The claim
A24 §6: “The context-dependent judgments through which communicative identity is constituted may exhibit structures that cannot be reconciled within a noncontextual probabilistic model.” A30’s H2: “The structure of those operations and their observable statistics cannot be represented by a common classical joint model satisfying explicitly specified compatibility and noncontextuality assumptions. This requires additional empirical constraints.”
A24 set it as the third of three layers: a pragmatic thesis (“well motivated”, now CLAIM-061), a theory of translation (“developing proposal”, the ancestor of CLAIM-115), and this. “I would resist collapsing these into one thesis. The first doesn’t establish the second, and neither establishes the third.” That discipline is the manuscript’s §11.
What it does not say
That meaning, the text or the reader is quantum. A21: “This is not evidence that the poem or brain is physically quantum.” Manuscript §9: “Quantum-like models may be useful in constrained cases, but are optional realizations”; §11: “not a presumption that pragmatic context effects are quantum”. The manuscript supposes it only so as to test it: §3 says “Whether pragmatic observables exhibit such patterns requires repeated, controlled measurement”, and §10 says to “report context effects separately from possible contextuality”. THEORY-013 counts against it for the data published so far (CLAIM-041).
The nearest test so far
Wang et al. (LIT-842) test lexical selection in ambiguous phrases, not pragmatic judgement. They report two contextual systems, and both fail the paper’s own parametric bootstrap. So the paper shows that the test can be run on language data. It is not evidence for this claim.
The BERT study
Lo, Sadrzadeh and Mansfield (LIT-844) test anaphora resolution, which is pragmatic, so they come nearer this claim than Wang et al. But the probabilities are BERT’s, not human judgements. And the contextuality is built into the schema: “the same one” and “the other one” make every instance a PR-prism. So it is not evidence about human pragmatic judgement.
Note of 2026-10-30: what holds it back
A supposition the manuscript makes in order to test it, and the test has not been run on human pragmatic judgements. The two language studies the record holds test lexical selection (LIT-842) or a language model’s anaphora resolution with the contextuality built into the schema (LIT-844). THEORY-013 reads the published behavioural data as showing none once direct influences are separated. It waits on a Contextuality-by-Default analysis of human pragmatic judgements that robustly finds or excludes contextuality.