The objection
CLAIM-151: which act an utterance performs “is partly constituted by its relations to participants, norms and expected uptake, so a transformation that changes those relations can change the act and not only its realization”. CLAIM-139’s ontological clause rests on it.
A translation changes those relations by definition. Its audience is the target language’s readers, not the source’s. It is read under the target culture’s norms and genres. Usually it has a second speaker, the translator. And it changes the words too, which are the intrinsic part that “partly” leaves the act. So in a translation neither the relational constituents nor the intrinsic ones are held fixed. On CLAIM-151, whether the act survives cannot be read off either structure. It depends on whether the changed relations count as the same relations.
The record’s statements of survival all say this, in one word each. A191 §1, in CLAIM-151: “Changing u while preserving the appropriate relationships may conserve the act’s identity.” TERM-040, quoting A218 §1: source and translation “may realize corresponding relational organizations”. “Appropriate” and “corresponding” name a map from the source’s relations to the target’s. Nothing inside the source’s organization, or inside the target’s, says which map that is. This is CLAIM-138’s point, that bare structure cannot pick out the relations that count, met at the level of CLAIM-151 by the move every translation makes. CLAIM-138 spares CLAIM-151 (“‘Partly constituted’ already grants a role to what is not bare structure”), and rightly for a single utterance. Under translation the non-structural part changes as well, so “partly” no longer helps.
Two consequences follow.
- QUESTION-005 changes kind. It asks how to operationalize “corresponding communicative environments and responses” without building the preferred notion of fidelity into the measurement. Under CLAIM-151 the correspondence is part of what the act is, so the question is about the act, not only about measuring it.
- The record’s model of translation does not cover the case. CLAIM-091 models translation as (u, s, a, k, g) → (u′, s, a, k′, g′), where a is the audience (TERM-031). The audience is held fixed. The case CLAIM-151 makes hard, a new audience, is outside the model.
CLAIM-004, which CLAIM-151 rests on, already lacks the criterion this needs. It says the manuscript “has no way to tell (ii) from (iv) in a given pair”, representation-changing equivariance from structural reorganization, and that a restated criterion would turn on “what the change of target context requires”. That is again a correspondence of contexts. CASE-018’s four speakers change one relation at a time with the words fixed. They do not show what happens when every relation changes at once.
What it does not say
It does not say that acts never survive translation, nor that CLAIM-151 is false. A teasing remark can plainly be translated as a teasing remark. It says that on CLAIM-151 such survival is a fact about a correspondence between two relational organizations, and that the record has no account of which correspondences preserve an act. It does not touch CLAIM-151 for transformations that keep the audience and the norms, such as CASE-018’s reattributions.
What would answer it
- An account of which correspondences of participants, audiences and norms preserve an act, fixed independently of judged fidelity. QUESTION-030 asks the nearest question, which interventions tell constitutive relations from accompanying ones.
- Or a restatement of CLAIM-151 in which the constituents are relations to roles (addressee, authority, co-participant), not to particular people. A new audience in the same role would then keep the constituent, and the open part would narrow to whether the target culture’s norms define the same roles.
Where it stands
On 2026-10-10 CLAIM-151 was restated (version 2) to take this objection’s second remedy: its constituents are relations to participant roles, TERM-040’s positions (“Promisor and promisee are positions in one kind”), so a new audience in the same role keeps them. The objection stays open on its residue: whether the target culture’s norms define the same roles. That is part of the correspondence QUESTION-005 asks for, and of re-performance fidelity (CLAIM-212). QUESTION-073 holds open the case where it bites hardest, teasing.
Note of 2026-10-10: a same-wording case
CASE-061 now illustrates the objection alongside CASE-064, with identical wording: the hymn phrase “wonder-working power” in the 2003 State of the Union, taken as an appeal to shared faith by those who sing the hymn and as secular praise by others, with both audiences’ uptake documented and the split measured experimentally (Albertson 2015; abstract read, stimulus and figures unverified). It shows the residue the objection keeps after CLAIM-151 version 2: whether a hearer holds the addressed role depends on norms the words do not carry, here inside one language and one broadcast. It does not settle whether the audience’s role constitutes the act or only evidences it, since the speaker designed both readings. The case also lists TV Tropes’s audience-split lists as a pool of items for CASE-055.
Note of 2026-10-10: an unintended-reading case
CASE-073 now illustrates the objection as well: the Kia Provo (March 2013), a concept-car name unchanged across audiences. To the motor press it was a coinage without meaning; in Northern Ireland, where “Provo” names a member of the Provisional IRA, six MPs’ motion said it “has caused deep offence”. Kia declared no intention to offend and withdrew the name for the UK and Ireland. It supplies what CASE-061 lacked: the speaker’s intention is fixed, declared and innocent, so the change of reading comes from the change of audience alone. It does not supply a change of language (the collision is with a regional English word), the original audience is the international motor press rather than a public, and the act concerned is a naming, so part of the change is a lexical collision. The objection’s residue, whether a new audience holds the role the speaker addressed, is shown again: nothing in the name told anyone which audience would hear what.