AGSK - Argument Analysis

Input: $ARGUMENTS

Step 1: Identify the Claim

CLAIM: [the central assertion being made]
CLAIM TYPE: [empirical | normative | definitional | causal | predictive]
SCOPE: [universal | general | specific | particular]

If the input contains multiple claims, identify the primary claim and list secondary claims separately. Analyze the primary claim through the remaining steps.

Step 2: Identify the Premises

PREMISES:
  P1: [first supporting reason]
      STATUS: [stated explicitly | implied | reconstructed]
  P2: [second supporting reason]
      STATUS: [stated | implied | reconstructed]
  P3: [third supporting reason, if any]
      STATUS: [stated | implied | reconstructed]

Reconstruct the argument charitably. If a premise is implied but needed for the argument to work, include it and mark it as reconstructed.

Step 3: Check Logical Validity

ARGUMENT STRUCTURE:
  IF [P1] AND [P2] THEN [CLAIM]?
  VALID: [yes | no]
  FORM: [deductive | inductive | abductive | analogical]

If deductive: Do the premises guarantee the conclusion? If any premise is true and the conclusion could still be false, the argument is invalid.

If inductive: Do the premises make the conclusion probable? How probable?

If abductive: Is this the best explanation? What alternatives exist?

If analogical: How strong is the analogy? Where does it break?

Step 4: Check Premise Truth

For each premise:

P[N] ASSESSMENT:
  TRUTH: [true | false | uncertain | partially true]
  EVIDENCE FOR: [what supports this premise]
  EVIDENCE AGAINST: [what undermines this premise]
  VERDICT: [strong | adequate | weak | unsupported]

An argument can be logically valid but have false premises. Both checks are needed.

Step 5: Identify Hidden Assumptions

HIDDEN ASSUMPTIONS:
  A1: [assumption the argument requires but doesn't state]
      PLAUSIBILITY: [high | medium | low]
      IF FALSE: [what happens to the argument]
  A2: [another hidden assumption]
      PLAUSIBILITY: [high | medium | low]
      IF FALSE: [consequence for the argument]

Common hiding places: definitions taken for granted, causal mechanisms assumed, scope limitations unstated, value judgments embedded as facts.

Step 6: Assess Strength of Evidence

EVIDENCE ASSESSMENT:
  TOTAL EVIDENCE CITED: [count]
  TYPES: [anecdotal | statistical | expert | experimental | logical]
  STRONGEST PIECE: [which evidence and why]
  WEAKEST PIECE: [which evidence and why]
  MISSING EVIDENCE: [what evidence would strengthen or weaken the argument]
  OVERALL EVIDENCE QUALITY: [strong | adequate | weak | absent]

Step 7: Find the Strongest Objection

STRONGEST OBJECTION:
  OBJECTION: [the single best counterargument]
  TARGETS: [which premise, assumption, or logical step it attacks]
  FORCE: [devastating | significant | moderate | minor]
  POSSIBLE RESPONSE: [how the arguer might reply]
  RESPONSE QUALITY: [strong | adequate | weak]

The strongest objection is not necessarily the most obvious one. It attacks the weakest load-bearing element of the argument.

Step 8: Overall Assessment

ARGUMENT QUALITY
================
CLAIM: [restate]
VALIDITY: [valid | invalid | partially valid]
SOUNDNESS: [sound | unsound | uncertain]
EVIDENCE: [strong | adequate | weak]
STRONGEST POINT: [what the argument gets most right]
WEAKEST POINT: [where it's most vulnerable]
OVERALL GRADE: [compelling | reasonable | weak | fallacious]
VERDICT: [should the claim be accepted, rejected, or held pending more evidence?]

Failure Modes

Integration

• Use with: /rlsk to analyze arguments occurring within relationships
• Use with: /cmpr to check if the argument is complete
• Use with: /difr to differentiate this argument from similar ones
• Use from: /claim when testing a specific claim’s argument structure
• Differs from /ht: agsk analyzes argument structure; ht tests hypotheses empirically