#ConservativeRetextualization - entityOfConcernRef-preserving textual re-expression

A.6.3.CR · Stable · Specialization pattern · Normative · Part A - Kernel Architecture Cluster

#About this pattern

This is a generated FPF pattern page projected from the published FPF source. It is canonical FPF content for this ID; it is not a FPF Reference product feature page.

#How to use this pattern

Read the ID, status, type, and normativity first. Use the content for exact wording, the relations for adjacent concepts, and citations to keep active work grounded without pasting the whole specification.

Type: Specialization pattern Status: Stable Normativity: Normative

Use this pattern when one already available source line about the same EntityOfConcern needs a second textual form: a report rewrite, summary, translation, or declared filtered restatement. The real job is still same-entity textual re-expression, not explanation, representation change, bridge work, retargeting, evidence, gate authority, or work authorization.

#Relations

#Content

#Problem frame

Use this pattern when one already available source line about the same EntityOfConcern needs a second textual form: a report rewrite, summary, translation, or declared filtered restatement. The real job is still same-entity textual re-expression, not explanation, representation change, bridge work, retargeting, evidence, gate authority, or work authorization.

Primary EntityOfConcern. The EntityOfConcern is one published textual rendering over the same EntityOfConcern line. It is not the whole source corpus, not an explanation face, not a downstream decision, and not a publication with a new authority-reference relation.

First useful move. Separate the source slice, the published slice, the omission or source-loss note, the admissible use, and the neighboring governing pattern that must take over if the rewrite stops being conservative.

What goes wrong if missed. A summary, translation, or manager-readable rewrite is treated as harmless editing after it has started hiding explanation work, bridge work, changed authority relation, or a narrower-use card.

What this buys. One honest same-entity textual rewrite with visible source tether, visible omission or loss notes, and a named governing pattern when the case stops being only conservative retextualization.

Ordinary use. If the rewrite is admissible only for orientation, source-finding, review, comparison, or planning preparation, one source-slice to published-slice sentence or mini-card with the admissible use and visible omission or source-loss note is enough.

Reliance-facing use. Open the fuller rewrite-admissibility record only when the rewritten text will be externally relied on, disputed, cited as a source-relation reason, used across context, or read as release, gate, work-preparation, engineering-justification, approval, or evidence justification.

Not this pattern when. Not this pattern when the case is primarily explanatory rendering (ExplanationFaithfulnessProfile), representation-scheme change (Representation-Scheme Transition), changed EntityOfConcern (A.6.4), comparative review (E.17.ID.CR), bridge or substitution use (F.9 or F.9.1), or a deliberately coarsened rendering whose narrower admissible use, non-admissible downstream use, and source-bearing return card has become primary. In that last case, use A.6.3.CSC Controlled Semantic Coarsening.

#Problem

Without a dedicated pattern for conservative textual re-expression:

1. report, summary, translation, and filtered rewrite cases are handled ad hoc;
2. authors treat textual simplification as if it were automatically conservative;
3. the boundary to explanation-facing renderings stays blurry;
4. correspondence-mediated rewrites are not distinguished from direct rewrites;
5. subsequent users cannot tell whether the result is still a view of the same EntityOfConcern or a new interpretive publication.

#Forces

• Same entity, different wording. Readers need different textual forms without reopening the EntityOfConcern.
• Compression vs loss visibility. Shorter or plainer forms are often useful, but omissions and source-loss modes must stay explicit.
• Direct vs correspondence-mediated rewrites. Some rewrites read from one source episteme; others depend on a declared CorrespondenceModel.
• Textual focus vs family creep. The pattern should cover same-entity textual re-expression, not explanation, not representation-wide shifts, and not retargeting.
• Publication discipline. Admissible MVPK faces and publication renderings still matter even when the transform looks like "just a rewrite."

#Solution — entityOfConcernRef-preserving textual re-expression under A.6.3

#Informal definition

ConservativeRetextualization is a named pattern specialized under A.6.3 U.EpistemicViewing for textual re-expression of the same EntityOfConcern.

It preserves entityOfConcernRef, keeps the transform effect-free, and allows only claim-preserving or explicitly loss-declared rewriting of already available content.

It may change register, ordering, textual density, language, emphasis, or local wording. It may not silently introduce new claims, new bridge licences, new work, evidence, gate, release, policy, assurance, adjudication force, or a changed EntityOfConcern.

#Pattern, case, and publication distinction

ConservativeRetextualization is a pattern description and a named specialization under A.6.3. Concrete entityOfConcernRef-preserving rewrites are passive episteme cases or publication texts reviewed under this pattern; the pattern itself does not act, decide, or publish.

This distinction matters because the pattern governs how a rewrite is recognised, justified, and checked. It does not require every short report paragraph, summary line, or translation sentence to carry a giant standalone record.

#Local working vocabulary

This pattern repeatedly uses a small working vocabulary.

• Source slice = the already available pinned or otherwise reviewable textual content being restated.
• Published slice = the resulting textual rendering that remains under entityOfConcernRef-preserving discipline.
• Ordinary case = a reviewable same-entity rewrite where source tether, omission notes, and neighboring-pattern conditions stay readable without a heavyweight review record.
• Claim-bearing case = a case where dispute, policy, assurance, required correspondence witness, or cross-context reliance makes a fuller record worth publishing.

sourceSlice and publishedSlice are local review labels for the source textual slice and the resulting textual rendering in one rewrite case. A publishedSlice is not automatically a U.EpistemePublication; it becomes one only when the governing publication discipline instantiates it as such.

These terms are only local review aids. They inherit the E.17:5.1e local-field rule: they do not create U.Kind, publication-face kind, RelationKind, evidence kind, project-side FPF kind and reference named by value, new governing pattern, new publication face, or a second semantic rule track.

#Scope and exclusions

In scope

• entityOfConcernRef-preserving report rewrite;
• entityOfConcernRef-preserving summary;
• entityOfConcernRef-preserving translation between natural-language textual forms;
• declared filtering or foregrounding of already-present claims in textual form.
• correspondence-witnessed textual synthesis where every receiving claim remains recoverable to one entityOfConcernRef-preserving source line or declared entityOfConcernRef-preserving correspondence witness.

Out of scope

• any change of entityOfConcernRef or hidden change of EntityOfConcern (A.6.4);
• explanation-facing renderings whose main purpose is explanatory rendering rather than same-entity rewrite (ExplanationFaithfulnessProfile);
• representation-regime changes such as text→table, text→diagram, or text→latent form (Representation-Scheme Transition);
• comparison, abductive-prompt, ranking, recommendation, bridge-mediated, substitution, or action-selection work that introduces new claims rather than restating available ones.

#Reader guidance

Use this pattern when the EntityOfConcern stays fixed and the published result still remains textual.

• If the main change is explanatory, apply ExplanationFaithfulnessProfile.
• If the main change is a representation-scheme shift, apply Representation-Scheme Transition.
• If the EntityOfConcern changes, apply A.6.4.

#What the user checks first

The user usually does not begin by filling every field name. The first useful questions are simpler:

1. Is the published result still about the same EntityOfConcern?
2. Is the result still textual, or has it become explanation or representation change?
3. Can the reader see what was omitted, softened, or foregrounded?
4. If several source slices or correspondence witness are doing work, can each receiving claim be traced to one entityOfConcernRef-preserving source line or declared entityOfConcernRef-preserving correspondence witness?
5. Is the source only pointed at, or is it actually used and still admissible for the intended use?
6. If any answer is doubtful, is the neighboring governing pattern named explicitly?

If omissions, softening, or filtering are admissible only because the published result is coarsened, tied to narrower admissible use, non-admissible for downstream use, and tied to source-bearing return, the case has crossed out of ordinary conservative retextualization even if the prose still looks like a summary. Use A.6.3.CSC Controlled Semantic Coarsening for that source-to-rendering relation.

Here, source-bearing return means returning to the source-bearing content, while changed governing-pattern claim means that the now-attempted explanation, representation-shift, retargeting, gate, evidence, work, assurance, or bridge claim is governed by a named pattern. A coarsened textual slice may need both.

Only after these questions are answered does a fuller claim-bearing review record usually become worth writing.

#Working-model first; explicit review record only when the case is claim-bearing

Most entityOfConcernRef-preserving textual rewrites should stay human-usable. This pattern therefore follows E.14’s working-model-first discipline: ordinary report, summary, or translation cases do not need a giant inline metadata block. What they do need is enough explicitness that the user can still tell what stayed the same, what was omitted, and when another governing pattern governs the case.

Ordinary case (default). For everyday entityOfConcernRef-preserving rewrites, it is usually enough that the text or its surrounding publication keeps explicit:

• which source U.Episteme claims are being re-expressed;
• that entityOfConcernRef remains preserved;
• whether the case is direct or correspondence-mediated when that is not obvious;
• what omissions or source-loss modes matter for the reader;
• which neighboring governing pattern applies if the case becomes explanation, representation shift, retargeting, gate, evidence, work, assurance, bridge use, or another non-retextualization claim.

Explicit review record (only for claim-bearing cases). A fuller record is warranted when the case is assurance-facing, gate-adjacent, cross-context, correspondence-heavy, policy-bearing, or likely to be disputed. The record may inherit pattern ids and already-pinned metadata instead of restating them inline. When published, that record normally captures:

• transform relation (patternSpecializationRef = A.6.3 specialization, governingPatternRef, sourcePublicationOrRecordForm, targetPublicationOrRecordForm, changeTargetRef);
• preservation context (entityOfConcernPolicy = preserve, boundedContextPolicy, viewpointPolicy, referenceSchemePolicy, representationSchemePolicy, groundingPolicy, referencePlanePolicy);
• claim and publication discipline (claimPolicy, claimScopePolicy, publicationScopePolicy, reliabilityTransportPolicy, pinningPolicy, provenancePolicy, lossProfile);
• continuity and bridge discipline (claimContinuityClass, microtheoryContinuityClass, onticContinuityClass, bridgeRequirement, conservativityWitness);
• downstream and admissibility discipline (worldContactPolicy, evidencePolicy, gatePolicy, workCrossing, upstreamGoverningPatternRef, downstreamGoverningPatternRef, admissibleFaces, admissiblePublicationRenderings, compositionRule, reopenCondition);
• naming and presentation discipline (publicNamePolicy).

The point of this record is not bureaucratic completion for every paragraph. It is to make claim-bearing cases reviewable without hiding meaning in style, topic familiarity, or editor intuition.

#Ordinary admissibility defaults

Default admissibility for ordinary entityOfConcernRef-preserving textual cases:

• primary admissible faces are PlainView and TechCard;
• bounded report-only use is admissible when source pins, provenance, loss notes, and entityOfConcernRef-preserving conservativity remain visible;
• InteropCard use is admissible only when the governing publication-face source explicitly permits source-pinned, text-preserving export without added semantics;
• AssuranceLane or gate-bearing use is not default and requires governing publication-face policy plus source-pinned conservativity without hidden strengthening.

#Direct and correspondence-mediated profiles

Direct ConservativeRetextualization

• source slice and published slice are textual re-expressions of one source episteme;
• no CorrespondenceModelRef is needed;
• the main required admissibility record is explicit loss and provenance discipline.

CorrespondenceConservativeRetextualization

• the receiving textual rendering is derived from a declared correspondence between epistemes or views of the same EntityOfConcern;
• CorrespondenceModelRef is required;
• the result remains under A.6.3 only if the correspondence witnesses entityOfConcernRef-preserving conservativity and no new claims are imported beyond the declared witness set.

Cross-language translation is not automatically direct. If the translation depends on declared correspondence, reference-scheme mediation, or bounded equivalence notes, it must be treated as correspondence-mediated rather than disguised direct rewriting.

#Recurring same-entity textual moves

The pattern covers a small family of recurring textual moves as long as the same EntityOfConcern remains explicit:

• Register shift — a technical statement is rewritten into plainer engineer-manager prose without changing what is being said about the same entity.
• Summary or filtered restatement — a source note is shortened or focused on one declared slice, with omissions stated rather than hidden.
• Cross-language restatement — the same source claim is restated in another natural language while the same source tether and same-entity line remain explicit.
• Correspondence-witnessed textual synthesis — one textual rendering is produced from declared same-entity correspondences without importing extra bridge or substitution admissibility record.

These are recurring move shapes, not separate governing patterns. The specialization relation remains the same: entityOfConcernRef-preserving textual re-expression under A.6.3.

#Shared conservative retextualization rule bundle

#A.6.3.CR:4.5.a. Preservation rule

A case under ConservativeRetextualization preserves the same EntityOfConcern line, the declared bounded context, and the already available claim-bearing source while changing wording, register, language, ordering, or density. It states what remains preserved about claim scope, publication scope, pins, provenance, grounding, and ontic scaffold, and it says whether the case is Direct or Correspondence.

#A.6.3.CR:4.5.b. Loss and reliability rule

A reviewed case makes explicit what is omitted, shortened, foregrounded, or carried only through a declared source-loss mode by the rewrite. Reliability transport may remain source-bounded or be explicitly downgraded, but it must never be silently widened by cleaner prose, more forceful rhetoric, or management-facing polish.

#A.6.3.CR:4.5.c. Authority and governing-pattern boundary rule

A case reviewed under this pattern stays same-entity and episteme. It does not govern explanation governance, bridge stance, retargeting, gate authority, or work enactment. If the rewrite becomes explanatory, bridge-bearing, gate-bearing, or world-facing, name the downstream governing pattern and the attempted claim explicitly.

#A.6.3.CR:4.5.d. Composition and reopen rule

Repeated direct rewrite over the same source line may be idempotent, but heterogeneous rewrites and correspondence-mediated rewrites are generally order-sensitive. A reviewed case must reopen whenever correspondence witness, source pins, provenance, admissible-face assumptions, or entityOfConcernRef-preserving conservativity stop being explicit.

#A.6.3.CR:4.5.e. Non-collapse note for correspondence

Correspondence-mediated retextualization does not by itself grant bridge licence, substitution licence, or comparative-review licence. If the case needs those required admissibility records, they must be declared separately rather than being smuggled in through correspondence language.

#A.6.3.CR:4.5.f. Local conservativity witness for borderline textual cases

For borderline textual rewrites, the user treats the case as no longer conservative under this pattern unless each point below remains visibly preserved or is explicitly loss-declared with the governing pattern for the changed claim stated.

• Modality and force. A rewrite may not silently turn possibility, uncertainty, permission, obligation, recommendation, decision status, bounded scope, temporal window, or hypothesis language into a wider commitment.
• Caveats and qualifications. A rewrite may not quietly remove conditions, exception notes, uncertainty markers, or temporal qualifiers that still matter for interpreting the same source.
• Reliability assessment. Cleaner prose, better ordering, or manager-facing polish may not silently raise confidence, warrant claim, or readiness for action.
• Bridge and substitution admissibility record. Same-entity textual fluency may not import cross-context equivalence, substitution, or comparative-review licence unless that admissibility record is declared elsewhere.
• Alternative preservation. A rewrite may not collapse open alternatives, rival hypotheses, or declared plurality into one apparently settled interpretation unless the loss is stated and still admissible under this pattern.

This witness is local to ConservativeRetextualization. It does not replace the broader conservativity invariants of A.6.3; it makes them inspectable for textual rewrites where fluent prose can otherwise hide strengthening.

#Archetypal Grounding

#Same-EntityOfConcern report rewrite

Source note slice. Service S exceeded the latency threshold in the evening batch window. Trace T-44 and dashboard pin D-17 show the spike. Two low-confidence hypotheses remain open.

Published report slice. Evening-batch latency for Service S exceeded the threshold. Source pins: Trace T-44, Dashboard D-17. Low-confidence hypotheses are omitted here and remain in the pinned source note.

This is an admissible direct ConservativeRetextualization because the EntityOfConcern stays fixed, the report remains textual, and the omission is stated rather than hidden. In ordinary internal use, this often needs only source pins plus visible omission notes rather than a full explicit review record.

#Ordinary inherited-pin summary

Pinned source cluster. Incident note N-14, trace T-44, and dashboard card D-17 are already published together under one incident review bundle.

Published stand-up slice. Evening-batch latency again exceeded the threshold for Service S. See N-14 / T-44 / D-17 for the pinned source cluster.

This is still an admissible ordinary case even though the short stand-up slice does not restate every pin and qualifier inline. The didactic point is that lightweight use may inherit already-published pins and provenance when the tether stays visible to the reader.

#Benign omission that stays ordinary

Source note slice. Service S exceeded the latency threshold in the evening batch window. Trace T-44 and dashboard pin D-17 show the spike. The note also lists two low-confidence hypotheses for separate investigation.

Published stand-up slice. Evening-batch latency for Service S exceeded the threshold. Source pins: T-44, D-17. Low-confidence hypotheses are omitted from this stand-up note and remain in the pinned source.

This stays ordinary ConservativeRetextualization because the omission is declared, the same EntityOfConcern remains visible, and no separate narrower admissible use, non-admissible downstream use, and source-bearing return card is doing the real work. Ordinary omission alone is not controlled semantic coarsening.

#Functional-description textual summary

Source note slice. The principle scheme says: choose method family MF-2 for small-batch mixing when material X remains below threshold T; selected method M-2 still requires work plan WP-17 and result measurement RM-4.

Published summary slice. For small-batch material X below T, method M-2 is the selected method. Work plan WP-17 and result measurement RM-4 remain required.

This remains ConservativeRetextualization because it is a textual restatement of the same source-episteme claims and it keeps the work-planning and result-measurement requirements visible. It is admissible for interpretation and source-finding. It does not by itself provide performed U.Work, evidence, gate passage, engineering justification, or control architecture. If the summary drops the work-plan and result-measurement requirements or makes the selected method look executable by summary alone, treat the text as A.6.3.CSC Controlled Semantic Coarsening or recover the project-side FPF kind and reference named by value that actually makes the requested use admissible.

#Generated-summary source-relation variant

A generated or machine-assisted summary may stay in ConservativeRetextualization only when it remains an entityOfConcernRef-preserving textual re-expression and its source relation is visible enough for the intended use. This is the ordinary LLM-generated-summary case: a model-produced paragraph over a pinned inspection note, method-selection note, safety note, incident note, or other source slice is not automatically ExplanationFaithfulnessProfile merely because it was generated; it remains ConservativeRetextualization only while it restates source claims and leaves omissions, loss, and non-admissible uses visible. Ordinary source-finding use can stay light; use the compact variant below when the summary will be reused, cited, disputed, or relied on.

When the generated-summary case needs the shared vocabulary rather than this CR-local question list, read the source relation through E.17:5.1b: source-pointer-only, source-available, source-retrieved, source-used, source-faithful, claim-admissible, claim-non-admissible, claim-contradicted, claim-plausible-only, source-omitted, source-loss-declared, claim-widened, added-linkage, independent-verification-present, admissible-for-this-use, downstream-use-forbidden, and reopen-trigger-present.

The summary may expose or cite the source slice it restates. It does not become that source slice by fluency, brevity, translation, layout, generated form, or reuse. If the source slice or required project-side FPF kind and reference named by value is missing, a repair request or source-gap note is only prospective; it does not retroactively make the earlier summary source-relation-admissible.

If the generated summary is source-pointer-only, merely plausible, claim-widened, or carrying added linkage, do not treat it as a conservative source-equivalent summary. Either keep it as source-finding or orientation, repair it against the source, or apply A.6.3.CSC, ExplanationFaithfulnessProfile, Representation-Scheme Transition, E.17.ID.CR, A.15, A.10, or another governing pattern according to the claim being made.

#Same-EntityOfConcern rewrite via declared correspondence

Source design slice. Cooling loop CL-2 preserves safe temperature margins during standard operating demand.

Source safety slice. Cooling loop CL-2 maintains the temperature condition required for hazard-control claim HC-7 during standard operating demand.

Published joint-review slice. For standard operating demand, Cooling loop CL-2 is described in both the design and safety views as maintaining the required temperature condition. This summary relies on CorrespondenceModel CM-12 and does not add claims beyond that declared overlap.

The synthesis may stay in this pattern only if the source relation remains explicit, every receiving claim remains recoverable to the design slice, the safety slice, or the declared CorrespondenceModel, and the text does not silently widen claims beyond the declared entityOfConcernRef-preserving overlap. Because correspondence witness is claim-bearing here, a claim-bearing review record is usually warranted.

#Cross-language re-expression without hidden bridge work

Source slice. The backup controller stays in passive watch mode until the primary loop fails two consecutive heartbeat checks.

Published slice. Резервный контроллер остаётся в режиме пассивного наблюдения, пока основной контур не пропустит две последовательные проверки heartbeat.

This remains in ConservativeRetextualization only if the translation is still tethered to the same source claim, preserves the same EntityOfConcern, and does not quietly add cross-tradition bridge claims such as "equivalent architecture role" or "same operational guarantee" beyond what the source actually states.

#Boundary to controlled coarsening

Source slice. Vendor bulletin VB-7 requires rollback when pressure drift exceeds 2.5%, and it keeps two equipment-specific exceptions in the pinned annex.

Published coarsened slice. Pressure drift above 2.5% is a warning condition in the bulletin. Check the pinned bulletin and annex before treating the note as rollback guidance.

This does not remain ordinary ConservativeRetextualization. The coarsened slice drops equipment-specific exceptions and remains only an orientation warning: it is not an executable rollback command. It can stay honest only through narrower admissible use, non-admissible downstream use, and source-bearing return to the source-bearing bulletin. Once that narrower-use card becomes primary, the case leaves ordinary same-entity rewrite and must use A.6.3.CSC Controlled Semantic Coarsening rather than being treated as a harmless summary.

#Boundary to explanation-facing renderings

A text is rewritten not mainly to restate the same source, but to explain why it matters, simplify reasoning for a learner, or narrate a mechanism. That move should leave ConservativeRetextualization and be reviewed under ExplanationFaithfulnessProfile.

#Boundary to representation-scheme transition

A prose note is rewritten as a table, matrix, diagram, latent representation, or distributed representation. Even if the EntityOfConcern stays fixed, this is not only a textual rewrite; it belongs with Representation-Scheme Transition.

#Bias-Annotation

Lenses tested: Arch, Onto, Epist, Prag, Did. This pattern intentionally biases toward same-entity conservativity and away from explanation or retargeting inflation. The main mitigation is to apply ExplanationFaithfulnessProfile, Representation-Scheme Transition, A.6.4, or the downstream governing pattern when the same-entity textual interpretation stops being honest.

#Conformance Checklist

1. CC-CR-1 — Same EntityOfConcern remains explicit. The case preserves entityOfConcernRef without special pleading.
2. CC-CR-2 — Textual re-expression remains the right family. The result stays a textual re-expression rather than explanation or representation shift.
3. CC-CR-3 — Loss, provenance, pinning, and reliability are explicit or inherited by pinned reference. The case states these explicitly or inherits them through already-pinned content that remains visible to review.
4. CC-CR-4 — Direct vs correspondence split is explicit. The direct-vs-correspondence split is explicit and justified.
5. CC-CR-5 — Correspondence witness is named where needed. If correspondence-mediated, CorrespondenceModelRef is declared.
6. CC-CR-6 — Local conservativity witness remains satisfied. The reviewed case does not silently widen modality, remove caveats, raise reliability assessment, import bridge or substitution licence, or collapse declared alternatives beyond stated loss notes.
7. CC-CR-7 — Governing pattern is explicit on failure. If the case fails any of the checks above, the governing pattern for the changed claim is named explicitly (ExplanationFaithfulnessProfile, Representation-Scheme Transition, A.6.4, B.5.2, or another governing pattern).
8. CC-CR-8 — Working-model first remains intact. Ordinary same-entity rewrites stay lightweight; fuller explicit review records are reserved for claim-bearing cases.

#Common Anti-Patterns and How to Avoid Them

#Consequences

• Textual same-entity rewrites get an admissible place without inventing a new heavy governing pattern.
• Direct and correspondence-mediated variants stay visibly separated.
• Loss, provenance, and reliability transport become explicit instead of implicit editorial judgement.
• Ordinary working-model use stays lightweight, while claim-bearing cases get a claim-bearing review record when risk warrants it.
• The pattern remains safely bounded by A.6.3, A.6.4, explanation-facing work, and representation-shift work.

#Rationale

This pattern is worth splitting out because same-entity textual re-expression is common, useful, and safer than many neighboring transform families when it stays explicitly conservative. Keeping it under A.6.3 as a named specialization preserves governing-pattern boundary while making a recurring authoring move easier to review, while still respecting E.14’s working-model-first discipline for ordinary cases.

#SoTA Alignment: Adopted Invariants, Adapted Invariants, and Rejected Shortcuts

SoTA alignment rule. Read each row here as source idea -> local FPF invariant -> practical local test -> popular shortcut rejected. A source citation governs nothing by reputation; it counts only when the cited idea is translated into the Solution, conformance checks, boundary rules, worked slices, and Relations of this pattern.

Traditions covered. This pattern binds itself to architecture-description governance, summarization factuality, translation-quality governance, and plain-language rewrite practice.

Architecture-description governance. A.6.3.CR adopts the discipline that rendered text must stay visibly tied to a declared source publication or U.View line. It therefore rejects same-topic textual polish as sufficient evidence of entityOfConcernRef-preserving conservativity.

Summarization factuality. A.6.3.CR adapts modern factuality concerns into a local conservativity witness: source pointer, source actually used, claim admissibility, contradiction, plausible-but-non-admissible claim, omission, declared source-loss mode, claim widening, added linkage, independent verification, admissible use, forbidden downstream use, and reopen trigger are treated as reviewable source-relation distinctions, not as style noise. The shared source-relation vocabulary is E.17:5.1b; the shared use-boundary terms are E.17:5.1c; the primary-boundary chooser is E.17:5.1d. This pattern uses them only for entityOfConcernRef-preserving textual restatement.

Translation and plain-language traditions. A.6.3.CR adopts the reader-oriented value of translation and plain rewrite, but rejects the still-popular habit of treating cross-language or plain-language textual fluency as automatic proof that no new claim has been introduced. The W3C MQM source is used for issue-type and evaluation discipline, not as a brand-level warrant that a translated or rewritten sentence is source-equivalent.

Local stance. Best-known current practice motivates a narrow rule: entityOfConcernRef-preserving textual restatement is admissible only when source tether, loss, provenance, and same-entity bounds remain explicit enough that the reader can still tell what was preserved, what was omitted, and when another governing pattern must be applied.

#Relations

• Builds on: A.6.3, A.6.2, A.7, E.10.D2, E.17.0, E.17, F.9, F.18, E.10
• Coordinates with: ExplanationFaithfulnessProfile, Representation-Scheme Transition, E.17.ID.CR ComparativeReviewUnit, A.6.4, B.5.2, A.15
• Impact radius: primary touch A.6.3; secondary review relation E.17.0, E.17, F.9; failed conservativity cases apply A.6.4, B.5.2, or A.15
• Boundary notes: explanation-facing cases apply ExplanationFaithfulnessProfile; representation-regime shifts apply Representation-Scheme Transition; bounded comparative review cases apply E.17.ID.CR ComparativeReviewUnit; EntityOfConcern changes apply A.6.4.

#A.6.3.CR:End

---

Last Updated: 2026-06-11 — this section last modified in upstream FPF commit 20c8a0a5 (github.com/ailev/FPF)