ARISTOTLE 515 latter that of realism. If Aristotle wrote the Categories long prior to the Metaphysics (which, if he wrote the book at all, he must have done), then we must suppose that at the outset of his independent career as a philosopher he began with an exteme reaction against the realism of Plato, and that in later life he returned from this to approxima tion towards Plato s views. The other hypothesis possible is, that a bias towards physical research and experiment and the collection of facts naturally led the Peripatetic school in the direction of nominalism, till at last some member of that school gave expression to this tendency by writing this little treatise called the Categories, derived, indeed, mainly from Aristotle s doctrines and teaching, but laying it down far more dogmatically than he woidd have done that the concrete individual is the unit of knowledge. 1 Considerations of style are insufficient to enable us to pro nounce in favour of one hypothesis or the other. For the school of Aristotle copied, and obtained a close resemblance to, their master s style. But whether this treatise was written by Aristotle or not, it has had a great influence. It led the world not only to think that Aristotle was a decided nominalist, but also that he classified existence under ten "categories" or summa genera. But this doctrine would hardly have been gathered from the undoubted writings of Aristotle. In his logical researches he natur ally much busied himself with the different relations which the predicate of a sentence can bear to the subject. And in the earliest of his extant works (Topics, i. 9), he enumerates " the classes of predications " (TO. yevrj TWV Karrjyopiwv) as ten, 2 and gives the same list as that given in the Categories; but the object of this enumeration is merely logical, in order to show what must be meant by the word " same " when it is predicated of any subject. Elsewhere, Aristotle does not adhere to the number ten ; he mentions 3 in one place eight, in another six, in another five, in another four, and very often three, Ka-nryopia.1, or modes of predication. There is no trace of his mapping out to himself the " Cosmos " under the divisions of ten or any other number of categories. In De Anima, i. 1, 7, he says that " it will be necessary, in discussing the soul, to define which of the categories it belongs to, and again whether it is a potentiality or an actuality." But, having said this, Aristotle does not further advert to the cate gories, while he obtains his whole definition of the soul by considering it as an " actuality." In fact, " Potentiality and Actuality," or " Matter, Form, and Deprivation," were the ontological " categories " of Aristotle, far rather than that logical list of ten kinds of predication, on which mediaeval and modern thinkers have laid so much stress. That they have done so is perhaps mainly due to this little work called the Categories, which (whether it was an early production of Aristotle himself, or was the compilation of some peripatetic follower) has had undue prominence given to it in relation to Aristotle s system. It has caused Aristotle to be misunderstood, severely criticised, and sneered at. At the same time it has given an impulse to 1 In De An., i. 1, 7, there is an apparent assertion of Nominalism rb 5e (<fov rb Ka66ou tfrot oi>6tv t<mv 4) vy-rtpov), but Torstrik points out (Ar. De An., p. 113) that Aristotle is here referring to the views of others, not stating his own. 9 Substance, quantity, quality, relation, place, time, posture (Ke?<r0cu), habit (or dress, x "), action, passion. Such a list would form a strange classification of all things in the universe. Some of these categories have an easily traceable affinity with the parts of speech, thus showing the relation between logic and grammar. But this is not their only source. As the individual man may sometimes be the subject of predication, his "dress" and "posture" were admitted among the classes of categories. Habit, from the Latinised form of
- X<iv, survives in modern languages, e.g., to be " habited" is to be
dressed, a lady s riding-" habit," habit (in French) = a coat. 8 See Bonitz s Index to Aristotle, in the Berlin edition (1870), sub vocc KUTTTyopiai, where the references are given. philosophers, from the Stoics 4 to Kant and J. S. Mill, to endeavour to frame an ultimate classification of all that exists or can be thought. The treatise On Interpretation (i.e., the expression of the thoughts in language) was pronounced spurious by Andronicus Rhodius. This we learn from some interesting discussions given by Brandis in his Scholia to Aristotle, p. 97. The opposite view seems to have been taken by Am- monius, Alexander of Aphrodisias, and Boethius. The arguments, however, both pro and con are inconclusive. Whether this treatise was by Aristotle or not, it contains a very full statement of the Peripatetic logic, so far as the Proposition, with its various characteristics, is con cerned. It quotes the treatise On Soul? and, therefore, was written later than Aristotle s undoubted logical works, more probably by one of his school than by himself. It is the source of much of the matter of the elementary logic of modern times, and contains many distinctions, at one time novel, but essential to clearness of thought, on Affirmation and Negation, the Different "Ways in which the Negative Particle may be used, Contrary and Contradictory Opposition, the Truth or Falsehood of Pro positions, Modality of Assertion, &c. Grote s 6 account of the contents of this work has opened it for the English reader. We now come to the undoubted works of Aristotle. These, as before said, were probably all actually written during his last thirteen years, but they must have, to a great extent, been prepared during the previous course of his life, during which he had thought out the divisions, the method, and the terminology of philosophy and science. The order of composition of these works, so far as it can be determined at all, must be determined by internal evidence. This internal evidence does not consist merely in references from one book to another (for these are not always re liable in some cases they are almost certainly interpolated), but still more in comparison of the thought in different books, and the various degrees of maturity exhibited by the same conception occurring in different books. For instance, in the first chapter of the Analytics the Topics are referred to ; therefore, either the Topics were written first, or else this reference is spurious. But the doctrine of the syllogism is worked out with far more precision in the Analytics than in the Topics, therefore the former hypothesis must be accepted. A similar combination of verbal and real internal evidence would seem to show 7 that the Topics (with the exception of the eighth book) were first written of all the extant works of Aristotle, next the Analytics (Prior and Posterior), next the eighth book of the Topics, next the Rhetoric, and then the Sophistical Refutations; and the same canon of criticism would lead us to believe that Aristotle next in order wrote the Nicomachean Ethics, and then (perhaps after an interval) the Politics and the treatise on Poetry. The above order of books may be considered as established with tolerable certainty. But the reasons seem rather far fetched which induce Valentine Rose 8 to lay down that Aristotle, in his 55th year, com menced a second series of writings with the Physical Lectures, which were followed by the works On the Heavens, On Generation and Destruction, and the Meteorologies ; and afterwards a third series, with the Researches about Animals, followed by the work On Soul, and his other psycholo- 4 See Grote s Aristotle, vol. i. p. 144. Trendelenburg, Kategorien- lehre. 5 De Interpret., i. 3. 9 Ar., vol. i. pp. 155-199. 7 See Aristotle on Fallacies; or, tht Sophistici Elenehi, with a Translation and Notes, by E. Poste (London, 1866), page 103, note 4. 8 Arist. Pseudepigraph. p. 3 ; cf. De Ar. Lib. Ord. el And., p.
204, sqq.