The claim
This is the record’s reply of 2026-10-10 to the second horn of CLAIM-243. It is decisive in the only way a disclaimer can be: it says what the account does not claim, and is checkable against the account’s own texts.
CLAIM-243 reads CLAIM-115 two ways. Read as a criterion, the thesis says which distinctions a translation ought to keep, and is a schema until the family of tasks Q (CLAIM-050) is fixed. The record answers by declining that reading. The account is descriptive: given stated tasks and receivers, it says how to tell which distinctions a rendering kept. Which tasks and receivers a translation ought to serve is not something the account decides. It is set by the purpose the translation serves, its brief: a subtitle for a comedy, a court interpreter’s rendering of testimony and a scholarly crib of a poem are owed different things.
The record already holds the split. CLAIM-085, from A45 §8: the paper’s two goals are “a normative theory of translation fidelity and a descriptive theory of how interpreters construct communicative significance. LLMs can help test the second. Human communicative practice remains essential for justifying the first.” The manuscript leaves the choice open: CLAIM-115’s summary says “the choice of observables and decision tasks is stated to be normative and contestable (§11)”. This entry draws the consequence for the thesis. The account supplies the second theory and not the first, so Q is an input to it, taken from the brief, and the revised thesis (CLAIM-220) states Q for the study that tests it.
What it answers
CLAIM-243’s second horn: that read as a criterion the thesis is a schema, and that a descriptive defeat condition cannot test it. Both are true of a normative thesis, and the account makes none. Its first horn, the descriptive reading, is met by CLAIM-220.
What it does not say
- It does not say the normative question is unimportant. It says this account does not answer it.
- It does not say any purpose licenses any rendering. A brief can be criticised, and a rendering can fail its brief; the account measures the second.
- It does not say the record holds a normative theory. Skopos theory (Reiss and Vermeer) states purpose-relative norms for translation; it is not held here, and CLAIM-023 records that it is absent.
Note of 2026-10-12: a counter on main that reads fidelity normatively, left to the owner
CLAIM-189, filed on main on 2026-10-11 as the counter to CLAIM-191, says the transport line’s “measures reward preservation” and that a fidelity profile “would prefer an argument that never changed”. This entry disclaims the normative part for the account: what ought to be kept is set by the translation’s purpose. CLAIM-189 does not cite this entry, and this entry does not consider a developing argument. CLAIM-189’s own “What would answer it” asks for a fidelity, “suitably defined”, that measures what the record secures. Whether the two conflict is left to the owner.
Note of 2026-10-12: the interlocutor’s reading, as advice
The interlocutor’s reading of 2026-10-12 reads this entry and CLAIM-189 as compatible: “A transformation loses some previously available distinctions. The loss was epistemically or pragmatically desirable.” CLAIM-250 files its proposal for what a developing argument should preserve. Recorded as advice.