Discourse on Method

Overview

The Discourse was published anonymously in Leiden in 1637 as the preface to a scientific book. Behind it came three treatises meant to prove the method worked: one on optics, one on meteorology, which contains the first correct explanation of the rainbow, and one on geometry, which invented what we now call coordinate geometry. Almost nobody reads the appendices now, and most editions omit them, but the Discourse was written to introduce them and its claims are less abstract if you remember what they were advertising.

It is short, informal and written in the first person, in French, which was a provocation. Descartes says he chose the vernacular so that people who use only their natural reason would judge him better than those who trust old books. He also keeps insisting that he is not telling anyone what to do, only describing what he did. That modesty is a tactic. Galileo had been condemned in 1633 and Descartes had shelved a treatise on the physical world in response, a decision he discusses in Part VI without naming the case.

Six parts, each a few pages: his education, the method, a provisional morality, the metaphysics, the physics and physiology, and the reasons for publishing. It can be read in an evening and rewards a second evening.

The argument

Part I is autobiography and complaint. Descartes was well taught at La Flèche and learned that mathematics alone gave certainty, while philosophy supplied endless disagreement and the other disciplines borrowed their principles from it. He left, travelled, served in armies, and decided that studying the book of the world was better than studying books, before concluding that even the world’s variety only taught him how little custom proves.

Part II describes the winter he spent alone in a stove-heated room in Germany working out his plan. He draws an analogy with town planning: cities that grew piecemeal are tangled, while those laid out by one engineer on open ground are regular. He is careful to say he would not tear down a state on this reasoning, only his own opinions. Then the four rules. Accept only what presents itself so clearly that there is no occasion to doubt it. Divide each difficulty into as many parts as possible. Begin with the simplest and easiest objects and ascend by degrees. Make enumerations complete enough that nothing is left out. The model is geometrical proof, extended to every subject.

Part III gives three or four maxims to live by in the meantime, since doubting everything is no way to conduct a life. Obey the laws and customs of your country and the religion you were raised in. Once you choose, act with resolution rather than wavering. And try to alter your own desires rather than the order of the world, since only your thoughts are entirely in your power. The last of these is Stoic, and Descartes knows it.

Part IV compresses the metaphysics that the Meditations would expand four years later. Doubting everything, he finds that his own thinking cannot be doubted, since doubt requires a doubter. He infers that he is a substance whose whole nature is to think, and that this substance is easier to know than the body and would be what it is even if no body existed. He then argues from the presence in himself of the idea of a perfect being to that being’s existence, and from God’s perfection to the reliability of clear and distinct ideas.

Part V summarises the suppressed physics: a world formed from matter in motion by laws God established, an account of the heart and circulation, and the claim that animal bodies are machines. Descartes then draws the line. Machines might imitate an animal exactly, but no machine could arrange words to answer sense put to it, and no machine acts from knowledge rather than from the disposition of its parts. Humans therefore have a rational soul, which he says cannot be derived from matter.

Part VI explains the delay and the decision. Descartes wanted to keep working privately, then reflected that knowledge of nature could relieve human beings of illness and labour, and that withholding it would be a failure of duty to others. He asks for experiments and objections, complains in advance about the ones he expects, and says he will publish no more if the reception is bad.

How to read it

Read it before the Meditations, not after. Part IV is the same argument in a quarter of the space, and having seen the compressed version makes the slow version legible.

Read Part II twice and write the four rules out. Then test them on something. They sound obvious until you try to apply them to a question that is not mathematical, at which point you discover how much work the word “clearly” is doing and how you would know when an enumeration is complete.

Do not skip Part VI. It is the most revealing section, because it is where Descartes talks about censorship, about what science is for, and about the social conditions that research needs, and it is where the ambition that drives the whole project is stated plainly.

The common misreading is that the Discourse is a sceptical book. It is the opposite: a manifesto by a man who thinks the sciences are about to be finished, and who treats doubt as a one-time clearing operation rather than a permanent condition. The other misreading is that the animal-machine argument is about cruelty. Descartes is making a claim about language and generality, not about feeling, and it is worth judging it on that basis.

Cress if you want it bound with the Meditations, Maclean if you want to know what the rainbow and the geometry had to do with it.

Questions it raises

• Can the four rules be applied outside mathematics, or do they presuppose a subject already made orderly?
• Descartes says he demolishes only his own opinions, not his society’s institutions. Is that distinction stable?
• If the provisional morality tells him to keep his country’s customs and religion while doubting everything, has he really doubted them?
• Is the cogito here an inference, as the word therefore suggests, or something else?
• The test for reason is the flexible use of language. Does that test still do what Descartes wanted it to do?
• Descartes justifies publishing by appeal to what knowledge can do for human welfare. Does that make the method answerable to its results?