The Grammar of Persuasion

Persuasion is usually taught as a matter of content — find the strongest argument, present it clearly — and this is why so much of it fails. The strongest argument, presented in the wrong order, does not merely underperform. It actively hardens the position it was meant to move.

The reason is that people do not evaluate arguments one at a time. They evaluate them against a position they already hold, and the first thing any new claim has to survive is not a test of its truth but a test of its threat. A claim that arrives as a challenge to identity is assessed by a different and much less generous faculty than one that arrives as an extension of something already believed.

This has an immediate consequence for order. The conventional structure — state your conclusion, then support it — is efficient and, for a hostile audience, close to self-defeating, because it announces the threat before any of the grounds that might make it acceptable. The classical alternative is to establish the shared ground first, at length, and to arrive at the disputed conclusion late enough that it reads as a consequence of things the listener has already agreed to rather than as a demand.

A second consequence concerns concession, which is chronically undersold. A concession is not a weakness in your case; it is the mechanism by which you demonstrate that you have understood the other one. An argument that concedes nothing signals, accurately, that it was constructed without reference to the opposing view — and a listener who detects that will discount everything that follows, correctly, as advocacy rather than reasoning.

The third and least comfortable point is about certainty. Overwhelming confidence is persuasive to people who already agree with you and repellent to everybody else, which makes it an excellent instrument for consolidating a base and a very poor one for changing a mind. Calibrated language — this holds under these conditions, this is the part I am least sure of — reads as weakness only to an audience that was never going to move. To everyone else it reads as the single most reliable signal that the speaker is describing the world rather than defending a position.

What all three points have in common is that they are about the relationship rather than the proposition. This is uncomfortable for anyone who believes that the merits should be enough. The merits are necessary. They have never been sufficient, and the pretence that they are is itself a rhetorical position — a comfortable one, because it locates every failure to persuade in the stubbornness of the audience rather than in the construction of the case.

The practical discipline that follows is narrow and demanding. Before you make the argument, be able to state the opposing view in a form its holders would accept as fair. Then find the largest thing you both believe, and start there. Then make your case in the order that follows from it, conceding what is genuinely true along the way, and claiming no more confidence than you actually possess.

That is slower than being right loudly. It is also, as far as anyone has ever been able to demonstrate, the only version that works.