Aristotle’s Categories

My first reading was Paul Studtmann’s article in the Stanford Encyclopedia of Philosophy (SEP), “Aristotle’s Categories.” The Categories basically lays out Aristotle's ontology — what exists and what you can say about what exists. Studtmann falls prey to a very common malady in academics, an inability to write for true beginners. I supplemented my reading with a much more amenable SEP article by Robin Smith, “Aristotle’s Logic,” which, along with Grim's explanations in email, helped me understand some unexplained terms and concepts in Studtmann.

There were a couple of interesting things to me about the Categories. The lesser interesting thing is Aristotle’s ontology itself. In essence, he claims there are substances and accidents. Substances exist independently while accidents only exist in substances. For example, a human being is a substance while skin color and height are accidents. You can’t have color by itself; color only exists as a property of a substance.

Substances and accidents are further categorized into universal and particular. Sadly, I discovered that universal accidents are not what happens when God goes on vacation and forgets to turn the gravity off. Instead, universal accidents are the concepts of accidents, such as redness. Universal substances are similarly concepts of particular substances. Humanity would be a universal substance, while Steve Earle would be a particular substance.

Aristotle prioritizes particular substances, calling them first substances; all the rest depend on them for existence: humanity doesn’t exist without individual humans, redness doesn’t exist without apples and other particular things that have that color. Here, he firmly disagrees with Plato, and I agree with him.

Although they use different and somewhat less precise terms, the Roman Catholic doctrine of transubstantiation is easily explained with this set of categories. The bread and wine maintain their accidents, but their substances are transformed into the body and blood of Christ.

But that’s old news. More interestingly on this occasion, Aristotle uses the term ‘said of’ for something that only exists in something else, an interesting use of grammar to describe reality. Studtmann informs us that some philosophers have tried to argue that Aristotle was really laying out a theory of language, but too much of the Philosopher’s work treats words as referring to actual things and most philosophers take the Categories as an ontological work, not a linguistic one. Even so, the Categories basically come down to substances and things you can say about substances. The Latin for the title is even the Predicamenta, The Predicates.

Language is a fascinating and important part of philosophy, and I’ll come back to it again.


Grim said...

A very great deal has been made of the point you realized right away, which is that Greek logic is tied very tightly to Greek grammar. As I told you in email, but repeat now for others, there seems to be no alternative to tying logic to grammar. Even where we try to move to a purely symbolic logic, to eliminate the ambiguities of natural language, we end up with 'syntax' and 'semantics' and so forth.

It turns out that we also end up with substances and accidents. An ordinary proposition (another grammatical term!) in symbolic logic might be this:


It turns out that "x" is functionally like a substance, while F and G are accidents. What this says is that "for every x," where x is any thing that is a member of your set, if that x has accident F it will also have accident G.

Now your set can be whatever you want: numbers, say ('for every x, if x ends in 4 (Fx), then x is divisible by 2 (Gx)'. Or it can be everything that exists.

But that's the real point: x is a thing, and it's the kind of thing that can have or express qualities. So we have something very similar to Aristotle's substance/attribute distinction assumed in the structure of the grammar.

There is a big difference, which is that F could be 'is a man,' so that things Aristotle believes are substances can end up being accidents. But I think Aristotle is right here, in a way that dilutes the criticism that he's just giving you a grammar. There's a bigger argument about what makes a substance a substance which is not entirely captured here. One thing that a substance can do is reproduce itself. Thus, man is a substance because man exists naturally in a self-sustaining way.

Aristotle therefore has a standard for what kind of a thing is an 'x,' whereas the modern logic is weaker. If 'being a raven' is an accident, then what is left of the 'x'? What is this x that is a raven?

Grim said...

I think a defender of the modern approach would object that this logic isn't troubled by ontology. It doesn't make any assumptions about what an "x" might be -- that's the business of another field of philosophy (and, to some degree, of the outgrowth of natural philosophy we call experimental science). The point would be that you can do the proofs without worrying about what is an "x" and what is an "F" or "G." Whatever it proves to be, the logic will hold.

But that's not quite right, as you can now see. The basic structure of the modern logic assumes that there will be an x, which has properties like F or G. The ontological structure is still there, assumed into the model. They just no longer have standards to guide them in thinking about what an "x" might be.

Tom said...

Right. You still have subjects, which must be nouns or noun-stand-ins (i.e., things) and verbs. I don't think that's something humans can escape, though I haven't thought a lot about it before this.

I'll need to think more about the comparison you are making (and probably read a lot more) to really make a call on the debate between modern logic & Aristotle's logic.