Conversely, an inexpensive discriminating observation can avoid an unnecessarily restrictive response. The comparison in C.11.DUA determines whether further inquiry can earn its burden for this receiving decision.
Compare whole continuations from the starting conditions
Make alternatives answer the same question. Align the starting conditions, time horizon, outcome descriptions and resource assumptions. Retain differences that matter to affected participants. Apply C.11.CRC to construct what a finite change adds relative to the current configuration, and C.11 to select among the warranted options.
A positive contribution relative to the present configuration is not yet a choice among available continuations. Another attainable use of the same means may be better. Conversely, a locally attractive component can fail when it requires unavailable complements, conflicts with another action or removes a later correction. Compare the connected consequences, including transition work, interactions and remaining choices.
The work of steering also changes this continuation. Further reasoning can change what an adviser can infer; presenting advice uses the recipient's attention; deciding to change can reserve means or establish a commitment; performing the change affects the subject.
Recover the contribution and burden of the move being considered. C.11.DUA:4.2.1 explains this comparison when optional advice to change a way of working leaves it unresolved. The same participant can perform several of these actions, and an existing sufficient answer can make further reasoning unnecessary.
This distinction changes which interventions are worthwhile and when. Cheap computation can justify frequent consideration while interruption and transition costs favor less frequent advice or change. A warranted continuation may retain an idea for a later useful occasion, make a smaller change, obtain a discriminating observation or keep the current practice. Compare these possibilities against what they enable for the recipient over the relevant horizon. Section 4.5.1 constructs their timing and shared-resource relations. Reuse an adequate comparison or established response where it already supplies this answer.
Value information through the decisions it can improve, together with any separately justified use of the knowledge. For one decision maker with a specified probability model, a fixed finite set of available actions A, integrable utility u, and current information I, the gross instrumental value of a signal Y can be written as:
$$ \mathbb E_Y\negthinspace\left[\max_{a\in A}\mathbb E[u(a,\theta)\mid I,Y]\right] -\max_{a\in A}\mathbb E[u(a,\theta)\mid I]. $$
The expression compares choosing after the signal with choosing without it. It assumes the signal can be ignored and the former actions retained. Acquisition, interpretation, delay and displaced opportunities enter the net comparison. If the experiment changes the system or available actions, compare those changed continuations rather than subtracting a price from an otherwise unchanged formula.
The existing information matters. Two signals that are useful separately can repeat one another; a weaker isolated signal can be more useful alongside what is already known. Brooks, Frankel and Kamenica's comparison of signals provides the source basis for this dependence. Entropy reduction is not by itself a substitute for the receiving decision's gain.
Keep feasibility, outcome preference and assurance distinct in the comparison. An option can be preferred but unavailable; an admissible option can be inferior; a feasible finite plan can leave no acceptable continuation. Use the strength of calculation or evidence needed for the conclusion. A local estimate supports a local choice under its assumptions; a safety guarantee or optimality claim requires the relevant stronger construction.
Connect choice to an effect that can occur in time
Follow the chosen continuation through the participants and means that must realize it. Establish what action is needed, who or what can perform it, what resources and permissions it uses, and when the effect is needed. The calculation, communication, agreement and execution consume time and capacity. Include them when they can change the option's availability or result.
Check what each action needs from a shared machine, specialist, reserve or decision maker. When one performance serves both commitments, count the time it occupies each required resource once. Work that must be done separately consumes time for each performance. Allow for the reserve needed for later action. A lower bound that exceeds capacity can rule out a deadline. A feasible schedule establishes possibility under its stated durations and access conditions. For a promise across uncertain conditions, show how the authorized participants can meet the deadline throughout that range, using information available before their commitments. A bound can establish this without fixing the eventual schedule.
Choose the organization of decisions where it is unresolved. One rapid feedback response may protect the system while a slower plan is revised. Alternatively, the relevant decisions may need to be made together. Different timescales do not by themselves require layers, and organizational titles do not establish update rates. Use C.32.LCA to construct or change the observation, decision and intervention arrangement when those relations are the limiting condition.
The effects of control can also change its future means. A command can reserve a resource, teach another participant a response, alter an incentive or expose a vulnerable interface. Include a strategic response by another participant when it can alter the assumed action-to-result relation. A participant's agreement to a proposal and their subsequent performance need explanation; neither is supplied by treating the person as a technical actuator.
When unauthorized intervention is a credible influence, identify which observations, selection rules, commands, resources or recovery mechanisms it could change. Compare protection by its effect on useful control as well as on the unwanted influence. Restricting access can also remove a necessary observation or delay an authorized response.
Specialized security and institutional methods supply the threat model and protective construction; the common question is whether control obtains the intended consequence under those changed conditions.
Coordinate recurring work and separately timed actions
Determine which temporal relations make the continuation work. A contribution may continue through an interval, recur at selected times, respond to an observed condition, or calculate its next useful moment. Identify the receiving action and the reference it needs: an order, a permitted interval, a deadline, a phase relative to another event or a recurring cadence. Use C.27.TA when the bearer, reference or temporal claim is unclear. A common repeating grid is useful only where the work needs one; distributed control can rely on bounded delays and local timing without synchronized clocks.
Choose how observations and responses are initiated from their useful effect and burden. A periodic review sets the next planned opportunity to examine a condition; whether it will detect that condition depends on the available observations and the review. The complete response also includes obtaining information, interpreting it, deciding, communicating and acting.
A condition-triggered response depends on a functioning detector and enough capacity to handle its notifications. Computing the next useful observation time is another option when a suitable prediction is available.
Combine these arrangements where their contributions justify the cost. Specify what continues between updates, which events take precedence when they coincide, and when an update changes the reference used by later work. Neither the word feedback nor the control mantra selects one universal schedule.
Construct the coupling that is required. Keep each contribution's needed timing while relating its preparation, execution and recovery to the other contributions and their shared means.
For periodic work, the period says how often an event recurs; its phase says where it occurs relative to a reference. An action at another phase need not wait for the next regular event. Conversely, adding that action must leave the resources and time required for the regular continuation.
Work backward from the required effect to its preparation and forward through recovery to the next required contribution. Change the means, allocation or commitment when those demands do not fit; changing the event's required time merely to fit an available routine changes the task.
When the combination fails, locate the lost relation. An individual action can be late while the recurring cadence remains correct; the extra action can instead delay the next regular contribution; or a late observation can make both appear jointly on time against an obsolete reference.
Correct the affected preparation, execution, information delivery or update rule. Preserve the intended reference while it remains appropriate, rather than treating an accidental late event as a new starting point.
When the reference or desired cadence changes, revise the dependent timing together. A reset is an explicit change to those dependencies, not an assumed consequence of every event.
Test the variation the intended use needs. If an extra action must be possible at several phases or under different arrival conditions, examine those required alternatives and the continuation after each. Repeating one successful combination establishes less. The common result is an arrangement that preserves its required temporal relations while allowing the warranted variation. Its realization can use learned bodily coordination, software scheduling, machines or organizational agreements; each needs a constructive account and qualification.
Use subsequent evidence to revise the relevant dependency
Observe what happened at the connection that matters to the next decision. Sometimes the result alone is enough. When it is not, distinguish the issued instruction, performed action and resulting change so that the next move addresses the relevant cause.
Compare the observation with what the continuation relied on. A mismatch can arise from changed conditions, an inadequate account of the state, a false causal expectation, unavailable execution or interference. Investigate the alternatives that lead to different responses. A bounded safe continuation may be available before the full diagnosis; complete explanation is not a universal prerequisite for action.
Use C.11.DBR to reconsider a decision whose relied-on basis has changed. The return depends on the failed dependency:
- When the purpose and causal account remain suitable but the situation has changed, update the estimate and select another action or plan.
- When performed action has different consequences from those expected, examine the observations and causal account; revise the model or Method where that difference warrants it.
- When execution is unavailable, change the allocation, provision, timing or commitment, or obtain the missing means.
- When required timing is lost, use §4.5.1 to distinguish action placement, recurring interval, information delay, shared-resource interference and a changed reference; repair the relation on which the continuation depends.
- When the result is no longer wanted or an accepted constraint has changed, return to the participant or practice that can revise the purpose.
- When the existing arrangement cannot obtain or use the required information and action in time, reconsider its architecture through C.32.LCA.
These are alternatives to distinguish in the explanation, not obligatory layers of escalation. A goal can change without prediction error; a better controller can be needed for an unchanged goal. Revise dependent conclusions and preserve those whose grounds remain valid.
These returns can also change the means of further action. When the receiving work develops knowledge, Methods or capability and chooses subsequent attainable problems, connect this control account to C.40.OD through §4.8. A bounded control result can close while that development continues; a completed result can provide the new possibility, without an execution failure to trigger it.
Recover specialized theories without flattening their differences
The common account makes theories comparable through the questions they answer. To use a specialization, identify its representation of the system and information, available actions, outcome criterion, execution assumptions and warranted conclusion. Preserve what its additional assumptions make possible and what it leaves to another contribution.
Compact improvement, decision and learning cycles. First establish whether the source prescribes a way, offers an explanatory heuristic or reminds a prepared practitioner of a known way. Relate a prescription to its operations, actual or proposed performance, provision and authority. Use a heuristic to investigate the relationships it explains before selecting an intervention. One performer may supply several control contributions, and several performers may share one. For a claimed control layer, identify the reference, limit or rule it sets and how this can affect the other controller’s action. For a feedback claim, also identify the returned observation and how it can change that setting. B.5.MR compares what models of those decisions specify or leave open.
Library.Reference:6.5 (source document not included in this publication: FPF-LIBRARY-REFERENCE.md#libraryreference65---compact-cycles-in-working-control-arrangements) develops the family connections, a sufficient short choice and the returns caused by changed conditions. Use an already adequate method; expand the arrangement where a missing dependency requires it.
Control engineering. A state-space model specifies how state changes with input and disturbance and how observations depend on state. An estimator supplies a decision-relevant account from observations and applied inputs. A controller selects input; constraints and continuation conditions determine which choices remain permissible. Linear dynamics, bounded disturbances and a specified observation model can support set-based state reconstruction and robust feasibility calculations. These assumptions make a technical construction possible; they leave goal legitimacy, unknown mechanisms and unmodeled interference outside that construction.
Model predictive control repeatedly solves a finite-horizon problem and applies a current part of the result. Its future input representation, feedback design and terminal conditions determine what responsiveness and continued feasibility the calculation includes.
Adaptive control adjusts controller parameters from observed behavior, directly or through updated parameter estimates or a model.
Dual control additionally selects action partly for how its informative consequences can improve later decisions. Updating a model after an incidental observation and choosing an informative action are therefore different constructive moves.
Economic MPC can optimize an operating cost rather than only track a fixed setpoint. It needs a defensible objective and constraints. The adjective economic does not identify whose consequences are included or supply an economic interpretation of every cost coefficient.
Economic reasoning. The common comparison becomes economic when it concerns competing uses of limited means and the consequences for participants with preferences and exchange opportunities. The current configuration is the starting condition. The opportunity cost concerns the best available use displaced by the choice; that use need not be the current practice.
A finite change may be an indivisible contract or a whole machine. A derivative is useful when the model and scale of change justify a local approximation. A shadow price describes sensitivity of an optimized objective to a modeled constraint under the relevant mathematical conditions. A purchase price states an available exchange. These answer related but different questions.
A payment already made is not recovered by rejecting the next action, while the acquired asset and surviving obligations remain part of the present situation. A future unavoidable payment can cancel from the difference between alternatives and matter to solvency.
Retain the participants when comparing consequences: one person's payment can be another's income, and a firm's gain is not an aggregate welfare conclusion. Ordinal preferences support ranking without making utility differences measurable or comparable between people. A scalar objective is a special representation, not a general consequence of using the word value.
Economic reasoning can expose a poor use of means; by itself it does not supply the feedback mechanism, authority or physical realization of control.
Active inference. In this formulation, a generative model relates hidden states, observations and actions or policies. Inference updates beliefs; prospective evaluation uses preferred outcomes and informational consequences to select action. The real process and its generative model remain different. A probability assigned to a preferred outcome is not the prediction that this outcome will occur.
Expected free energy combines preference and uncertainty terms under a declared model. Its technical uses of risk and ambiguity should be recovered from that formulation. They are not automatically monetary risk or every kind of uncertainty. Sophisticated inference considers how future observations change future choices. This supplies one way to represent the conditional continuation in §4.3; it does not make every active-inference calculation closed-loop within its planning horizon.
A preference distribution proportional to exp(−αc(o)) can connect a cost c to negative log preference when the normalization exists: −log P(o) = αc(o) + log Z. That authored correspondence requires a chosen scale α, outcome domain and normalization Z. It does not equate all of active inference with economic maximization: the informational terms, action representation and grounds for the cost matter.
Organizations and biological inquiry. Organizational control adds commitments, participation, authority and strategic responses. A promised resource can become unavailable to others before it is physically consumed. A production forecast, permission to make a promise and the promise itself support different conclusions. Connect them through the actions and obligations they establish.
For a natural organism, an engineer's constructive prescription becomes a research question about a mechanism. A model may explain behavior through feedback or inference without establishing that the organism computes the modeler's objective. Compare competing models by observations and interventions that distinguish their predictions. Useful relations can transfer between engineered and biological accounts while their implementation and evidence remain different.
Direct development through related methods and capabilities
The controlled continuation can change what an agent will be able to do, including how it obtains further development. The agent may be a person, an AI system, a team or an organization. Begin with the wanted change in possibilities: which work or problem is beyond the attained repertoire, why it matters, and what would make it attainable. C.40.OD supplies the full construction of open-ended development. The present operation connects its changing possibilities to observation, recommendation, choice and intervention.
Recover the achieved configuration. Include the agent's available Methods and capabilities, retained assistance, instruments, authority, obligations and conditions of use. Use actual work to establish what the agent can obtain under those conditions.
A record of training or a recommender's estimate can inform this inquiry; the resulting estimate must say what performance and assistance it represents. This starting account makes a useful existing capability available as a means of developing another, through C.40.JD:4.2, without presuming the capability still to be acquired.
Find what must change in the connection to the wanted result. Reason backward from that result to the required contributions and forward from the achieved configuration to what it already makes possible. The unmet connection can concern a constituent Method, the way adequate Methods are combined, an unavailable capability, access to help, or the allocation of work and authority. This identifies where an intervention could change the continuation; construct and compare the resulting alternatives before selecting one.
For instance, additional practice can develop an absent judgement; a changed division of work can instead obtain that judgement from someone already capable. The two interventions can deliver the same immediate answer while leaving different capabilities and dependencies.
To decide how to develop a Method or capability, identify the relation through which the proposed change could affect the agent’s subsequent work.
One organization can remain the developing subject while it changes a working Method, the preparation of its performers and its way of choosing further development. These changes do not require one universal hierarchy. A service provider can affect an encompassing result without becoming part of its recipient. A Method that improves another can have that Method as its subject rather than its constituent. Before claiming an improvement in the receiving Method, explain what the development work changes in that Method and how this affects its use.
Construct a complete intervention. Explain the causal and constructive connections: the proposed work changes a named subject; that change makes a contribution obtainable; the contribution enables receiving work; the receiving result changes further possibilities. Include the preparation, assistance and transition that make those connections feasible.
Retain sufficient parts. If a better instrument removes a difficult operation, reconsider whether acquiring that operation serves the development aim. If the aim is independent judgement, having an external expert supply every judgement may meet today's delivery need while leaving that aim untouched.
Recommend and choose from the attained position. A development recommender performs the work of forming grounded advice about these alternatives. Its useful result can be a preferred direction, several incomparable directions, a discriminating trial, a request for a missing contribution, or an explained reason to retain the current practice. Use C.40.DC to construct the directions and C.11.CRC to compare their finite contributions; the chooser uses C.11 with the affected interests and authority.
The comparison is marginal in the sense of starting from what is already attained. It includes the best obtainable competing use of scarce means, the transition burden, complementary resources, retained assistance and future options. Buying a result, acquiring a capability and changing the need for that capability can all be alternatives.
At a later stage, compare from the then-current configuration: capability already acquired and help already secured change what the next intervention can obtain. Do not repeatedly credit the whole programme's gain to each later step. The economic specialization in §4.7 explains the relation to opportunity cost, finite changes and local marginal measures.
Apply the distinction in §4.4 to the development advice. Identify which later work could use the proposed capability or Method and when that contribution becomes valuable. Examining an improvement can be worthwhile while the agent continues successful work; presenting it now additionally requires a worthwhile continuation through the recipient's interpretation, choice and development effort.
Use C.11.DUA when that demand remains unclear. An explanatory advance or a newly attainable problem can warrant development before an operational intervention is selected. Its contribution is judged at that research or development horizon.
Observe the performed development and its receiving use. A recommendation passes through a choice, available means and work. Recover what was done and under which conditions, then establish what the receiving work obtained, what the agent can now do with the retained help, and what the attempt revealed about the Method and provision.
These are the connected results of C.40.JD:4.3. The mechanism of acquisition remains specific to the agent and subject: human practice, machine training, software modification and organizational learning can share this signature while requiring different bodies and evidence.
Use the resulting evidence at the relevant return. Revised evidence can change the recommender's estimate while leaving capability unchanged. Changed practice or assistance can change capability; a redesigned working Method can change what capability is needed. A changed purpose can change the preferred continuation. Selecting among those returns is the same dependency-based operation as §4.6, now applied to development.
Reconsider the recommender or the development Method when repeated misdiagnosis, ineffective interventions or excessive burden make it the consequential object of change. A new observation or a cheaper way to form sound advice can also make a better arrangement attainable while the current one succeeds. Explain how the proposed change would affect subsequent development, compare the transition from the existing arrangement, and test the improvement through the development it helps obtain.
When a useful Method repeatedly fails to enter consideration, A.15.11 can connect it with an occasion in the work, make its relevance recognizable and provide access to its operation.
Establish what happens after that encounter. Recognition followed by an unavailable instruction, missing capability or an unsuccessful application calls for a different repair. Count the preparation and interruptions of a proposed cue in its comparison with the present arrangement. Its uptake and its benefit to the receiving work require different evidence.
Continue from the new possibility. A completed project can supply both its promised result and the means to formulate a previously unattainable problem. Use C.40.NP to construct that next worthwhile question, including the knowledge, Methods and assistance it needs.
It need not be another rung of a fixed ladder. Consolidation, use of the acquired result and ending one local programme can be appropriate. Open-ended development continues the construction of further attainable problems and means of addressing them; an indefinitely repeated operation or a fixed curriculum does not supply that continuation by duration alone.
Multilevel-learning research contributes a useful question: which changes are fast relative to the present decision, which are retained long enough to shape subsequent work, and how does information pass between them?
Evolutionary search adds a different construction: create variants, examine their consequences, and preserve or reproduce selected variants. Use it when such variation and retention are parts of the Method.
Those temporal and selection relations can be connected to the table above, but the connection must be constructed for the work. A richer trajectory or a larger diversity score is useful only through the development result it helps obtain.
Use the control mantra to hold the reasoning together
To explain how an existing control arrangement can behave, recover the installed rules, available signals and applicable dynamics. Follow their connections to the response or stability property in question. To explain a particular episode, establish what information reached which participant or rule, what it did and what followed. Keep the alternatives the evidence leaves open. For a commission to choose or change a continuation, use the questions below. Use §4.5.1 to construct the timing that the work requires.
Assess the situation → construct possible continuations → compare and choose → realize the action → examine what happened → revise the continuation.
Each phrase recalls a relation developed above. Assessment is sufficient relative to the choice. Alternatives include future conditional actions. Comparison includes the consequences and means of the whole continuation. Realization connects a decision to a change in the subject. Examination supplies evidence for the next use. Revision follows the dependency whose basis changed.
Use the mantra to locate the unresolved connection, then work there. A new observation can send reasoning to the model, a resource constraint to the alternatives, or a changed purpose to the comparison basis. Several answers can be refined together. The arrows give a teaching order, not a one-pass workflow or a required first step. Use the fuller explanation and subject Methods when a remembered phrase no longer supplies the needed operation.
Archetypal Grounding
A measurement is useful only through an available response
A workshop can run a batch quickly or cautiously. For a constructed comparison, suppose an agreed net outcome score is 100 for fast operation in a good material state, −20 for fast operation in a bad state, and 40 for cautious operation in either. The two material states initially have equal probability. These are illustrative decision values after the included operating costs, not observed production data or interpersonal utility measurements.
Without another observation, either action has expected score 40. A test costing 5 has accuracy 0.8 in each state and does not change the material or delay the decision. A good test result gives probability 0.8 of good material: fast operation then has expected score 76. After a bad test result, its expected score is 4, so cautious operation is preferable at 40. Each result has probability one half. The conditional continuation obtains 0.5 × 76 + 0.5 × 40 − 5 = 53, a gain of 13.
The construction uses more than the test's accuracy. It uses the conditional future choice. If the future action is fixed before the test, either fixed action gives 40 − 5 = 35. If testing instead consumes the only fast-operation window, the available post-test action is cautious and again gives 35. The same informational accuracy then has a different practical value because the continuation changed.
Now suppose outsourcing the batch is feasible, gives 55 under the same basis, and remains available after the test. Without testing, outsourcing is the best available continuation. With testing, choose fast operation after a good result and outsource after a bad one: 0.5 × 76 + 0.5 × 55 − 5 = 60.5. The net informational gain is now 5.5 relative to the best available continuation, not 20.5 relative to current cautious operation. Adding a feasible alternative changes both the conditional policy and the value of testing.
If the outsourcing offer instead expires before the test returns, that combined policy is unavailable; compare its immediate 55 with the original test-based 53.
During use, the operator checks the assumption that test delay leaves both actions available and that the observed test characteristics apply to this material. If either premise changes, revise the dependent comparison. The result does not require redesigning the workshop's control architecture.
A feasible local correction can defeat the next commitment
A service team has one qualified specialist available from 09:00 to 15:00. Under the stated conditions the repair takes four hours and the team’s other accepted obligation takes three. Each is ready to start and can be completed on its own in that window. If their work is distinct and the specialist is needed throughout, seven hours are required. The team cannot deliver both by 15:00.
Now suppose both durations include the same one-hour inspection. Its result meets both recipients’ acceptance conditions, and the remaining work is distinct. The specialist can inspect from 09:00 to 10:00, complete the repair from 10:00 to 13:00 and use the last two hours for the other obligation. If material for those last two hours arrives only at 14:00, six hours of capacity in total no longer secures both deliveries. If four and three hours are only lower bounds, subtracting the shared hour gives a six-hour minimum. For a promise covering uncertain durations, establish how the deadline can be met throughout the promised range. When the stipulated durations are fixed, the exhibited schedule settles this case.
When the distinct work requires seven hours, possible changes include obtaining another qualified hour usable on the work in time, agreeing with the affected recipient on a different scope or deadline, or asking the authorized participant to choose which consequence to accept.
If another qualified specialist becomes available, the resource conflict can close without changing the goal. If the client withdraws the obligation, the goal and commitment basis changes even with no forecasting error. If neither remedy can be arranged inside the decision window, the team may need a different standing allocation or authority arrangement. The changed dependency selects the different return.
Comparing explanations of an organism's response
A researcher observes rapid corrections of a person's hand movement and wants to distinguish a response to delayed visual position from a response that also predicts the consequences of recent motor commands. The result sought is an explanation of the obtaining control, not a recommendation that the person perform the researcher's algorithm. Use C.28.CM:4.3–4.6 to construct alternative mechanisms, derive their different consequences and qualify the evidence that could distinguish them.
Consider two simplified hypotheses for a hand moving along one axis toward a target at zero. Positive positions and commands are rightward. The display supplies an earlier position y; let c be the net displacement that the second model predicts from the motor commands issued since that displayed position was sampled.
The hypotheses share a corrective gain k = 0.5 and a next command expressed as displacement over the same short interval. Their command-to-movement model, observation delay and absence of other corrective inputs are stipulated conditions of this comparison, to be qualified before interpreting an experiment.
The display is the same, but the proposed information-to-action relations predict opposite signs. The difference follows from the intermediate state estimate. With no predicted intervening displacement, both command -1 cm; that condition alone cannot distinguish them.
The useful experimental contrast therefore combines a controlled display delay with a qualified history of intervening movement, while observing the direction and timing of the ensuing correction. Increasing delay without constructing those different predictions would leave the research question unanswered.
Suppose a rightward correction is observed under the contrasting conditions. It disagrees with the visually driven model's leftward prediction and is compatible with the second model.
Before attributing the response to internal prediction, examine how the commands and observations were established and what other information the person could use.
A third model that uses current proprioceptive information can also produce a rightward correction. That alternative requires a further discriminating condition or leaves the internal mechanism unresolved; merely fitting the second model cannot remove it. Conversely, an inaccurate command-to-movement model can defeat this predictor without defeating prediction as a whole class of explanation.
The common Method has supplied a connected research result: two proposed observation-and-action arrangements, a condition under which their consequences differ, an observable contrast and the remaining alternative that changes the inference. The subject investigation determines which relations obtain. This application uses the same control construction with an evidential conclusion instead of an instruction to the organism.
Develop a capability rather than repeat a successful delivery
Consider a constructed case. A team produces sound diagnoses with one expert's help. Its other members can calculate candidate explanations, but cannot yet recognize when an important modeling assumption has failed. The current work is successful; the wanted development is the team's ability to handle that judgement when the expert is unavailable.
The recommender recovers the connection: calculation is available, assumption recognition is not, and the expert currently supplies it. It constructs three proposals to examine against the team’s aims and commitments. Continue buying qualified diagnosis as a service; practise contrasting cases and receive feedback on the missing judgement; or change the working Method so ordinary cases are handled locally and exceptional cases obtain a reserved expert contribution.
The first can be sufficient if the only retained aim is delivery. The second aims to develop local judgement. The third proposes a different Method for handling ordinary cases and securing help with exceptions. A combination can protect current delivery while development proceeds.
Compare the alternatives from the team's present repertoire and commitments. Practice consumes time that could deliver existing work; the expert must have time to give the proposed feedback; continuing external help remains part of the result where required. Choose the combination only if its additional attainable capability and future uses warrant that burden. An AI assistant can supply contrasts or perform calculation, but the comparison asks which consequential judgement the resulting team can obtain and under what continuing help.
Suppose qualified expert support remains available for six weeks. Two practice periods each week and the expert's feedback time can be reserved without displacing the team's existing delivery commitments. For its planned work after those six weeks, the team values the expected gain in independent judgement more than the other available uses of that time.
The completed recommendation is: "Retain expert-supported delivery and undertake the bounded practice programme. Use unseen cases to establish what judgement becomes dependable and what help must remain." The responsible team chooses that combination and reserves the means. Capability has to be established through the ensuing work.
During practice, a new observation makes an exercise variation a plausible way to improve the next session. An AI recommender can compare it at low cost within its available means. Suppose the available comparison supports no improvement in attainable learning or later use over the selected exercise, while discussing the change would interrupt the team's work.
Retain the selected programme: the attainable comparison was worth making, but its result gives no reason to present this variation or require the team to investigate it. Reuse that comparison while its basis holds. A newly feasible exercise that develops the missing judgement sooner could instead justify advice even while delivery remains successful.
During implementation, team members examine previously unseen cases, decide which assumption matters and explain the resulting change of diagnosis. The receiving work then checks those judgements with appropriate subject evidence.
Suppose they become dependable with a short independent review supplied by the expert, but not without it. The achieved configuration now includes that review. The next choice can retain it, reduce its burden through further development, or use the supported capability in a new family of diagnoses.
Now suppose the expert announces an earlier departure and the retained review will become unavailable before the next planned reconsideration. Waiting to present this change would leave the team relying on unavailable support.
The recommender raises it in time for the responsible chooser to obtain a qualified replacement, restrict the work to an independently supported scope, or stop the unsupported work. The earlier decision to limit routine advice does not govern this changed continuation. A proposed replacement becomes available only when its required capability, agreement and time are established.
A new family of diagnoses may expose a limit shared by all performers. The next useful intervention can then be development of the diagnostic Method rather than more repetitions of the same exercises.
If the recommender repeatedly prescribes exercises for this shared methodological gap, its diagnostic distinction becomes a candidate for repair. The observed limitation can therefore redirect development to the team's capability, the diagnostic Method, or the recommender's diagnostic distinction. A change to the recommender earns its place through better subsequent development and its full burden, including any additional observations or human interpretation it needs.
Explain both effective and ineffective feedback in the same account
Consider a stipulated discrete-time model x_(t+1) = x_t + u_t. Its controlled quantity is deviation from zero, and its available input is u_t = -k*x_t. Assume current state observations and immediate execution, without saturation. The resulting update is x_(t+1) = (1-k)*x_t.
At k = 0.5, successive deviations are one half and one quarter of the initial deviation. At k = 2.5, they are -1.5*x_0 and 2.25*x_0: the sign alternates and the magnitude grows. Both settings realize observation-dependent feedback in this model. The second is unsuitable for reducing the deviation, even though the description contains a sensor, rule and actuator.
At k = 0, the input is zero and this correcting dependence disappears. The equation describes that limit; whether holding a zero input is a useful chosen action is a separate question about the situation.
Suppose an updating component and parameter memory are distinct parts of the controller assembly. The component selects a new gain from observed behavior and writes it to the memory used by the acting controller. A.12 can expose this internal action.
If the update changes the gain from 0.5 to 2.5, its being controlled and traceable does not make the result an improvement. The changed rule must be compared through its consequences for the subsequent operation. A recovery to an earlier qualified setting is available only if the implementation retains and can restore it under the current conditions.
Now change the observation condition: the rule receives a delayed state. The original immediate-observation update no longer describes the joined operation. Recover the delayed information and the implemented rule, then evaluate that model or change the arrangement. Do not fix the explanation by declaring the missing timely observation a zero state. These returns preserve the common control relations while changing the subject calculation needed.
Obtain a decisive check before a difficult-to-reverse change
Consider a constructed archive migration. Removing the old service would also remove the team's ordinary recovery route. The responsible operator has authorized removal only after compatibility of the required formats and operations is established. A qualified ten-minute check on an isolated copy can settle that compatibility question under the stipulated migration conditions; it finishes while both migration and retaining the old service remain available.
The present uncertainty distinguishes continuations with different consequences. A compatible result allows the operator to consider migration under the remaining established conditions. An incompatible result keeps the old service in use while the migration Method is repaired. Running the check preserves that conditional choice at a burden the team accepts.
The completed advice is: "Perform the compatibility check before removing the old service. Continue the migration only after compatibility has been established and the other conditions are satisfied. If incompatibility is found, retain the old service and repair the migration Method. If compatibility has not been established in time, retain the old service." This advice selects inquiry and a conditional continuation. The check and migration remain to be performed.
If the selected check is unavailable, other qualified evidence may establish the same compatibility in time. If compatibility has not been established in time, removal remains unavailable under the authorization. Difficulty obtaining evidence does not establish compatibility. The team can retain the service or pursue a feasible revision through the responsible authority. The qualified check supports its stated compatibility conclusion; other migration properties require a basis.
Change the receiving use. The team now needs only an explanation of the formats' conceptual differences and makes no migration commitment. Suppose the available specifications suffice for that explanation and the check's possible outcomes would leave it unchanged. Give the supported explanation with its scope. A later migration reopens the compatibility question. The worthwhile inquiry changed with its receiving use, while the requirement governing removal retained its force.
Bias-Annotation
The construction favors explanations whose consequences can be connected to action. That emphasis can conceal values that are not captured by the selected outcome measure, or treat a participant's resistance as a disturbance rather than a legitimate difference in purpose. Retain the affected interests and authority behind the comparison.
A technical control model can encourage unwarranted confidence in observation, actuation or common objectives when transferred to organizations. A biological analogy can obscure different mechanisms. Keep the added assumptions of each specialization visible and revise the model when they no longer serve the receiving use.
Conformance Checklist
A practitioner or researcher can use the account when the questions relevant to its explanatory or practical purpose have content-bearing answers at the needed depth:
- What system and consequential change are being considered, for whom and over what span?
- For explanation, what information-to-action relations are supported, what consequences follow under stated conditions, and which alternatives remain?
- What information can reach the rule or chooser in time, and which remaining uncertainty can change the explanation or action?
- For a practical choice, what continuations are available, including conditional future responses and retained adequate practice?
- For that choice, how do the consequences, interactions, costs and remaining possibilities compare?
- How does the observed response take effect, or how could a proposed response do so, through the means and any applicable authority?
- Which later difference changes which relied-on conclusion?
- Where a specialized formalism is used, what assumptions give it its result, and what question remains outside it?
- When development is intended, which intervention changes a capability or Method, how is that change established in receiving use, and which further worthwhile problem becomes attainable?
These answers establish a connected proposal or working account. Effective implementation, empirical performance and formal guarantees require the evidence appropriate to those further claims.
Common Anti-Patterns and How to Avoid Them
Consequences
The practitioner can combine independently useful contributions into a continuation whose assumptions agree, and can revise that continuation without discarding sound work. The account also exposes an unavailable contribution instead of hiding it in a plausible overall narrative.
The cost is attention to dependencies that a familiar routine can leave implicit. Apply the Method at the unresolved connection and reuse an adequate whole. More elaborate representations are worthwhile when their contribution to the receiving explanation, action or assurance justifies their burden.
Architectural Rationale
The recurring work is constructing and revising a connected control account. It has a selectable problem and result even when every component Method already exists. C.11.CRC supplies a finite contribution comparison; A.15.7 selects an action within current work; C.11.DBR reopens a changed decision basis; C.32.LCA constructs a control arrangement. None is required to change its independent purpose to explain this encompassing use.
The common construction belongs at the transdisciplinary level because its relations survive across different controlled systems and implementations. Their mathematical, economic, organizational and biological fillings retain their Methods and warranted conclusions.
The mantra makes the developed relations easier to retain while each operation remains usable at its point of need. A sufficient existing continuation can be kept; a changed dependency reopens the contribution that relied on it. The general explanation supports that selective use across the different subject constructions.
SoTA-Echoing
The practice question is how to connect current knowledge, consequential choice and continued action without silently importing one field's assumptions. The selected line is the explained composition in §4: retain the specialist constructions, make future conditional action explicit, compare the whole continuation, and revise the dependency whose basis changes.
It is an authored methodological synthesis. Its conditional calculations support the distinctions illustrated here, not an empirical claim that this exposition outperforms every existing practice.
For the active-inference formulation in §4.7, the source basis is Friston and colleagues, Active Inference: A Process Theory, 2017, §§2 and 8; the latter already distinguishes naive from sophisticated future inference. Friston and colleagues' Sophisticated Inference, 2021, develops conditional prospective inference. This pattern adapts that constructive distinction while retaining other control and decision formulations as alternatives; it does not infer a unique objective for all living or artificial systems. 2017 paper, 2021 paper.
Relations
- Apply C.11.CRC to construct the contribution of a finite configuration change and C.11 to choose among warranted alternatives.
- Apply C.11.DUA to decide whether further advice, observation or computation can improve the receiving choice enough to justify its burden.
- Apply A.15.7 for the immediate action within an understood working Method; use this pattern when its connected control basis is unresolved.
- Apply C.11.DBR when subsequent evidence changes a relied-on decision premise.
- Apply C.30.LCA to recover a control structure and C.32.LCA to construct or change it. Those results can supply the arrangement used here.
- Apply C.39 when a needed way to obtain a contribution is missing, C.28 for a causal question and C.29 when a mathematical representation's choice, transfer or supported use remains unresolved.
- Apply C.40.OD for the whole construction of open-ended development; use C.40.DC, C.40.JD and C.40.NP to construct directions, realize the joined development and continue from its results. §4.8 connects these results to control and recommendation.
- Use B.1.5.RS for the consequences of replacing a constituent Method, C.32.MWA.PP for whole performance arrangements, and C.30.ILC with C.32.MLAO when the intervention addresses an architecture residual between the identified scopes.
- Use A.22.CGUS when the availability of possible continuations and the conditions that block them need to be made explicit.
C.43:End
Clarify Cultural-Evolution Wording for a Claim or Action
Tech-name:
CulturalEvolutionWordingUsePrecisionRestorationPlain-name: cultural-evolution wording-use precision restoration Type: Precision-restoration companion pattern (C) Status: Stable Normativity: Normative unless marked informative Placement: Part C -> C.36 companion Builds on:E.10,E.10.ARCH,E.10.DEV,E.10.ROLE,C.36,F.17,F.18,F.9,A.3.1,A.3.2,A.15,C.18,C.19,G.5, andG.11. Purpose: recover the object, relation, or claim hidden by culture, style, tradition, genre, scene, practice, technique, platform, regime, attractor, or developmental-machinery wording, then return to the project question without turning the label into ontology.
Use This When
Use this pattern when source or project prose uses cultural-evolution wording and a current claim or action depends on what that wording means here. If the word is only ordinary or quoted language and no FPF claim relies on it, leave it alone.
When generic development, evolution, or lineage wording hides the changed or represented subject, continuity or membership, posture, direction or value basis, or direct owner, use E.10.DEV first. Continue here only when that recovery exposes a cultural-population, discipline, style, tradition, recognition, selection, transmission, retention, or mediation question whose wording is unclear. Use E.10.MOVE only for a separately relied-on trajectory or path ambiguity.
Trigger expressions include, for example, culture, cultural evolution, style, tradition, genre, scene, technique, practice, platform, platform regime, measurement regime, attractor, developmental machinery, cultural lineage, canon, and school. They are recognition cues.
What Goes Wrong If Missed
The repair becomes a synonym swap. Style becomes method, platform regime becomes context, practice becomes a generic process label, or attractor becomes dynamics before the sentence says which object, relation, or claim is current. The result looks cleaner but carries an accidental ontology.
What This Buys
The first result is a short ordinary statement: what the expression means in this use and which pattern supplies the next needed definition or test. For example: Here “platform” refers to the short-video recommendation System and to the visibility and recognition relations around it. The claim is that changes in those relations altered which dance variants were copied; use C.36 for the cultural-evolution case and C.18 only if archive retention is the next question.
Use C.36 for a cultural-evolution case. For a Method, Work, discipline, bridge, archive, pool, selected-set result declaration, publication, architecture, dynamics, measurement, choice, or refresh claim, use the pattern that defines or tests that claim.
First Useful Move
Write the short ordinary statement first. Stop when it makes the next project action clear.
When a handoff or repeated use needs durable memory, keep the same result in this optional line:
Fill only the fields the receiving use needs. If the object, relation, claim, or applicable rule cannot yet be recovered, keep the label as quoted source wording, ordinary prose, or a blocked-use cue. Do not choose a smoother umbrella word merely to fill the line.
Problem Frame
Cultural-evolution sources and project documents use compact labels because ordinary language has to move quickly. A word such as style, genre, practice, platform, or technique may be a useful local sign. It may also hide one or more distinct objects, relations, or claims—for example a Method family, Work family, local system-role kind, System-classification judgment, assignment species or occurrence, direct participation relation, discipline, canon or memory episteme, recognition relation, selected set, archive, front, mediation System, architecture, measurement relation, publication label, or mathematical-lens claim. Ambiguity alone does not prove that several of these claims are present.
Use C.36 for the cultural-evolution case. This companion recovers only enough meaning to write the ordinary claim and use the pattern that defines or tests the next needed distinction.
Problem
Without recovering the object or relation, a local wording repair can produce four failures:
- source labels become root kinds by spelling;
- platform and regime labels become hidden Systems, scope containers, or authorities;
- style and tradition labels become genre trees or single trajectories;
- developmental-machinery and practice labels become method, work, or process labels by taste rather than by current relation.
The repair must keep useful local labels while stopping them from carrying unearned ontology.
Forces
Solution
Recover the assertion whose meaning changes the decision. Write it so that a colleague can tell what observation or rule would settle it. Saying that a method was “adopted” might report an authorized decision to begin using it, a first use by named practitioners or continued use across a set of teams. Name the one meant here. If the record supports a different, narrower assertion, state what it already permits someone to do. Qualifying the words identifies the claim to examine. Use C.36 or the governing discipline to establish whether it holds.
Before these cultural routes, send unresolved generic development or evolution wording to E.10.DEV. Return here only when its repaired claim has a cultural-evolution subject and a cultural wording question remains; preserve its ordinary non-use, missing-information result, non-cultural population or lineage architecture gap, and one-pass stop otherwise.
-
Cultural-evolution case. Use C.36 when the claim concerns how a collective or discipline generates, transmits, recognizes, selects, retains, or deliberately changes variants. Method and Work families, system-role facts, canons or memory epistemes, mediation, style or tradition terms, variant sets, and interventions remain separately identifiable inside that case.
-
Term and bridge work. Use F.17 when the needed identification or qualification of a source-local or project meaning is unresolved, and F.18 when a durable name is needed. Preserve the claims made under each source-local meaning. Use F.9 only when a relation between distinct source-local cells is current. Keep C.36 only for the cultural-evolution case that makes the term matter.
-
Practice, technique, and developmental-machinery wording. Ask what the phrase names:
- For a reusable way of doing, use A.3.1. A description qualifies as U.MethodDescription under A.3.2 when its exact EntityOfConcern is an already admitted Method and it makes at least one substantive claim about that Method as a way of doing.
- For a plan, use A.15.2. For a particular dated U.Work occurrence, establish the A.13 basis, independently admit Work under A.15.1, and use F.6 only when precise assignment-bound attribution is required.
- If several Methods are related or assembled, use A.3.1's method-relation route; it keeps ordinary relations, order-sensitive composition, and a selected Structure distinct.
- If the wording is about who acts or how someone is assigned, E.10.ROLE separates direct participation, a local system-role kind or classification under A.2, an assignment under A.2.1, and a relation between system-role kinds under A.2.7.
- Use A.1.1 only when the decision depends on bounded model use or one of its direct model-applicability, actual-use, or fixed-content-coherence relations. Use A.10, C.20, or C.23 only when the claim is about evidence use, discipline composition, or method-family maturity.
-
Variant-set, archive, front, pool, selected-set, publication, and refresh wording. Use C.18, C.19, G.5, E.17, E.24.PUB, G.11, or E.18.1 according to whether the next question concerns generation or retention, current-pool treatment, a selected-set result declaration, publication, refresh, or problem-to-work carry-through.
-
Platform, regime, and mediator wording. Do not presume a System. The phrase may identify, for example, a mediating System or architecture, recognition or selection relation, measurement or visibility relation, publication or source-currentness relation, episteme, bounded model-use structure, direct model-use relation, claim scope, or ordinary project label. Admit a System only when that System is the object. Keep any local system-role kind, classification, and assignment separate and optional. If the current governing rule needed to state or test the intended claim for the named participants and use is absent, return A.6.RCD's
missing-governor. -
Meta-holon-transition (MHT), level, boundary, feedback, model-use or scope reframe, and frustration wording. For a claim about more abstract or more concrete models, name the model and what is preserved or newly determined under
B.5.MR:4.3. For a claim about more teams or a longer interval, name the population or period. Recover whether any further claim concerns a new holon or level, whole reidentification, a System boundary, a relation that crosses a holon boundary, supervisor-subholon feedback, bounded model use or claim scope, cross-scope architecture residual, mathematical-lens use, or interlevel ethical conflict. Use A.1, A.1.1, B.2.P, B.2, B.2.2, B.2.3, B.2.4, B.2.5, C.30.ILC, C.29, D.2, D.3, or D.4 according to that recovered claim. Keep C.36 only for the cultural-evolution case. -
Attractor and dynamics wording. Recover the model supporting a basin, stable-dynamics, state-transition-law, or other mathematical claim. Use an adequate domain result directly; use A.3.3 for a state space and transition law, C.27 for temporal adequacy, and C.29 when a separate mathematical representation, transfer, or reliance question remains. Keep a loose style or tradition meaning in ordinary wording.
-
Architecture wording. Use C.30 for an architecture question or ArchitectureClaim, C.30.ASV for a structural-view question, and C.30.AD for an architecture-description question. A selected structure, ArchitectureRelation, claim, description, view, representation, and publication remain distinct. Treat
ArchitectureOf@Contextonly as a legacy retrieval phrase resolved by the current C.30 edition, not as a current record or ontology.
A C.36.P repair closes when the short ordinary statement names the needed object, relation, or claim and makes the next use or stop visible. The optional recovery line is not required for a clean local repair.
Use the patterns for development-loop semantics, archive or front relations, pool policy, selected-set result declarations, Method-family semantics, measurement, refresh, publication use, or architecture use. C.36.P recovers the wording needed to enter them.
Recovery Result Table
The trigger phrases are examples. Recover the intended object, relation, or claim first; then use the pattern that defines or tests the distinction needed by the next action.
Worked Micro-Examples
"The Platform Changed The Style"
The editor finds two records. The recommender changed the relative visibility of two dance variants, and the later published clips show a different mix. Those records can support the visibility and publication claims over the clips covered. “The platform changed the style” could also mean that dancers copied a different movement because of the recommendations, or that classes retained the change. Either is a further claim. To establish copying from this source, trace the relevant exposure and the dancers' use of the movement. For causal influence or continued class use, examine the corresponding comparison or later work.
An editorial note for someone deciding how to present the evidence can read as follows.
"This Tradition Is An Attractor"
If attractor is a loose metaphor for a stable recognizable style, state that local meaning and use C.36 for the cultural-evolution case. Use F.9 only when a relation between distinct source-local cells is current. For a basin, stable-dynamics, or state-transition-law claim, recover the model, correspondence and conditions that support it. Use an adequate domain result directly, A.3.3 for its state space and transition law, C.27 for a temporal-adequacy question, and C.29 when a separate representation, transfer, or reliance question remains.
"Technique As Developmental Machinery"
If technique names a way of doing work, use A.3.1 U.Method. If it names a training plan, use A.15.2 U.WorkPlan. For a claim about a particular rehearsal or production occurrence, establish A.13's basis and independent A.15.1 admission; use F.6 only for precise assignment-bound attribution. Ordinary guidance about rehearsing or producing needs no invented occurrence. If the same term carries different source-local meanings, preserve each meaning and its claims. Use F.17 when their identification or qualification is needed, F.18 only when a durable name is needed, and F.9 only when a relation between distinct cells is current. Use C.36 only when the technique participates in a cultural-evolution case.
"The Scene Became A New Level"
If a music or dance source says a scene, platform circulation, or canon “became a new level”, first recover the claim. Have participants become constituents of a new performing collective, has circulation reached more of the existing groups, or does a model retain fewer distinctions for the same question? Identify the actual composition, coverage or model change and the evidence for it. Wider circulation can itself be the whole claim. Use C.36 for a changed recognition regime, C.18 for an archive relation, G.5 for a selected-set result declaration, G.11 for refresh, and F.17, F.18, or F.9 for term and bridge work. A whole-reidentification or MHT claim uses B.2.P and then B.2 or its System or episteme specialization when current. A feedback-down claim uses B.2.5. A frustration or residual claim uses C.30.ILC for an architecture residual, C.29 for a mathematical lens, and D.3 or D.4 for ethical level conflict or mediation.
Boundaries
This pattern does not define the cultural-evolution case or intervention. Use C.36 for those questions.
Use C.36.P to recover one ordinary claim. Keep the optional recovery line only when a handoff or repeated use needs it. Once the meaning and applicable rule are clear, stop the wording repair and return to the project question.
If a source uses CulturalEvolutionWordingRecoveryLine@Context, read it as CulturalEvolutionWordingRecoveryLine qualified for local retrieval; recover any required scope or relation separately.
Cultural labels remain usable as local signs for the recovered objects, relations, and claims.
Relations
Builds on: E.10, E.10.ARCH, E.10.DEV, E.10.ROLE, C.36, F.17, F.18, F.9, A.3.1, A.3.2, A.15, C.18, C.19, G.5, and G.11.
Coordinates with: E.10.MOVE for a remaining independent trajectory or path ambiguity, A.1, A.1.1, A.6.RCD, B.2, B.2.P, B.2.2, B.2.3, B.2.4, B.2.5, A.3.3, C.16, C.20, C.23, C.27, C.29, C.30, C.30.AD, C.30.ASV, C.30.ILC, D.2, D.3, D.4, E.17, and E.18.1.
C.36.P:End
Sustain and Renew Shared Ways of Working
Type: Method pattern
Status: Stable
Normativity: Normative unless marked informative
Problem frame
Use this when changing participants, tools or questions puts a shared way of working out of reach, or when a useful discovery needs to become available to other participants. A team may receive correct answers while losing the ability to apply the method under changed conditions. A researcher may discover a construction that colleagues cannot yet use. An ensemble may preserve a recording while newcomers cannot vary the phrase.
Begin with the work participants need to perform or develop. Follow one available result back to the contributions that make it usable, and locate the contribution they cannot obtain. Restore access, arrange a capable contributor, support acquisition or develop a better way, according to the difficulty found.
The first result is a means of obtaining and using the missing contribution. When the assignment includes carrying out the change, continue through its use in the receiving work. That use can expose a further difficulty or open a new question.
This pattern concerns the availability and renewal of shared methods. A shared repertoire is the collection of ways participants can obtain and use in their work, including ways learned from predecessors or made accessible through other agents. People, AI agents and collective agents can contribute. Descriptions, examples, tools and working relationships help make those methods available.
Use an already adequate method directly when its means of use remain available. Improving one person's performance within a stable arrangement can use the relevant development method. Use C.36 for a broader question about cultural transmission, selection or change.
Problem
A result and a way of obtaining and applying it have different uses. A stored equation can answer a repeated question while leaving its reader unable to determine which device it describes. A machine can produce a proof that another contributor can verify but cannot yet adapt. Reproducing one rhythmic phrase leaves open how to make a related phrase for another entry point.
Changes in the practice expose these differences. A departing colleague may take away an interpretive skill. A replacement tool may produce the same output through different assumptions. An AI may introduce a useful method whose operations are unfamiliar to the receiving participants.
Preservation also involves choices. Requiring every participant to learn every operation can consume the resources needed for the work. Keeping an obsolete procedure can prevent improvement. A practice needs to retain or replace the means of obtaining useful contributions, while leaving room for methods whose later uses are being discovered.
Forces
Solution
Make the required methods obtainable through an arrangement that participants can use and, where needed, develop.
Recover the receiving work → locate the unavailable contribution → choose how it will remain obtainable → arrange acquisition, access or redistribution → use it in the receiving situation → retain the means of use. Develop a further continuation when its expected value justifies the work.
These are dependencies of the work. An available demonstration, tool or newly discovered method can be the starting point.
Start with work worth sustaining or opening
Name a situation in which participants need to obtain or apply knowledge, make something, intervene, decide or investigate. Identify what has changed or what discovery could open useful work. Take an existing example, a recent failure or an intended variation that makes the need concrete.
For a continuing activity, recover which result must be obtainable and what the receiver must be able to do with it. For development, identify the promising construction or operation and an attainable way to explore it. A present clue can justify retaining a method before its eventual application is known.
Bound the effort by this use. If the existing arrangement already supplies the needed contribution, use it and stop. A suspected limitation matters here when it can change the receiving work or a selected development opportunity.
Locate what participants cannot obtain
Follow the chosen case through its inputs, operations and receiving results. Examine the explanations, tools and capable help that participants can already obtain. Use existing work and knowledge when they settle what is available; perform a revealing attempt when an unresolved difference matters to the next action.
Locate the first consequential gap. For example:
- The method is understood, but the replacement instrument's interpretation is missing.
- A calculation can be obtained, but nobody receiving it can connect its variable to the physical quantity.
- The relevant contributor exists, but others cannot obtain that help when the work needs it.
- A demonstration can be copied, but its operation cannot yet be used under the changed condition.
These gaps call for different repairs. E.23.CAE distinguishes unavailable access, recognition of applicability, activation, expression and capability change. Restore access or the missing input when that is sufficient. Use E.23.CDI and the relevant learning method when a participant needs to acquire a capability.
For connected physical, mathematical and computational work, B.5.MPC helps locate the missing relation. B.5.RC and B.5.RA help recover a construction or argument. The practice's subject methods supply the corresponding operations in other fields.
Return the missing contribution in terms a provider can act on: “relate this indication to position for the replacement device” is enough to direct the next work. A label such as “knowledge loss” leaves that work unspecified.
Choose how the method will remain obtainable
Decide which contributions the receiving participants will perform themselves and which they will obtain through other people, AI or tools. Keep the capability where the later use needs it. A person who must recognize a faulty physical interpretation needs that understanding; an available specialist or AI may supply the algebra.
Compare retaining the existing arrangement, adding a contributor, teaching or training, providing a tool, changing the method, and retiring a superseded way only where those alternatives answer the found difficulty. C.11 supplies a consequential choice among formed options. C.40 supplies construction of alternatives when the way is still to be found.
Retain the material that makes the selected arrangement usable. Depending on the gap, this can be a readable construction, its inputs and conditions, an explanation of a decisive step, an executable model or an example that shows how to change the method. Keep access to the needed performer or executor as part of the arrangement. If several contributions are promised, trace the allowed uses through the people, tools and handovers they need. Establish whether the provisions can meet their respective commitments together. One handover can be assessed on its own delivery conditions. Select a form that the receivers can use.
When the method changes, make the changed operations and conditions recoverable so participants can obtain the version their work needs. Keep the earlier variant when continued use, comparison or recovery needs it. An access repair can leave the method itself unchanged.
Arrange the selected acquisition or transfer
Use a form of learning or provision suited to the participant and operation. A demonstration may expose a visible procedure; explanation may supply the missing relation; practice with feedback may develop performance; a configured tool or obtainable specialist may supply the contribution directly.
Choose what the receiver needs to recover from the method. Carrying out a stable procedure may require recognizing its entry conditions and interpreting its result. Adapting it can require reconstructing a dependency or derivation. Developing it can require changing an assumption or inventing another operation. Use B.5.RC/RA for the corresponding reasoning and the relevant subject development method for acquisition.
For a human learning a demonstrated method, one useful arrangement is to observe its application, attempt the work with available help, and then use the operation in the receiving situation. Choose examples and assistance from the required transfer. An AI learner or a supplied tool can need a different training, configuration or interaction method.
An explanation can use C.2.8 to recover what its reader needs to extract, and a subject explanation method to construct it. Introducing a method into an organization can require coordination of its work and resources. Use the corresponding practice's methods for those contributions; the result here is an obtainable way to perform the previously missing work.
If an action requires another participant's agreement or access to a controlled resource, obtain it through the arrangement that governs that action. Keep the independently usable contributions available while the missing part is being arranged.
Apply the method and use the resulting difference
Perform the receiving work when implementation is part of the assignment. Have the receiver interpret and use the contribution under the conditions that matter there. A changed device, requested result or available resource can reveal whether the method supports the intended adaptation.
Compare the result with the question that led to this work. If the result is usable, continue the activity it enables. If it fails, locate the missing operation, interpretation, access or capability and repair that part. Reuse contributions whose conditions hold.
State what the handoff was to deliver. Receipt of the agreed procedure and working access can complete a handoff of access before the recipient's first shift. To establish transmission through use, identify what the recipient obtained through the arrangement and where they used it. If a specialist supplied a calculation and the receiver interpreted it correctly, the result can establish that they completed that assignment together. To rely on the receiver to reconstruct the calculation after the specialist leaves, examine their work under that changed condition. Use C.36 for claims about spread or persistence across a population.
Keep the means of renewal available
Retain a way for later participants to obtain the method and its useful explanation. This can be through colleagues, teaching, a maintained tool, accessible source material or a combination. The choice follows the continuity problem.
When development is wanted, use a consequential variation to open the next question. Change a premise, recipient, constraint or operation; examine the result through C.40. Preserve useful material and the means of further variation. C.18 supports an exploration archive; C.19 helps choose which directions in an active search pool to continue, retain or stop.
Reconsider the arrangement when a relevant participant leaves, a tool or interface changes, an explanation ceases to support use, or a new question needs an unavailable operation. End the current work when its receiving use is supplied. Later renewal can begin from a later need.
Archetypal Grounding
A replacement sensor and a changed interpretation
The following is a constructed laboratory case. A group infers position q from a sensor indication r under a linear working account. The earlier device has reference readings q=0, r=2 and q=1, r=5. They give r=3q+2, so indication 14 corresponds to position 4.
A replacement device has available reference readings q=0, r=4 and q=2, r=10. A saved notebook still returns 4 for indication 14. The group needs the method that relates a device's indication to position.
The missing contribution is the replacement device's relation. A technician supplies its reference readings and conditions. An AI obtains the slope (10-4)/(2-0)=3 and intercept 4. An engineer connects the resulting equation r=3q+4 to the device and infers q=(14-4)/3=10/3. The saved answer changes because the physical interpretation changed.
The group retains the two-point construction with its device and reference meanings and an obtainable way to perform each contribution. The technician need not derive every formula independently; the receiving engineer needs to recognize which device and account the formula concerns. The linear relation is a working assumption: two points determine its coefficients but leave empirical adequacy across a wider range open.
A later question about another range can use that retained construction to ask where the linear account is usable or what further account is needed. The current case finishes at the interpreted conditional position result; a new inquiry depends on the use that needs it.
Retain a rhythmic operation
In this constructed case, an ensemble uses an eight-pulse cycle with onsets at 0, 3 and 6. New participants can copy the recording. The next arrangement requires shifting each onset position by +2 modulo 8.
Adding two modulo eight gives 2, 5 and 0, or the ordered onset set 0, 2, 5. The operation depends on knowing the cycle, what an onset position means and how positions wrap at its end. Retaining those meanings lets a contributor produce the changed phrase.
One musician demonstrates both phrases while counting the cycle. If the ensemble's commission is to perform the supplied arrangement, the musicians may use the marked parts and count together. A rehearsal can establish that result. If a musician is to arrange future rotations, give that musician a new offset and the original phrase. Leave them to determine the new onsets using the support the ensemble normally provides. Their calculation and its use in the changed phrase can show that they applied the rotation to the new offset. A performance of the already demonstrated phrase can leave that question open.
The new phrase can become material for another variation. That continuation uses the operation and its means of performance. A claim about which phrase listeners find interesting adds the listener and musical-use question through E.10.INT.
Learn a strategy discovered by AI
Brinkmann and colleagues studied a machine-discovered strategy in a constrained reward-network task. Learners selected a demonstrator using available performance, watched and repeated its moves, attempted a solution, and then solved other networks. Nine of fifteen human-machine populations retained the strategy through the tested generations; six later lost it. This supports arranging receiving use while keeping later persistence open to observation. See Brinkmann et al., 2026.
Bias-Annotation
The arrangement can unintentionally select participants by who already understands its representations or can access its tools. Inspect that effect when it changes who can contribute or benefit, and provide a different representation, help or access where the work calls for it.
A familiar human learning method can be a poor choice for an AI learner, and an AI's efficient procedure can be difficult for a person to acquire. Recover the required contribution and the receiver's means before choosing transfer. Preserve a human understanding requirement when the work depends on it.
Conformance Checklist
- The receiving work or development opportunity explains what needs to remain obtainable.
- The found gap names an operation, interpretation, access or capability that can change that work.
- The arrangement makes each required contribution obtainable under the receiving conditions, including any dependence on the other contributions.
- Acquisition, explanation or provision is suited to the participant and the required use.
- Judge the handoff against the commission. Receipt of the agreed procedure and working access establish delivery when those are the promised results. A claim that the recipient used the transmitted method identifies the use and its connection to the source. A proposal states what still has to be put in place.
- Retained material and help support the intended later use or variation.
- Continuation, renewal or stopping follows the work's value and remaining difficulty.
Common Anti-Patterns and How to Avoid Them
Consequences
The method makes continuity a question about usable contributions. A group can retain mathematical, physical or computational understanding while changing who performs particular operations. A promising discovery can enter shared work through an attainable form of acquisition or provision.
The arrangement consumes resources. Explanation, practice, tools and access have costs, and some methods remain hard to transfer. The result can therefore be a useful redistribution, a narrower continuation or an identified contribution that still cannot be obtained.
Architectural Rationale
C.36 describes cultural transmission, recognition, selection, retention and change. This member provides the coordination needed when participants must keep a method obtainable and usable through a change. Its work joins access and capability, the method's material and explanation, the receiving application and the next opportunity to develop it.
C.40 supplies development through variation of usable material and coupled changes to problems and ways. That contribution is valuable after the practice can obtain the means of working with the material. A practice can also need continuity without a present search project, so the common coordination starts from the receiving work.
The repertoire can be distributed across people and AI. Keeping it usable depends on the relations between contributions, including understanding where the result must be interpreted. The same arrangement can obtain knowledge, apply it and develop its method. This connects the epistemological and methodological sides of inquiry without prescribing one allocation of work for every practice.
Descriptions and examples help because they can make operations recoverable. They work alongside trained participants and tools. A changed question tests a different use from repeating an answer; choose that variation when adaptation is part of the work being sustained.
SoTA-Echoing
What must the receiving participants learn?
The question is how to keep applying and adapting a useful method when its original contributor or tool changes. Adapt: locate the understanding or operation that the receiving work needs, then obtain it through a capable contributor, a tool or acquisition. This changes the allocation in §4.3 and the transfer in §§4.4–4.5.
A serious alternative is to keep access to a capable provider and retain its usable results with their conditions. At the same application deadline and with the same available help, this can cost less than teaching the method to its receivers. Choose it when the provider can supply the expected variations and the receivers can interpret and use them. Where the receiver must recognize a changed physical interpretation, the answer alone leaves that operation unavailable. In §5.1 the group therefore keeps physical interpretation with the receiving engineer and obtains algebra through AI.
Klowden and Tao (2026), §4, supplies the contemporary distinction between checking a mathematical result and understanding that supports further work. Adopt that distinction; adapt the allocation to other practices through the operation they need. The mathematical argument does not establish which human or AI allocation is best in every field. The demonstrated acquisition in §5.3 also separates receiving use from later persistence.
The accepted trade-off is the time and support needed to make a missing operation obtainable. Compare that burden with continued provision before assigning learning. Reconsider the allocation when a cheaper provider supports the same use, access becomes unreliable, or the receiver still cannot perform the needed variation. Sections 4.2 and 4.5 then locate the part to change.
What should remain available before its later use is known?
Adapt: keep a promising method with the means of trying a worthwhile variation, as in §§4.1 and 4.6. The serious alternative is to retain only methods needed by currently specified work. That alternative protects time and attention when continuity of a stable activity is the task. For exploration, it can discard the very operation from which a new question could be developed.
Stanley and Lehman's 2015 stepping-stone argument supplies this challenge to selecting solely by a known destination. Enhanced POET (2020), §§2–3, supplies a computational construction in which difficulty relative to available agents, novelty and transfer work together. Its experimental environments support that combination under their conditions; they do not determine how a laboratory or ensemble should teach and retain its methods.
With a fixed retention and practice budget, preserving the possibility of variation consumes resources that could support present work. Accept that trade-off where an attainable continuation is worth trying; §4.6 uses C.40 to develop it and C.18/C.19 for the relevant archive and live-pool choices. Retain an earlier variant when continued use, comparison or recovery needs it (§4.3). Reconsider retention when the variation becomes unattainable, its means become too costly, or another arrangement supports the useful continuation with less burden.
Relations
C.36.MC qualifies a cultural claim about a Method repertoire and develops the explanation that lets another practitioner reconstruct and adapt a selected way. Use its result here when retention or renewal depends on that reconstructibility; an available description alone establishes neither acquired capability nor cultural retention.
- Specializes C.36 for sustaining and renewing shared ways of working.
- Uses E.23.CAE and E.23.CDI to distinguish unavailable use from capability change and to construct a needed development intervention.
- Uses B.5.RC, B.5.RA and B.5.MPC for reconstruction, argument use and connected foundational reasoning where those contributions are needed.
- Uses C.40 to develop a method and its next questions through usable material and variation.
- Uses C.18 and C.19 for retained exploration material and the treatment of directions in an active search pool.
- Uses C.2.8 for the understanding a receiving explanation must support.
- Uses E.10.INT when wording about interest or curiosity leaves the proposed continuation unclear.
C.36.RP:End
Qualify and Develop a Method Culture
Type: Method pattern Status: Stable
Primary working result: a bounded cultural account and a warranted
continue,revise,branch,replace,stop, orunknowndecision. It retains the admitted-Method or status-preserved candidate/observed-practice branch; names the generation, transmission, recognition, selection, memory, retention, or loss claim with its participants, applicability, evidence, and interval; and separates that claim from its episteme and practitioner consequences. Sufficient current grounds can complete the account. Credible action-changing rivals remain visible; a new intervention or observation is selected only when its contribution, whole burden, feasibility, and authorization warrant it.
Use This When
Use this pattern when a group intends to continue or change how a bounded population generates, transmits, recognizes, selects, remembers, retains, or loses the ways of working in that practice—not merely improve one Method, publish an edition, teach one person, or change one project choice. Recover the change the group needs to understand or bring about. One testable claim about transmission, selection or retention can settle it. Where the practice must continue through changes of membership, trace how successive participants can acquire, use and pass on the relevant way under those conditions. Explain how the links together support continuation. Use an adequate existing predicate and its test; derive a missing predicate as in §4.1.
When transmission or development is the live question, follow one recipient from the available account to the work they need to do. Can they explain the important choice, perform it with the intended support, and change it when the situation changes? Section 4.5 connects the contributions needed for that receiving work. Identifying a card, teacher or repository does not supply the connection.
For a first pass, say in ordinary language what is being passed, selected, remembered, or lost; by whom and for whom; what the available evidence supports; and whether that supports continuing, revising, branching, replacing, stopping, or leaving one claim unknown. Keep the identified card, rule, Work occurrence, System, or direct relation and uncertainty that change this use. If that answers the question, stop. Add predicate, claim-episteme, occurrence-identity, structure, or architecture apparatus only when a later decision must rely on it.
When the result must be retained or handed on, use a small cultural account. Name the C.20-recognized Discipline or smaller unresolved practice boundary and the subject-status branch: an admitted Method, with an independently grounded A.15.1 enactsMethod occurrence when the cultural claim relies on actual performance of that Method, or a candidate lineage/source-described observed practice without claiming Method admission or enactment. Keep the testable cultural claim, named participants, available evidence, relevant interval and uncertainty, and the supported decision together. Add the lightest A.6.RCD disposition, separate subjects, authorization, hypotheses, consequences, or selected-inquiry details only when the current claim or receiving use needs them. A continued practice needs no new intervention row merely to finish.
Here culture is Plain practice wording for bounded claims about how a population generates, transmits, recognizes, selects, remembers, retains, changes, or loses practice. A school label, framework edition, repository, training event, local Method choice, repeated task, or institutional publication does not establish those claims together.
Do not use this pattern for a bounded introduction whose primary result is target changes, later Method use, and outside consequences; use A.3.1.MI. Do not infer cultural adoption, retention, relation occurrence identity, selected structure, architecture, or authority from attendance, publication, tool access, one successful project, an organization label, or a stated preference.
Working Distinctions
Problem Frame
A method culture is easy to claim from visible carriers. A specification has editions, a card set is taught, a Method is placed in a checklist, or a community uses a shared label. These facts may support generation, institutional selection, publication, memory, local use, or a candidate transmission claim, but they do not all support the same predicate, population, occurrence, or history.
Deliberate change is harder. An authorized team may change its own cards or rules, yet have no authority over a field or individual practitioners. Later participants may perform Work with those subjects without demonstrating long-term recognition or retention. First settle what the available grounds support for this practice and receiving use. When a change or inquiry is needed, derive its testable cultural claim, separate its episteme and evidence, compare relevant intervals without assuming occurrence continuity, and retain serious explanations that could change the decision.
A continuing practice also needs a way to obtain and use new contributions. Practitioners may repeat the demonstrated example while remaining unable to adapt it. An experienced colleague silently supplies the missing interpretation; when that colleague leaves, the description still exists but the useful practice does not. Conversely, successful work can reveal a valuable variation without showing that the earlier way was defective.
Problem
Training is called adoption. Publication is called institutional memory and practitioner retention at once. A mandatory field is called Method enactment. Individual card selection is reported as population cultural selection. Several cases, organizations, and time periods are combined into one diffusion story. Authorization to edit project material becomes authority over the practice.
These compressions make cultural claims impossible to falsify. The team cannot tell which relation changed, whether the intended population received or selected anything, what consequence followed, who was authorized to intervene, or what evidence would justify continuation.
Correct classification alone does not repair this difficulty. A team can distinguish teaching, a description and later Work yet leave unexplained what the recipient must learn or obtain, how the contributions work together, and how a useful new result returns to the repertoire. The practice then depends on private reconstruction by the few people who already know it.
Forces
Solution
C.36 governs cultural-generation, transmission, selection and retention claims; C.36.RP obtains continuing access to shared ways of working. This Method uses those contributions to qualify a claim about a population's Method repertoire and to make a selected explanation reconstructible and adaptable. It neither infers culture from publication nor makes every access repair a learning programme.
Bound the practice and name the decision a cultural claim is to support. For a transmission claim, recover both the recipient's use and its connection to a particular source. For a claim about capability, identify the choices the recipient can make with the support expected in the next job. For a claim about the repertoire of a group, establish how the named participants can obtain the needed method under the conditions the group will face. A reliable specialist, peer or service may provide it. Further observation is worth its cost when the remaining possibilities would lead to different actions. Use the appropriate subject-status branch and keep the cultural relation, evidence and practical consequence identifiable.
For transmission and development, connect the recipient's needed action, the available explanation and support, a warranted adaptation, and the result worth keeping through section 4.5. These are connections among existing Method-development and capability-development contributions, not compulsory stages for every cultural account.
Pattern-Use Unfolding
- Bound the practice and Discipline claim. Use a current C.20 result when the named practice is recognized as a Discipline for this use. Otherwise name the smaller practice, collective, population, place, and period under C.36 and keep the Discipline claim unresolved.
- Choose the subject-status branch. For the admitted-Method branch, name an A.3.1-admitted Method or variant. When the cultural claim relies on actual performance of that Method, independently ground the A.15.1
enactsMethodrelation in a particular dated Work occurrence. For the candidate branch, name the status-preserved candidate lineage or source-described observed practice, its candidate account or source claim, identified card, template, description episteme, or other claim-bearing material, and any Work facts on which the claim relies. Ground a Method claim through A.3.1 and an enactment claim separately through A.15.1. ClaimU.MethodDescriptionmembership only when A.3.2 establishes that the episteme has an already admitted Method as its exactEntityOfConcernand makes at least one substantive claim about that Method as a way of doing. - State the cultural claim and apply its predicate. Identify the value said to be generated, transmitted, recognized, selected, remembered, retained, or lost, its participants, and the conditions in which the claim is being made. If a current direct predicate states the claim, apply its test to the available facts and return the supported result. Otherwise use A.6.RCD to derive the needed local compound claim, reusable predicate definition, or relation-kind question, including a positive and a discriminating negative or failed case for that construction. A label or C.36 category alone is not an obtaining relation.
- Separate the world claim, episteme, evidence, and intervals. Keep the direct relation or bounded relation-bearing claim separate from its claim episteme and evidence reliance. Name the interval the available evidence covers. Compare later claims or intervals when the current use concerns later change, retention, or loss; current evidence need not be replaced by a new observation. Open A.6.REL occurrence identity only when a later use must distinguish one obtaining occurrence from another; repeated participants do not establish the same occurrence.
- Name every participating subject. Identify the cards, descriptions with their recorded kind and status, publication occurrences, repositories or other Systems, teaching, tutoring or event Work, decision-rule epistemes, fields, forums, assignments, permissions, authority, and other direct relations relied on. Select an A.22 structure only when one organization of identified constituents, obtaining relations, constraints, and use frame changes the decision; open C.30 only for a current architecture claim.
- Recover the authorization the intended action needs. For a proposed intervention, name the change decision-making Agent, A.13 basis, assignment, permission, or authority predicate, participants, permitted change, scope, positive basis, and protected conditions. Return
missing-change-authorizationwhen that intervention exceeds it, retaining independently supported cultural claims. A factual account creates no intervention to authorize. A continuation or change satisfies the authority conditions that govern it; ownership, project position, expertise, assignment, permission, and authority do not substitute for one another. - Keep the practical consequence separate. State the practitioner Work, understanding, burden, result, constructed-Method position, or other consequence that matters to the continuation. Distinguish its current supported observation from a benefit expected of a proposed change. Keep the subject, relevant baseline, evidence, and any causal reliance separate from the cultural claim; a present account requires no invented future-effect observation.
- Retain the explanations that can change this use. Keep credible non-equivalent hypotheses and their consequences when they could change
continue,revise,branch,replace, orstop, using B.5/B.5.2 when a durable abductive result is needed. There is no fixed hypothesis count. Unresolved explanation can remainunknownalongside a supported current continuation. Before designing or prescribing a probe, compare its attainable contribution with the whole cost, delay, participant burden, and displaced work; use C.11.DUA when that demand needs appraisal. Recover feasible performers, access, resources, window, and protection when they can change the answer. Use A.3.3 only for a state-space and transition-law claim, C.27 for an interval or persistence claim, C.28 for causal reliance, A.15.7 during ongoing Work, and C.11 after a chooser and OptionSet exist. - Choose the warranted continuation or change. When a receiving difficulty requires a new contribution, develop the usable alternatives through section 4.5 before comparing names of carriers or interventions. Compare retaining current subjects, changing one named subject or direct relation, branching for part of the population, replacing the current means, and stopping or reverting where material. Preserve burdens, exclusion risks, authorization, reversibility, and evidence limits. Continue on sufficient current grounds without a new experiment. Select an intervention or observation when its result could change the receiving decision or warranted claim enough to justify its obtainable work; state the distinction it is intended to resolve.
- Perform a selected authorized change through its Method. When the task includes performing the change, admit the intervention Work, performers, enacted admitted Method, action history, temporal extent, containing System, changed subjects, used Systems, named direct relations, and result. A recommendation may finish at a warranted proposal; it does not report that Work or a transformation.
- Support a claimed tested change or later cultural relation. When the result claims such a test or later relation, use the observations at the named interval, applying the same predicate definition to the stated participants or a deliberately changed population. Record positive, negative, inapplicable, unknown, or missing-information. Keep later Work, non-use, burden, adaptation, rejection, and consequences separate. A favourable consequence establishes neither the cultural predicate nor causality. An unavailable later observation withholds that later claim, not an independently supported current account.
- Keep other cultural claims within their evidence. Generation, recognition, selection, memory, retention, loss, or transmission beyond the primary predicate needs its own value, participant meanings, applicability, test, evidence, interval, and disposition. Retain a gap when it limits the receiving use.
- Finish the decision and return the useful contribution. Return
continue,revise,branch,replace,stop, a retained set, orunknown, with the grounds and limits needed to use it. Include a next observation only when selected for a feasible, worthwhile inquiry. Distinguish a defeated maintained claim from a useful new contribution. Return a contradiction to the particular claim or edition at its owner. When its receiving use warrants keeping a reusable difference, A.3.1.MV can retain it as a status-preserved candidate without declaring the parent false; A.15.8.MS can consider a changed support proposal. Name the changed operation or relation, origin, intended use and evidence limits. A.3.2.DE can author or revise MethodDescription or candidate-account content without changing its subject status. Other source claims return to their own episteme; an observation that warrants no maintenance can remain a bounded Work fact. Reconsider the affected result when its subject status, predicate, population, interval, evidence, credible hypothesis, consequence or authorization changes.
Record the Result
Keep only the rows the claim and receiving use need. A later receiver gets the substantive grounds and limits needed to use the same result.
What Changes in Practice
Teams stop calling every publication, teaching session, mandated field, or repeated use “culture”. They can identify one testable cultural predicate and value, use an admitted-Method or candidate branch without laundering status, and separate the world claim from its episteme, evidence, intervals, participating subjects, and consequences.
A supported practice can continue on sufficient existing grounds. When a new inquiry is selected, it is informative and worth its obtainable work. Rival explanations and unknown remain where they matter, without turning a bounded transmission claim into retention, selection, occurrence continuity, or causality.
Minimal Constructed Transmission Replay
Calibration-Method-Engineering-Practice-B is a bounded project practice; no C.20 Discipline recognition is claimed. Admitted Method Engineering Method M-ME-One-Defect-One-Result-1 says: for each trial finding, identify the changed object, separately governed maintained result, particular contradicted claim or edition, and next maintenance or stop action, then rerun the relevant check. Two method engineers and four calibration reviewers are independently admitted under A.13 for the named practice-session Work family. In this constructed case their dated review-practice Work independently satisfies A.15.1 enactsMethod for this admitted Method. Correctness of a particular disposition is assessed from the reviewer’s work on that finding. The card states the rule, and the teaching event is a separate occurrence.
The value under test is the four-part defect-disposition rule. In this constructed practice, CAL-TransmissionPredicate-1 concerns a reviewer who obtained that rule from the engineers and used it during D30–D60. Link the receiving work to their edition or instruction. A reviewer might identify the passage in the engineers' edition that they consulted and show where they used it to change the assignment of a repair. An explanation of a rule also taught elsewhere can establish its application while leaving the immediate source open. Applying the same rule after a class, without evidence of its source, still establishes the observed application. The local performance criterion also calls for correct dispositions on both seeded defects without the transmitter supplying the owner or next action. Merging two independently maintained results fails it. The claim episteme records source use, what the reviewer was given and each outcome. A.6.RCD stops at this local compound claim. No reusable relation kind or individually reidentified relation occurrence is needed here.
The participating subjects remain separately identified: one-page card CAL-Decision-Card-1; two paired teaching Work occurrences; coached-practice Work; two seeded-finding epistemes; and each reviewer's later Work. The card has admitted Method M-ME-One-Defect-One-Result-1 as its exact EntityOfConcern and states the defect-disposition rule as a way of doing; A.3.2 therefore classifies it as a U.MethodDescription. Permission PERM-CAL-CARD-EDIT-1(ME-Agent-B1, ME-Agent-B2, edit, CAL-Decision-Card-1) permits the two engineers to edit that project card. Assignment ASG-CAL-TEACH-1 assigns the paired sessions. Neither predicate authorizes changing the wider field or a reviewer's later Method choice.
The evaluator uses the following answer key for two seeded findings adapted from A.3.1.CC §4.4. The reviewer receives the case evidence and the current teaching card. Record which layers the card identifies and which owner and action the reviewer must still select.
The existing card identifies the two defect layers. The reviewer still has to connect the particular defeated claim to the result that owns it and select the maintenance action. In this constructed trial, dated returns from three reviewers show where each used the engineers' rule and how each chose the owner and next action on both cases. The fourth return merges the two maintained results. Keep the arrangement for the three reviewers under these conditions. Work on the merger with the fourth. The four returns settle that the all-four criterion failed in this trial. They leave open what the fourth reviewer can do after a repair.
A proposed teaching card will give the complete dispositions for these two examples. It can support correct work on those examples. To find out whether reviewers can decide which owner should act on a further finding, give them evidence for that finding and leave its disposition open. Record which decisions the card supplied. If who made them matters, trace them to the person or process that selected them. An editor may have copied the dispositions from an approved answer key.
Now suppose the receiving team needs all four reviewers to handle equivalent findings without immediate coaching. A supervised preparation session is available before that use, the reviewers and seeded cases are accessible, its burden is acceptable, and the existing permissions cover the selected card edit and practice-session work. Replace only the card's label-only worked contrast with the two owner/action dispositions above, then give each already coached reviewer one equivalent case without immediate help and before further coaching. The obtainable answer is whether these reviewers, with this revised card and their prior learning, handle that case correctly or make errors or request help.
Choose the further case for the intended work. If the reviewer must identify the defect-bearing layer, use unfamiliar evidence from which to infer it. A title that gives the layer away leaves the owner and maintenance action for the reviewer to decide. The evaluator checks the contradicted claim, its owner, the action and which other claims remain sound. A successful disposition on a single further case establishes performance on that case.
That answer does not separate the card's contribution from earlier coaching. H-CAL-1, that the card's explicit description-versus-representation contrast is sufficient, and H-CAL-2, that coached feedback supplies decisive discrimination, remain unresolved: later success by already coached reviewers is compatible with both. Current results—three positive cases and one negative merger—do not discriminate them either. A causal use of either explanation requires C.28 and an appropriate comparison basis; this bounded continuation claims none.
Return revise for the changed receiving conditions. The original two-case test succeeded for three reviewers and failed for the fourth. Whether all four will satisfy the criterion in a future trial remains open. In the new trial, one wrong disposition on a required case defeats the claim that all four correctly handled both kinds of defect in that trial. If all four succeed on the single case given to each, their success settles those cases. To confirm the two-kind criterion, obtain evidence for each reviewer's other kind under the conditions of the new trial. A later repair opens a new question about performance under the revised conditions.
Make a practice recoverable and changeable by its recipients
Begin with the receiving work. Recover what the practitioners need to obtain and which decisions make the difference between a usable result and imitation of an example. A newcomer who must choose and adapt an exercise needs more than the instruction for running one selected exercise. Someone who only needs to perform a familiar operation may need a reminder rather than another course. State which contribution must be available from the practitioner, material, tool or another person under the receiving conditions. Distinguish a required change in the recipient's own capability from a result that adequate available support can supply. Preserve sufficient existing capability and help; use A.2.2 for the relied-on capability and E.23.CDI when a capability-change question is current. A.3.1.MI handles a bounded introduction when no cultural question is being asked.
Recover the reason for the way before preserving its form. Trace the available explanation from its inputs, through the operations and their dependencies, to the needed result. Ask what each consequential operation makes possible, which conditions must remain true, and which features belong only to the demonstrated situation. A condition protecting the result is not interchangeable with the historical choice of an instrument or presentation. If the source or a competent contributor cannot supply a necessary reason, retain that gap. Do not turn permission to improvise into missing domain knowledge. A.3.2.DE supplies development of the needed explanation; C.39.MF, C.39.RP, C.39.RQ, A.3.1.DR, A.3.1.QM, C.32.MWA.PP, B.1.5.CM and C.39 supply construction and comparison when the way itself is missing.
Construct a bounded adaptation from the changed condition. Compare the receiving task with the demonstrated one. Preserve the requirements and dependencies that apply; alter the operation, allocation, instrument or support whose condition changed. Follow the proposed replacement forward to its result and backward to its inputs, preparation, participants and shared resources. A simpler local action may move work to a provider or create preparation and cleanup elsewhere. If two contributions must be combined, make the result of the first usable by the second and check the joint conditions; compatible names do not establish this. Use C.39.RQ and C.32.MWA.PP for the situated alternative and B.1.5.CM/B.1.5 only when claiming a composite Method. A missing input-producing operation calls for that contribution or a stop, not an invented claim that the alternative works.
Give the recipient a usable way to make the same decision. Explain a consequential choice and contrast it with a plausible wrong choice. Let the recipient recover the reason, construct or select the next action and follow its consequence in a representative task. Where learning is needed, obtain a development operation appropriate to the holder and the required later performance. It must supply the conditions, practice or other learning operation, feedback and variation needed for that change, together with usable support. Retain the mechanism's human, organizational or technical applicability rather than treating every recipient as a human learner. The feedback must identify the failed interpretation, operation or condition, rather than merely provide a finished answer. The source may need a fuller explanation for preparation and a short reminder during performance. C.37.MP preserves the necessary correspondence between those uses; A.15.8.MS configures their availability. Effort allocated to a probe does not also establish provision for explanation repair, any needed teacher preparation and feedback, or later maintenance. When that provision is missing, obtain a feasible allocation or another qualified contributor, reduce the offer, or retain the supported arrangement while deferring expansion. The teaching activity, material and acquired capability remain different results.
Judge the receiving result under the conditions that matter. Use matching current evidence when it suffices. Choose another task only for a worthwhile unresolved question. Observe what the practitioner did, what help supplied a necessary operation, and whether the result supports the next use. Retain help that belongs to the intended practice. Test without it only when the later use requires independence. A changed task can reveal whether the practitioner uses the reason or merely repeats the example, but learning a new method from a source and transferring an already learned method remain different questions, with different starting capabilities, permitted help and receiving tests. A correct result with assistance does not establish unsupported performance; a failure under a new requirement does not erase a supported earlier use.
Return what experience makes worth keeping. First recover the input, action, result and changed condition. A.3.2.DE receives a needed explanation or candidate-account change. A changed one-action representation stays with its C.37 result; C.37.MP maintains a profile only when several unlike Method-related uses must remain related. A changed access, tool or support arrangement goes to A.15.8.MS; a proposed change in reusable operations, dependencies or applicability goes to A.3.1.MV with its origin and evidence limits. A successful variation can be a candidate contribution without a defect in its parent. Local improvisation can instead remain one Work observation. Keep the older way for situations it still serves, and make the reason for selecting either way available to the next recipient. Generation of that contribution, its later transmission and its retention are separately supported claims.
The common move is to preserve the useful result and its enabling reasons while changing what the new situation requires, then make that decision available for another use. The domain supplies the requirements, operations and tests. A movement-design exercise and a coffee-making procedure do not share those domain rules merely because both can be adapted and taught; sections 5 and 5.4 show the shared move and this boundary.
Follow transmission through human and technical participants
Use this profile of sections 4.1–4.5 when the cultural claim crosses people and technical agents, or when a change in a participant's mechanism can defeat continuation. Retain the practice, value, transmitting and receiving participants, interval and evidence. The population can include human–human, human–AI, AI–human and AI–AI relations. “Human–AI culture” alone does not establish which of them obtains.
Identify what is meant to become available to the receiver: a distinction, rule, operation, explanation, coordination convention or way of obtaining help. Determine how it enters the receiving work. For a person, the relevant change may require explanation, an attempt, feedback, variation and later use. For an agent system, it may require a selected source, changed procedure, retrieval, retained state, a tool or a changed model. Use the development or engineering Method. Do not infer a common acquisition mechanism from a common value or a successful interaction.
Distinguish receipt, use and the source of the use. A recipient may have known the procedure before the teaching session, followed the same third-party template as the demonstrator, or obtained a consequential choice directly from a tool. A dated explanation, revision history or observed handoff can show what the recipient took from whom. Check which decisions the recipient made and which a partner or tool supplied. Accept the joint result when the commission is for supported work. Entrusting a choice to the recipient alone calls for evidence that the recipient can make it with the support available in that later situation.
When the participants must work together, examine their common interpretation as well as each participant's individual result. Ask for a consequential continuation or use a contrasting case. A shared word can conceal different objects or criteria. Engineering.Reference:19.6 supplies repair of the partner interpretation. A disagreement about authority or obligation instead needs the corresponding coordination result; it is not necessarily a failed transmission.
Consider a constructed practice in which a reviewer must preserve a distinction between an observed fact and an inferred explanation. A person learns it through contrasting reports and feedback. An agent is configured to retrieve the current criterion and expose its evidence and inference. Both correctly classify the initial cases. This supports those performances under their stated conditions. It does not yet establish retention after a gap, transfer to another report family or successful transmission to a newcomer.
Now the agent restarts without the relevant task state, or the person encounters a report whose familiar visual cues have changed. Recover which contribution is missing. A.15.10 and the agent-engineering Methods address the first condition; human development and transfer Methods address the second. Preserve the earlier supported performances while keeping the changed-condition claim unresolved. If a repaired procedure or learning contribution later supports the relevant use, record that bounded result without claiming that either mechanism caused every improvement.
Follow the return into the shared repertoire. A useful clarification may need to change the public account, the teaching material, an agent instruction or the access arrangement. Change the dependent use and preserve its correspondence to the governed value. A local shorthand can remain useful among familiar participants while requiring translation or explanation for newcomers. A technically retained memory can preserve an obsolete rule; preservation is not correctness.
For an agent result, retain the model and operative configuration, context and memory conditions, tools, permitted assistance and interval that qualify it. A changed model or tool can reopen a relied-on result without proving failure. For a human result, retain the actual person or supported population, task conditions and assistance. Do not import a model benchmark as a ceiling on every agent, or a human-learning mechanism as a universal explanation of technical adaptation.
Finally separate the cultural judgement from work output, individual capability and team capability. The team can continue supported work while one broader transmission or retention claim remains unknown. A successful team result is not a substitute for that claim, and the unresolved claim does not automatically require stopping the useful work.
Archetypal Grounding - MeCaMinD Transmission
The MeCaMinD record enters through the candidate/observed-practice branch. The source describes movement-design practices and later facilitation Work but does not establish A.3.1 Method admission or an A.15.1 enactment occurrence for a candidate whole. The receiving action is to turn a design challenge into a usable session, not merely to retrieve a card. The investigation concerns what five novice facilitators obtained from the core team's materials or the teachers and used in their sessions, from card development through the August 2023 school.
The source's 4 M framework connects the intended movement concept, the participants' required mood, the selected activities and possible modifications. These contributions serve different functions: an activity that warms up a group does not by that fact produce the wanted design insight. Cards provide instructions and selection information; the Game Board helps arrange a coherent session and retain alternatives. Teachers helped newcomers overcome literal adherence to instructions and consider combining activities. Thus the receiving problem includes obtaining the decision about adaptation, not only making the text legible.
The source reports that the core team tried methods, changed card content and layout, reduced the set, and developed the board. The useful selection conditions included practical usefulness and whether facilitators could communicate and instruct the activity from its description. Five novices subsequently prepared and ran 45-minute sessions with warm-up, design activity and debriefing. Prior self-study used by most, teaching, individual tutoring, background and local tailoring participated in that result. These are reported combined conditions, not an isolated effect of the cards or proof that every later task can be performed independently.
One reported session addressed reluctance to suggest or accept walking meetings. After introducing that challenge, the facilitator used different ways of walking as preparation, then office-persona role play. Between enactments, participants discussed and tried ways to signal that a walking meeting had begun. They came to understand the initiating difficulty more clearly, but reported less ideation than the facilitator had intended. The useful result was sensitization and an initial basis for finding solutions, not demonstrated resolution of the design challenge. For Method construction, the case distinguishes preparation, exploration of the problem and generation of solutions: choosing a coherent set of activities does not make those contributions interchangeable.
A facilitator therefore needs a connected account: which intended result guides selection, how the activities prepare for or contribute to that result, what may be changed, and how the plan remains available during performance. Rich information can help preparation while a concise reminder is more useful during action. A.3.2.DE develops the missing account; C.37.MP relates its preparation and performance uses; A.15.8.MS obtains the corresponding support. The applicable capability-development Method supplies learning and feedback when the recipient cannot yet make the required decision. Their joint contribution is not replaced by listing those providers or by declaring the final plan correct.
For this case, MEC-TransmissionPredicate-1 concerns a way of preparing or conducting a session that a facilitator obtained from a particular source and used. Identify the source and the receiving action. The authors document how one facilitator selected a warm-up from the cards and used a second group game suggested by a teacher. The session plan and the account of that suggestion connect the teacher to this particular choice. The teacher could pass on a game developed elsewhere. A plan and a successful session on their own establish performance under the observed conditions. If the facilitator already knew the procedure, these observations alone leave transmission from the named teacher open. A.6.RCD keeps the claim local to the identified source, recipient, practice and interval. The project designers' account of revising their materials establishes their role in that past work. Any decision about changing a future course requires the corresponding authority. Study of the novice facilitators, Table 1 and discussion of warm-up activities
Two explanations can matter to a later choice. H-MEC-1 says the revised cards and board support a usable plan after self-study, before tutoring. H-MEC-2 says teaching, tutoring, background or local tailoring supplies indispensable support. Reported later facilitation is a bounded positive observation; overload and reliance on the board expose conditions a simpler account must address. The available observations do not isolate the explanations.
Return branch. All five facilitators prepared and ran sessions with the support available at the school. That result supports retaining the arrangement for a comparable course. Credit particular materials or teachers with the choices that the study traces to them. How the same facilitators would have planned using the cards alone remains open. If a later course needs less tutoring, name the planning or adaptation the facilitator must undertake with the reduced support and examine that work under the changed conditions. Seek a comparison that accounts for prior practice and differences between tasks if the decision turns on the effect of tutoring.
A useful new adaptation need not contradict the earlier description. When a receiving repertoire or future support use warrants keeping it, recover the changed operation and conditions, then use A.3.1.MV for a proposed reusable difference or A.3.2.DE, C.37.MP, A.15.8.MS for an explanation, representation or support contribution. A future card intervention stops at missing-change-authorization until the needed authorization is recovered. Neither this account nor a proposed inquiry establishes Method admission, independent capability, long-term retention, wider recognition or selection, adoption beyond the five facilitators, or causality; causal reliance opens C.28.
Separate Organizational-Selection Replay
The telecom SRA record is a different population, organizational System, and history and also enters through the candidate/observed-practice branch unless independent FPF Method admission is supplied. The value under test is the requirement that feature pre-study decisions include a performed security risk assessment. The selecting participant is the named telecom organizational System, the selected value is that requirement, and the receiving loci are its bounded feature-prestudy decisions during the reported period. Local compound predicate SRA-SelectionPredicate-1 is positive only when the current decision rule requires a named security-assessment Work result and the sampled feature decision designates such a result; the discriminating negative is a completed mandatory field with no recoverable assessment Work. A.6.RCD stops at this local claim, with no relation kind or occurrence identity admitted.
Distinguish the mandatory security-impact field, definition-of-done rule episteme, release checklist, SRA-forum Work and System, training Work, expert-help Work, stale source-description epistemes and templates, reorganizations, tool linkage, and appointed-guardian assignments. Candidate or source labels do not turn those materials into U.MethodDescription. None of these subjects alone establishes organizational selection or enactment.
Two hypotheses remain live. H-SRA-1 says mandatory fields, definition of done, and checklist are sufficient to select the practice for feature pre-studies; it predicts completed fields backed by assessment Work even without forum or expert intervention. H-SRA-2 says forum review, training, expert help, current descriptions, tool linkage, and guardian assignments are needed to turn nominal compliance into performed assessment; it predicts empty, copied, or unsupported fields where those conditions are absent. Later use and non-use in a 45-feature release, 41 respondent reports, forum use across eight projects, identified risks, requests for help, stale material, and reorganization effects supply positive and negative observations but do not isolate either hypothesis.
First return the supported organizational-selection account with its non-use and evidence limits. A named assessment Work result distinguishes assessment from nominal field completion; it does not establish that requiring another assessment is worthwhile for every feature. Appraise the disputed requirement through C.11.DUA: which security consequence it protects against, what the existing assessment basis covers, what a new assessment could change, and the burden, delay, and protective work it would displace. The reported presence of a forum, training, expert help, current material, tools, or guardians settles neither the merits nor authorization for a new rule.
For a constructed change with a new externally accessible function and an uncovered security consequence, suppose the specialist assessment is obtainable in the decision window and can change the protective design. Retain the assessment requirement for that scope and, if current records do not distinguish nominal compliance from performed assessment, select the bounded linkage to the Work result when its contribution warrants the cost. A record-linking proposal is not the assessment or evidence of its protective effect.
For the opposed constructed condition, suppose repeated low-consequence feature changes stay within an adequate existing assessment basis, and repeating the full assessment adds no useful distinction while displacing work on an uncovered exposure. Recommend a narrower requirement that reuses that basis and reopens assessment when the protected premise changes. A favourable merits judgement does not relax the rule: the binding requirement remains until its authorized revision or exception is available. If the current process provides no timely permissible route, return that impasse for the affected feature, not permission to bypass it.
Select any new support-condition comparison only when it can change the current decision and warrants its feasible work. The historical source does not supply the assignment, permission, or authority predicate for the proposed rule change; performing it stops at missing-change-authorization until that basis is recovered. The qualified historical account remains complete. Any causal conclusion about linkage, support, or protection requires C.28.
Preserve the gaps: not every engineer used the practice; some found it not worthwhile; several conditions covary; no universal transfer follows. Do not join this population, predicate, source-described practice, or consequence history to MeCaMinD.
Essence Institutional Boundary
The Essence record supports bounded institutional claims only: contributor organizations generated and revised language content; OMG selected formal editions; issue-handling Work, specifications, and machine-readable material supplied publication subjects; retained formal versions and the 2.0 beta can support an institutional-memory claim when its value, participants, predicate, and interval are stated. These claims preserve the source-described status of the participating practice and materials. A Method claim needs independent A.3.1 admission; an enactment claim additionally needs its own A.15.1 basis.
The source does not establish recurring practitioner enactment, population recognition or selection in practice, long-term retention, practical superiority, improved results, or fitness for every Method family. Keep the supported institutional claims and their limits as a completed bounded account. A proposed change to one participating subject is a separate question; select its inquiry only when it is useful and attainable, retain credible action-changing hypotheses, and recover the named assignment, permission, or authority predicate for the change. A claimed tested intervention needs Work and the later observation it asserts.
EC-417 Missing Basis
The full EC-417 application supplies the receiving problem, alternatives, calculations and the distinct limits used below.
EC-417 supplies no C.20 recognition, bounded Method-development population, admitted Method-development Method with enactment, or derived cultural predicate and evidence. Its candidate release, candidate accounts, description epistemes, support results, WorkPlan, and task observations do not establish culture. Because no admitted Method is identified as the descriptions' EntityOfConcern, they are not U.MethodDescription. A.3.2.DE can correct candidate-Method-account content when a defect is found; the absence of a cultural result is not itself such a defect. A status-preserved candidate branch may carry those identified accounts, epistemes, and Work facts, but it asserts neither Method nor enactment. Return the missing value, participants, and predicate when a cultural result is needed; do not invent an intervention, rivals, or future observation merely to complete EC-417's separate release task. Use MeCaMinD, SRA, and Essence only for the unlike claims each source supports.
A Coffee School Teaches Adaptation, Not One Machine Sequence
Consider a constructed school case. Its taught café routine uses ground coffee, a steam wand and a particular presentation. A home user instead has a capsule machine and a separate milk frother, wants one acceptable drink with manageable preparation and cleanup, and does not require the café's presentation. The equipment instructions and a competent practitioner supply the applicable quantity, temperature, milk-texture and safe-operation requirements; this example supplies none of those domain rules by inference. The old café routine remains useful in its original situation.
The school wants its instructors and trainees to pass on the ability to adapt a procedure, not just reproduce a sequence tied to one machine. Start by distinguishing the operations' contributions from their familiar names. Dosing and preparation have to be supplied when a capsule moves some of that work to a manufacturer. Replacing the steam wand with the available frother changes the milk-preparation operation and its cleanup, not the need to meet the applicable milk requirements. Removing a presentation step is allowed here because that result is outside the stated home use; it would be a wrong omission in an order that requires it. Preparation, waiting, cleanup, equipment access and practitioner capability belong in the comparison, not outside the price of the convenient change.
The instructor asks a trainee to explain those substitutions and build the home procedure from the supplied operations. Changing the card heading from “café” to “home” leaves an unavailable steam-wand step. Deleting that step leaves no operation producing the required milk. A supported proposal instead names the available frother operation, its input and result, when that result can be combined with the coffee, and the instructions and support needed to perform it. C.39.MF, C.39.RP, C.39.RQ, A.3.1.DR, A.3.1.QM, C.32.MWA.PP, B.1.5.CM supply this situated construction. A.3.2.DE makes the reasoning and instruction recoverable; A.15.8.MS ensures that the required support is available. Where the trainee cannot make the substitution, the instructor demonstrates the missing decision and uses representative practice and feedback through the applicable learning Method, rather than quietly completing the plan and crediting the trainee.
Suppose the constructed observations show that one trainee can explain and carry out the home proposal with the equipment instructions, while another can only repeat the old café sequence. Preserve the first bounded receiving result and the second's learning need. Do not report population retention or unaided competence for either. The school can retain the home procedure as a candidate variation with its origin, changed operations and unresolved evidence through A.3.1.MV; the café parent need not be declared false. If only the diagram became easier to read while the operations stayed the same, maintain that representation instead. If borrowing a tool made the unchanged operation possible, maintain the support arrangement.
Now change the receiving order: a café needs its specified presentation and repeated service within a short window. The home proposal's omission and equipment throughput no longer have the same basis. Recover those requirements and compare feasible operations, allocation and support; do not transfer the home result by the shared word “coffee”. This can preserve the café procedure and the bounded home candidate while reopening only the latter's proposed use at the café.
What carries across from MeCaMinD is the connection from receiving purpose to a justified selection or adaptation, from that decision to usable explanation and support, and from actual receiving work to a contribution worth preserving. What does not carry across is a movement-card typology, a coffee preparation rule, a teacher's authority or an empirical claim of successful cultural transmission. The school case is constructed; the MeCaMinD observations remain their own source record.
Bias-Annotation
Conformance Checklist
- A C.20 Discipline result is used, or a smaller C.36 practice and population boundary is named with the Discipline claim unresolved.
- The subject's status is preserved: admitted Method, with independently grounded enactment when the cultural claim relies on it, or candidate/observed practice plus identified candidate accounts, source claims, cards, templates, description epistemes, and any relied-on Work facts without Method,
U.MethodDescription, or enactment inflation. - The claims and their tests address the change sought by the group. A conclusion that depends on connected relations has grounds for those connections in the same practice.
- An existing direct predicate is used when sufficient; otherwise A.6.RCD returns the lightest disposition. A.6.REL occurrence identity is opened only when later use must distinguish occurrences.
- The world-side claim, claim episteme, evidence and reliance, and named-interval observations remain separate.
- Cards, descriptions with their recorded kind and status, publications, Systems, teaching/tutoring/event Work, rules, fields, forums, assignments, permissions, authority, and other direct relations are separately identified as subjects.
- A.22 is used only for one selected structure and C.30 only for a current architecture question.
- When the intended action requires authorization, its assignment, permission, or authority predicate, participants, permitted action, scope, positive basis, and protected conditions are explicit.
- The supported current account or continuation can finish without a fixed hypothesis count or new intervention. Credible action-changing rivals and relevant uncertainty remain; a selected inquiry has an attainable contribution worth its whole burden, including design effort, and feasible participants, access, resources, and window.
- C.27 is used only for a temporal adequacy claim, A.3.3 only for a transition law, C.28 only for causal reliance, A.15.7 only during ongoing Work, and C.11 only after a chooser and OptionSet exist.
- A claim of tested change or later retention uses the Work and interval observations it asserts, without automatically reidentifying one occurrence across intervals. Missing later evidence withholds only the dependent claim.
- Practitioner or constructed-Method consequences, non-use, burden, adaptation, and rejection remain separate from the cultural predicate.
- A transmission claim identifies what the recipient obtained from the claimed source and used. A capability or authorship claim identifies the decisions supplied and those the recipient made. The observed result can support continued joint work with the available help.
- An unresolved explanation remains
unknownwhere relevant without erasing an independently supported continuation, narrower claim, or stop. - Separate cases, populations, organizational Systems, histories, predicates, and effects are not joined.
- A return names either the particular defeated claim or edition, or a useful new contribution and its receiving use. A.3.1.MV preserves candidate status for a proposed reusable semantic difference; explanation content, a one-action representation, a multi-use profile and support retain their respective A.3.2.DE, C.37, C.37.MP and A.15.8.MS results. Origin, changed content and evidence limits remain visible. Mere novelty or population spread establishes none of these needs.
Common Anti-Patterns and How to Avoid Them
Consequences
Accounts of cultural change identify testable relations and their evidence. Teams can continue a sufficiently supported bounded practice, keep a qualified account, or choose an authorized worthwhile intervention. A claim of changed transmission, selection, memory, retention, or another relation uses its own observations rather than claiming the rest.
Practitioners can also turn a recipient's difficulty into the right contribution: a better explanation, a usable support arrangement, needed learning or a proposed change to the way itself. The reasons for the change become available to later recipients instead of remaining with the original teacher or expert. A successful variation can enrich the repertoire without invalidating an adequate parent.
The cost is patience and narrower conclusions. Long-term retention and loss need later observation, and visible carriers or successful local Work may leave population-level relations unresolved.
Rationale
Cultural continuity and change require testable predicates, not a bundle of labels. A card can participate in one transmission claim without being recognized, selected, or retained by a whole population. That claim also does not establish that the population enacts a Method described on the card; enactment concerns independently admitted Methods in Work. A mandatory field can participate in an organizational selection claim while training Work, later practitioner Work, Method admission, and enactment remain separate.
A practice can remain recognizable while its means change because the required result and relevant dependencies can be preserved across different operations or support. Making that reasoning available lets a recipient do more than imitate a surface sequence. Which differences are permissible remains a domain question. A cultural account then asks who obtained, selected or retained the resulting way; the constructive proposal does not answer that question by itself.
Separating subject status, world claim, claim episteme, evidence, interval observations, authorization, and consequence prevents candidate laundering and invented occurrence continuity. A bounded decision can use sufficient existing grounds while preserving unresolved explanations. When a discriminating inquiry is worth its feasible work, it supports deliberate change without fictitious authority or causality. Institutional, organizational, and project cases contribute without being merged into one history.
SoTA-Echoing
Practice question. When recipients can work with a combination of descriptions, tools and teaching, what warrants continuing that arrangement, developing a more adaptable explanation, or first studying which contribution makes the difference?
Selected answer. Preserve the bounded result established under the receiving conditions. Spend on further construction or discrimination only when a different receiving requirement, unavailable provision, credible failure or prospective gain makes its result consequential.
One serious alternative is to retain the complete demonstrated arrangement: provide the cards, board, teaching and tutoring together, without first making every adaptation decision independently recoverable. That is the better answer when recipients need the same result, the provision remains obtainable, and its whole burden is acceptable. Section 4.5 must not turn such a success into a compulsory reconstruction project. If another recipient must adapt the practice under changed conditions or cannot obtain a necessary teacher's contribution, repeating the arrangement no longer answers that use. The additional construction makes the consequential choice, its grounds, the changed operation and the remaining provision recoverable. It costs explanation, practice and feedback work; it earns that cost only through the receiving action it enables. Sections 4.5 and 5.4 work through this difference without assuming that a shorter card or a new example teaches adaptation.
A second serious alternative is to isolate the effects of material, teaching and tutoring before deciding how to continue. Such a study can be preferable when a consequential provision choice depends on their separate effects and a discriminating comparison is feasible. It is unnecessary for a claim restricted to successful work with the observed combination. MeCaMinD therefore yields the combined-conditions account in section 5 while card-only transmission remains unresolved. This accepts weaker causal knowledge without weakening the stated observation. Section 4.1, steps 8–11, preserves that boundary and requires the whole cost, delay, participant burden and displaced work to enter any new inquiry decision. The opposed constructed security-assessment cases in section 5.1 show why an additional inquiry or repeated assessment can either protect an uncovered consequence or merely displace useful work.
Waern et al. supply observed combined provision and difficulties with literal use; they are comparison evidence, not an experiment isolating that provision. Ardi et al. supply observed use, non-use and burden; the opposed requirement choices are this pattern's constructed applications, not an empirical ranking reported by the study. Adapt those distinctions; retain a sufficient arrangement. Reopen this comparison when the receiving use requires a stronger claim, the necessary provision changes, a recipient cannot make the consequential adaptation, or a feasible rival offers a better result at comparable whole burden. A new causal or retention claim needs its own evidence.
Reopen the pattern when a field case supplies decision-changing evidence for another named cultural predicate, when practitioners cannot derive a testable claim or bound authorization and population, when repeated probes reveal an independent measurement problem, or when current C.20, C.36, A.6.RCD, or A.6.REL semantics change the action.
Relations
- A.3.1.MV supplies either an admitted Method variant or a status-preserved candidate lineage. A.3.1.MI can supply bounded introduction and later-practice observations. An admitted-Method cultural claim needs independent A.15.1 enactment when its predicate relies on actual performance of that Method; the candidate branch preserves its candidate/source accounts and description epistemes under their own maintained results and asserts neither Method nor enactment. Neither input establishes a cultural predicate.
C.20governs Discipline recognition andC.36the bounded cultural-evolution question.A.6.RCDderives the needed bounded relation-bearing claim and stops at the lightest disposition.A.6.RELsupplies occurrence identity only when a later use must distinguish obtaining occurrences.A.3.1governs Method admission,A.3.2U.MethodDescriptionmembership,A.13precise Agent bases,A.15.1intervention and receiving Work plus a namedenactsMethodoccurrence, andA.10evidence and reliance. Candidate accounts, source labels, carriers, descriptions without that membership basis, and teaching events substitute for none of them.C.11.DUAappraises a questionable inquiry or requirement and connects its receiving contribution to its whole burden and feasible continuation. It does not supply permission, domain experimental design, or performed Work.B.5andB.5.2govern durable hypothesis generation and consequences.C.27applies only to a temporal adequacy question,A.3.3only to a state-space and transition-law claim,C.28only to causal reliance,A.15.7only to a next action in ongoing Work, andC.11only after a chooser and OptionSet exist.- A.19 comparison and selected-set semantics govern non-trivial alternatives.
A.22applies only when one selected organization of identified constituents, obtaining relations, constraints, and use frame changes the decision;C.30only when an architecture claim is current. A.3.4governs any separate claim that a continuing card, description, System, selected structure, organizational System, or other subject underwent one bounded change. Intervention Work and a revised edition do not establish thatU.Transformationby themselves.- Description, publication, teaching, tool, support, rule, field, forum, assignment, permission, authority, Work and direct-relation changes retain their direct governors. Section 4.5 connects C.39.MF, C.39.RP, C.39.RQ, A.3.1.DR, A.3.1.QM, C.32.MWA.PP, B.1.5.CM, A.3.2.DE, C.37.MP, A.15.8.MS and A.3.1.MV with the separately qualified capability and support results for the recipient; it does not replace the Methods that obtain those results. A return names a defeated claim or a useful new contribution with its receiving use, origin, changed content and evidence limits. A.3.1.MV preserves the candidate status of a proposed reusable difference. A.3.2.DE can author or revise MethodDescription or candidate-account content. C.37 retains a one-action representation; C.37.MP retains an applicable profile of unlike uses; A.15.8.MS retains the support result. A cultural claim or population spread alone creates no such change.
C.36.MC:End
Last Updated: 2026-11-22 — upstream FPF commit 7a93d8f8 (github.com/ailev/FPF)