---
id: "20260919-0229-bipartite-two-things-this"
title: "Resource-rational analysis's own target article cites Oaksford & Chater's Wason-task paper by name — the bipartite pair is a direct citation, not a shared-ancestor inference"
type: "capture"
status: "promoted"
promoted_to: []
not_promoted: ["Claim 1 (Lieder & Griffiths 2020 cite Oaksford & Chater's 1994 Wason-task paper by name as a worked example) — fully duplicates the already-promoted [[claim-lieder-griffiths-2020-cite-oaksford-chater-1994-as-worked-example]], written the following night (2026-09-20) from the sibling capture 10-inbox/raw/2026-09-20-bipartite-two-things-this-vault-knows-in-different.md, which independently investigated the identical seed pair against the identical primary (same source_sha). That note's quote is the fuller, cleanly-grounded version of the same sentence this capture only captured in two fragments, and it has since been cross-model audited (claude-opus-5, 2026-09-22) and stamped verified_verbatim (2026-10-01, seek_verify). Nothing this capture adds survives retrieve-before-write.","Claim 2 (Lieder & Griffiths 2020 attribute the naming of 'rational analysis' to Chater & Oaksford's 1999 retrospective) — fully duplicates the already-promoted [[claim-lieder-griffiths-2020-name-chater-oaksford-1999-as-rational-analysis-source]], from the same 2026-09-20 sibling capture, same primary, same in-text sentence. The existing note is additionally more mature than this capture's version: a 2026-09-22 audit caught and corrected a scope overclaim (the term is also attributed to Anderson 1990 alone twice elsewhere in the same article) that this capture's own phrasing happens not to make but also never checked for. Verified_verbatim as of 2026-10-01. Not promoted.","Claim 3 (the bipartite resemblance is a direct citation chain, not a shared-ancestor coincidence) — fully duplicates the already-promoted [[observation-oaksford-chater-wason-resource-rational-pairing-is-confirmed-bridge-via-citation]] from the same 2026-09-20 sibling capture: identical verdict, identical reasoning (both citations sit on facing pages 3-4 of the same target article), already cross-model audited and confirmed. This capture's commentary callout makes a distinct small point — that retrieval almost gave the answer away by stopping at 'shared ancestor' — which is a reflection-worthy observation but not a new claim; recorded here rather than re-promoted.","Further leads (whether Oaksford & Chater's probabilistic framework ever gets a resource-rational treatment in later work; which open-peer commentators tagged CD/ESD also cite Oaksford & Chater 1994; Oaksford & Chater 2007's Bayesian Rationality content) — genuine unchased leads, but they substantially overlap the 2026-09-20 sibling capture's own further-leads list (Bayesian Rationality 2007 unopened, Hahn & Oaksford 2007 unopened, whether resource-rational analysis's five-step method has been applied to the Wason task directly), already logged and left unchased there. None is load-bearing for any claim kept this session (none were kept), so none is routed to 50-questions/ per the intake-discipline rule against promising minor nice-to-verify work.","Entity candidates (John R. Anderson, Mike Oaksford, Nick Chater, Falk Lieder, Thomas L. Griffiths, Peter Wason) — all already have hub pages. The specific citation-chain fact is already recorded: entity-john-anderson.md, entity-mike-oaksford.md, and entity-nick-chater.md each carry a dated 2026-09-20 Updates line naming it; entity-falk-lieder.md and entity-thomas-griffiths.md already list both claim-notes in References. One real gap found and fixed this session: entity-resource-rational-analysis.md, the central concept hub this capture itself flagged, had never been updated with this sharper citation fact despite being the obvious place for it — appended a dated Updates line there. Peter Wason's page was judged not to need updating: the citation-chain fact is about Lieder & Griffiths' and Oaksford & Chater's citation practice, not a new fact about Wason himself or his task, and the page's existing description of the Bayesian-camp recast already stands uncontradicted."]
origin: "batch"
writer_model: "claude-sonnet-5"
date_created: "2026-09-19T00:00:00.000Z"
provenance: "Batch run 2026-09-19, bridge-seed check against topic pairing [[claim-oaksford-chater-recast-wason-task-as-optimal-data-selection]] × [[claim-resource-rational-analysis-formalizes-utility-minus-compute-cost]] (cosine 0.84). 10-inbox/raw/ was searched for a capture mentioning both note slugs by exact string match; none exists (a different capture, 2026-09-17, investigates a different bipartite pair — the Werbos/Oaksford-Chater optimization-as-theory-of-mind bet vs. Gigerenzer's ecological-rationality flag — and is not a sibling of this pair). This is fresh primary research, not a bridge-seed duplicate."
derived_from: []
tags: ["rational-analysis","resource-rationality","oaksford-chater","wason-selection-task","lieder-griffiths","bayesian-cognition","cognitive-science","cross-domain-bridge","bridge-seed-check"]
source_url: "https://cocosci.princeton.edu/papers/lieder_resource.pdf"
source_sha: "6c61b43dc361ea3fe39d1c33bc945ba1c32916920b3a45f8f19f0c0574503680"
source_title: "Resource-rational analysis: Understanding human cognition as the optimal use of limited computational resources"
source_author: "Falk Lieder and Thomas L. Griffiths"
source_date: "2020-01-18T00:00:00.000Z"
source_venue: "Behavioral and Brain Sciences 43, e1:1-60, Cambridge University Press (target article + open peer commentary + authors' response)"
source_tier: 1
seek_code_commit: "200835c"
recovered: true
recovered_on: "2026-10-01T00:00:00.000Z"
recovered_from: "Seek session transcript a7c674ff-f92c-49d3-82f1-914124b80338.jsonl on the mini - the session's Write at 2026-09-19 09:36 UTC, replayed verbatim"
recovery_note: "Overwritten the same night: another bipartite session was told to write the same filename (2026-09-19-bipartite-two-things-this-vault-knows-in-different.md). Collision fixed in seek-code c516ae8. Restored unpromoted at Cali's instruction, 2026-10-01. Before promoting, read the same-pair captures listed below: several sessions researched this pair, and only one per night survived."
same_pair_captures: ["2026-09-20-bipartite-two-things-this-vault-knows-in-different.md"]
---


**BRIDGE-SEED DISCIPLINE:** `10-inbox/raw/` was searched for any capture
mentioning both `claim-oaksford-chater-recast-wason-task-as-optimal-data-selection`
and `claim-resource-rational-analysis-formalizes-utility-minus-compute-cost` by
slug. No file mentions both. A superficially similar-looking capture exists
(`2026-09-17-bipartite-two-things-this-vault-knows-in-different.md`), but it
investigates a *different* named pair (the Werbos/Oaksford-Chater
"optimization as theory of mind" bet vs. Gigerenzer's "not optimization"
flag) — not this one. This capture is genuine new research, not a duplicate
check.

**Retrieve-before-write context.** The vault already knows, via three
existing notes, that both ends of this pair independently trace back to
[[claim-rational-analysis-anderson-optimal-solution-marr-level|Anderson's
1990 rational analysis]]: Oaksford & Chater's Wason reanalysis is described
in the vault as rational analysis's "canonical demonstration," and
[[claim-resource-rational-analysis-extends-andersons-rational-analysis|resource-rational
analysis self-describes as a direct extension of Anderson's rational
analysis]]. That is a real but *structural* link — two programs sharing a
common ancestor, established independently in two different capture
sessions, never directly cross-checked against each other. This session did
the direct check: does Lieder & Griffiths's own paper actually cite Oaksford
& Chater's specific Wason-task work, or only Anderson? It cites both, by
name, in the target article's own text and reference list.

**Verdict: not a false friend, and tighter than the existing structural
link.** Lieder & Griffiths (2020) — the primary source behind
[[claim-resource-rational-analysis-formalizes-utility-minus-compute-cost]] —
directly cites Oaksford & Chater's 1994 optimal-data-selection paper (the
paper behind
[[claim-oaksford-chater-recast-wason-task-as-optimal-data-selection]]) as a
worked example of rational analysis explaining a classic bias, and separately
cites Chater & Oaksford's 1999 retrospective as the source of the term
"rational analysis" itself. This is a direct citation chain, not merely two
notes converging on a shared ancestor from opposite directions.

---

## Claim: Lieder & Griffiths (2020) cite Oaksford & Chater's 1994 Wason-task paper by name as a worked example of how rational analysis explains classic reasoning biases

**Claim type:** technical-mechanism / historical (a specific citation
relationship between two named primary sources) — Tier 1 required and met.

In the target article's review of rational analysis's successes, Lieder and
Griffiths write that such models have "provided surprisingly good
explanations of cognitive biases by identifying how the environment that
people's strategies are adapted to differs from the tasks participants are
given in the laboratory," and give as one of the named examples "the
confirmation bias (Austerweil & Griffiths 2011; Oaksford & Chater 1994)."
"Oaksford & Chater 1994" resolves, in the paper's own alphabetized reference
list, to: "Oaksford, M. & Chater, N. (1994) A rational analysis of the
selection task as optimal data selection. *Psychological Review*
101(4):608–31," tagged [aFL, CD, ESD] — cited by the target-article authors
themselves (the "a" prefix), not only by a commentator. This is the exact
1994 paper that [[claim-oaksford-chater-recast-wason-task-as-optimal-data-selection]]
already identifies as the source of "the specific expected-information-gain
mathematics" behind the Wason-task reanalysis — the paper is the Wason
selection task, reframed as optimal (information-maximizing) data selection,
which is the standard reading of Wason's task as a test of confirmation-bias
behavior.

**Provenance:**
- source_url: https://cocosci.princeton.edu/papers/lieder_resource.pdf
- source_sha: 6c61b43dc361ea3fe39d1c33bc945ba1c32916920b3a45f8f19f0c0574503680 (fetched via extract_pdf this session; tls: verified)
- source_title: "Resource-rational analysis: Understanding human cognition as the optimal use of limited computational resources"
- source_author: Falk Lieder and Thomas L. Griffiths
- source_date: 2020-01-18
- source_venue: Behavioral and Brain Sciences 43, e1, Cambridge University Press
- source_tier: 1
- exact quote (quote_check-grounded against the extracted text): "the confirmation bias" ... "Austerweil & Griffiths 2011; Oaksford & Chater 1994" (these two fragments are quote_check-grounded individually; the pdftotext extraction of this two-column PDF breaks the full sentence across a column/line boundary that the grounding tool cannot bridge as one string, so both fragments are recorded separately rather than as one over-assembled quote)
- exact quote, bibliography entry (quote_check-grounded verbatim): "Oaksford, M. & Chater, N. (1994)" and "rational analysis of the selection task as optimal" (both grounded; the full entry reads, per direct visual read of the same extracted text at the reference-list line: "Oaksford, M. & Chater, N. (1994) A rational analysis of the selection task as optimal data selection. Psychological Review 101(4):608–31. doi:10.1037/0033-295X.101.4.608. [aFL, CD, ESD]")

---

## Claim: Lieder & Griffiths (2020) attribute the naming of the "rational analysis" paradigm itself — the framework resource-rational analysis explicitly extends — to Chater & Oaksford's 1999 retrospective

**Claim type:** technical-mechanism / historical (citation attribution) — Tier
1 required and met.

Introducing Anderson's program, the target article states: Anderson (1990)
developed the idea of cognition as rational adaptation to environmental
structure, "creating a cognitive modeling paradigm known as rational
analysis (Chater & Oaksford 1999)." The bracketed citation resolves in the
reference list to "Chater, N. & Oaksford, M. (1999) Ten years of the
rational analysis of cognition. *Trends in Cognitive Sciences* 3(2):57–65,"
tagged [aFL] — cited by the target-article authors. This is the same paper
already in the vault as the source of
[[claim-chater-oaksford-1999-list-marr-integration-as-outstanding-question]].
So the paper that names, for Lieder and Griffiths, the very paradigm they
extend into resource-rational analysis is a paper co-authored by one half of
this topic's other seed note (Oaksford & Chater's Wason work) — reinforcing
that the connection is not incidental.

**Provenance:**
- source_url / source_sha / source_title / source_author / source_date / source_venue / source_tier: as above (same primary, same fetch)
- exact quote (quote_check-grounded): "known as rational analysis" (grounded verbatim); the surrounding sentence, read directly from the extracted text at the relevant lines, is: "creating a cogni-\ntive modeling paradigm known as rational analysis (Chater &\nOaksford 1999) that derives models of human behavior from structural environmental assumptions" — the line-wrapped hyphenation ("cogni-/tive") and the column-boundary split around "(Chater &/Oaksford 1999)" mean quote_check cannot ground the whole sentence as one string; "known as rational analysis" is the load-bearing grounded fragment, and the citation "(Chater & Oaksford 1999)" plus its full reference-list resolution are confirmed by direct read plus a separate quote_check hit on the reference-list title "Ten years of the rational analysis of cognition" (grounded verbatim).

---

## Claim: the bipartite resemblance is real and is a direct citation chain, not a false friend and not merely a shared-ancestor coincidence

**Claim type:** structural/historical claim about the relationship between
the two seed sources — Tier 3–4 acceptable for this kind of claim, but
grounded here at Tier 1 since it rests on the two citation facts above.

The two seed notes were flagged as sharing no vocabulary by construction.
That is true at the level of each note's own text, but the primary sources
behind them are not independent: the paper that grounds
[[claim-resource-rational-analysis-formalizes-utility-minus-compute-cost]]
cites, in its own body text and reference list, both the specific 1994 paper
behind
[[claim-oaksford-chater-recast-wason-task-as-optimal-data-selection]] (as a
worked example of rational analysis explaining a classic bias) and the 1999
paper that supplies the very name "rational analysis" for the paradigm
resource-rational analysis extends. This sharpens the vault's existing
"both descend from Anderson" structural finding
([[claim-resource-rational-analysis-extends-andersons-rational-analysis]],
[[observation-resource-rational-analysis-confirms-anderson-leg-not-werbos-leg]])
into something more specific: Oaksford & Chater are not just adjacent to
Anderson in the same tradition, they are directly cited, by name, inside the
paper that formalizes resource-rational analysis's compute-cost objective.
The resemblance the embedding model picked up (cosine 0.84, no shared
vocabulary) corresponds to a real, documented, single-hop citation — not a
coincidence and not a false friend.

> [!note] Seek's commentary:
> The nice part of this one is how the vault's own retrieval almost gave the
> answer away and then undersold it. Two separate notes already said "both of
> these trace to Anderson" — true, but that's a two-hop, retrospective
> observation built by a synthesis note, not a citation either author actually
> made. Going to the primary text turned "these two both belong to the same
> tradition" into "this paper cites that paper, by name, in its own reference
> list" — a stronger and more falsifiable claim than the vault had. It's a
> reminder that "shared ancestor" and "direct citation" are different
> strengths of evidence even when they end up pointing at the same two texts. — Seek

---

## Further leads

- Neither Oaksford & Chater 1994 nor 1999 nor 2007 is cited anywhere in the
  target article's discussion of the *compute-cost formalism itself*
  (Equations 2–4) — they appear only in the historical/motivating sections
  (§1–2), not in resource-rational analysis's own mathematical machinery.
  Whether Oaksford & Chater's probabilistic framework ever gets a
  resource-rational (compute-cost-penalized) treatment in later work by
  either camp is untested here — a natural next hop.
- The reference list's [aFL, CD, ESD] tags mean at least two open-peer
  commentators (initials CD, ESD) also cited Oaksford & Chater 1994
  independently of the target article authors — worth identifying which
  commentaries these are and whether they engage the Wason-task material
  more directly than the target article does (not read this session; the
  commentary section is ~45 pages and was not searched exhaustively for
  these two specific citation tags).
- Oaksford & Chater (2007), *Bayesian Rationality*, is also cited by the
  target article [aFL] but not examined for content in this session — only
  its bibliography entry was confirmed present.

## Safety flags

None. The only source fetched this session was
https://cocosci.princeton.edu/papers/lieder_resource.pdf via `extract_pdf`
(tls: verified, no elevated-suspicion transport) — a standard academic PDF
hosted on the author's own institutional site. No addressed-to-AI language,
override language, claimed authority, tier self-assignment, file-system
instructions, credential requests, or urgency framing was encountered in the
extracted text (target article, commentaries, or references).

## Entity candidates

- John R. Anderson — person — the foundational figure both seed notes'
  research programs measure themselves against and explicitly extend
  (Anderson 1990's rational analysis); already covered by
  [[claim-rational-analysis-anderson-optimal-solution-marr-level]] in the
  vault, flagged first here per the ancestry-priority convention rather than
  left implicit.
- Mike Oaksford, Nick Chater, Falk Lieder, Thomas L. Griffiths — people —
  already have entity pages in the vault
  ([[entity-mike-oaksford]], [[entity-nick-chater]], [[entity-falk-lieder]],
  [[entity-thomas-griffiths]]); this capture adds a direct citation edge
  between their work, not new entities.
- resource-rational analysis — concept — already has an entity page
  ([[entity-resource-rational-analysis]]); this capture's finding (the direct
  citation to Oaksford & Chater 1994/1999) could be added to that page's
  genealogy section as a sharper claim than "extends Anderson" alone.
- Peter Wason — person — already has an entity page ([[entity-peter-wason]]);
  the selection task itself is the object both seed notes and this capture's
  citation chain run through.