The claim
This is the record’s reply of 2026-10-10 to CLAIM-219, and a narrowing of CLAIM-105. The derivation below is decisive and checkable; the claim that a change of cover is the headline’s surviving form is a proposal, to be tested by the discriminating study (CASE-055).
CLAIM-105’s headline is “Translation may preserve local meanings while changing the global structure of meaning”, in the regime A78 named: “preserve most local pragmatic judgments while substantially changing their global compatibility structure”. The objection’s dilemma takes “global compatibility” as either contextuality proper or overlap consistency.
The fixed-scenario form is sharp, but needs contextuality
Within one scenario, global compatibility can change maximally while most local judgements are kept exactly. Take n odd, n ≥ 3, and observables a_1, …, a_n with outcomes ±1. The contexts are C_i = {a_i, a_(i+1)} for i = 1, …, n, indices mod n, so that C_n = {a_n, a_1}. Weight every context 1/n.
- Model e. In every context the two observables disagree: the outcomes (+, −) and (−, +) each have probability 1/2. A global assignment consistent with every context’s support would need a_(i+1) = −a_i for every i, so going round the cycle a_1 = (−1)^n a_1 = −a_1, since n is odd. No ±1 value satisfies that, so no global assignment is consistent with the supports. The model is strongly contextual, and its contextual fraction is 1.
- Model e′. The same, except in C_n, where a_n = a_1: the outcomes (+, +) and (−, −) each have probability 1/2. Now the constraints are a_(i+1) = −a_i for i = 1, …, n − 1 and a_n = a_1. The first n − 1 give a_n = (−1)^(n−1) a_1 = a_1, since n − 1 is even, so they agree with the last. The two assignments a_i = s(−1)^(i−1), s = +1 or −1, satisfy every constraint. Their equal mixture reproduces every context: in C_i for i < n it gives (s(−1)^(i−1), −s(−1)^(i−1)), which is (+, −) or (−, +) with probability 1/2 each; in C_n it gives (s, s), which is (+, +) or (−, −) with probability 1/2 each. So e′ has a global distribution, and its contextual fraction is 0.
- What changed. In both models every observable is +1 or −1 with probability 1/2 in each context that measures it, so neither signals, and overlap consistency is the same. The contexts C_1, …, C_(n−1) are identical. C_n changed from support {(+, −), (−, +)} to the disjoint support {(+, +), (−, −)}, a total-variation distance of 1. With equal weights the weighted observational distortion is D_obs = (1/n) · 1 = 1/n.
For n = 3 this is CASE-037 with the outcomes of one observable relabelled. So within a fixed scenario the contextual fraction can fall from 1 to 0 at a weighted local distortion of 1/n, which is as small as one likes for large n. THEORY-165 says the fraction is Lipschitz within a scenario and that its constant “depends on the scenario and is not bounded here”. Measured against weighted D_obs, the example shows the constant is at least n.
The objection’s bound, that moving every context by at most ε moves each overlap discrepancy by at most 2ε, is correct, and the example respects it: no marginal moves at all. But it concerns overlap consistency and every context moved a little, not global extension with most contexts kept exactly. So the second horn does not close the fixed-scenario form.
The first horn does. Every such case needs a contextual model on one side: here e. The record’s reading of the behavioural data has none. NOTE-235: “Behavioural data have plenty of the first and, so far, none of the second”, where the second is contextuality proper. So, on the evidence the record holds, the fixed-scenario form has no case but a constructed one, and that part of the objection is conceded.
The form the data leave open: a change of cover
CLAIM-105’s own note says where else the headline can live: the contextual fraction does not jump within a scenario, and “Only a change of scenario can make it jump”. The manuscript says it in §11: “A transport may preserve overlaps but change the available measurement cover.” A change of cover needs no contextual data, because it is a change in which questions a reader can answer at all.
Operationally, a context is a question whose answers readers give determinately: agreement among them above a threshold fixed in advance. A rendering changes the cover when a question that is determinate for source readers is indeterminate for target readers, or the reverse.
- The target decides what the source left open. French tu and vous, or Japanese honorific marking, oblige the target to settle the speaker–addressee relation that an English source can leave open. Target readers answer “Are these two intimates?” determinately; source readers do not.
- The source decides what the target cannot. Villon wrote in criminal argot, and Henley rendered it in English underworld slang that is now “deliberately opaque” (CASE-010). Readers inside such a milieu can answer an insider question, who among the speakers belongs, that readers outside it cannot.
In neither case need any one context’s distribution move: the change is in which contexts exist for which readers. CLAIM-257 gives the measure, with the questions only one side can decide reported apart from the distortion summed over the questions both can.
What it answers
CLAIM-219. Its first horn is conceded: no data the record holds are contextual, so the fixed-scenario headline has no case but a constructed one. Its second horn’s bound is right but beside the headline’s regime, as derived above. Its own second remedy, “a restatement of CLAIM-105 about a change of cover or scenario”, is taken.
What it does not say
- It does not say contextual pragmatic data cannot exist. LIT-264 offers their absence as a working hypothesis, and the record has read one paper’s data sets.
- It does not say a change of cover is contextuality. A rendering that decides a question the source left open has changed which questions can be asked, not shown that the answers cannot be glued.
- It does not fix the threshold or the rate in its defeat condition. Both belong to the discriminating study’s preregistration.
Note of 2026-10-10: a documented case
CASE-072 gives this claim a documented case for one half of a change of cover, in place of the invented tu/vous illustration. Hafez’s “Shirazi Turk” leaves the beloved’s sex and nature open in the Persian, and four English renderings decide it three ways: a maid (Jones 1771, Bell 1897), God (Clarke 1891, “His dark mole”) and a boy (Davis 2012), with Davis saying in print that his choice is arbitrary. The question decided is the one this claim needs, who is addressed and in what relationship.
It does not supply the other half, a question the source decides that a rendering makes undecidable; that half still rests on CASE-010. Nor does it measure reader agreement: the openness of the source is attested by translators and scholars, and a defender can reply that genre convention makes the beloved determinately male for Persian readers. The case’s reader test, which could be items in CASE-055, would settle that.
Note of 2026-10-10: the reverse half
CASE-059, found by a search the owner asked for, is a documented candidate for the half CASE-072 left open: a question the source decides that a rendering makes undecidable. Tatiana’s letter in Eugene Onegin III moves from вы to ты at «я твоя» and back to вы at «ваша честь», with nothing but the pronouns to mark it. Kline (2009) has “you” throughout, so neither move is in the English; Spalding (1881) moves to “thou” and never returns. The question is this claim’s kind, the speaker–addressee relation, and the three texts are graded: the Russian decides both moves, Spalding one, Kline none. The lead first given, Pushkin’s «Ты и вы», does not serve, because the poem names its pronouns and every rendering must keep the switch.
It does not supply reader evidence. That Russian readers take the moves determinately is attested by one essayist and unchecked school notes; Lotman and Nabokov were not read. That Kline’s readers cannot decide is inferred from the text. A defender can say the passion decides the intimacy anyway, which leaves the return to вы as the cleaner item. The reader test in the case could be items in CASE-055. With CASE-072, the claim now has a documented candidate for each half of a change of cover, and measured reader agreement for neither.
Note of 2026-10-12: a second sense of “cover” on main, left to the owner
CLAIM-197, filed on main on 2026-10-12, uses “cover” in the sheaf sense: a rule saying which families of contexts count as covering. CASE-048 version 2 applies it to one viewer’s readings as a child and as an adult, as sections over two contexts that do not overlap, so that they glue. This entry’s operational sense makes a context a question readers answer determinately, and in that sense the two readings may share a context. The record does not say whether the two senses are one notion. Which sense the record uses, or whether it keeps both under separate names, is left to the owner.
Note of 2026-10-15: split under ADR-043
ADR-043’s classification found this claim compound: a demonstrated odd-cycle construction, the empirical finding that the behavioural data the record has read have no contextuality proper, and the interpretive proposal that a change of cover is the form of CLAIM-105’s headline those data leave open. The construction, with its bound on the Lipschitz constant and the point that any change of global extension needs a contextual model, is now CLAIM-319. The empirical finding is CLAIM-325, split from CLAIM-219, whose dilemma rests on the same premise. This claim rests on both and keeps the proposal and its operational definition of a change of cover; it stays interpretive. Its title names the two premises as what the proposal starts from, and its conclusion is the proposal, so it was kept.