1 My reconstruction of the traditional position is based on the following sources: Ross, W. D. Aristotle, 5th ed. (London: Methuen, 1949): 23–24,Google Scholar 24, n. 1 (hereafter cited as Arist.); Jones, J. R. “Are the Qualities of Particular Things Universal or Particular?” Philosophical Review 58 (1949): 152–156,CrossRefGoogle Scholar 162-163; Anscombe, G. E. M. and Geach, P. T. Three Philosophers: Aristotle, Aquinas and Frege (Ithaca: Cornell Univ. Press, 1961): 8–10;Google Scholar Allen, R. E. “Individual Properties in Aristotle's Categories,” Phronesis 14 (1969): 31–32;CrossRefGoogle Scholar Angelelli, Ignacio Studies on Gottlob Frege and Traditional Philosophy (New York: Humanities Press, 1967): 12–15.CrossRefGoogle Scholar Angelelli's account is especially important because it reflects the opinion of scholars from late antiquity to the present.

2 Owen, G. E. L. “Inherence,” Phronesis 10 (1965): 97–105.CrossRefGoogle Scholar

3 This claim is based on the following considerations. (1) Aristotle often uses the particular instances of individuative kinds, such as man, horse, bronze ball (or sphere) and bronze statue, as paradigmatic illustrations of substances (cf., e.g., Cat. 1b4-5, 27-28, 2b11-14, 3b15, Top. I 103b29-31, On the Gen. of ani. IV 767b30-34, Met. XII 1069a30-31, VII, VIII, passim). (2) Only the instances of individuative kinds would seem to meet the criterion of being “a certain this” (tode ti) required of substances (e.g., Met. VII 1029a 28). For, among other reasons, it is difficult to see how such a phrase could be used to refer to anything but a discrete particular: cf. J. A. Smith, “TODE Tl in Aristotle,” The Classical Review 35 (1921): 19. (3) Aristotle explictly denies that masses, such as earth, fire and air, are substances (Met. VII 1 040b5-10, cf. 992a6-8). On occasion he does, however, speak as if masses were substances (Top. I 1 03a 14 ff., IV 120b38, Met. Vll 1028b10-11, V 1017b10-11; compare Met. V 1016a4 with b8-9, Phys. I 185b9-11, V 227a 10 ff.). But at least on two of these occasions he is recounting the various philosophical, and even perhaps the non-philosophical, views regarding substance (Met. VII 1028b10-11, V 1017b10-11, cf. VII 1040b5 ff.). Accordingly, each time he speaks of masses as substances he might not be expressing his own Judgment. Yet even if it is not true that Aristotle always restricts substances to individuative kinds, and that at some point he also believes mass kinds should be counted as substances, he still seems to treat them as if they were individuative kinds. This is revealed, I believe, in Topics 1.7 where he discusses how water from the same spring should be regarded as the same in species. There Aristotle treats water from a particular spring as a species and all the water from that spring as the same in species. He also seems to treat the water from the spring as if it were composed of numerous amounts of water, each enclosed in some way — perhaps in buckets. According to Aristotle, these amounts of water, like any items said to be the same in species, form a “family“ and “resemble” one another (Top. I 103a18-19); thus, like particular men, they are to be regarded properly as instances of the same species and so the same in species. He goes on to make the same point about water in general. Clearly, Aristotle is treating the mass kind water as if it were an individuative kind: as if it were a kind having instances that are distinct and countable particulars capable of being grouped together through their resemblance to one another.

4 According to Ackrill, what is in a subject must be inseparable from that subject and so it must be shared by no other subject; consequently, as an unsharable item it is to be properly regarded as an unsharable particular. Since a certain white is in a subject, it must, then, be an unsharable particular: Ackrill, J. L. Aristotle's ‘Categories’ and ‘De lnterpretatione’ (Oxford: Clarendon Press, 1963): 74–75.Google Scholar

5 I agree with Owen, against Ackrill, that at Categories 2a36-b3 Aristotle maintains that universals may be in a subject: Owen, “Inherence,” pp. 100-101; cf. Jones, p. 155.

6 Owen, “Inherence,” pp. 98-99.

7 Although Owen never uses the expression “genus,” and he uses the expression “species” only once, his single use of the latter clearly reveals he holds that species are found in every category: Owen, “Inherence,” p. 99.

8 My account is based on the following sources: Johnson, W. E. Logic, Part I (Cambridge: Cambridge Univ. Press, 1921)Google Scholar: chapt. xi; Prior, Arthur N. “Determinables, Determinates and Determinants,” Part I, Mind 58 (1949): 1–20;CrossRefGoogle ScholarPubMed Searle, John R. “Determinables and the Notion of Resemblance,” Proceedings of the Aristotelian Society, Suppl. Vol. 33, (1959): 141–158.Google Scholar

9 There is general agreement concerning an early date for the Topics: Pamela M. Huby, “The Date of Aristotle's Topics and its treatment of the Theory of Ideas,” Classical Quarterly 12 (1962): 72-80; in n. 1 on p. 72 Huby lists several scholars who regard the Topics as early. Some scholars, however, have doubted the authenticity of the Categories, and have argued that it was written by one of Aristotle's students: e.g., E. Dupréel, “Aristote et le Traité des Catégories,” Archiv für Geschichte der Philosophie 22 (1909): 230-251. For excellent discussions of the various arguments against its authenticity and also excellent defences of its authenticity and early date, see: Isaac Husik, “On the Categories of Aristotle,” Philosophical Review 13 (1904): 514-528; “The Authenticity of Aristotle's Categories,” The Journal of Philosophy 36 (1939): 427-431; L. M. De Rijk, “The Authenticity of Aristotle's Categories,” Mnemosyne 4 (1951): 129-157. Ross always believed that chapters 1-9 of the Categories were authentic and early (Arist., pp. 9-10), and he was convinced by Husik's arguments, which he did not discover until well after their publication, that 10-13 also belonged to the Categories and that 14-15 were at least by Aristotle: “The Authenticity of Aristotle's Categories,” The Journal of Philosophy 36 (1939): 431-433. A few of the numerous, more recent scholars who also believe that the Categories is authentic and early include: Joseph Owens, “Aristotle on Categories,” Review of Metaphysics 14 (1960-61): 73-90; Ackrill, p. 69; G. E. L. Owen, “The Platonism of Aristotle,” Proceedings of the British Academy 50 (1965): 125-150.

10 This applies explicitly to both substantial and non-substantial items; e.g.: “Thus… ‘animal’ is the genus of ‘raven’, and so is ‘bird’. Whenever therefore we say that the raven is a bird, we also say that it is a certain kind of animal. so that both the genera are predicated of it” (Top. I 107a23-26); ”… the man who said ‘X is a man’ has also said that it is an animal and that it is animate …” (Top. II 112a 17-19); ” … whoever employs the term ‘virtue’ employs the term ‘good’, seeing that virtue is a certain kind of good …” (Top. VI 142b16-17, cf. IV 124b15-21). Translations of the passages from the Topics are from the Oxford translation.

11 “…if the genus…be known it does not follow of necessity that the species is known as well…” (Top. Vl 141b32-33, cf. 143a26-28).

12 “For the genus ought to divide the object from things in general. and the differentia from any of the things contained in the same genus” (Top. VI 140a27-29); “… for a specific differentia, if added to the genus, always makes a species…” (Top. Vl 143b8-9).

13 Several passages support this interpretation; the most important of them is found in Aristotle's discussion of the claim that the genus and differentia are prior in intelligibility to the species: “They [i.e., genus and differentia) are also more intelligible; for if the species be known, the genus and differentia must of necessity be known as well (for any one who knows what a man is knows also what ‘animal’ and ‘walking’ are), whereas if the genus or the differentia be known it does not follow of necessity that the species is known as well: thus the species is less intelligible” (Top. VI 141b29-34). It is clear from this passage that since knowledge of the genus alone or the differentia alone does not yield knowledge of the species — a combination of genus and differentia — genus and differentia must be logically independent of one another; if this were not the case, knowledge of one of them would introduce knowledge of the other and so also the species. Aristotle also suggests that the genus and differentia are logically independent when he holds that they are not to be “predicated” in name and definition of one another (Top. VI 144a28-b3; cf. IV 122b20 ff., 121a11-12, V 132b35 ff.). The genus should not be predicated of the differentia because, among other reasons, this would be to treat the differentia as a “species” falling under the genus. This would seem to mean that the differentia is logically independent of the genus, for if it were logically dependent on its genus, so that an analysis of it would introduce its genus — an analysis of biped, for instance, would introduce animal: whatever is a biped is also an animal — then it would clearly be in danger of assimilation with the class of items that are the species of the genus — biped, like man, would appear to be a kind of animal — and the distinction between species and differentia would be destroyed. Similar considerations would seem to hold in the case of the differentia's being predicated of the genus, although Aristotle does not actually maintain that if this were done, the genus would be treated as a species of the differentia. Much the same point concerning the genus-differentia relation seems to be made when Aristotle says the differentia cannot be “inside” the genus nor the genus “inside” the differentia, for this too would be to treat the differentia as a species of the genus and the genus as a species of the differentia (cf. Top. IV 122b18-20, 123a1-2). Furthermore, the mere fact that Aristotle insists so often that a definition of a species must be composed of two distinct elements, a genus and a differentia (e.g., Top. I 103b15-16, VI 141b25-27, VII 153a15-18). supports the claim for logical independence, for if the genus and differentia were not independent, one of them alone would surely be sufficient to define the species. There are, however, a few passages from the Categories and the Topics that conflict with my interpretation of Aristotle's genus-differentia relation (Cat. 1b16-20, Top. I 107b19-26, VI 144b12-30). These, I believe, can be accounted for as later additions that reflect a change of mind on Aristotle's part concerning the nature of his genus-differentia relation. For a more complete discussion of the issues raised in this footnote and the nature of Aristotle's genus-species relation, see my paper, “Aristotle and the Genus-Species Relation,” The Southern Journal of Philosophy, 18 (1980), 37-50.

14 Top. VII 153a36-b1, I 1 07b29-31, cf. Met. X 105 7b8-11; the qualities of “justice” and “injustice” are composed of the differing genera “virtue” and “vice” but of the same differentia “of the soul” (Top. VII 153b6-10); the relative “knowledge“ must have relative differentiae (Top. VI 145a13-15); the quality of “virtue” is composed of the genus “state” and the differentia “good” (Top. VI 144a9-16); “astonishment” is made up for the genus “wonderment” and the differentia “excess” while “conviction” of the genus “conception” and the differentia “vehemence” (Top. IV 126b 13-19).

15 Of course, Aristotle need not have held to the view that non-substantial items are susceptible to a genus-species analysis to the end of his life. I think it reasonably evident that in chapters four and five of Metaphysics VII Aristotle gives an analysis of non-substantial items that cannot be construed in terms of the genusspecies schema: non-substantial items are said to contain the substantial subject of which they are predicated as an “additional” element in their nature, and for that reason they are not definable or at least not in the way substantial items are definable. But this consideration need not detain me, for I am obliged only to argue that at one period Aristotle did apply a genus-species analysis to the nonsubstantial categories and that there is good reason to believe that the Categories was written in that period.

16 For a discussion of how “one in number” can be used to support the traditional position at Owen's expense, see: Matthews, Gareth B. and Cohen, S. Marc “The One and the Many,” Review of Metaphysics 21 (1968): 640–643.Google Scholar

17 When Aristotle first introduces the notion of the genus in the Topics, his example is “animal” (Top. I 101 b31). In his first discussion of the nature of the genus he uses “animal” as his example of a genus and “man” and “ox” to illustrate the species (Top. I 102a31 ff.). In a paragraph summarizing what Aristotle regards as some basic distinctions between genus and differentia, “animal” is his example of a genus and “man” his example of a species (Top. IV 128a 20 ff.). Balme, D. M. has pointed out that in the logical works Aristotle's “stock” examples of genus and species are animal and man: “Genos and Eidos in Aristotle's Biology,“ Classical Quarterly 12 (1962): 84;Google Scholar this too suggests that the species and genera of kinds were his paradigmatic species and genera. Aristotle's genus-species relation was obviously developed in terms of the Platonic method of division, since the relationships uncovered by division are similar to genus-species relationships. Plato applies division indifferently to kinds and characteristics (for characteristics: Sophist and Statesman, passim; for kinds: Statesman 264A-265D, 279C ff., 287E-291C, Sophist 220A), and although in the dialogues characteristics receive far more attention from him than do kinds, and so it would appear that he developed division through an analysis of characteristics, there is reason to believe he might have developed division through an analysis of kinds, as one would have expected. For Plato had a serious interest in the classification of natural kinds. A fragment of the Middle Comedy poet Epic rates depicts Plato instructing young boys in the classification of botanical kinds (Athenaeus 2.59c-f), and this activity would probably not have been selected for ridicule unless it were well known that Plato was interested in such classifications: cf. Balme, p. 81, Guthrie, W. K. C. A History of Greek Philosophy, 4 vols. (Cambridge: Cambridge Univ. Press, 1975), 4: 22–23,Google Scholar Field, G. C. Plato and his Contemporaries: A Study of Fourth-century Life and Thought, 3rd ed. (London: Methuen, 1967): 38.Google Scholar Moreover, Speusippus’ clear interest in the classification of natural kinds, which is shown by the fragments from his Similars, reveals at least that there was interest in such a program in the Academy (e.g., Athenaeus 3.86c, 3.105b).

18 I wish to express my gratitude to Alexander Mourelatos, Kenneth Winkler, Fred Miller and an anonymous referee of the Canadian Journal of Philosophy for their helpful criticisms of earlier versions of this paper. I presented a version of this paper at the 1978 Western Division Meetings of the American Philosophical Association.