Build the strongest possible version of a position you disagree with, then find the real weaknesses. Use when someone says 'I disagree with this but I want to be fair', 'help me argue against this properly', 'steel man this for me', 'what am I missing about the other side', 'I want to understand this position before I criticize it', 'am I attacking a straw man', or 'help me find the real flaws in this argument.' Applies the Principle of Charity, then identifies genuine weaknesses using Kline's 7 objection types.
Help the user build the strongest possible version of a position they disagree with, then identify its genuine weaknesses — not straw men. From Chapter 3 of How to Think by Zakery Kline.
"The principle of charity holds that you should consider the strongest possible version of an objection, not just easily dismissed weak versions."
Most people argue against weak versions of positions they oppose. They pick the dumbest advocate, the worst argument, the most extreme example — and then congratulate themselves for demolishing it. This is intellectually lazy and practically useless. If your counter-argument only works against the weakest version, it doesn't work at all.
The steel man approach inverts this: build the position up to its strongest form FIRST, then look for genuine cracks. If you can defeat the strongest version, you've actually accomplished something.
The user disagrees with a position — political, philosophical, business strategy, personal — and wants to argue against it properly. Or they suspect they might be attacking a straw man and want to check. Or they want to understand why intelligent people hold a position that seems obviously wrong to them.
Ask: "What position do you disagree with? State it as clearly as you can."
Then ask: "What's your current understanding of WHY people hold this position? What do you think their best arguments are?"
Listen carefully. The gap between their stated understanding and the actual strongest version of the position is where the work happens.
Now reconstruct the position in its strongest possible form. This is not about agreeing with it — it's about representing it as a thoughtful, intelligent advocate would.
"Rather than dismissing the objection superficially, the author engages with its core concerns and examines what would follow if the objection were valid."
Walk through four questions:
1. Best Arguments FOR the Position What are the most compelling logical arguments? Not the talking points — the actual reasoning that a well-informed advocate would use. Strip away rhetoric and find the structural argument.
2. Supporting Evidence What empirical evidence, data, historical examples, or lived experience supports this position? What would a researcher cite? What patterns does this position correctly identify?
3. Motivating Values and Concerns What legitimate values drive people to this position? What real problem are they trying to solve? What fear, aspiration, or moral commitment underlies the position? These are often more important than the arguments themselves.
4. The Thoughtful Advocate's Version If the smartest, most well-read, most honest person who holds this position were making the case — not a pundit, not an ideologue, but someone who has genuinely thought it through — what would they say? This is the steel man.
Deliver the reconstructed position back to the user:
STEEL-MANNED POSITION
Your original understanding:
[What the user said they thought the position was]
The strongest version:
[The steel man — the position as its best advocate would state it]
What you were missing:
[Key arguments, evidence, or values the user hadn't considered]
Legitimacy check:
[What's genuinely right or insightful about this position, even if you ultimately disagree]
Ask: "Does this feel like a fair representation? Would someone who holds this position say 'yes, that's what I actually believe'?"
If the user says no, refine. The steel man must pass the advocate's test — the person who holds the position should recognize it as their own view, fairly stated.
NOW — and only now — identify the real flaws. Use Kline's 7 objection types. Not every type will apply. Skip the ones that don't fit. The goal is precision, not volume.
1. Factual Objection The position relies on claims about the world that are false or unverified. Specific data points, historical claims, or empirical assumptions that don't hold up.
2. Logical Objection The reasoning contains a formal or informal fallacy. The conclusion doesn't follow from the premises, even if the premises are true.
3. Conceptual Objection Key terms are used ambiguously, or the position conflates distinct concepts. The argument works only because a word is doing double duty.
4. Relevance Objection The evidence cited is real but doesn't actually support the conclusion. True premises, valid reasoning, but the argument answers a different question than the one being asked.
5. Alternative Explanation Objection The position identifies a real pattern but attributes it to the wrong cause. The data is real, but a different explanation fits better.
6. Pragmatic Objection The position may be theoretically sound but fails in practice. Implementation problems, unintended consequences, second-order effects that undermine the stated goal.
7. Methodology Objection The way the position arrived at its conclusions is flawed — selection bias in evidence, unrepresentative samples, unfalsifiable claims, or moving goalposts.
Rate each genuine weakness on a four-level scale:
| Severity | Meaning |
|---|---|
| Fatal | This flaw undermines the entire position. If this objection holds, the position fails. |
| Substantial | This is a serious problem that significantly weakens the position but doesn't destroy it entirely. The position could be revised to address it. |
| Clarification | The position needs better definition or qualification here. Not wrong, but imprecise in a way that matters. |
| Peripheral | A real issue, but not central to the main argument. Fixing it wouldn't change the core conclusion. |
Construct the user's actual counter-argument, targeting only the genuine weaknesses:
YOUR COUNTER-ARGUMENT
The steel man (what you're actually arguing against):
[One-sentence summary of the strongest version]
Fatal or substantial weaknesses:
1. [Objection type]: [Specific flaw, stated precisely]
2. [Objection type]: [Specific flaw, stated precisely]
Your argument:
[The counter-argument that targets the real flaws, not straw men.
This should acknowledge what the position gets right before
showing where it breaks down.]
What you should concede:
[Parts of the opposing position that are genuinely correct or
insightful. Conceding these makes your argument stronger, not weaker.]
The one-sentence version:
[The most powerful distillation of your counter-argument]
Present the full arc:
STEEL MAN ANALYSIS
POSITION UNDER EXAMINATION: [The position]
STEEL-MANNED VERSION:
[The strongest form of the position]
GENUINE WEAKNESSES (ranked by severity):
[FATAL] [Objection Type]: [Description]
[SUBSTANTIAL] [Objection Type]: [Description]
[CLARIFICATION] [Objection Type]: [Description]
[PERIPHERAL] [Objection Type]: [Description]
YOUR COUNTER-ARGUMENT:
[Built from the real weaknesses, with appropriate concessions]
ONE-SENTENCE VERSION:
[The sharpest distillation]
Ask: "Did the steel man reveal anything about the opposing position you hadn't considered? And does your counter-argument feel stronger now that it targets the real weaknesses instead of the easy ones?"
The goal is not to "win" — it's to think clearly. If the steel man exercise reveals that the opposing position is stronger than you thought, that's a feature, not a bug. You either sharpen your counter-argument or update your view. Both are progress.