Skip to content

Argument Mapping

In prose, an argument’s structure is hidden: the main claim, the reasons for it, the unstated co-premises each reason needs, and the objections against it are blended into fluent text where a broken inference reads as smoothly as a sound one. Argument mapping makes the structure explicit: the contention, the reasons that support it, the co-premises each reason depends on, and the objections and rebuttals, laid out as a tree so every link is visible. The output is an argument map. Important boundary: a valid structure does not make the premises true; the map shows structure, not truth.

  • An argument or recommendation must be evaluated for soundness before it is trusted.
  • A fluent, persuasive case may be hiding a broken inference or an unstated assumption.
  • A debate needs its logical structure made explicit so people argue the same point.
  • Simple claims with no real argumentative structure to map.
  • To judge how persuasive something is (this analyzes logical structure, not rhetoric).
  • To generate ideas or options (wrong tool).
  • As proof an argument is sound: a tidy map can still rest on false premises.

When asked to map an argument, follow these steps:

  1. State the contention. The single main claim the argument is trying to establish.
  2. Lay out the reasons. For each, give the premise that directly supports the contention.
  3. Surface co-premises. For each reason, make explicit the unstated assumption it silently needs to actually support the contention. This is where most hidden weakness lives.
  4. Add objections and rebuttals. The strongest objections to the contention or to specific reasons, and any rebuttals.
  5. Flag the weak links. Mark the inferences that do not hold and the premises that are load-bearing but unsupported.
  6. Emit the argument map per references/TEMPLATE.md.

Use the template in references/TEMPLATE.md. The deliverable is the structured map with flagged weak links, not prose.

Before finalizing, verify:

  • The contention is a single, clearly stated claim.
  • Each reason has its co-premises made explicit, not left unstated.
  • Objections and rebuttals are included, not only supporting reasons.
  • The weakest links and unsupported load-bearing premises are flagged.
  • The output distinguishes valid structure from true premises (it does not claim soundness from structure alone).
  • The output is the argument-map artifact, not prose.

Tier S, with a scope caveat. Argument-mapping-based instruction produces among the largest measured critical-thinking gains in the field (van Gelder, effect sizes ~0.7-0.85). That evidence is for sustained practice (course-length), not for a single map fixing one argument, so the strong claim is “learning to map improves reasoning,” not “this one map carries that effect.” Evidence is transferred from human contexts, not AI-validated. Full grading: evidence/dossier.md.

See references/EXAMPLE.md for a completed argument map.

A full worked run (the shared Northwind scenario)

A completed run of think-argument-mapping, on the shared Northwind scenario. This is the quality bar a generated map should meet.

Northwind is a B2B SaaS. Here the skill maps the argument in the memo advocating the free tier, to test whether it actually holds together.


  • Northwind should launch a self-serve free tier to hit the Q3 growth target.
  • Reason 1: A free tier will sharply increase signups.
    • Co-premise (unstated): The people who sign up for free are the people we want (ICP-fit), not tire-kickers.
    • Objection: Competitor free tiers attract mostly non-ICP users. -> Rebuttal: We can gate it (but then it is not really “self-serve free”).
  • Reason 2: More signups will increase revenue.
    • Co-premise (unstated): Free-to-paid conversion at our ICP is high enough to outweigh free-tier cost.
    • Objection: Our current trial-to-paid conversion is falling, which predicts low free-to-paid too.
  • Reason 3: Competitors all have a free tier.
    • Co-premise (unstated): Their economics and ICP resemble ours (so imitation is valid).
  • Objection to the contention: The Q3 growth shortfall may be a funnel problem, not a packaging gap, in which case a free tier adds cost without fixing the cause. -> Rebuttal: None offered in the memo.
Link or premiseProblemWhat it would need to hold
Co-premise of Reason 2 (ICP free-to-paid conversion)Load-bearing and unsupported; the whole revenue case rests on itA pilot showing ICP free-to-paid covers free-tier cost
Reason 3 -> contentionImitation is not an argument; assumes comparable economicsEvidence competitors’ free tiers actually pay off, and that ours would too
Unaddressed objection (funnel vs packaging)The argument never rules out the cheaper alternative explanationData that packaging, not the funnel, drives the shortfall

The structure is not yet valid: the contention follows only if the Reason-2 co-premise (ICP conversion economics) holds and the funnel-vs-packaging objection is answered, and the memo supports neither. Reason 3 is imitation dressed as logic. For the argument to hold, run the conversion pilot and rule out the funnel explanation - i.e. the case is currently fluent but unsound.


Note: the value is exposing that the entire revenue case rests on one unstated, untested co-premise (ICP free-to-paid economics) and that the strongest objection was never addressed - both invisible in the smooth prose of the original memo.

What the research does and does not show, with graded sources

Single source of truth for the argument-mapping skill. The SKILL.md, sidecar, and evals derive from this. One of the library’s strong-evidence anchors.

Skillthinking-framework-skills.argument-mapping (installable name think-argument-mapping)
Familyreasoning-clarity
Evidence tierS (strong, with a scope caveat)
ConfidenceHigh that explicit structure improves reasoning quality; the measured effect is for sustained practice
Statusdraft (authored 2026-05-31 from the discovery corpus)

1. The mechanism (what actually does the work)

Section titled “1. The mechanism (what actually does the work)”

In prose, an argument’s structure is hidden: the main claim, the reasons for it, the unstated co-premises each reason needs, and the objections against it are all blended into fluent text, where a broken inference reads as smoothly as a sound one. Argument mapping makes the structure explicit: the contention at the top, the reasons (premises) that support it, the co-premises each reason silently depends on, and the objections and rebuttals against it, laid out as a tree so every inferential link is visible. The work is done by exposing the load-bearing-but-unstated premises and the links where support is weakest, which prose hides.

Boundary: a tidy map shows the argument’s structure, not the truth of its premises. Structure being valid does not make the premises true.

  • Informal logic and critical-thinking tradition; the Toulmin model is an ancestor. Tim van Gelder’s work (Reason!/Rationale) operationalized computer-supported argument mapping and produced the strongest effect estimates.

No trademark. Named descriptively.

3. What the evidence shows, and what it does NOT show

Section titled “3. What the evidence shows, and what it does NOT show”

Strongly supported (the S): argument-mapping-based instruction produces among the largest measured gains in critical-thinking skill in the field - van Gelder and colleagues report effect sizes around 0.7 to 0.85.

The honest scope caveat: those effect sizes are for sustained practice (typically a semester-length course building the skill), not for a single one-shot map magically improving one decision. So the strong evidence is that learning to map arguments improves reasoning; the claim that producing one map fixes one argument is weaker and practitioner-level. Grade S for the method, but do not imply a single use carries the course-length effect.

The evidence is from human learners and analysts, not AI-augmented use. Transferred, not AI-validated. The AI value: a model produces fluent prose arguments where bad inferences hide; forcing it to externalize the contention, co-premises, and objections as a structure is a direct counter, and the map is inspectable.

Works best when: an argument or recommendation must be evaluated for soundness; a fluent case may be hiding a broken inference or an unstated assumption; a debate needs its logical structure made explicit.

Fails or misleads when (poor-fit / anti-patterns):

  • Simple claims with no real argumentative structure to map.
  • Mapping rhetoric or persuasion as if it were logic (it analyzes logical structure, not how convincing something is).
  • Treating a tidy map as a sound argument (valid structure does not make premises true) - the central failure mode.
  • Generating ideas or options (wrong tool).
  • Claiming a single map carries the course-length learning effect.

An argument map: the contention; each supporting reason with its co-premises made explicit; the objections and rebuttals; and a flag on the weakest links and the unsupported or load-bearing-but-unstated premises that most need support.

  1. van Gelder, T. (2015) and the Reason!/Rationale argument-mapping studies - effect sizes ~0.7-0.85 for critical-thinking gains.
  2. Toulmin, S. (1958) - the model of argument structure (claim, grounds, warrant, rebuttal) that underpins mapping.

Verification status: the van Gelder effect-size range is well-attested for course-length instruction; keep the “single map != course effect” caveat visible in any public claim. Do not attach the 0.7-0.85 figure to a one-shot use.

Thinking Framework Skills v0.3.0 · 38 frameworks