Organon

Overview

There is no book called the Organon. The name is a later editorial label, Greek for “instrument,” and it carries a claim: logic is not a part of philosophy but a tool you need before you can do any of it. Six separate treatises were gathered under it, and they are not a course. They were written at different times, they overlap, and the order they are printed in is a teaching order imposed afterwards, running from the simplest unit of speech to the ways arguments go wrong.

Five hundred pages is more than anyone reads at once, and almost nobody does. Programs assign one or two: usually the Categories with On Interpretation, because they are short, or the Posterior Analytics, because it is the one with a philosophical question in it. The Prior Analytics is where the machinery is, and it is technical in a way no other ancient text is, closer to a mathematics textbook than to a treatise.

What makes the whole thing worth confronting is the level Aristotle is working at. He is not arguing about justice or the soul. He is asking what an argument is, and answering in a way that lets you check whether one works without knowing anything about its subject.

The argument

Categories. What kinds of thing can be said about a subject? Aristotle lists ten: substance, quantity, quality, relation, place, time, position, state, action, and being acted upon. Everything else in the list depends on the first. A particular man or a particular horse is a primary substance; species and genera are secondary substances; and colours, sizes and locations exist only as belonging to something. The book is four pages of classification with two thousand years of metaphysics hanging off it.

On Interpretation. From terms to statements. A statement is the kind of speech that can be true or false, which prayers and commands cannot, and every statement has exactly one contradictory. Aristotle sets out how affirmations and denials, universal and particular, relate, which later becomes the square of opposition. Then chapter 9, the shortest famous passage in the corpus. If it is true today that there will be a sea battle tomorrow, it seems that nothing anyone does can prevent it, and deliberation is pointless. Aristotle wants to keep the principle that every statement is true or false without conceding that the future is already settled, and the chapter has been read a dozen incompatible ways.

Prior Analytics. The syllogism, and the first formal system in history. Take two premises sharing a middle term; a conclusion follows or it does not, and which it is depends only on the arrangement. Aristotle sorts the arrangements into three figures, proves which moods are valid by reducing them to the obvious ones in the first figure, and shows the rest invalid by counterexample. He then extends the whole apparatus to premises about necessity and possibility, which is much harder and much less successful. The achievement is the method: validity has been made a matter of form.

Posterior Analytics. Having a valid argument is not yet having knowledge. Scientific knowledge, epistēmē, is knowing the cause of a thing and knowing it could not be otherwise, so a demonstration must proceed from premises that are true, primary, immediate, and explanatory of the conclusion. This produces the problem the treatise exists for. If every premise must itself be demonstrated, proof never starts; if some are not demonstrated, how are they known? Aristotle rejects both an infinite regress and circular proof, and answers in Book II that the first principles are reached by induction from perception, and grasped by nous, a direct intellectual seeing. The last chapter, with its image of a routed army re-forming one man at a time, is the closest thing in Aristotle to an account of how experience becomes understanding.

Topics. Argument from endoxa, the things generally accepted or held by the wise. This is dialectic rather than demonstration: you are arguing with someone rather than proving something. Aristotle organises it around the four predicables, definition, genus, property and accident, and supplies hundreds of topoi, lines of attack and defence for use in debate.

Sophistical Refutations. A catalogue of bad arguments that look good, divided into those that depend on language, such as ambiguity and bad word division, and those that do not, such as begging the question and the fallacy of accident. Aristotle ends by pointing out that on this subject there was nothing whatever before him.

How to read it

Do not read it straight through. Pick the entry the course needs and go deep.

If you want the philosophy, read the Posterior Analytics, Book I chapters 1 to 6 and then Book II chapter 19. That is perhaps thirty pages and it contains the whole problem of first principles.

If you want the logic, read Prior Analytics I.1 to I.7 with pencil and paper, and work the reductions yourself. It is unreadable passively and reasonably quick actively.

If you want the metaphysics, read the Categories entire, then On Interpretation 9, and note how fast the questions get deep.

Learn four terms before you start and the rest will follow: term, premise, universal and particular, and middle. Aristotle writes his syllogisms with letters standing for terms, and if you are not converting them into examples as you go, the pages will slide past without leaving anything.

The common misreading is to treat this as an early, clumsy version of modern logic and to grade it accordingly. Aristotle is not attempting a formal language for mathematics. He is analysing scientific explanation in ordinary speech, and the question that most interests him, where the premises of a science come from, is one that formal logic deliberately does not ask.

Questions it raises

• Are the ten categories a classification of words, of thoughts, or of what there is? Does Aristotle need to decide?
• If a statement about tomorrow is already true or false today, is anything up to us? Does chapter 9 of On Interpretation escape that, or only postpone it?
• Demonstration must start somewhere. Is Aristotle’s appeal to induction and intellect an answer to the regress, or an admission that the regress has no answer?
• Knowing why is said to be better than knowing that. Is that true of every science, or only of the ones that have causes to give?
• Dialectic argues from what people accept, demonstration from what is true. What do you do when a science is new and there is nothing yet to accept?
• Logic was called an instrument rather than a part of philosophy. Is it neutral between philosophies, or does the syllogism already commit you to a picture of what the world is like?