Logic Checklist¶
Checklist — instantiates Deductive Chain Validation
Runs a fixed set of prompts over any argument — hidden premises, equivocal terms, invalid steps — so common reasoning faults are caught by routine rather than by luck.
A Logic Checklist is a short, fixed list of yes/no prompts a reviewer runs over any argument: Is a premise missing? Does a key term shift meaning? Does the conclusion actually follow if the premises are granted? Its defining trait is breadth over depth — it does not deeply investigate any single claim or gather evidence; it guarantees that a standard set of common reasoning faults gets looked for every time, by routine rather than by the reviewer happening to notice. It is portable and content-agnostic, and its whole value is that it catches the faults that slip past precisely because nobody thought to look for them. It is a scanning instrument, not an investigator.
Example¶
A product team makes a confident pitch in a strategy meeting: "Our biggest competitor shipped an AI assistant and their signups jumped, so we should ship one to grow." Rather than argue by vibes, a reviewer runs the logic checklist over it. Missing premise? — yes, two: the argument silently assumes the assistant caused the signup jump, and that the competitor's users behave like ours; both get written down as premises the pitch was riding on unstated. Equivocation? — grow means "signups" in this sentence, but the quarter's goal doc defines the target as "paid conversions," so the conclusion answers a different metric than the goal. Does it follow? — even granting both premises, jumping from "the assistant helped them" to "we should ship one" skips an unstated cost-and-fit step. In ninety seconds a persuasive-sounding pitch has become three specific, answerable objections — none of which required any new data.
How it works¶
The checklist is a fixed battery of prompts, each aimed at one recurring fault: missing or hidden premises, equivocal or drifting terms, invalid inference steps, overbroad conclusions, and unscanned exceptions. The reviewer walks the argument against the list in order, forcing a conscious yes/no on each — which converts "I didn't spot a problem" (the absence of a red flag) into "I checked for this specific problem and here is what I found." It works precisely because it is dumb and complete: it does not rely on the reviewer's insight or mood, only on their willingness to run the list. That makes it fast, teachable, and repeatable across reviewers who would otherwise each miss a different fault.
Tuning parameters¶
- List length — a three-prompt pocket version versus a twenty-item audit. Longer lists catch rarer faults but invite rushed, box-ticking passes that catch nothing.
- Prompt specificity — generic ("is this valid?") versus fault-named ("is a term equivocating between two senses?"). Named prompts find more but must be maintained as the argument domain shifts.
- Forced-answer strictness — whether every prompt demands an explicit answer or may be skipped. Mandatory answers prevent silent skipping but slow the pass and breed rote responses.
- Independence — run by the argument's author or by a second reader. A second reader catches the author's blind spots; self-review is faster but shares the author's frame.
When it helps, and when it misleads¶
Its strength is turning error-catching from a matter of luck and talent into a routine — a cheap forcing function[n1] that surfaces the boring, common faults (a suppressed premise, a slippery word) that experts miss because they are looking for something cleverer. It scales across reviewers and needs no domain expertise to run.
Its failure mode is checklist theater: the list is walked ritually after the decision is made, every box ticked, nothing actually examined — the form of validation with none of the substance. A related misuse is treating a passed checklist as proof of soundness, when the list only prompts for faults and never verifies that the surfaced premises are actually true. The guarding discipline is to require a written answer to each prompt (not a checkmark) and to remember the checklist's output is a list of questions to pursue, not a verdict — a clean pass means "no common fault was spotted," never "the argument is correct."
How it implements the components¶
premise_list— its prompts force the reviewer to surface missing and hidden premises and write them down as explicit inputs, so nothing load-bearing stays unstated.definition_alignment— a dedicated prompt scans for equivocation: does each key term hold one stable meaning across the rule, premises, and conclusion.validity_check— the core yes/no prompt asks whether the conclusion follows if the premises are granted, testing the support relation without judging the premises' truth.
A checklist only prompts; it does not itself verify that the surfaced premises are true (premise_verification) or label the leftover evidential uncertainty (residual_uncertainty_label). That case-specific, empirical work belongs to Diagnostic Logic Check, which applies fixed criteria to one verified case rather than scanning an argument in the abstract.
Related¶
- Instantiates: Deductive Chain Validation — the Logic Checklist is the lightweight, general-purpose pass that operationalizes the archetype when full formality is overkill.
- Sibling mechanisms: Syllogism Template · Diagnostic Logic Check · Legal Syllogism Review · Policy Eligibility Review · Rule-Engine Validation · Requirements Traceability Check · Proof Checking
Editorial Notes¶
Form Classification¶
Form family: Assessment, Review & Assurance
Rationale: Logic Checklist operates as a bounded evaluation of existing evidence or work that produces a finding or disposition because it runs a fixed set of prompts over any argument — hidden premises, equivocal terms, invalid steps — so common reasoning faults are caught by routine rather than by luck.
Independent corroboration: The frozen evidence defines Logic Checklist as 'Runs a fixed set of prompts over any argument — hidden premises, equivocal terms, invalid steps — so common reasoning faults are caught by routine rather than by luck', so its operative form is Assessment, Review & Assurance.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Philosophy
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Universal
Rationale: Validity, premise identification, and equivocation are classical concerns of philosophical logic.
Related originating lineages:
- Rhetoric — Practical argument appraisal and audience-facing fallacy checks materially shape the checklist form.
Review resolution: Both independent reviews assign primary provenance to philosophy. The queued secondary differences (encyclopedia_synthesis_disagreement) are reconciled by retaining rhetoric only as formative or independently established lineage(s), not merely as application domains. origin_mode=cross_disciplinary_synthesis records the provenance relationship, while domain_reach=universal separately records applicability breadth. confidence=high preserves the more cautious assessment, and encyclopedia_synthesis=true records whether either reviewer identified a corpus-specific synthesis.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] Equivocation is the informal fallacy in which a single term is used in two different senses within one argument, so a conclusion that looks to follow actually trades on the shift. A named prompt for it is why a checklist catches the slip a fluent reader glides over. ↩