Steelmanning

Working Definition

Steelmanning is the practice of reconstructing the strongest possible version of an opposing position before criticizing it: stripping away weak phrasings, supplying missing premises, and bringing in the best available evidence even if the original speaker failed to mention it. It goes further than the philosophical principle of charity, which asks what the speaker most plausibly meant, by asking instead which version of their thesis is hardest to refute; and it differs from the Ideological Turing Test by not requiring that actual adherents recognize the reconstructed argument as their own. In popular discourse the word often denotes a gesture of rhetorical fair play—a polite summary before launching into a rebuttal. In its strict methodological sense, however, it is a stress-test for one's own beliefs, measured by whether one's critique still holds up against the reinforced version.

Genesis and Global Lineage

The term was coined by inverting the straw man fallacy: the phrase “man of straw” as an easily defeated imaginary opponent dates back to the 1620s, and the straw man entered logic textbooks as an informal fallacy with Stuart Chase's Guides to Straight Thinking (1956). The underlying epistemic norm is much older: in Chapter II of On Liberty (1859), John Stuart Mill argued that anyone who knows only their own side of a case knows little of that, and must hear opposing reasons from the people who actually believe them and defend them in their most plausible form. A second lineage runs through conflict psychology: in Fights, Games, and Debates (1960), Anatol Rapoport adapted Carl Rogers's client-centered therapy into a “Rogerian strategy” of debate, requiring disputants first to state the opponent's position so fairly that the opponent agrees, next to list the contexts in which that position is valid, and finally to invite an assumption of similarity. Daniel C. Dennett popularized this protocol in Intuition Pumps and Other Tools for Thinking (2013) under the name “Rapoport's Rules.” Meanwhile, the word steelmanning crystallized in the rationalist blogosphere: Scott Alexander urged readers in 2009 to consider the “least convenient possible world,” economist Bryan Caplan proposed the Ideological Turing Test in June 2011, and Chana Messinger wrote the classic blog post “Knocking Down a Steel Man: How to Argue Better” in December 2012 (though the term already appeared on LessWrong in 2011). The opposite tradition—winning disputes at any cost—had been codified by Arthur Schopenhauer in his Eristische Dialektik (1864). By the mid-2010s, steelmanning had migrated from rationalist forums into mainstream journalism and management advice.

Central European Lineage

We found no documented link between this concept and the intellectual history or historical experience of Central Europe.

Why Now

Three developments between 2017 and 2026 turned a niche blogosphere term into a contested fixture of public life. First, intense polarization on social media made steelmanning a watchword for intellectual hygiene: Conor Friedersdorf praised it in The Atlantic (June 2017) as the highest form of disagreement, and Julia Galef's The Scout Mindset (2021) popularized it as a hallmark of the “scout” who wants an accurate map of reality rather than the “soldier” who wants to win. Second, artificial intelligence safety research institutionalized the concept as a technical protocol: in AI Safety via Debate (Irving, Christiano, and Amodei, May 2018), two AI models debate a question before a human judge on the premise that truth prevails when each side must surface the strongest counterargument, and after 2023 the prompt “steelman the opposing view” became routine practice for querying large language models. Third, 2024 brought a sharp backlash: an influential March 2024 essay on LessWrong called steelmanning “the nadir of charitable discourse,” philosopher Ben Burgis (July 2024) questioned the demand that disputants constantly rewrite their opponents' arguments, and economist Noah Smith (October 2024) linked public steelmanning to “sanewashing”—the media habit of translating incoherent political rhetoric into respectable policy proposals.

Explanatory Power

Steelmanning separates defeating a position from merely defeating its clumsiest spokesperson. In their taxonomy of the straw-man fallacy, Scott Aikin and John Casey (2022) distinguish the classic straw man (distorting an opponent's argument), the weak man (selecting the weakest real argument made by an opponent), and the hollow man (attributing a fabricated argument to a vague group). Steelmanning is the exact inverse of this triad and explains why so many public controversies end in feelings of triumph without changing anyone's mind. Consider three domains. In an engineering architecture review, rejecting a rival design only after strengthening its assumptions produces a decision that will survive when the problem resurfaces a year later; rejecting a caricature only guarantees another meeting. In climate policy debates, a degrowth advocate who can reconstruct the strongest case for green technological innovation immediately sees which of their own premises carry real empirical weight and which are merely tribal slogans. And when working with a large language model, asking it to steelman both sides of a strategic choice reveals whether an initial recommendation reflected deep evidence or merely the framing of the user's prompt.

Axes of Debate

Underappreciated Angle

Discussions of steelmanning rarely notice that Anatol Rapoport designed his Rogerian rules for conflicts in which both parties share an interest in mutual understanding—and that in 1968, at the height of the Vietnam War, Rapoport himself judged the Rogerian approach largely inapplicable to debates over US intervention, noting that every method, like every proposition, has a bounded region of validity. Steelmanning inherits that boundary condition. In an information environment where one actor is conducting strategic disinformation or flooding the zone with bad-faith slogans, publicly steelmanning their claims performs free intellectual labor on their behalf: the conscientious critic constructs coherent doctrines that the speaker never intended to stand behind, while the cost of refutation balloons in accordance with Brandolini's law. The practical lesson is rarely stated plainly: steelmanning should be practiced against oneself—to test the robustness of one's own conclusions against the best possible counter-case—rather than performed as public PR on behalf of actors who are not playing a game of truth.