In the Categories Aristotle defined substance as that which is neither predicable of nor in another.1 In saying that a substance is not predicable of another, Aristotle meant to exclude genera and species from the category substance.2 A man is a substance but not man. In saying that a substance is not in another, Aristotle meant to exclude property particulars from the category. A man is a substance, not his color.3
The Categories treats substances as simples. Though a particular substance, Bucephalus the horse, has parts, it is nevertheless a single entity in the category substance and, hence, incomplex in the way a black [End Page 289] thing or a running man are not.4 Black things and runners are complex because they are aggregates of substances and property particulars. Even if a horse is one substance and, thus, an entity (unlike a substance cum some of its attributes or a group of related substances), a horse is made of parts and one may wonder how it is related to its parts, as well as how its being made of parts coheres with the definition of substance given in the Categories. The Categories does not tell us how a complex substance is related to its parts; it only tells us that its parts are not in it in the way its properties are.5 This makes sense; while the particular properties of a horse ontologically depend on it, it would seem ontologically dependent on its parts. If that is the case it is hard to see how a horse could be a substance, if a substance is neither predicable of nor in another (or others).
It is well known that Aristotle tried to solve the problem of how composite substances are true unities in such works as the Physics and the Metaphysics. Aristotle’s proposed solution to the problem of how his definition of substance allows that composites exist is fascinating, as are later interpretations of it. However, we will not pursue any of these solutions. Instead we will explore a theory of composite substances that is based on a different definition of substance from Aristotle’s. This theory was worked out over many centuries by philosophers of the Nyāya-Vaiśeṣika school, one of the ‘orthodox’ schools of Indian thought.6 We shall argue in this paper that the Nyāya-Vaiśeṣika theory of substance is plausible in itself, provides clear criteria for distinguishing substances from non-substances, and poses no problems in principle for accepting the existence of composite substances. We think that these features of their theory give it advantages over Aristotle’s. Common sense holds that, if there are substances, such things as trees and cats are, but it also holds that such things are not only composite substances but are made [End Page 290] of other substances. It is this last intuition that Aristotle’s definition of substance seems to deny and, for this reason, we believe that a theory of substance that allows that some substances are made out of other substances has certain prima facie advantages over Aristotle’s.
Every metaphysical theory is beset with difficulties, so it will come as no surprise that this holds of the Nyāya-Vaiśeṣika theory of composite substances. In this paper we do not have the space to pursue all the difficulties that beset that theory and will concentrate on the one we think gravest. It is that the Nyāya-Vaiśeṣika theory entails that no composite substance in fact endures for more than a few moments. Since the intuition that at least some composites endure for quite a length of time is at least roughly as strong as the intuitions that there are composites and that they are made of other substances, we believe that this entailment of the Nyāya-Vaiśeṣika theory of composites is quite grave, perhaps as grave as those that face any Aristotelian theory of them. Fortunately, the Nyāya-Vaiśeṣika theory...