--- title: "Steelmanning" description: "Steelmanning means restating an opposing argument in its strongest and most charitable form before responding to it — the deliberate opposite of a straw man. Its value is not only fairness: defeating the strongest version secures your conclusion, while defeating a weak version leaves the real arg..." last_updated: "2026-08-02" --- # Steelmanning Steelmanning means restating an opposing argument in its strongest and most charitable form before responding to it — the deliberate opposite of a straw man. Its value is not only fairness: defeating the strongest version secures your conclusion, while defeating a weak version leaves the real argument standing untouched. Engaging with the strongest, most charitable version of an opposing argument rather than a weakened caricature of it. The opposite of straw-manning. *Example: Before rebutting a proposal, you restate it in its strongest form — one its supporters would happily sign off on.* ## What it looks like **In a policy discussion** > Before I disagree — the strongest case for this is that it protects people with no other options, and the cost falls mainly on those who can absorb it. Is that fair? *Why it fails:* This is steelmanning done correctly, and the check at the end matters: it invites correction, so you are engaging their argument rather than the one you find easiest to answer. **In a technical review** > I think the case for this architecture is that it isolates failures at the cost of latency, which is the right trade if outages are expensive. Assuming that, here's my concern. *Why it fails:* Naming the trade-off the other side is actually making moves the discussion to the real disagreement — whether outages are that expensive — rather than to whichever weakness is easiest to name. **In a disagreement about evidence** > The best version of your position isn't the study you cited, it's the mechanism behind it. Let me address that instead. *Why it fails:* Improving the opponent's argument before answering it is the strongest form: if it fails even at its best, the conclusion is secure in a way that refuting a weak citation would not be. ## Practice ### Question 1 What is the strategic reason to steelman rather than just the ethical one? A. It makes you appear more reasonable to onlookers B. Defeating the strongest version of an argument secures your conclusion, whereas defeating a weak version leaves the strong version unaddressed and available C. It shortens arguments D. It prevents the other person from getting angry **Answer: B.** The epistemic payoff is what makes steelmanning worth the effort. If you refute the weakest form, your opponent can simply produce the stronger one and you have gained nothing. Appearing reasonable is a real side benefit, but it is not the reason the technique works. ### Question 2 Which is a correct steelman of 'we should delay the launch'? A. 'You want to delay forever and never ship anything.' B. 'The strongest case is that a bad first impression costs more than the delay, since re-acquiring churned users is expensive. Is that the argument?' C. 'You're just being cautious because you're risk-averse.' D. 'Delaying is always wrong, so this can't be right.' **Answer: B.** Option B improves the argument by supplying the reasoning that makes it strongest, then checks whether that is what was meant. Option A is a straw man, option C attacks disposition rather than argument, and option D simply asserts a conclusion. Note the check at the end — a steelman the other person rejects is not a steelman. ### Question 3 Does steelmanning require you to end up agreeing? A. Yes — that is the point of engaging charitably B. No. It requires engaging the best version, which is compatible with concluding that even at its best the argument fails C. Yes, at least partially, in every case D. No, but you must concede at least one point **Answer: B.** Steelmanning is about what you engage with, not what you conclude. Rejecting an argument after stating it at full strength is a much more credible rejection than one aimed at a caricature. Requiring partial agreement would make it a rhetorical courtesy rather than a reasoning tool. ### Question 4 What is the most common way steelmanning goes wrong? A. Making the argument too strong to answer B. Constructing a version the other person does not actually hold, so you address a position nobody is defending C. Taking too long D. Being seen as insincere **Answer: B.** A steelman the other person disowns is just a different argument — sometimes called a 'weakman in reverse.' This is why the confirming question matters: stating your version and asking whether it captures their view keeps the exercise anchored to the actual disagreement rather than to your model of it. ### Question 5 How does steelmanning relate to the principle of charity? A. They are unrelated B. Steelmanning is a practical application of it — the principle says interpret others' arguments in their strongest plausible form, and steelmanning is doing that explicitly and out loud C. The principle of charity applies only to written texts D. Steelmanning is the opposite of charitable interpretation **Answer: B.** The principle of charity is the interpretive norm — read an ambiguous argument in the way that makes it most reasonable. Steelmanning operationalizes it by stating that strongest reading aloud before responding, which also makes the interpretation checkable by the person whose argument it is. ## Frequently asked ### What is the difference between steelmanning and straw-manning? They are exact opposites. A straw man substitutes the weakest or a distorted version of an argument and refutes that; a steelman constructs the strongest version and engages it. One makes refutation easy and worthless, the other makes it hard and meaningful. ### Isn't steelmanning just helping your opponent? It helps their argument and it helps you. If the strongest version fails, your conclusion is secure against anything they can produce later. If it survives, you have learned something you would have missed by attacking a weaker form — which is the outcome worth wanting. ### How do I steelman an argument I find genuinely bad? Ask what someone intelligent and well-intentioned would have to believe for this to make sense, and what concern it might be tracking even if poorly expressed. Then state that version and check it. If no such version exists, saying so after the attempt is far more credible than saying so before. ## Related - [Straw Man](https://www.criticalthinkingexercise.org/pages/glossary/straw-man.md) - [Epistemic Humility](https://www.criticalthinkingexercise.org/pages/glossary/epistemic-humility.md) - [Argument](https://www.criticalthinkingexercise.org/pages/glossary/argument.md) - [Identifying Hidden Assumptions](https://www.criticalthinkingexercise.org/pages/exercises/hidden-assumptions.md) - [Advanced Argument Evaluation](https://www.criticalthinkingexercise.org/pages/exercises/advanced-arguments.md) - [Lesson: mastery/dialectical-thinking](https://www.criticalthinkingexercise.org/pages/learn/mastery/dialectical-thinking.md) ## Sitemap See the full [sitemap](https://www.criticalthinkingexercise.org/sitemap.md) for all pages.