Changing a mind without a fight — the architecture of an argument that actually moves people.
Lesson 4 of The Art of Being Understood · see all five lessons →
Read each part, then pause for the questions. The whole idea reverses an instinct almost everyone has.
Here is the counter-intuitive move at the heart of all serious persuasion. When you want to defeat an opposing view, your instinct is to find its weakest version, caricature it, and demolish that. This is the straw man: you build a flimsy effigy of your opponent's position, set fire to it, and declare victory over an argument nobody actually held.
The masterstroke is the opposite. Before you answer the other side, you state their case more strongly and more fairly than they could state it themselves. You grant every reasonable point. You find the version of their argument that an intelligent, honest advocate would be proud of. Then — and only then — you answer it. This is the steel man.
The effect on a listener is disproportionate. By showing you have understood the opposing view at its strongest, you earn something no amount of force can buy: the suspicion that you might actually be right. A mind that feels caricatured digs in. A mind that feels accurately understood opens, just a little — and a little is all you need.
Two and a half thousand years ago, Aristotle laid out three ways a speaker moves an audience, and nobody has improved on them since. Ethos is persuasion through credibility — the listener's sense that you are competent, fair and worth trusting. Pathos is persuasion through feeling — engaging what the audience cares about, fears, hopes for. Logos is persuasion through logic — the structure of claim, reason and evidence that makes a case hold together.
Most people, asked to argue, reach straight for logos: they pile up reasons and assume the reasons will do the work. But reasons rarely move anyone on their own. We are persuaded by people we trust (ethos), about things we feel (pathos), and then reassured by logic that lets us justify the move we were already inclined to make.
And there is a fourth element, absent from the textbook triad, that the strongest persuaders never omit: concession. Granting openly what is true in the other side. It looks like a weakness — you are conceding ground — but it is the engine of credibility. The speaker who concedes nothing sounds as though they have not listened, and a listener who feels unheard cannot be persuaded.
Argument goes wrong in four recurring ways. Learn to hear them and you are halfway to avoiding them.
1. The straw man. Attacking a weakened, distorted version of the opposing view — which convinces no one who actually holds it. 2. The unearned assertion. Stating a claim as obvious or settled when it has been neither argued nor evidenced; volume and confidence standing in for reasons. 3. The absence of concession. Granting the other side nothing, which reads not as strength but as not having listened. 4. Volume substituting for structure. Repetition, emphasis and intensity deployed where a clear architecture of claim, reason and evidence should be — loudness where there ought to be logic.
The vocabulary of persuasion — the metalanguage you need to talk about argument, and the phrases you need to build one. Tap each card for a precise definition and a usable example.
The same view, argued three ways. Notice that only the rhetorical approach changes; the underlying claim is identical.
| Register | The same position |
|---|---|
| Bludgeoning assertion (no concession) | "Remote work obviously doesn't work. Anyone who thinks otherwise is just avoiding the office. Productivity collapses — end of story." |
| Straw-man attack | "So the remote-work crowd want us all sitting at home in pyjamas forever, never seeing a colleague again? That's absurd." |
| Concession-led persuasion (with a steel man) | "It's true that remote work gives people focus and saves hours of commuting — and for deep, solo tasks it can be genuinely better. And yet the spontaneous conversations that spark new ideas are hard to replicate on a screen. So the question isn't home or office; it's which work belongs where." |
These are not debating tricks. They are the precision tools of a C2 persuader: the difference between an argument that is correct and one that moves people. Try each before you reveal the model answers.
The most powerful structure in persuasion is to give ground and then pivot. You concede what is true, which earns trust, then turn on a hinge word — but, and yet, nonetheless — to your own point. "Granted, X. But…" / "It is true that… and yet…" / "I take the point that… nonetheless…". Complete each pivot.
1. "Granted, the new policy will cost more in the short term. "
2. "It is true that the critics have a point about the risks, "
3. "I take the point that change is unsettling; "
1. "Granted, the new policy will cost more in the short term — but the alternative costs us far more, and for far longer."
2. "It is true that the critics have a point about the risks, and yet the risk of doing nothing is the one we keep failing to count."
3. "I take the point that change is unsettling; nonetheless, staying still has stopped being the safe option."
The move: the concession must be genuine for the pivot to work. A fake concession ("I hear you, but…") fools no one. Grant something that actually costs you.
Sound is part of sense. Ideas arrive in threes land harder than in twos or fours, and a balanced antithesis — "not X, but Y" — makes a contrast unforgettable. Build the rhythm.
1. Turn into a triad — "We need to act because it is urgent":
2. Turn into an antithesis — "This is about progress": "This is
3. Build a "not only… but also…" — on any topic:
1. "We need to act because it is urgent, because it is right, and because no one else will." (triad — note the rising weight to the third item)
2. "This is not about looking backwards, but about deciding what comes next." (antithesis — the balance carries the force)
3. "Not only does it save money, but it also wins back the trust we lost." (escalation — the second clause outweighs the first)
The move: put your strongest item last in a triad, and let the antithesis do the contrast so you don't have to spell it out.
Credibility cannot be asserted ("trust me, I'm an expert") — it must be demonstrated. You earn ethos through precision, through fairness to the other side, and through admitting the limits of your own case. Rewrite each boast as earned credibility.
1. "I know more about this than anyone here." →
2. "My argument is completely watertight." →
1. "I've spent ten years on exactly this problem, and here is the one thing that surprised me most…" (precision and specificity earn what the boast only claimed)
2. "This argument has one real weakness, and I want to put it on the table myself before anyone else does…" (admitting a limit makes the rest more believable, not less)
The move: the fastest way to build ethos is to be visibly fair to the side you oppose. Naming your own argument's weak point is a power move, not a concession of defeat.
Before you rebut, build the opposing argument so well that an honest advocate of it would nod. Then answer it. Steel-man each position below in one strong sentence — as its own supporters would wish it put.
1. Steel-man the case against a policy you happen to support:
2. Steel-man the case for a view you find unconvincing:
1. "The strongest case against is not that the goal is wrong, but that this particular method has failed everywhere it's been tried — and good intentions won't change that record." (notice: no caricature, a point a real opponent would be glad to claim)
2. "At its strongest, the argument is that the costs fall hardest on the people least able to bear them — and that's a serious objection, not a trivial one." (you've made the opposing case sound reasonable, which is the whole point)
The test: would someone who holds the view say "yes — that's exactly it, only better"? If not, you've built a straw man, not a steel one.
A persuasive case has visible architecture. State the claim, give the reason it holds, back it with evidence, then land the so-what — why it matters to this listener. Keep a bank of concession and persuasion phrases ready and reach for them out loud.
Here is one argument — that a city should invest in cycling infrastructure — made twice. Read the bludgeoning version, then watch concession and the steel man transform it.
"We obviously need more bike lanes. Anyone who opposes them is just a driver who doesn't care about the planet. The car lobby will tell you it'll cause traffic chaos — that's nonsense, they always say that, they just want to keep their precious parking spaces. The science is settled, everyone sensible agrees, and frankly if you're against this you're against the future. End of discussion."
"Let me start with the strongest objection. (steel man) It is true that taking road space for bike lanes can, at least at first, slow traffic and squeeze out parking — and for a shop owner whose customers arrive by car, that is a real and immediate cost, not an imaginary one. (concession) I take that seriously. And yet (pivot) the cities that held their nerve tell a consistent story: within two years, the lanes fill, some of those car trips simply vanish, and the streets that local businesses feared would empty became the busiest in town. (evidence) So the real choice isn't cars versus bikes. It's whether we want streets that are cleaner, safer, and better for the very shops that worry about them most. (so-what, with a triad)"
Same position. The one that concedes is the one that converts.
Now move to the last tab. Everything above is about to be tested out loud.
Do all four, in order. The first warms up the idea; the rest put the toolkit under pressure. The timer is there to stop you escaping into vagueness — fill the time with structure instead.
Before anything else, explain the architecture of persuasion in your own words, as if to an intelligent colleague who missed today's session. Cover ethos, pathos and logos; explain what concession does and why omitting it is fatal; and describe the steel man and how it differs from the straw man.
Use at least five of the Key Language terms (the three appeals, to concede, the straw man, the steel man, rhetoric, antithesis, the rule of three, to rebut, compelling, to appeal to).
Take a position you genuinely believe and argue it to a sceptical listener. You must build in, deliberately and audibly, three things:
a concession-pivot ("Granted, X — but…"), the rule of three somewhere in your case, and a steel man of the opposing view before you rebut it. Notice which feels least natural — that's the one most worth practising.
Choose a position you genuinely disagree with — and argue for it. State it more persuasively, more fairly, and more compellingly than its own advocates usually manage. No caricature, no winking to the audience, no "of course this is wrong".
The test: would a sincere believer in that view say "yes — that's exactly it, only better"? Only then have you steel-manned rather than straw-manned.