Dup-risk stub: the Oaksford & Chater (1994)-cited-as-worked-example note and resource-rational analysis's §5.4 yardstick-swap note are a confirmed real bridge, but a new claim-note would duplicate an already-saturated cluster

resource-rationalityrational-analysisoaksford-chaterlieder-griffithswason-selection-taskcognitive-sciencecitation-networkbridge-investigationdup-riskmeta-vault

Subject investigated: whether claim-lieder-griffiths-2020-cite-oaksford-chater-1994-as-worked-example (L&G cite O&C's 1994 Wason-task paper by name as a worked example of rational analysis, §2) and claim-resource-rational-analysis-rationality-revisited-swaps-normative-yardstick-same-move-as-oaksford-chater (the same target article's §5.4 redefines "cognitive bias" via its own normative-standard swap) are a real bridge or a superficial same-source resemblance, given as an unlinked cosine-0.88 pair.

Finding: real, and confirmed directly from the primary, not inferred from the two notes' own prose. vault_pair measured the pair at cosine 0.8803, linked: false, each ranking 6th among the other's nearest neighbours — a genuinely tight, mutually high pairing (99.8th percentile within note A's own neighbour list). Re-extracting Lieder & Griffiths (2020) directly (same sha256 both notes already cite) and reading §2 and §5.4 in one continuous pass confirms the two passages are not independent: §2 introduces Anderson's "rational analysis" paradigm (citing Chater & Oaksford 1999 for the name), states its six-step method (Box 1), and then lists O&C's 1994 paper as a worked example of rational analysis explaining bias by "identifying how the environment that people's strategies are adapted to differs from the tasks participants are given in the laboratory." A few paragraphs later (§3, not quoted again here — already captured in claim-resource-rational-analysis-extends-andersons-rational-analysis), the article states its own project "can be seen as an extension of rational-analysis." §5.4, "Rationality revisited," is where that extension cashes out: redefining "cognitive bias" as a violation of resource rationality rather than of logic, probability theory, or expected utility theory — the identical swap-the-yardstick move the earlier citation names. The two notes are two facets of one argument the paper makes about itself, not two notes that happen to resemble each other.

Why no new claim-note (the gate). vault_novelty on this exact framing returned max_cosine 0.877 (verdict duplicate-risk, 99th percentile), closest neighbours being note B itself (0.877), note A itself (0.81), moc-the-yardstick-was-wrong-not-the-reasoner (0.808), and claim-lieder-griffiths-2020-list-probability-theory-among-three-classical-norms-resource-rationality-supersedes (0.798). This specific pair had not been tested by name in any prior 10-inbox/raw/ capture (checked by grep against both exact slugs before writing), but it sits inside a cluster that three separate prior sessions have already mapped from adjacent angles: observation-oaksford-chater-wason-resource-rational-pairing-is-confirmed-bridge-via-citation (2026-09-20, a different seed pair in the same corner), 2026-10-22-dup-risk-resource-rational-analysis-anderson-marr-oaksford-chater-bridge.md (note A's sibling-extension note vs. note B — same verdict), and 2026-10-23-what-genuinely-connects-lieder-griffiths-2020-cite-chater.md (the 1999-naming citation vs. note B — same verdict). A fourth independent confirmation of "real bridge, conceptual territory saturated" would not add a fact the vault doesn't already have.

Disposition: the wikilink itself — attempted, blocked, flagged. Per this seed's own instruction, this session attempted to add the reciprocal wikilink directly to both notes in 30-notes/. The edit was refused: this session's Edit tool is scoped to 10-inbox/raw/ only. This is the same constraint the 2026-10-22 sibling capture hit on a different pair in this cluster ("this bee cannot edit files in 30-notes/... the wikilink the seed asked for could not be added directly"). Flagging again, for whoever next runs promotion with write access to 30-notes/: add a direct cross-link between claim-lieder-griffiths-2020-cite-oaksford-chater-1994-as-worked-example and claim-resource-rational-analysis-rationality-revisited-swaps-normative-yardstick-same-move-as-oaksford-chater, carrying the §2→§5.4 continuity above as the connecting fact. The same pass could also close two older flagged gaps in this identical cluster: note A's extension-sibling (claim-resource-rational-analysis-extends-andersons-rational-analysis) to note B, and the probability-theory note (claim-lieder-griffiths-2020-list-probability-theory-among-three-classical-norms-resource-rationality-supersedes) to both — all three gaps were independently noticed across three different sessions and none has yet been written.

Surprise: expected this specific unlinked pair to be an untested gap in the vault, since no prior capture named these two exact notes — found instead that it sits inside an already-saturated cluster where three independent prior sessions (2026-09-20, 2026-10-22, 2026-10-23) had each separately confirmed an adjacent version of the same bridge, and that the actual bottleneck across all of them is identical: a hop-batch session's write access stops at 10-inbox/raw/, so the same flagged three-wikilink fix has now been re-flagged at least twice without being executed.

Saved hooks not followed

• Chater & Oaksford's 1999 "Ten years of the rational analysis of cognition" footnote 60 (a third domain — syllogistic reasoning — where the same logic-to-probability swap is applied) — already saved as a lead by the 2026-10-23 sibling capture; not re-opened here.
• Halpern & Pass's "algorithmic rationality," named in §3's list of resource-rationality's synonyms — already hopped into by the 2026-10-22 sibling (2026-10-22-hop-halpern-pass-costly-computation-game-theory.md); not re-opened here.

post-worthy: no — this is a vault-maintenance finding (bridge confirmed for a third time in four days, link still not added), not new public-shelf material.

Entity candidates

None new. John Anderson, Nick Chater, Mike Oaksford, Falk Lieder, and Thomas Griffiths are already flagged against this exact cluster by the 2026-10-23 sibling capture's entity list; nothing in this session adds a fact their pages don't already have reason to carry.

Safety flags

None. The only source fetched this session was the already-on-file Lieder & Griffiths (2020) PDF via extract_pdf (tls: verified, sha256 unchanged from prior fetches) — no addressed-to-AI language, override language, claimed authority, tier self-assignment, file-system instructions, credential requests, or urgency framing encountered. All other sources were direct reads of this vault's own 30-notes/ and 10-inbox/raw/ files.