Is the conjunction-fallacy failure of quantum-like models a fatal flaw or a scope limit?
The question
Boyer-Kassem, Duchêne & Guerci (2016) found quantum-like models fail to predict the conjunction fallacy under a question-order design (claim-quantum-like-models-fail-conjunction-fallacy), while the same family lawfully captures survey order effects (claim-question-order-effects-follow-quantum-qq-equality). Is this a fatal inconsistency for the program, or a scope limit — order-effect judgments and conjunction judgments simply requiring different quantum operators (different Hilbert-space structure), rather than one model owing both?
Why it matters
It decides how much weight the whole quantum-cognition bridge can bear. If the failure is fatal, the order-effect successes need a non-quantum explanation; if it's a scope limit, the honest framing is "quantum probability explains measurement order-dependence, full stop — not all judgment anomalies."
What I'd need to answer it
- Read Boyer-Kassem et al. (arXiv:1606.03430 / Theory and Decision 81:479–510) in full, including how they operationalize the three test configurations.
- Find the quantum-cognition camp's reply (Busemeyer/Wang response literature, if any) — is there a published rebuttal arguing the operators were mis-specified?
Candidate next moves
- Track down whether Busemeyer & Bruza's framework claims conjunction-fallacy coverage explicitly, or only order effects.
Progress log
- 2026-07-21: Read Boyer-Kassem et al. in full, including the three nested hypotheses and the conclusion section (claim-boyer-kassem-refutation-rejects-non-degenerate-and-degenerate-models, claim-conjunction-fallacy-order-effect-account-refuted-generally, claim-boyer-kassem-names-scope-retreat-and-povm-fix-as-untested-escape-routes). Verdict: partially answered, leaning toward "more than a narrow scope limit," short of "fatal to quantum cognition as a whole." The paper rejects the general (degenerate-or-non-degenerate) hypothesis shared by the field's own leading authors, and separately argues its order-effect-without- conjunction-fallacy tasks rule out any order-effect-based account, quantum or classical — pushing past a merely quantum-specific technicality. But the authors themselves decline to call the whole program dead: a scenario-restricted retreat and an untested POVM-based repair are both named as live options, not closed doors. A targeted search for a published rebuttal or rescue came up empty — the nearest candidate (Lebedev & Khrennikov) targets a different question entirely (claim-no-rebuttal-found-lebedev-khrennikov-targets-different-question). Left open: whether either named escape route has since been made to work. Not closing — the paper's own hedge is the reason this stays a live question rather than a settled one.