抄録
Aristotle deals with the method of definition by division mainly in his Analytica Posteriora, Metaphysics Z12, and De Partibus Animalium A2-3. Scholars have found an inconsistency in his treatment of the method and have tried to use it to determine the chronology of those works. The inconsistency is that in Anal. Post, and Metaphysics, he seems to argue for the consequence of adoption of the method, where the species defined is the final difference, while in Part. Animal., this consequence is the very point that is criticised. The aim of my paper is to show that the inconsistency does not exist by distinguishing two aspects of Aristotle's treatment in which he considers the characteristics of the method. The first step is to make it clear that he maintains a consistent attitude toward it throughout the works. A clue to our question is to ask what is the status of that which is defined by the method of division. For in Anal. Post., Aristotle criticises the thought of Speusippus, who is said to have given up the theory of Ideas because it is incompatible with the theory of division. We can find a debate on the status of the definiendum in the Early Academy in Plato's Theaetetus and Aristotle's Metaphysics Z15. Aristotle's conclusion is that individuals cannot be defined in terms of differences found among them and Speusippus agrees with him on this point. The fundamental opposition between these two philosophers can be seen in Aristotle's treatment of Speusippus' claim that one who defines something must know everything. The basis of the claim is that it is impossible to know things' differences from something without knowing that thing, but that without the differences one cannot know that thing-for it is the same as that from which it does not differ and different from that from which it does differ. From this claim and various evidence found in his extant fragments, we can infer that Speusippus identifies that which is defined with the bundle of differences which we attain by the method of division. Aristotle argues that this claim is simply false because things are not different from each other in virtue of every difference, and that the use of dichotomy sets us free from the requirement of omniscience. The criticism derives its force from his so-called essentialism which admits two kinds of difference, substantial and accidental. His definiendum is not the bundle of differences but the substance to which the substantial differences belong. Now we can see more clearly the contrast in their attitudes toward the consequences of adoption of the method that the species defined is the final difference and that the number of the species is equal to that of final differences. Speusippus must accept these consequences as true, while Aristotle still has a right to reject them because his definiendum is independent of the system of division. Those who define things by division have the intention of completing the system of division as a system of classification, which must be hierarchial in that it does not contain anything undefined nor any cross-classification. In Anal. Post. Aristotle concedes that if we use dichotomy, anything undefined will not remain, but he never considers the problem of cross-classification. It is treated in full in Part. Animal., which asserts the inevitability of cross-classification. Aristotle's treatment is, therefore, not inconsistent. In Anal. Post, and Metaphysics he considers the logical consequences which the method of division implies and in Part. Animal, he tries to estimate them.
