A.6.4 - EntityOfConcern Retargeting: Judge a Bounded Use Across Different Entities
Status: Stable Type: Definitional pattern
One-line summary. Use EntityOfConcern retargeting to classify an arrow between epistemes about different entities. A claimed receiving use needs a separate bounded-use assertion; a current-case judgement compares that assertion with the relevant facts.
Retargeting in plain terms. The two epistemes concern different things. One named receiving use is admitted only when facts establish the stated invariant and show that the visible loss remains acceptable under named conditions.
Use this when. Use this pattern only after C.2.1 identifies the two epistemes and shows that their EntitiesOfConcern differ. A changed model kind, ontology frame, predicate set, coordinate system, or notation is a cue to repeat that identity test, not proof of retargeting.
What goes wrong if missed. A changed EntityOfConcern is treated as “the same thing in another form”, so claims, evidence, gate results, work authority, or currentness are carried forward without their own direct predicates or current results. The opposite error is to demand a semantic Bridge or reversible mapping when the case needs neither.
First useful move. Identify r by its endpoint epistemes, arrow rule or designator, and the selected formal substrate’s equivalence criterion. Name both EntitiesOfConcern. For a claimed receiving use, state the invariant, visible loss, named use and conditions in a separate bounded-use assertion q. When judging a current case, compare its facts with q.
What this buys. The reader can decide one receiving use from the bounded-use assertion q and current facts; source-claim survival, performed Work, and EntityOfConcern identity retain their separate bases.
Not this pattern when. If the EntityOfConcern is preserved, use the pattern for the change that occurred: A.6.3.CR for faithful restatement, including translation, A.6.3.RT when a change of representation language or reasoning medium enables the intended operation, A.6.3.CSC for controlled coarsening, or E.17.EFP for explanation mode. A normal time-to-frequency description of the same signal is an A.6.3.RT case. Use C.29 when a separate mathematical representation or transfer question remains; bring its correspondence and limits back to that representation change. Use F.9 only for a separately claimed Bridge between two local senses. Keep every other independently current claim under its direct pattern; the Relations section names the main neighboring owners once.
A.6.4:1 - Problem frame
Several familiar moves can hide a change of EntityOfConcern, but none proves that change merely by its name or notation:
- Physical cabinet and selected functional structure. An episteme about cabinet
Cab-7and one about selected functionalU.StructureRoute-Aconcern different entities when C.2.1 identifies them independently. A behavior-test result can supply facts for a bounded retargeting judgement. The expressionRealises(Cab-7, Route-A)contributes only if its own relation and obtaining conditions are established; §5 shows a case whose judgement needs no such relation. - Signal and spectrum. The ordinary Fourier case often concerns one signal in two representations. Use A.6.3.RT when the EntityOfConcern is preserved. Add C.29 only when the needed mathematical correspondence or its limits remain unsettled. A.6.4 opens only if the receiving episteme concerns a separately identified spectrum object and the use explains why that object, rather than the original signal under another representation, is current.
- Observations and fitted model. A dataset and a learned model can be different entities. An affirmative q may state a predictive invariant and name a prediction use; the fit and held-out test outcomes then supply current-case facts for the separate judgement. Individual observations and unmodelled distinctions remain visible losses. Model fitting is a separate activity.
For a case used positively, the local arrow r relates two epistemes with different EntitiesOfConcern, q affirmatively states the bounded-use proposition, and the case facts satisfy it. Any claimed operation application or Work is identified separately under its own conditions.
A domain relation, mathematical transform, or F.9 Bridge contributes only the claim established by its own governor. C.2.1 independently identifies the two EntitiesOfConcern.
A.6.4:2 - Problem
Without this discipline:
- Notation decides ontology. A changed coordinate system, mathematical domain, model kind, or predicate vocabulary is treated as proof that the EntityOfConcern changed.
- Retargeting is confused with viewing. A move from one independently identified entity to another is called another view, so its loss and receiving-use boundary disappear.
- The invariant and loss remain rhetoric. Phrases such as “energy is preserved” or “the model summarizes the data” do not say which claim survives, which distinctions disappear, or which use remains sound.
- A mapping apparatus replaces the case. A generic kind bridge, score, diagram, or reversible optic is demanded even when the endpoints, q, and current facts already let the reader judge the named use.
- Arrow, claim, and execution collapse. A mathematical arrow is treated as if it granted a use, performed an operation, or produced an episteme.
- A transformation-flow position is mistaken for a use judgement. Placing r and q in E.18 does not establish the invariant or acceptable loss for the receiving use.
- Neighboring changes disappear into one label. Grounding, representation, scope, operating conditions, viewpoint selection, publication, operation application, and performed Work are folded into retargeting instead of being identified only when they occur.
A.6.4:3 - Forces
- Different subject versus unsupported claim. The endpoint EntityOfConcern may change while every receiving assertion remains subject to A.6.2 P2. A separate q states which invariant and loss support one receiving use.
- Useful loss versus hidden loss. Retargeting may discard information, but the loss boundary and the use that tolerates it must be visible.
- Direct case versus universal apparatus. Use a domain relation, mathematical map, semantic Bridge, or diagram as current-case basis only when its own governor establishes the needed claim. Identify r from its endpoints, arrow rule or designator, and formal equivalence; judge q separately from current facts.
- Composition versus accidental equivalence. Compatible retargetings may compose. Equality of two evaluation routes, reversibility, idempotency, or semantic correspondence requires its own stated conditions; it does not follow from the word retargeting.
- Modularity. The retargeting arrow relates two epistemes, q states one bounded-use proposition, and a separate current-case judgement says whether the facts satisfy it, fail it, or leave it undecidable. Grounding, publication, Work, evidence, assurance, gate, flow structure, and cross-local-sense correspondence remain separate.
A.6.4:4 - Solution — separate the arrow, use claim, current-case judgement, and any application
A.6.4:4.1 - Informal definition
Definition. An EntityOfConcern-retargeting morphism is a local
EpMorphism r : X -> Ywhose endpoint epistemes concern different entities. A separate bounded-use assertion q affirms or denies that one declared invariant makes the stated loss acceptable for one named receiving use under named conditions.
EntityOfConcernRetargetingMorphism is a local mathematical subtype in the selected formal substrate, not a durable kind. This pattern defines that subtype and the practical discipline for claims about its use.
Keep four things distinct:
- The arrow
r. Within the selected formal substrate, its domain X, codomain Y, arrow rule or designator, and declared formal equivalence identify it. The two endpoint epistemes and their different EntitiesOfConcern are recoverable. A changed use claim does not create another arrow. - The bounded-use assertion
q. This is a C.2.1 episteme about arrow r. Its ClaimGraph states the invariant, visible loss, named receiving use, conditions, and affirmative or negative polarity. Its complete claim content, EntityOfConcern, and effective ReferenceScheme identify q. A citation inside q can point to case facts; it does not decide whether those facts satisfy the proposition. - The current-case judgement. Compare the current facts with q’s conditions and proposition, and report
satisfies,fails, orcannot decide. That result is not q’s polarity and does not reidentify q or r. Use A.20 only when the case raises an internal-constraint check, A.10 only for a current evidence-use claim, and B.3 only for a current assurance claim or its material-reliance threshold. Otherwise the named rule and direct case facts are enough. - Any application occurrence. If a system computes, authors, or otherwise produces or changes an episteme by using the declared operation, identify that A.6.1 application, its argument and result bindings, the performing system, and any Work separately. The mathematical statement
r : X -> Yalone names no occurrence.
For a current-case judgement of a receiving-use assertion, ask:
| Question | What it recovers |
|---|---|
| Which arrow relates the source and receiving epistemes? | r’s endpoints X and Y, arrow rule or designator, and selected formal substrate’s equivalence criterion |
| Which different entities do they concern? | the independently identified EntityOfConcern pair |
| What does q affirm or deny? | invariant, visible loss, named receiving use, conditions, and polarity |
| Which current facts bear on that proposition? | the direct case basis |
| What do those facts show? | satisfies, fails, or cannot decide |
| If the case cannot be decided, what is missing? | the missing fact and reopen condition |
These answers may be one short paragraph; they require no new record form or assurance package. Add a separately governed commitment change, neighboring claim, or durable result only when it changes q, the judgement, or the receiving action; ER-1 and CC-A.6.4-5 name the values to inspect. Add an F.9 Bridge only when the same case separately claims a semantic relation between two F.17 local senses.
When the judgement of an affirmative q is fails, retain q as the stated proposition but do not admit that case. When a current-case judgement is cannot decide, keep the source material, name the missing fact and what would reopen the question, and stop. Failure of an affirmative q does not by itself establish a negative q; a negative assertion needs its own claim content and case basis.
A.6.4:4.2 - Formal declaration and object boundaries
When a receiving use needs a reusable formal declaration, an A.6.0 U.Signature(profile=FormalSubstrate) episteme can declare the local subtype EntityOfConcernRetargetingMorphism. The declaration remains distinct from the subtype, an arrow, a use claim and an application.
SubjectKind = local formal subtype EntityOfConcernRetargetingMorphism of EpMorphism
RangedValueKind = admitted ordered-pair range over U.Episteme values satisfying the declared endpoint-kind constraints
ResultKind = omitted; r is the declared subject, not an operation result
Applicability = selected formal substrate and endpoint and arrow-family conditions
X and Y are C.2.1 epistemes. r : X -> Y is one local mathematical arrow in the selected formal substrate. Its identity uses the endpoints, arrow rule or designator, and the selected substrate’s equivalence criterion; the endpoints alone do not identify it. The declaration states which parts of X and Y’s claim content, EntityOfConcern, and effective ReferenceScheme remain the same or differ. If r’s rule reads a representation or another separately obtaining relation, it names the occurrence and compares endpoint facts without changing that occurrence.
A.6.4 reuses the one A.6.2 formal model: category Ep, endpoint-only thin category EoCBase, dom, cod, identities, compose, and the declared mapping α. For retargeting arrow r, α(r)=u_{α(X),α(Y)} is the unique formal endpoint arrow between the independently different EntitiesOfConcern. It records only that endpoint difference and forgets r’s arrow rule; it is not an independently declared domain or world-side relation. The local classification function entityOfConcernChangeMode returns retarget for r and records the same endpoint difference. It classifies only the endpoint-change mode; it adds no domain-function evaluation or second retargeting calculus.
The bounded-use assertion q, current-case judgement, and any application occurrence remain separate. Add A.6.5 SlotSpecs only inside a reusable direct-relation declaration; they are not fields of r, X, Y, or q.
A.6.4:4.3 - Laws (ER-0…ER-6)
These laws refine A.6.2 for the local retargeting subtype. Every assertion in Y remains subject to the parent P2 support condition, including assertions outside the invariant named by q. The laws do not assert durable U-kind membership.
ER-0 - Arrow class and endpoint basis.
An arrow r : X -> Y is in the local retargeting subtype only when X and Y are C.2.1 epistemes, entityOfConcernChangeMode(r)=retarget, and their EntitiesOfConcern differ. A shared label, kind name, diagram, implementation, use claim, or F.9 card identifies neither r nor its endpoints by itself.
ER-1 - Arrow identity and neighboring facts.
- The selected formal substrate supplies r’s arrow rule or designator and equivalence criterion; same endpoints alone do not identify an arrow.
- The declaration states which parts of X and Y’s claim content, EntityOfConcern, and effective ReferenceScheme remain the same or differ. It names any separately obtaining representation or other relation that r’s rule reads and the endpoint facts compared; r does not change that occurrence.
- Grounding, scope, operating condition, representation, and any viewpoint selected for a describing use remain separate values or relations.
- A different scheme, scope, context, or plane does not by itself create an F.9 Bridge. Cite F.9 only for a claimed direct relation between two F.17 local senses.
ER-2 - Separate use proposition and current-case judgement.
For each named receiving use, one separate C.2.1 assertion q states an affirmative or negative proposition about whether the source claims preserve the declared invariant in the receiving episteme and whether the visible loss is acceptable under the named conditions. The same r may have different q assertions for different uses without changing arrow identity.
A separate current-case judgement applies q to the current facts and returns satisfies, fails, or cannot decide. A direct fact, proof, test, or obtaining relation can enter the ordinary case basis through its own governor. Open A.20 only for an internal-constraint claim, A.10 only for evidence use, and B.3 only for assurance or its material-reliance threshold. Each retains its own identity; q’s polarity stays as written, and cannot decide is reopened when its named missing fact becomes available.
ER-3 - Composition and separately claimed final use.
Two retargeting arrows with an exact matching middle episteme compose in the parent Ep category; A.6.2 category closure supplies the composite and requires it to satisfy the parent laws. The composite remains in the A.6.4 retargeting subtype only when its final endpoint EntitiesOfConcern differ and its other subtype laws hold. A round trip whose final endpoints concern the same exact entity is a preserve-mode EFEM arrow in the parent class, not an A.6.4 retargeting arrow.
A claim that an admitted composite is suitable for a final use is another q: it states the final source and receiving entities, preserved invariant, accumulated loss, receiving use, conditions, and polarity. A separate judgement applies that proposition to the final current case.
No universal SquareLaw follows. A consumer that claims two evaluation routes equivalent, or relies on a correspondence between epistemes, identifies the routes or correspondence, comparison rule, tolerated difference, and witness under the direct governor of that claim.
ER-4 - Determinism and repeat boundary.
Functionality, reversibility or equivalence of routes requires the selected formal substrate’s domain, equality or equivalence and applicable witness. A retargeting arrow r : X -> Y has different endpoint EntitiesOfConcern, hence X ≠ Y; it is not self-composable, so compose(r,r) cannot express an idempotence law for that arrow. A repeat property of a construction operation is a separate claim about its rule and inputs. A.6.3:4.7 illustrates the first-arrow/next-arrow form when that repeat returns the same receiving episteme. Ambient time, randomness, solver state or external services must be accounted for when they can change the operation’s result.
ER-5 - Applicability and optional semantic-Bridge branch.
The formal declaration states admissible endpoint families and material mathematical conditions. Each q separately states the invariant, loss boundary, receiving use, case conditions, and affirmative or negative polarity; the current-case judgement states whether the facts satisfy it. F.9 is triggered only for a separately claimed relation between two local senses. Optional CL summarizes evidence about that Bridge; it is neither a retargeting threshold nor a participant in r or q.
A named mapping or relation without a governing definition supplies no additional claim. If the receiving use needs that claim, identify its direct governor or leave the missing-governor question unresolved.
ER-6 - Separate application, Work, and resulting episteme.
An arrow that preserves the EntityOfConcern remains in A.6.2 preserve mode. A.6.3 additionally requires its viewing construction and laws. Describe measuring, fitting, translating or authoring in ordinary practitioner terms. Identify an A.6.1 application and its bindings when a declared operation is current. A dated Work claim follows A.13 and independent A.15.1 admission, with F.6 only for precise assignment-bound attribution. Identify any resulting episteme under C.2.1; the mathematical arrow relates endpoints without performing that activity.
A.6.4:4.4 - Boundary with representation, explanation, transformation-flow structure, and neighboring claims
A.6.4 is triggered only by an independently established change of EntityOfConcern. A changed kind, ontology frame, predicate set, mathematical domain, or notation is a recognition cue that reopens the C.2.1 identity test; none decides the branch by itself.
Boundary rules:
- If the EntityOfConcern stays the same, use
A.6.3.RTwhen the intended operation depends on changing the representation language or reasoning medium. - if the EntityOfConcern is preserved and the main change is explanation mode, explanatory stance, or explanation-facing publication, use
E.17.EFP; - if the same case also asserts a semantic relation between two local senses from different semantic contexts, test
F.9separately and cite a Bridge only when its predicate obtains; useF.9.1only for an optional stance note about that already constituted use claim. A domain correspondence, mathematical rule, or direct case fact enters the current-case basis only through the claim established by its governor; that claim alone does not open F.9; - if the receiving use requires a mapping, correspondence or route-equivalence claim whose direct governor is missing, resolve that missing basis before relying on the claim;
- if the EntityOfConcern is preserved and the receiving item is useful only under narrower declared use with visible loss and source-bearing reopen, use
A.6.3.CSC; - if a
StructuralReinterpretation,PathSliceId,CrossingRef, orDecisionLogRefis present, useE.18,A.20, orA.21for graph, path, constraint, and gate relations. Those references do not prove semantic continuity or retargeting admissibility by themselves; - for decoded or latent output, use C.2.6 for source and use claims; use C.2.7 when representation-factor readings matter. For the named use, establish the source or model-state relations, decode or probe relations, and recoverability evidence it requires. Retain supported claims and uses; resolve a missing premise before relying on a use that depends on it. State any source-return action that use requires under its direct governor;
- B.5.2 governs changes within its abductive loop, including its prompt, candidate and rival sets, selected prime hypothesis, plausibility filtering, and reopen. Apply A.6.4 additionally only for an independently established EntityOfConcern change. State q when a receiving use is claimed, and judge a current case when that result is needed;
- keep any other independently current claim under the pattern that defines or tests it; the Relations section names the main neighboring owners.
A.6.4 defines arrow r, bounded-use assertion q, and the separate current-case judgement that E.18 may place at a StructuralReinterpretation locus. That placement identifies none of them and does not make the judgement satisfies.
A.6.4:5 - Archetypal Grounding (Tell-Show-Show)
Tell. Retargeting relates epistemes about different EntitiesOfConcern. For a claimed receiving use, state the invariant, visible loss and conditions; judge a current case from its facts when that judgement is needed.
Show 1 — Physical cabinet to selected functional structure. Take Cab-7 and functional U.Structure Route-A as independently identified entities. Under an explicit finite-routing scheme, source episteme X states two observations of Cab-7 in configuration K: input L routed to output 1, and input R routed to output 2. X also identifies cabinet layout Layout-7 and manufacturer Maker-7. A separate definition episteme Z specifies Route-A with those pairs and a third pair M→3. Receiving episteme Y concerns Route-A and selects only its declared L→1 and R→2 pairs. C.2.1 distinguishes X, Y and Z by their complete claims, concerns and effective schemes.
In the selected finite-table substrate, r relates X to Y by the rule TestedRowProjection: match the named input rows and use Z’s definition for the receiving structure assertions. Arrow equivalence uses the same endpoints and rowwise equality of that rule. This does not infer Route-A’s definition from cabinet behavior alone.
Affirmative q states: Under K, Cab-7's observed outputs for {L,R} match Route-A's declared outputs. Dropping cabinet layout, manufacturer details and Route-A's M branch is acceptable for preparing a fault-isolation check list restricted to the observed inputs L and R. The current case compares L→1 with L→1 and R→2 with R→2; both match. The resulting check list is “for L, inspect output 1; for R, inspect output 2”; that use needs neither the omitted cabinet details nor the M branch. The judgement is therefore satisfies for this bounded use. A changed configuration or another input requires new case facts. Missing output observations yield cannot decide; an observed mismatch yields fails without changing q’s polarity.
The unused expression Realises(Cab-7, Route-A) has no supplied direct predicate or governor in this fixture. It contributes nothing to the judgement. Placement of r and q in E.18 likewise supplies no case fact.
Show 2 — Fourier near-miss and positive branch. In the ordinary case, X and Y both concern sampled signal run Signal-17; X uses a time-domain representation and Y a frequency-domain representation. Use A.6.3.RT for this representation change. Under its declared conditions, Parseval’s relation gives the energy-preservation equality; the spectrum notation still represents the same EntityOfConcern. If that mathematical correspondence or its conditions are unsettled for the receiving use, use C.29 to establish their adequacy before relying on the equality.
A positive A.6.4 branch opens only if C.2.1 separately identifies, for example, signal run Signal-17 and spectral-distribution object Spectrum-17 as the two EntitiesOfConcern. The receiving use must concern Spectrum-17—for example, comparing its peak distribution with another spectrum—rather than merely read another representation of Signal-17. Then r relates the two epistemes; affirmative q states the spectral-comparison proposition and may cite the Fourier relation and Parseval test. The current-case judgement is satisfies only when the named facts meet q’s conditions and criterion while lost time localization remains visible.
Show 3 — Dataset to model. X concerns dataset D; Y concerns fitted model M, independently identified under the applicable model pattern. Affirmative q states the predictive invariant, visible losses, named prediction use, and conditions. The fit result and held-out test outcome are current-case facts; the judgement is satisfies when those facts meet q’s conditions and invariant criterion. Individual observations and unmodelled distinctions remain visible losses, and any other dataset claim needs its own transfer basis. Identify any fitting application and performed Work separately under ER-6.
A.6.4:6 - Bias-Annotation
A.6.4 foregrounds the source and receiving entities when the EntityOfConcern changes. Recover that pair. For a claimed receiving use, state q’s invariant, visible loss, use and conditions. When judging a current case, recover the facts needed for that judgement. Open a neighboring branch only when a separately current claim needs its direct pattern.
A.6.4:7 - Conformance Checklist (normative)
CC-A.6.4-1 - Endpoints and changed EntityOfConcern. C.2.1 identifies X and Y and their EntitiesOfConcern; the two entities differ. A changed kind, frame, predicate set, domain, or notation alone does not pass this check.
CC-A.6.4-2 - Arrow identity. The selected formal substrate supplies r’s endpoints, arrow rule or designator, and equivalence criterion. Same endpoints, a diagram, or a use claim alone does not identify r.
CC-A.6.4-3 - Separate use proposition and case judgement. For a claimed receiving use, one C.2.1 assertion q names r, the receiving use, invariant, visible loss, conditions, and affirmative or negative polarity. When evaluating a current case, a separate judgement reports satisfies, fails, or cannot decide from current facts. The same r may have another q and judgement for another use.
CC-A.6.4-4 - Conservative receiving claim. Every assertion in Y satisfies A.6.2 P2. For an affirmative q, satisfies additionally requires enough case basis for q’s invariant, loss condition and stated use. For either polarity, facts contrary to q yield fails; a missing deciding fact yields cannot decide and names that fact and the reopen condition. Neither result changes q’s polarity.
CC-A.6.4-5 - Triggered additions only. For a Description or specification-use episteme, name every material change to claim content, effective scheme, grounding, scope, operating condition, or selected viewpoint under A.7 and E.10.D2. Add those values, evidence, currentness, a route-equivalence test, or a reopen condition only when they change q or the reader’s action.
CC-A.6.4-6 - Separate semantic correspondence. Test an F.9 Bridge only when the case also claims a relation between two local senses. The Bridge, its bounded-use claim, optional CL, evidence, and reliance remain separate from r and q.
CC-A.6.4-7 - Separate application and Work. For measurement, computation, actuation, model fitting, authoring, and other effects, identify the operation application and its argument and result bindings when an A.6.1 operation declaration is current; identify any performed Work separately. The arrow statement r : X -> Y neither identifies an application or Work occurrence nor proves a production relation.
CC-A.6.4-8 - Fourier boundary. A same-signal time/frequency change uses A.6.3.RT; C.29 is added when a separate mathematical representation or transfer question remains. A.6.4 is used only after the receiving spectrum or other mathematical object is independently identified as a different EntityOfConcern.
CC-A.6.4-9 - StructuralReinterpretation boundary. E.18 governs structure position, path, crossing, and gate relations. A.20 tests q’s proposition only when an internal constraint is current, and A.21 governs any gate decision. None identifies r or supplies a satisfies judgement merely by reference or placement.
CC-A.6.4-10 - Stop and ordinary use. A missing deciding fact yields cannot decide and names the fact and reopen condition; contrary facts yield fails. Otherwise one short paragraph answering the six practical questions is enough. No separate evidence, assurance, publication, currentness, or reusable declaration is required unless its own use condition is current.
A.6.4:8 - Common Anti-Patterns and How to Avoid Them
| Anti-pattern | Why it fails | Correct action |
|---|---|---|
| Retargeting as viewing | A changed EntityOfConcern is treated as the same object under another viewpoint. | Resolve EntityOfConcernRef to the entity: same-entity arrows remain in A.6.2 preserve mode and qualify for A.6.3 only under its viewing laws; use A.6.4 when the independently identified entities differ and its retargeting laws hold. |
| Retargeting as publication rendering | A diagram, export, or face is treated as the arrow or as support for its use. | Keep publication forms in E.17 and E.24.PUB; state r and the separate use claim q only when each is current. |
| Universal Bridge as admission | A KindBridge, F.9 Bridge, CL, mapping, or optic is required or used to inherit every downstream claim. | Use the A.6.4 minimum basis; add F.9 only for a separate local-sense relation and state every neighboring claim under its own rule. |
| Mathematical notation decides retargeting | A Fourier, graph, path, or category representation is treated as proof that the EntityOfConcern changed. | Use C.29 for the mathematical lens and repeat the C.2.1 identity test. Use A.6.3.RT when the entity is preserved; use A.6.4 only for independently different entities. |
A.6.4:9 - Consequences
- Viewing and retargeting separate. A viewing arrow preserves the EntityOfConcern. A retargeting arrow relates epistemes with independently different EntitiesOfConcern; q states one bounded-use proposition, and the separate current-case judgement says whether the facts satisfy it.
- StructuralReinterpretation receives one core rule. E.18 can place r and q without duplicating their identities; the A.6.4 current-case basis remains the facts and separate judgement.
- Loss becomes usable information. A lossy mapping can serve one bounded purpose; reversibility and semantic identity remain separately established claims.
- Optional apparatus stays optional. F.9 enters only for cross-local-sense correspondence; route-equivalence, evidence, assurance, gate, publication, and Work branches enter only when their own claims are current.
- Description boundaries remain visible. Claim content, scheme, grounding, scope, operating condition, and viewpoint changes do not disappear into one retargeting bundle.
A.6.4:10 - Rationale
A.6.4 exists because some mathematical arrows relate epistemes that concern different entities. A changed use claim does not create another arrow; the reader decides each receiving use from its q and current case facts.
A.6.4:11 - SoTA-Echoing
Practice question. How can a transformation definition, its execution and a correctness claim remain separate, and which independent test establishes that the source and receiving epistemes concern different entities?
| Source or practice | Contribution | Limit |
|---|---|---|
| Zhao et al., KBX: Verified Model Synchronization via Formal Bidirectional Transformation (2024) | KBX separates formal bidirectional-transformation definitions, generation of a synchronizer, and consistency verification. | This supports the declaration, application, and use-claim split. KBX synchronizes models; it does not decide FPF EntityOfConcern identity or make one bounded use sound. |
| He and Zan, BIT: A template-based approach to incremental and bidirectional model-to-text transformation (2024) | BIT distinguishes a usable surface language, a formally defined core, executable printer/parser behavior, round-trip properties, and empirical cases. | This supports keeping readable first use, formal declaration, execution, and well-behavedness evidence distinct. BIT’s model/text synchronization does not decide whether two FPF epistemes concern different entities. |
| FPF C.2.1, C.29, and A.6.3.RT | C.2.1 identifies each episteme and EntityOfConcern; C.29 bounds the mathematical lens; A.6.3.RT handles representation change with preserved EntityOfConcern. | These are the direct identity and routing rules. |
| Fibrations, cospans, Fourier transforms, and data/model mappings | These provide mathematical lineage and stress cases for endpoints, composition, invariants, and loss. | None proves that the EntityOfConcern changed or that a receiving use is sound. |
The A.6.4 split among r, q, and any application occurrence is a bounded FPF synthesis from these distinctions, not an externally established retargeting ontology.
A.6.4:12 - Mini-checklist (for use)
Start with questions 1–2 to classify the arrow. Questions 3–4 concern a claimed receiving use; question 5 concerns its evaluation in a current case. Questions 6–7 apply when those additional claims are current.
- Does the EntityOfConcern change? If no, use A.6.3 or another preserving pattern.
- Which arrow, endpoint epistemes and EntitiesOfConcern are involved? Name them before naming a mapping technology.
- Which invariant does q state? Name its proposition and case assumptions.
- What is lost, and which receiving use tolerates that loss? A broad “same meaning” answer is insufficient.
- What witnesses the invariant and loss judgement? If the witness is missing or contradicted, stop or reopen.
- Is a relation between two local senses also claimed? Only then test F.9 separately; no Bridge follows merely from retargeting.
- Was computation or another activity performed? When an A.6.1 operation declaration is current, identify the application and its argument and result bindings. Identify any performed Work separately from r and q.
A.6.4:13 - Relations
- Builds on. A.6.0 for a reusable FormalSubstrate declaration; A.6.2 for the local arrow discipline; A.6.3 for the preserved-EntityOfConcern neighboring branch; C.2.1 for episteme, EntityOfConcern, and use-assertion identity; C.29 for mathematical-lens use; A.6.3.RT for preserved-EntityOfConcern representation transitions; A.6.5 for SlotSpecs inside a reusable direct-relation declaration; A.7 and E.10.D2 for Description and specification-use boundaries; C.2 and C.3 or the relevant domain pattern for the invariant; and F.9 only for a separately claimed relation between local senses.
- Consumed by. E.18 may place r and q at a
StructuralReinterpretationlocus; A.20 may test the proposition carried by q; E.17 may publish an episteme that describes the case; KD-CAL and LOG-CAL may reason over a stated invariant. None redefines r or q. - Neighbor boundaries.
A.6.1andA.15govern an application and Work;A.10governs evidence;B.3governs assurance and reliance;A.21governs gate decisions;C.27governs temporal adequacy;A.3.3governs dynamics; the applicable direct control pattern governs control; andE.24.PUBgoverns publication. C.2.6 recovers source and use claims, and C.2.7 governs representation factors; B.5.2 governs its abductive-loop state.