The Socratic Method

Ask most people what the Socratic method is and you get some version of "teaching by asking questions." That is not wrong so much as it is uselessly vague — it describes a seminar, a job interview, and a police interrogation equally well. The actual method is far more specific, and far more hostile. It is a procedure for destroying a claim using nothing but material the claimant has already agreed to.

Socrates wrote nothing. Everything we have comes through Plato, with a different account from Xenophon and a hostile cartoon from Aristophanes, which means the historical man is permanently out of reach. What survives intact is the machine he ran. It is worth taking apart, because it still runs.

The procedure

The Greek word is elenchus — refutation, or cross-examination. It has four moves, and they always come in the same order.

First, the interlocutor asserts something, usually a definition: courage is this, piety is that. Second, Socrates asks for agreement on some further claims, each of which looks harmless and unrelated. Third, he shows that those claims together entail the negation of the original definition. Fourth — and this is the part that gets skipped in every summary — the interlocutor is left with no replacement. The definition is gone and nothing stands where it stood.

Watch it run

In the Laches, two generals are asked what courage is. Laches — a soldier, answering as a soldier — says courage is standing firm in the ranks and not running away. It is a good answer. It is the answer his whole life supports.

Socrates does not disagree. He asks about the Scythians, who fight by retreating and wheeling around to attack again. Are they cowards? Laches says no, obviously not. So a man can retreat and still be courageous. So courage is not standing firm.

Then it widens. What about courage in illness? In poverty? In politics, or at sea? Laches agrees that courage appears in all of them. But now the definition has to cover a man holding a line, a man withdrawing in good order, and a man dying quietly of a disease — and "standing firm in the ranks" covers only the first.

Note what did not happen. Socrates never said courage is something else. He offered no theory. He asked whether the Scythians were cowards, and the answer that destroyed the definition came out of Laches' own mouth.

Why it ends in nothing

The early dialogues almost all end in aporia — impasse, literally "no way through." The definition is dead, no successor is proposed, and everyone goes home. Readers who arrive expecting a lesson find this maddening, and reasonably so. What is the point of a method that reliably produces failure?

The Apology gives the answer, and it is stranger than the usual paraphrase. The oracle at Delphi is said to have declared that no one was wiser than Socrates. He took this as an insult he needed to disprove, so he went looking for a wiser man — a politician, then poets, then craftsmen. He kept finding people who knew something real and, on the strength of it, believed they knew about the large questions too. His conclusion was not that he knew nothing. It was narrower: he did not think he knew what he did not know. That is the whole of the advantage.

So the impasse is not a failure of the method. It is the output. Elenchus does not produce knowledge; it produces an accurate map of where knowledge is absent — which, if you were about to act on a confident false definition, is a considerable service.

Why it got him killed

In 399 BCE Socrates was tried on two charges: impiety, and corrupting the young men of Athens. He was convicted and executed with hemlock. The standard reading is that Athens could not tolerate a gadfly, which is true and a little too flattering to everyone involved.

The sharper reading is that the charge had a factual basis. The method is trivially easy to imitate badly. A young man who has watched Socrates dismantle a general's definition of courage has learned a transferable skill: he can now make any respected adult look foolish in public, using only that adult's own admissions. What he has not learned is anything to put in place of what he tore down. Athens had spent a generation watching well-born young men acquire this trick, and some of them — Alcibiades and Critias among them — had gone on to do real damage to the city.

Refutation is much easier to teach than judgment. That is a permanent property of the method, not an accident of Athens.

What it is good for now

Strip away the togas and elenchus is a tool for a specific job: testing whether a proposed definition survives contact with the cases it claims to cover. The moves transfer directly.

The counterexample hunt is the core of it. Someone proposes a rule — a policy, a threshold, a definition of "done," a security boundary. You do not argue with the rule. You look for the case that the proposer will immediately agree is on the wrong side of it, and you ask about that. Common-law reasoning works this way, code review works this way, and every good design review is an elenchus with better manners.

The manners matter more than the technique, though, and this is where imitation usually goes wrong. Socratic questioning done in bad faith — where you already have your conclusion and you are walking someone into it — is not elenchus. It is a trap with a philosophical costume, and everyone can feel the difference. The genuine version requires that you not know the answer either. That is the expensive part, and it is why the method is far more often invoked than run.