CLAIM-419: The record's grammar is fixed at any moment but amendable by decision, and in practice a move that fit no scheme or relation has repeatedly been treated as a defect of the grammar and answered by amending it, so a move the grammar cannot hold is not only forced or lost; whether every such move is caught is not shown

The claim

CLAIM-181 says the record “fixes the grammar an argument is expressed in”, and that “a move that fits no CLAIM, TERM, ARG or declared relation is either forced into one or lost”. It asked to be answered by “a move that mattered in the exchange and that the record could not hold”, or by showing that every case of drift the record caught was drift by the exchange’s own standards.

The owner, at U18, quoting A17’s “epistemic accountability requires not only traceability within the current conceptual scheme but also some means of contesting the scheme used to make inquiry traceable”: “this definitely aligns with how we’ve been using luria: when we’ve encountered representational challenges, we’ve generally interpreted those as a symptom of a represntational weakness of the record and used that opportunity to mature the luria configuration, featureset, or both.”

The record’s decisions bear it out. Each was a move the grammar could not hold, recognized as such and answered by amending the grammar:

• ADR-044 (2026-10-15): a claim that gave evidence for another without the other depending on it could only be written as supports, and the link fixer then wrote a false rests_on. The grammar asserted an inference nobody made, and corroborates was added (CASE-131).
• ADR-043 (2026-10-15): claims of different kinds of support were indistinguishable, and warrant was added.
• ADR-042 (2026-10-14): a thought experiment, a construction and a real specimen were told apart only by prose, and CASE gained kind.
• ADR-046 (2026-10-18): a term that rivalled or narrowed another could not be filed as Proposed, and could be.
• ADR-049 (2026-10-21): a question partly answered had no status, and partly answered questions split.

So the grammar is fixed at any moment and amendable over time, which is CLAIM-195’s point that the constraint set is “fixed by an amendable decision”. A move that fits nothing is not only forced or lost: a third outcome is that the failure to fit becomes evidence against the grammar. A18 named the condition: “A record becomes second-order corrigible when failures of its representational distinctions can themselves become objects of inquiry, argument, and justified revision”, and its test of openness, whether failures of the categories can become “recognizable, contestable, and actionable without requiring the failed categories to validate the challenge in advance”. A18 found the bootstrap in the record’s practice: “ordinary prose, source files and observed discrepancies can provide that provisional external language.”

What it does not say

• Not that the grammar catches everything. The evidence is the amendments made. A move the grammar forced into a wrong shape without anyone noticing leaves no decision behind, so the record of decisions cannot show its absence. CLAIM-190’s point stands: measures computed in the grammar cannot certify it alone.
• Not that every misfit was amended. ADR-032 kept the literature roles (foil, method, analogue, boundary marker) in prose though they “recurred”, which is CLAIM-181’s own evidence. That was a decision not to amend, and it is recorded, but the move is still held only in prose.
• Not that amendment is free of the risk A18 named. Its anti-ad-hoc constraint: an amendment should answer a recurring misfit and not be made to save a particular claim.
• Warrant. Empirical: it rests on the record’s own decision history, observed, not on an argument about grammars in general.

Note of 2026-10-27: found again in luria, without this claim

At U7 of the exchange of 2026-10-27 the owner sent the assistant to luria (“I think I may have inadvertently operationalized this notion of apparatus already”). A7 found this claim’s loop in luria’s own record, the principle that an awkward structure reports a distinction the model has stopped expressing (LU-DP-016), which it called “the smoking gun”, and promotion of a vocabulary to a scheme (LU-ADR-124), and named it “a form of reflexive epistemic apparatus”. It did not find this claim, filed from the owner’s U18 of 2026-10-22 with the record’s five amendments as evidence, nor CASE-131. CLAIM-476, the apparatus claim filed from that exchange, rests on this one. Nothing here changes.

Note of 2026-10-27 (evening): the readings of the U15–U23 works

CEGAR, read today (NOTE-1231), was offered by the exchange as luria’s DP-016 in formal dress. The pattern matches: a failure under a fixed specification is located and repaired by refining the abstraction. Two disanalogies bear on this claim’s limit.

• CEGAR’s misfit is certified. Lemma 1 decides whether a counterexample is spurious against the concrete model, so every wrong merge that matters eventually surfaces. A record has no concrete model to check against. That is why this claim’s limit, that a silently forced move leaves no trace, has no counterpart there.
• CEGAR only splits. The amendments this claim lists (ADR-043’s field, ADR-044’s relation) are additions, not splits.

And even with a certified misfit, the least amendment can be hard to find: the coarsest split that removes a counterexample is NP-hard (Thm 3).

Note of 2026-10-31: the misfit as a nonprototypical case, and as disindividuation

Two of the readings filed for Bencherki, Brummans and Vézy bear on this claim’s third outcome.

• Tsoukas and Chia (LIT-1913, pp. 573–575; THEORY-1279). Categories are radially structured. Applying one to a nonprototypical case is “an imaginative extension” that widens “the radius of application of the concept, thus transforming it” (p. 574). That act is normative, and “The indeterminacy of extension does not indicate arbitrariness”. The amendments listed above are such cases: support that was not dependence (ADR-044), and a thought experiment among specimens (ADR-042). What the record adds is where the extension happens. For Tsoukas and Chia it happens in use and may never be institutionalized (p. 579). Here the closed vocabularies rule out the silent form: a missing word “is added by decision … never replaced by the nearest wrong one” (CLAUDE.md). So the extension is made at the case that needed it, as a decision, and institutionalized at once (CASE-178). Their open question, how a local extension becomes organizational, has a designed answer in this record. This is the record’s own case. It does not test THEORY-1279, because nobody classified the cases as prototypical before seeing what changed.
• Disindividuation (LIT-1902, p. 1356): “the experience of needing to reorganize in order to deal with a preindividual element that resists incorporation”, read at p. 1365 as an “inflection point” that calls “for creativity”. This has the structure of the third outcome: a move that does not fit becomes the occasion for reorganizing the grammar. It is analogy only. The paper identifies disindividuation by felt hesitation (THEORY-1280), and the record identifies a misfit by a filer’s judgement that a structure is awkward (LU-DP-016). Neither is a check, so this claim’s caveat stands: a misfit absorbed without being noticed leaves no decision behind.

No status, title or warrant changes.