An argument not tested against its strongest opponent is a preference with vocabulary
The argument you have not hardened against its strongest opponent is not an argument. It is a preference with vocabulary.
The default mode in an argument is to find the weakest version of the opposing position and dismantle it. This is easy, gratifying, and counter-productive, because defeating a weak version of the opposition does not strengthen your position at all. It only prevents you from discovering the places where the real opposition has actual force. You remain unaware of your genuine vulnerabilities, which means you remain unable to address them.
Why attacking weak versions of opposing arguments makes your own position weaker
There is a mechanical relationship between the quality of the opposition you engage with and the quality of the position you develop. A position that has been tested against a genuinely strong version of the counter-argument has been forced to address its actual weaknesses, sharpen its claims, and locate the conditions under which it does and does not hold. A position that has been tested against a weak version has undergone none of this pressure and contains the vulnerabilities that the weak version did not surface.
This is not a moral argument for fairness. It is a practical argument about how strong thinking is produced. The position you defend in the real world, in a boardroom, a negotiation, or a competitive market, will be tested against people who are not performing the weakest version of the counter-argument. It will be tested against people who have their own interests and their own analyses, and who will find the genuine vulnerabilities if they exist. Whether your position holds under that pressure is determined by whether you found those vulnerabilities first, which requires engaging with the strong version, not the convenient one.
The discipline of steelmanning is therefore not about being nice to your opponents. It is about producing positions that are actually defensible rather than positions that feel defensible inside a vacuum of unchallenged reasoning.
What presenting the opponent's position back to them actually requires
Presenting an opposing argument back to its holder in terms they recognise and endorse is harder than it sounds. Not because the words are difficult, but because it requires temporarily suppressing your own framing and inhabiting a different one. Your brain wants to describe the opposing position in terms that make your own look better, to characterise the motivation as lower than it is, the evidence as weaker than it is, and the implication as more extreme than the opponent intended. Each of these is an act of cognitive self-protection that corrupts the process.
The test is specific: can you state the opposing position clearly and completely enough that the person who holds it would say "yes, that is my argument"? Not "yes, roughly" or "yes, but you have missed the main point". Actually yes. This is a high bar, and reaching it is genuinely difficult, which is exactly why it is valuable. The difficulty is proportional to the cognitive work being done, and the cognitive work is the compression of two positions against each other until the genuine point of difference is located.
Most disagreements persist not because the positions are fundamentally incompatible, but because neither party has yet correctly understood the other's strongest version. Once the steelmanning is done correctly, the actual disagreement often turns out to be smaller, more specific, and more tractable than the surface argument suggested. This is not always a welcome outcome: disagreements that feel large and principled often become small and empirical under the pressure of honest steelmanning, and some people would prefer the larger version.
The internal version: writing as steelmanning with yourself
The same operation applied internally is the well-constructed essay. When you think through a hard problem by writing it out, the process that produces genuine clarity is not running one voice until it reaches a conclusion. It is running two voices, a proponent and a critic, and allowing each to apply full force to the other's position. The proponent builds the case. The critic finds the objections. The proponent responds to the best objections, not the easiest ones. The critic pushes harder.
What results, when the process is conducted honestly, is not simply a document with a conclusion. It is a record of the positions that could not be sustained, the qualifications that had to be added, the conditions under which the conclusion holds and the ones under which it does not. The essay is strong because it has already absorbed the strongest opposition and either answered it or acknowledged it as a genuine tension that the conclusion requires you to accept.
The failure mode is easy to recognise: the essay or analysis that reaches its conclusion in the opening paragraph and then spends the rest of its length supporting that conclusion, filtering evidence and characterising opposition in ways that always support the pre-decided conclusion. This is not thinking. It is confirmation with the structure of argument. The steelmanning test it fails is simple: where in this document is the strongest opposing case presented in the terms its proponents would endorse?
What changes once steelmanning becomes a habit
The quality of your own positions improves, because they have been stress-tested against something real. You stop holding views that feel obviously right but collapse under examination, not because your views become more timid, but because they become more accurately held, with the conditions and qualifications that the genuine counter-arguments forced you to add.
The quality of your disagreements changes too. Arguments that would previously have continued indefinitely, each side restating their position more loudly, often resolve once each party has correctly understood the strongest version of what the other is actually saying. The residual disagreement is smaller, sharper, and actually useful: it tells you precisely where the genuine tension is, which means it tells you precisely where the investigation should focus.
And the relationship with your own certainty changes. Once you have developed the habit of testing positions against their strongest opposition, you become more calibrated about which claims genuinely survive the test and which ones depend on the comfort of not having been seriously challenged. The positions that have survived are ones you can hold with real confidence, because you know where the objections land and why they do not, in the end, hold. The positions that have not been tested you hold provisionally, which is the honest epistemic status for things you have not yet examined under pressure.