Steelman

You're at a team meeting and a colleague proposes reorganizing the department. Before she's finished her second sentence, you've already spotted three flaws. By the time she stops talking, you dismantle the proposal efficiently and convincingly. The room moves on. But here's the question nobody asks: did you engage with what she actually meant, or with the weakest version of her idea — the version that was easiest to defeat? If you're honest, you probably argued against a straw man. And in doing so, you may have killed a genuinely good idea because you never gave it a fair hearing.

This is the territory of steelmanning and strawmanning — two opposing disciplines that shape not just the quality of your arguments, but the quality of your thinking.

The Core Distinction

Strawmanning is engaging with a weakened, distorted, or oversimplified version of someone's argument. You swap out what they actually said for something easier to attack, then defeat the substitute. The term comes from the practice of training against straw dummies — opponents that look real from a distance but crumble on contact.

Steelmanning is the opposite: deliberately reconstructing someone's argument in its strongest, most charitable, most logically sound form — and then engaging with that. Not the version they stumbled through in the meeting, not the version that's easiest to ridicule, but the version they would present if they had unlimited time, perfect clarity, and their best day of thinking.

The distinction matters because it determines whether you're pursuing truth or pursuing victory. Strawmanning wins arguments. Steelmanning wins understanding. Only one of those leads to better decisions.

This is NOT the same as playing devil's advocate — a performance of skepticism designed to test ideas in conversation. Devil's advocacy is rhetorical: you argue a position you may not believe to probe someone else's reasoning, and everyone in the room knows it. Steelmanning is epistemic: you construct the best possible case for a view because you genuinely want to understand whether it's right. One is a technique for stress-testing. The other is a commitment to truth over victory.

Why We Default to Straw

If steelmanning is so obviously superior, why does almost everyone default to strawmanning? The answer lies in several converging psychological forces.

The most powerful is identity protection. When someone presents an argument that challenges our beliefs, our brains process it less like an intellectual proposition and more like a threat. Naomi Eisenberger's neuroimaging research at UCLA shows that social rejection and ideological challenge activate the dorsal anterior cingulate cortex — the same region that processes physical pain. The brain is not distinguishing between 'this idea threatens me' and 'this situation threatens my body.' The threat response is the same, which is why ideological challenge feels like attack. Strawmanning is a defensive reflex: by weakening the threat, we protect our existing worldview from genuine challenge. This is closely related to confirmation bias, our tendency to seek and favor information that confirms what we already believe. Strawmanning is confirmation bias weaponized — instead of ignoring inconvenient evidence, you actively distort it.

The second force is cognitive load. Steelmanning is genuinely hard work. It requires you to temporarily inhabit a perspective you disagree with, to think generously about someone else's reasoning, and to construct the best version of an argument that might undermine your own position. Your brain resists this because it's metabolically expensive — the prefrontal cortex, responsible for this kind of deliberate analytical reasoning, consumes significant energy. Strawmanning, by contrast, is fast, cheap, and satisfying. It gives you the dopamine hit of a "win" without the cognitive cost of genuine engagement.

The third force is social incentive. In most environments — political debates, social media, workplace meetings — audiences reward confident takedowns more than nuanced engagement. Strawmanning plays well to crowds that already agree with you. It signals tribal loyalty. Steelmanning, on the other hand, can look like weakness or disloyalty to your "side," because it involves publicly acknowledging that opponents have legitimate points. This is the same dynamic that drives groupthink — the pressure to conform to group consensus makes it socially costly to take opposing arguments seriously.

These three forces compound rather than merely coexist: identity protection makes the opposing argument feel threatening before cognitive load has a chance to engage, and social incentive then reinforces the strawman precisely when it occurs, making the pattern harder to interrupt with each repetition. Together they explain why strawmanning isn't a character flaw — it's a default setting. Which means overriding it requires deliberate effort and specific technique.

What Steelmanning Actually Looks Like

Steelmanning is more than being polite or saying "I see your point" before disagreeing. It's a specific intellectual discipline, and it's worth being honest about what it feels like from the inside — because nobody tells you this part.

First, you separate the person from the position. Arguments arrive tangled up with tone, delivery, personality, and your feelings about the speaker. Steelmanning requires you to extract the core logical claim from all that noise — to hear what someone means, not just what they said. This is where active listening earns its keep: you can't steelman an argument you didn't actually hear.

Second, you supply the missing pieces. People rarely present their arguments in perfect form. They leave premises unstated, skip logical steps, and choose their words imperfectly. Steelmanning means filling in those gaps charitably — asking "what would this argument look like if the strongest possible reasoning supported it?" This is the opposite of what most people do instinctively, which is to exploit the gaps as weaknesses.

Third — and this is the hard part — you argue against the strong version. Once you've built the best possible case for the opposing view, you engage with that. If you can defeat the steelmanned version, your counter-argument is genuinely robust. If you can't, you've learned something important.

Here's what nobody warns you about: the third step is uncomfortable in a way that's hard to describe until you've experienced it. When you steelman an argument you genuinely disagree with — when you do it properly, not as a performance — you will feel your certainty shift beneath you. You'll be constructing the best case against your own position, and at some point a voice in the back of your mind will say, wait — that's actually a decent point. Your first instinct will be to undercut it, to add a qualifier that weakens it, to subtly rig the steelman so it's not quite as strong as it could be. That impulse is the whole game. Resisting it is what separates genuine steelmanning from its theatrical imitation.

Charlie Munger — Warren Buffett's long-time business partner and one of the most respected investors of the twentieth century — turned this discomfort into an explicit rule. He maintained that he was not allowed to hold an opinion on a topic unless he could state the opposing case better than its advocates. Not as well as — better than. This meant he routinely spent more time building the case against his own positions than building the case for them. The result wasn't indecisiveness. It was that when Munger committed to a position, it had survived the harshest possible scrutiny — his own. His track record suggests the practice worked.

When Steelmanning Changed History — And When Its Absence Was Catastrophic

The consequences of steelmanning and strawmanning are easiest to see when the stakes are highest.

In the early 1860s, Abraham Lincoln assembled a cabinet composed largely of men who had opposed him — rivals who thought they were better suited for the presidency. His Secretary of State, William Seward, had called Lincoln unfit for office. His Secretary of the Treasury, Salmon Chase, actively schemed against him. Lincoln's approach wasn't masochism. It was institutional steelmanning. He deliberately surrounded himself with the strongest possible critics of his own thinking, because he understood that governing during a civil war required decisions that had survived adversarial challenge. The resulting administration wasn't harmonious, but the quality of its reasoning under extreme pressure was remarkable. Lincoln didn't just tolerate disagreement — he engineered it into his decision-making process.

Now consider the opposite. In the run-up to the 2003 invasion of Iraq, the U.S. intelligence community faced significant internal dissent about the evidence for weapons of mass destruction. Analysts at the State Department's Bureau of Intelligence and Research and at the Department of Energy raised substantive objections to key claims. These weren't fringe dissidents — they were credentialed experts flagging specific problems with specific evidence. But the prevailing political momentum strawmanned their objections. Dissent was treated as insufficient confidence rather than engaged as a substantive counter-argument. The strongest version of the skeptics' case — that the evidence was ambiguous at best and that the intelligence was being shaped to fit a predetermined conclusion — was never genuinely engaged with at the decision-making level. The consequences of that failure to steelman are measured in hundreds of thousands of lives and trillions of dollars. This wasn't a failure of intelligence gathering. It was a failure of critical thinking — specifically, a failure to subject the dominant narrative to the strongest possible challenge before acting on it.

These aren't edge cases. They illustrate a pattern: the more consequential a decision, the more dangerous it is to engage only with weak versions of opposing arguments.

Steelmanning as an Epistemological Tool

The deepest value of steelmanning isn't rhetorical — it's epistemological. It's a method for getting closer to truth.

When you only engage with weak versions of opposing arguments, you create an illusion of certainty. You "win" every debate, your confidence in your existing beliefs grows, and you gradually lose the ability to distinguish between being right and being unchallenged. This is how smart people end up holding increasingly extreme or poorly examined positions — not because they lack intelligence, but because they've never subjected their beliefs to a worthy opponent.

Consider what happened when Ignaz Semmelweis proposed, in 1847, that physicians should wash their hands before delivering babies. His evidence was clear: mortality rates in the maternity ward staffed by doctors were three to five times higher than in the ward staffed by midwives, and doctors moved directly from performing autopsies to delivering babies. The medical establishment of the 1840s did not engage with this evidence. They engaged with the implication — that doctors were killing their patients — and found it intolerable. They strawmanned the argument by attacking the man and the implication rather than the data. A steelmanned engagement would have meant examining his actual claim: that a specific hygienic intervention (handwashing with chlorinated lime solution) reduced a specific outcome (puerperal fever mortality). That case was not defeated, because it was not engaged. Semmelweis died in a mental institution in 1865. Germ theory vindicated him a decade later. The consequence of two decades of strawmanning was measured in preventable deaths across European hospitals. The opposition did not lack intelligence. It lacked the discipline to engage with the strongest version of an argument that threatened something they held dear.

The philosopher John Stuart Mill captured the principle: "He who knows only his own side of the case knows little of that." You cannot fully understand why you believe something until you've genuinely grappled with the best reasons to believe otherwise. Steelmanning is the mechanism that makes this grappling possible.

Steelmanning Yourself: The Preemptive Version

Most discussions of steelmanning frame it as something you do to other people's arguments. But its most powerful application is one you can practice alone, on paper, before any conversation happens: steelmanning the opposition to your own ideas.

Before you present a proposal, pitch a project, or commit to a strategy, ask: what is the strongest possible case against this? Not the obvious objections — the ones you've already prepared answers for. The objections that would come from the smartest, most informed, most good-faith critic in the room. Write them down. Try to make them genuinely persuasive.

This practice does three things simultaneously. It reveals blind spots you didn't know you had. It prepares you for challenges you'll actually face, rather than the easy ones you were hoping for. And it often improves the original idea, because the process of stress-testing forces you to strengthen weak points before anyone else finds them.

Lincoln did this institutionally by filling his cabinet with rivals. Munger did it individually by refusing to hold opinions he couldn't argue against. You don't need a cabinet or a decades-long investing career. You need fifteen minutes and the discipline to argue against your own best case before you bring it to anyone else.

Where Steelmanning Breaks Down

Steelmanning has real limitations, and pretending otherwise would be — ironically — a failure of intellectual honesty.

Not every argument deserves steelmanning. Some positions are made in bad faith, and constructing the best version of a deliberately dishonest argument lends it unearned credibility. Steelmanning assumes good faith on both sides. When that assumption is clearly violated, the tool doesn't apply.

The credibility laundering problem is real. This is steelmanning's sharpest edge and it deserves serious attention. When a sophisticated thinker constructs the "best version" of a harmful argument, they can inadvertently make it more persuasive and more dangerous than the original advocate ever could. A fringe conspiracy theory, articulated clumsily by its proponents, might convince nobody. But steelmanned by a respected voice — cleaned up, given logical structure, stripped of its obvious absurdities — it can reach an audience the original never would. Media environments face this constantly: the instinct to "present both sides fairly" can function as a laundering mechanism for positions that don't deserve the elevation. Knowing when steelmanning shades into platforming is a judgment call with no clean formula, and getting it wrong has consequences. The tobacco industry's decades-long campaign against smoking-cancer research illustrates this at scale. Industry-funded scientists in the 1950s through 1980s did not produce obviously false science — they produced sophisticated, peer-reviewed-looking research that gave major media outlets a "both sides" peg. The steelmanned version of the doubt argument, cleaned up and given academic structure, reached audiences the crude denial never would have. This is credibility laundering as a deliberate strategy, and it worked for decades.

Steelmanning can become a form of avoidance. Some people use the practice to avoid ever taking a firm position. They become so focused on seeing every side that they never commit to a view. This is epistemic humility taken past its useful range — at some point, you've heard the arguments, weighed the evidence, and need to decide. Perpetual steelmanning without eventual judgment is intellectual paralysis, not wisdom.

It doesn't work well under time pressure. In fast-moving conversations, negotiations, or emergencies, the cognitive overhead of steelmanning is a liability. The practice is most valuable for important decisions where getting the right answer matters more than getting a quick one — the kind of reversible vs. irreversible decisions framework that helps you know when to deliberate and when to act.

Power dynamics matter. Asking someone to steelman the arguments of a person or system that is actively harming them can be dismissive and tone-deaf. "Have you considered the strongest version of your employer's argument for not paying you fairly?" is a question that prioritizes intellectual elegance over the legitimate experience of the person in front of you.

Building the Habit

The most practical way to develop steelmanning is to practice where the stakes are lowest first. When you read an opinion piece you disagree with, pause before forming your response and ask: what would this argument look like if the smartest person you know were making it? Write down the strongest version. You'll often find that the exercise changes your response — sometimes subtly, sometimes dramatically.

In conversations, a single phrase does most of the work: articulate your best reconstruction of the other person's point — "so what you're saying is..." — and then ask whether you've captured it accurately. This practice of checking your understanding before responding eliminates most strawmanning, because strawmanning requires you to not verify that you've understood correctly. The simple act of seeking confirmation forces you to construct a version of the argument good enough that the speaker could recognize it as their own.

For your own ideas, build a pre-mortem into your process before committing to a decision. Spend ten minutes writing the strongest possible argument for why you're wrong. If you can't construct a compelling counter-case, either the decision is genuinely robust or you haven't thought hard enough — and both are useful to know. The pre-mortem connects the individual practice of steelmanning to institutional design: Lincoln's cabinet of rivals was a pre-mortem engineered into a decision-making structure, ensuring that challenges arrived before the commitment rather than after it.

Over time, steelmanning becomes less effortful and more automatic. You start hearing the strong version of arguments natively, rather than having to reconstruct them deliberately. This doesn't make you a pushover — it makes your disagreements sharper, your positions more defensible, and your thinking more resilient.

Back to That Meeting

Remember your colleague and her reorganization proposal? Imagine replaying it, but this time you steelman before you respond. Instead of attacking the three obvious weaknesses, you reconstruct the best version of her idea: "If I understand correctly, you're arguing that our current structure creates bottlenecks between teams X and Y, and that reorganizing around projects rather than functions would reduce handoff delays and improve ownership. Is that right?" She confirms — and clarifies a point you'd missed entirely. The conversation that follows is harder, slower, and less satisfying to your ego. But it addresses the real problem. Maybe her idea still doesn't work. But now you know why it doesn't work, not just that you could make it sound bad. And that's a fundamentally different kind of knowing.

The Steel-or-Straw Test is a single question you run after any disagreement: did I argue against what they actually meant, or what was easiest to defeat? The answer tells you whether you're building with steel or straw.