Steelmanning
.jpg?width=1200)
A reader's summary of the practice of arguing against the strongest form of an opposing view: what it asks of you, where the habit comes from, and why it is more often admired than done online.
The idea at a glance
To steelman a position is to rebuild it so that it is as strong as it can honestly be made, including fixing weak wording and filling gaps its defenders left, and then to engage with that. The term is a play on “straw man”, the flimsy figure set up to be knocked down. In a comment thread the practice shows up as a sentence like “I think your best point is…”, which is rare enough to be noticed when it appears.
Origin
The word “steelman” is a product of the internet. It circulated on blogs and discussion forums in the early 2010s, and the exact first use is hard to pin down, so this page does not credit it to one person. The practice it names is far older. Its best-known modern statement is the set of rules attributed to Anatol Rapoport, which the philosopher Daniel Dennett set out in his 2013 book Intuition Pumps and Other Tools for Thinking.
History and context
Mill made the case in On Liberty in 1859: a view held without ever meeting its strongest objections is held as a prejudice, not as knowledge, and the opinions we reject deserve a hearing from the people who hold them rather than from their enemies. Medieval scholastic writers built the habit into their method, stating the objections to a thesis at length before replying.
Rapoport, a mathematician and game theorist known for the winning “tit-for-tat” entry in Robert Axelrod's prisoner's dilemma tournaments, applied the same cooperative reasoning to debate. In the twentieth century, analytic philosophers gave a related idea the name principle of charity, discussed by Willard Van Orman Quine and Donald Davidson in the context of interpreting what other people mean.
Main ideas
Answer the best version, not the easiest one
A steelman restates the other side's position more clearly and more persuasively than they did, then replies to that. The test is whether the person you disagree with would say, yes, that is what I meant, and better put.
The mirror image of the straw man
A straw man swaps the opponent's view for a weaker one that is easy to knock down. A steelman does the reverse. Both change the target before the argument starts, and only one of them is honest about it.
Rapoport's rules give it a procedure
The game theorist Anatol Rapoport is credited with a four-step order for criticism: restate the other view so well they say thanks, list what you agree with, list what you learned from them, and only then say where you disagree.
Mill's version: know the other side
In On Liberty (1859), John Stuart Mill wrote that anyone who knows only their own side of a case knows little even of that. Steelmanning is the working habit that follows from his claim.
Charity toward the speaker, effort from the listener
Steelmanning is close to the philosophers' principle of charity, which asks that you read a speaker so their statements come out as reasonable as possible. It is also a cousin of Hanlon's Razor: both start by assuming the other person is not a fool.
Critique
- Who decides what is strongest? The steelman is built by the critic, so it can drift away from what the opponent actually believes. An argument no one holds, polished until it is unbeatable, is a different kind of straw man: it flatters the person who built it.
- It can launder bad positions. Spending effort to improve a claim lends it respect it may not have earned, and the improved version can then circulate apart from the original, which was weaker. Steelmanning a view is not the same as endorsing it, but readers do not always see the difference.
- It assumes a good-faith audience. The rules work between two people trying to understand each other. In a thread where the other side is performing for an audience, restating their case generously may buy nothing and cost time.
- It is slow, and platforms reward speed. A fair restatement takes paragraphs. A quote-post with a caption takes a second and travels further, which is why the practice thrives in essays and long forums and thins out in feeds. It sits against Brandolini's Law, which already says that careful answers cost more than careless claims.
Impact
Steelmanning has become a stated norm in several online communities that prize argument, where it is held up as the polite way to disagree, and it is a common opening in essays that survey a contested question. Its value is as much for the person doing it as for the person it is aimed at: you either find the objection holds up better than you expected, or you can say with some right why it fails. As with echo chambers, the point is less to win than to leave the exchange knowing more than you did.