---
title: "Human conditional-reasoning 'fallacies' can be realized as closed-world inference in Answer Set Programming"
type: "claim"
status: "seedling"
audit_status: "capture-verified — the bee read arXiv:2311.04412 (Sakama, TPLP 2024) via extract_pdf, TLS-verified, at capture time; the queen's independent re-check was not run in this headless promotion. Treat as capture-verified, not verified-verbatim."
writer_model: "claude-opus-4-8"
source_url: "https://arxiv.org/abs/2311.04412"
source_title: "Human Conditional Reasoning in Answer Set Programming"
source_author: "Chiaki Sakama (Wakayama University)"
source_date: "2024-01-01T00:00:00.000Z"
source_quote: "Inference in ASP is deduction based on default logic (Reiter (1980))."
source_tier: 1
provenance: "Promotion from 10-inbox/raw/2026-07-11-hop-closed-world-human-reasoning.md, 2026-07-12"
origin: "batch"
derived_from: ["10-inbox/raw/2026-07-11-hop-closed-world-human-reasoning.md"]
date_created: "2026-07-12T00:00:00.000Z"
tags: ["reasoning-psychology","logic-programming","non-monotonic-reasoning","closed-world-assumption","wason-selection-task","cognitive-science"]
---


In the psychology of conditional reasoning, two inference patterns are canonically labelled fallacies: *affirming the consequent* (AC — inferring P from "if P then C" together with C) and *denying the antecedent* (DA). They are logically invalid under the material conditional, yet people perform them routinely. Sakama's framing is that this is not brokenness but pragmatics: "humans are more likely to perform logically invalid but pragmatic inference," and "humans often perform AC or DA as pragmatic inference in daily life" — driven by *conditional perfection*, the tendency to hear "if" as "if and only if" (Geis & Zwicky 1971, cited but not read in primary here).

Sakama (TPLP 2024) formalizes AC, DA, and DC by realizing them in **Answer Set Programming**, chiefly through program *completion*. The load-bearing engine is non-monotonic: "Inference in ASP is deduction based on default logic (Reiter (1980))." Reiter's default logic is the same lineage that produced the **closed-world assumption** for incomplete databases in 1978 — "what is not currently known to be true must be false" ([Wikipedia, CWA](https://en.wikipedia.org/wiki/Closed-world_assumption), Tier 4, definitional pointer; the primary is Reiter 1978) — and its operational cousin *negation-as-failure*, which "amounts to believing false every predicate that cannot be proved." Reading "if" as "iff" is exactly what a closed-world completion of a rule set does.

The significance is that database machinery built for incomplete knowledge doubles as a model of how people reason with conditionals. This mechanism is the concrete engine underneath the broader claim that a logic-programming account of human reasoning rivals the probabilistic one ([[claim-logic-programming-account-rivals-probabilistic-account-of-conditional-reasoning]]), and it grounds the "fallacies" in the same [[claim-inference-logical-types|deduction/induction/abduction]] vocabulary the vault already tracks.

> [!note] Seek's commentary:
> The elegant part is the identity between a *linguistic* move (conditional perfection, "if" → "iff") and a *computational* one (closed-world completion). The ASP realization is Sakama's own Tier-1 contribution and cleanly quoted; the conditional-perfection driver rests on a 1971 primary I have not opened, so I have stated it as Sakama frames it rather than asserting the pragmatics independently. — Seek