A logic-programming account of human conditional reasoning rivals the probabilistic (Oaksford–Chater) account of the same data

reasoning-psychologylogic-programmingbayesian-cognitionwason-selection-taskclosed-world-assumptioncognitive-science

The "irrational" results in the psychology of conditional reasoning — people affirming the consequent, denying the antecedent, and choosing the "wrong" cards in Wason's selection task — have more than one formal rehabilitation. The vault already holds the probabilistic lineage: Oaksford and Chater recast these tasks as optimal, calibrated inference under uncertainty rather than failed logic (claim-oaksford-chater-recast-wason-task-as-optimal-data-selection, claim-oaksford-chater-redefine-rationality-as-reasoning-about-uncertainty). Sakama (TPLP 2024) sets that lineage beside a distinct, non-Bayesian one: a computational-logic / closed-world account in which "relatively little attention has been paid for realizing such pragmatic inference in computational logic or logic programming (Stenning and Lambalgen (2008); Kowalski (2011))," explicitly juxtaposed against "(Oaksford and Chater (2001))."

The two accounts explain the same behavioural data through different machinery. The probabilistic account changes the objective function — from logical validity to Bayesian information gain. The logic-programming account keeps a logic but changes which logic — from the monotonic material conditional to non-monotonic closed-world reasoning, where reading "if" as "iff" is the completion of a rule set (claim-conditional-reasoning-fallacies-as-closed-world-inference-in-asp). (The specific operator label negation-as-failure is Sakama's, not S&vL's own term in their 2001/2004 papers — see claim-stenning-van-lambalgen-do-not-use-negation-as-failure-in-2001-2004-papers.) Stenning & van Lambalgen's frame distinguishes reasoning for an interpretation from reasoning from it (claim-stenning-van-lambalgen-distinguish-reasoning-for-vs-from-an-interpretation), so "many seemingly irrational choices [are] better understood as applications of alternative logics." This makes it a sibling to other "the yardstick was wrong, not the reasoner" moves the vault tracks, and to other alternative formalisms for reasoning anomalies such as quantum-cognition models (claim-quantum-like-models-fail-conjunction-fallacy).

The existence of the logic-programming lineage is Tier-1 attested via Sakama's citations, and the specific characterization that Wason results are closed-world reasoning is now confirmed directly from S&vL's own 2001/2004 primaries (claim-stenning-van-lambalgen-model-wason-fallacies-as-closed-world-reasoning). The one residual: whether the specific negation-as-failure operator is S&vL's own term appears (from these two papers) to be no — it is Sakama's gloss — with the fuller 2005/2008 development still unread. See question-verify-stenning-van-lambalgen-closed-world-wason.

Correction history.

• 2026-09-05 — Original note (2026-07-12) rested on Sakama's secondary paraphrase of "Stenning and Lambalgen (2008)," carried an [unverified-claim — 2008 primary unread] flag, and (a) attributed negation-as-failure to S&vL as their own term and (b) paraphrased their two-step frame as reasoning "first to an interpretation and then from it." A direct read of S&vL's 2001 JoLLI and 2004 Cognitive Science primaries (promotion of 10-inbox/raw/2026-08-31-do-stenning-van-lambalgen-actually-model-the-wason.md) confirms the closed-world characterization in their own words (claim-stenning-van-lambalgen-model-wason-fallacies-as-closed-world-reasoning) but corrects both details: negation-as-failure is Sakama's operator label, absent from both S&vL papers (claim-stenning-van-lambalgen-do-not-use-negation-as-failure-in-2001-2004-papers), and their actual frame is reasoning for vs from an interpretation, not "to… from" (claim-stenning-van-lambalgen-distinguish-reasoning-for-vs-from-an-interpretation). Body and flag updated to match; status seedling → budding. The 2008 book itself remains unread, but the characterization no longer depends on it.