The objection
Footing and stance are what the line’s theses say survives or is lost. CLAIM-115’s defeat condition measures “pragmatic observables (stance, footing, alignment, expected uptake)”. CLAIM-061 says renderings can differ “in footing and illocutionary force”. CLAIM-084 makes the primary outcome of a reconstruction study “pragmatic-frame retention judged blind to condition”, and the A70 passage it quotes says that “requires independently elicited judgments about footing, stance, and communicative function”. The designs fix the response categories in advance: CASE-021 elicits “condemnation, solidarity, amusement and irony”, and CASE-040 collects interpretive judgements of “tone, stance, solidarity, and authority”.
The only footing judgement the record has made twice of the same texts disagreed with itself. CASE-012: “The speaker labels drifted. A38 §2.2 called the four ‘affectionate teasing, cynical complicity, resigned observation, and social criticism’; A39 §1.2 made them ‘A fellow participant teasing an acquaintance’, ‘A worldly insider articulating an informal rule’, ‘A sympathetic observer recognizing a recurrent failure’ and ‘A critic of consumption and lifestyle inflation’, so that ‘The party always gets its cut’ went from cynical complicity to an informal rule without comment.” Its conclusion: “The speakers are the analyst’s attribution, not a property of the wording.” The judge was one analyst labelling texts it had itself written and knowing what each was meant to do, and its labels did not hold between two passes.
No design in the record supplies what a measure of footing needs:
- a definition given to judges (“footing” is a technical term, Goffman’s, LIT-780, which the record holds unread);
- a response format beyond the analyst’s category list;
- a reliability target, between judges or between passes of one judge;
- a validation against a contrast whose answer is fixed in advance, such as renderings built to differ only in the attributed speaker.
The manuscript’s split annotators (§10, as QUESTION-005 reports it) guard against fitting a measure on its own test data. They do not show that the judgements are reliable. And the readings do not reach the footing half. CLAIM-061’s “Support from dynamic semantics” ends: “None of the three addresses footing.” Austin and Goffman “are cited for it and unread here” (CLAIM-061; LIT-785, LIT-780).
The record concedes neighbouring points and not this one. CLAIM-014’s defeat condition allows that performative updates may not be “elicited as observables with stable response distributions”, but only for commitments, obligations and permissions. QUESTION-005 notes that A49’s admissibility condition “takes the observable distinctions as given”, which is the analyst’s category list; it asks how to fix a correspondence of contexts, not whether one context’s footing can be judged reliably.
What it does not say
It does not say footing or stance cannot be measured, nor that lay readers cannot tell teasing from reproach. It says the record holds no evidence that they can do so reliably in the categories the designs use, and one piece of evidence that the analyst could not. It does not object to CLAIM-061 or CLAIM-084 as stated, only to the evidence they can presently draw on: until a measure is validated, a defeat test run on these categories tests the categories as much as the thesis.