The claim
A35 §7: “Interpretive fidelity: Does the translation lead audiences to analogous understandings? Performative fidelity: Does the translation support analogous changes in the speaker–audience relationship and communicative commitments? The latter might be much harder to preserve.” The reason: “A speech act doesn’t merely update an audience’s beliefs about the speaker. It can change their relationship … These are not all reducible to belief revision.” Hence “the more appropriate state might be a joint communicative state” of speaker, audience, common ground and relationship. The case is CASE-026: “a translation can preserve interpretive content while failing as a social act.”
Where it went
A38 §10.3 kept the point as “the social constitution of communicative meaning”, and A39 §3.4 kept the joint state with an illustrative factorization. Neither named the two fidelities. The manuscript keeps social uptake only as observables (“entitlement to criticize”, “a plausible continuation”, §2; whether a statement “licenses criticism or shared laughter”, §5), which measure the performative side without saying it can fail while the interpretive side holds. Krifka’s informative/performative updates, which A35 drew on, are not cited there or in the record.
What the speech-act readings say
- Partly supported. Krifka (LIT-835) treats the speaker’s commitment and the uptake of the content as separate updates, which succeed or fail separately. But his “performative” is commitment or a change in the world, not the speaker–audience relation this claim’s case turns on.
- In tension. Brandom (LIT-794, skimmed) holds that the same content with the same force gives the same deontic significance.
Note of 2026-10-30: what holds it back
Its one ground, CASE-026, is a stipulated thought experiment. The performative side it turns on, complicity in the speaker–audience relation, is footing, for which CLAIM-227 (Active) finds no validated measure. CLAIM-061 was promoted only after dropping footing to concede CLAIM-227, and on a documented case (CASE-056). The missing result is a documented rendering in which readers’ understanding is kept and the relation is judged lost, by a measure of footing.