Almost everything said about this book gets its nature wrong. It is not a treatise. It is a preface.

In 1637 Descartes published a volume containing three scientific essays — on optics, on meteorology, and the Geometry that introduced coordinate geometry and changed mathematics permanently. The Discourse on the Method is the introduction he wrote for them: an explanation of how he had arrived at the results that followed. It was published anonymously, and the full title runs Discourse on the Method of Rightly Conducting One's Reason and Seeking Truth in the Sciences.

And it is in French. That is the other thing routinely missed, and Descartes explains it in the final part: he is writing in the language of his country rather than in Latin because he hopes to be read by those who use only their natural reason, rather than by those who trust only ancient books. In 1637 that means people without a university education — which, at the time, means most people, and effectively all women. It is a deliberate act.

What reading it is like

Sixty pages, six parts, in plain conversational French, written in the first person as the story of an education. It is the easiest entry to Descartes by a distance and it is genuinely enjoyable, which very little seventeenth-century philosophy is.

The cost is that the arguments are sketched rather than made. The cogito appears in Part IV, in a paragraph, with the proofs of God's existence compressed into a page or two beside it — and Descartes himself said this was too brief and was why he wrote the Meditations. Read as philosophy, it is thin. Read as what it is, an account of how a person came to think this way, it is very good.

Where to start: Part II for the method, Part III for the provisional morality, which is the most underrated thing in it. Part IV if you want the famous sentence in its original French. Part V is the one modern readers find hardest to forgive.

The six parts

One. The autobiography, opening with the line everybody quotes: good sense is the best distributed thing in the world, since everyone thinks they have enough of it. He describes his education at La Flèche, praises it, and reports that at the end of it he found he had learned nothing certain.

Two. The four rules. Accept nothing not clearly and distinctly known; divide each problem into parts; proceed from simple to complex; enumerate completely. Preceded by the image of rebuilding a house rather than patching it, and the observation that a city planned by one mind is better ordered than one grown by accretion.

Three. The provisional morality, adopted while the rebuilding is under way, because you cannot suspend judgment about how to live while you are working. Obey the laws and customs of your country and the religion you were raised in. Be as firm and resolute in action as you can, and follow even a doubtful opinion consistently once chosen. Conquer yourself rather than fortune, and change your desires rather than the order of the world.

That third maxim is Stoic and he says so. The whole section is a better piece of practical philosophy than its neighbors, and it is skipped because it is not epistemology.

Four. Je pense, donc je suis. The cogito, the proofs of God, the distinction between mind and body — all in a few pages. See cogito for the argument and the Meditations for the version Descartes thought adequate.

Five. The physics, the account of the circulation of the blood, and the argument that animals are machines. He holds that a sufficiently complex automaton could imitate an animal indistinguishably, but not a human — because no machine could use language to respond appropriately to anything said in its presence. The passage is now read as the first statement of a test later associated with Turing, and the treatment of animals is the part of Descartes modern readers most dislike.

Six. Why he published, and why he had not published earlier — a careful passage about Galileo's condemnation four years before, and about his own suppressed treatise on the world.

Key passages

Part I — good sense is the most equitably distributed thing in the world.

Part II — the four rules; the planned city.

Part III — the three maxims of the provisional morality.

Part IV — Je pense, donc je suis; the proofs of God; the real distinction.

Part V — animals as automata; the language test.

Part VI — the reasons for publishing in French, and the account of why the physics was withheld.

What is contested

How much of the autobiography is true. It is a constructed narrative with a philosophical purpose, and the dates and the tidiness of the sequence do not all survive checking against the correspondence.

Whether the cogito here is the same argument as in the Meditations. The French je pense, donc je suis has the inferential donc; the Latin of the Meditations avoids it. Descartes insisted the cogito is not a syllogism, and this phrasing is the main evidence against him.

The animal machine doctrine. Whether Descartes held that animals feel nothing at all, or only that their behavior requires no mind to explain it, is genuinely disputed in the scholarship — and the harsher reading was taken up enthusiastically by people who found it convenient.

Its relation to the three essays. It was written to introduce them and is now published alone, so it is almost always read without the work it was explaining. The Geometry it prefaces is the most consequential of the four texts and the least read.