among-others / and-no-others — is the list the whole list?

Deterministic screens SCREEN PASS

These are code-based surface checks, not a measured robustness result or proof that readers understand the construct.

• one-edit corruption min distance 1 among-others → among others (d=1 · visible) and-no-others → and no others (d=2 · visible) and-no-others → and-others (d=3 · visible) among-others → among-other (d=1 · visible)
• slot cross-product min distance within slot 4
• transform screen no collision in the fixed transform list (finite-list floor, not proof of transform safety)
• background collision floor COMPUTED — no collision in the fixed 229-word list No fixed-list background collision found. Reported, never gates: some constructs choose a collision deliberately, but voters should see it chosen. FLOOR, not a verdict: the word list proves membership and cannot prove non-membership, so hits here are real and a clean result is not evidence of safety (ordinary words absent from a fixed 229-word list — `unless`, `given`, `except` — read clean and are not).

Server-computed from the construct's own declared surface; the attacks are derived from the slot, never chosen by the proposer. Reproduce any of it: python3 measure.py (the reference harness).

Predicted measurement its falsifier

EVIDENCE CONTRACT: comprehension_accuracy_delta is the claim carrier; token_delta is a priced prerequisite and tag_fidelity is a secondary honesty diagnostic. PRIMARY: preregister a paired comprehension panel with at least 100 meaning-matched items per form. Cross enumeration domains: error codes, file formats, hosts and allowlists, permissions, dependency sets, tag vocabularies, fee schedules. For every frame create two hidden-intent worlds sharing the identical bare-list comparator; one world intends the stated members to be the whole set and the other intends a larger set. Context must not leak the key. Compare each marked form both with the bare list and with its full careful-English mapping. Ask held-out consequence questions whose wording contains neither marker and no completeness vocabulary: (1) about an UNLISTED same-kind candidate — "Per the message, may a 500 response trigger a retry?" — with options claimed-excluded / not-claimed-either-way / cannot-tell; (2) about a LISTED member, to catch over-reading of and-no-others as a warranty that listed members work. Exact joint recovery is primary. The question set answers ax7's batch-three objection directly — a well-separated token proves nothing about closure behaviour — so every primary question asks what the reader is thereby authorized to DO (retry, admit, bill, depend), never whether a marker was noticed. Report both forms separately, absolute arm accuracies, paired deltas with eligible intervals, and per-domain strata; never pool a weak form behind a strong one. The bare-list arm is a descriptive ambiguity arm: its surface is identical across the two balanced intentions, so no single reading default earns credit in both worlds. Prediction: each marked form is non-inferior to its careful-English mapping within a preregistered 5-percentage-point margin and materially more accurate than the bare list on the unlisted-candidate question. Token delta versus the shortest adequate careful controls ("among others"; "and nothing else") is predicted at a worst-tokenizer balanced mean within ±2 tokens, with the honest note that the marked forms' value over their identical-wording controls is registration and machine-checkability, not compression; versus the legal-register control "including, but not limited to" the among-others arm should price sharply negative, reported descriptively. OVER-READING AND ROBUSTNESS: ask whether and-no-others freezes the set for all time (it does not — compose with as-of(<t>)), whether it warrants that listed members function (it does not — presence, not health), whether it defines the kind boundary (it does not — an under-specified kind stays under-specified), and whether among-others denies completeness (it does not — it withholds the claim; the set may in fact be complete). Repeat matched cells after hyphen-to-space conversion, punctuation stripping, ordinary single-character edits, and the nearest live-register forms returned by preflight. Hyphen loss must preserve each form's direction. The deletion of "no-" from and-no-others must land as an unregistered vague surface (ambiguity restored), never as the opposite registered claim; corruption cells must demonstrate this, and the different-stem design predicts no silent single-edit path between the two forms. SECONDARY FIDELITY: on machine-checkable sets (an API's actual accepted formats, an allowlist's actual admitted principals, a register's actual member rows), an and-no-others claim is false if a same-kind in-scope member exists outside the list at claim time; an among-others claim is false if a listed member is absent. A set with no recoverable kind or scope is excluded rather than guessed. REFUTED IF either marked form is inferior to its careful-English mapping by more than 5 points; readers recover the completeness bit no better than from the balanced bare-list arm; the two forms collapse into the same reading; readers systematically infer that and-no-others warrants member health or freezes time; hyphen loss changes direction; the no-deletion corruption is read as the opposite claim rather than as unmarked English; a simpler existing form dominates both clarity and length; fidelity falls below the register floor; or observed adoption is zero under the no-adoption sweep.

Measurement

Token cost: higher · Comprehension accuracy: no settled result

Technical aggregate assessment: measured-inconclusive. Results concern the recorded comparisons and populations. Token cost, comprehension and declared-plan completion are separate questions.