Why it is a question
The record’s audit of 2026-10-10 found that the criterion in every central defeat condition is a human judgement with no task and no receiver named. CLAIM-244: “Human judgements of communicative fidelity” (CLAIM-115), “Human judgements of shared communicative kind” (CLAIM-139), while the line’s own theses say fidelity is task-relative (CLAIM-128) and relative to the interpreter (CLAIM-074). CLAIM-226: averaging a convenience sample “measures the act for an unnamed population”, and for interlingual cases the source and target judges are different communities.
The record’s replies answer part of it.
- CLAIM-220 indexes the manuscript’s thesis to a named population and stated tasks, and CLAIM-204 does the same for the October thesis.
- CLAIM-240 says a holistic judgement is well defined as fidelity under a population’s default task, if the population is named and the weighting fixed in advance.
- CLAIM-217 makes competent participants’ uptake the standard of a kind, indexed to a population.
- CASE-055 names populations, tasks and an invariance test for one study.
None of these says which tasks and receivers the line’s claims are about in general, as opposed to in one study. Three things are open.
- Tasks. Choice of act and choice of continuation are the study’s tasks. Whether they exhaust what “recover and act upon” means, or stand for a wider family, is not argued.
- Receivers. Whether the claims are about the target language’s ordinary readers, about the translation’s intended audience, or about competent participants who know the standings in play (CLAIM-217), and whether these come apart.
- Across populations. How fidelity for one population bears on another: whether a rendering faithful for one community can be unfaithful for another, and what then is the fidelity of the rendering (CLAIM-074).
An answer names, for each central claim, a task family and a receiver population, and states how a result for one population is carried to another, or that it is not.
Note of 2026-10-12: filed positions on main
Since PR #121 branched, main has filed two positions on whose receiver counts. CLAIM-160 separates the writer’s projected receiver, which fixes intended meaning at the write, from the actual reader, at whose read fidelity is measured. CLAIM-171 objects that the projected receiver is one more reading. Neither names a task family or a receiver population, so neither answers this question as posed. CLAIM-160’s projected receiver is the source-side counterpart of “the translation’s intended audience” named above.