8. The editions have a comma or colon after évuTapyougay, but Kai kara ToUTOV TOY Tpdrov appears to go closely with évurapyovady and to mean kal évreAexela evurapyovody. Possibly we should print a colon after tpdrov and read 6 with T.
1O-II. ta yap... movet, ‘for he identifies substances with the atoms’, from whose atomic nature it follows that one atom cannot con- tain two atoms. Cf. De Caelo 30326, De Gen. et Corr. 325% 35.
12, 6 dpiOpds odvOeors povddwy. This is practically the same as the earliest recorded Greek definition of number, povédwv-cvornua, which Thales is said to have borrowed from the Egyptians (lambl. 2% Necom. Ar. Introd. p. 10. 8). Cf. A. 10208 13 n.
15-16, pyre... onpatver, cf. 1038 23-29 ; 16-17. pyr é§ odcidv .. . otvGetoy, cf. 1039% 3-11.
1g. éhexOy mada, 10319 12.
22. tav totepov, Z. 15, H. 6. Aristotle is not very successful in solving the problem.
- The Ideas are not substance (ch. 14).
1039" 24. Observe the consequences for those who say the Ideas are separately existing substances, and at the same time resolve the species into genus and differentiae. ‘Animal’ will be either the same
Bq
202 Commentary
or different in ‘man’ and in ‘ horse ’—i.e. numerically ; in definition it is clearly the same. If man is a separate ‘this’, animal must be so too.
33. (1) If ‘animal’ be supposed the same in horse and in man, (a) how can it be the same in things that exist.apart? Will it not be divided from itself?
ba. Further, (2) if it shares in ‘ two-footed’ and in ‘many-footed’, it, though one individual, will have contrary attributes; if it does not, in what sense can it de two-footed? It is absurd to say ‘ by com- position ’, i.e. by contact or by mixture.
7. (2) If ‘animal’ is different in each species, (a) there will be practically an infinity of things whose substance is ‘animal’. Further, (2) each of several things will be ‘animal itself’. For (i) the ‘ animal ’ in each species will be the swds/ance of that species, for it is that and not anything else that each species is called after ; otherwise that other would be the genus of man ; and (ii) all the elements in the essence of man will be /deas ; and therefore, since what is the substance of one thing cannot be the Idea of another, the ‘animal’ in each species of animal will be ‘animal itself’.
14. Further, (c) from what will the ‘animal’ in each species be derived? how can it be derived from ‘animal itself’? How can the ‘animal’, whose essence is just to be animal, exist apart from ‘animal itself’?
16. (3) These and even greater difficulties arise if we consider the relation of Ideas to senszble things. Evidently, then, there are not Forms of sensible things such as some suppose.
1039? 26- 19 is very similar to P]. Parm. 131 a—k, and in particular the language in 1 recalls that in 131 B ev dpa dv Kal tabrov év roAXors Kal xwpls ovow ddov dpa évérrat, Kat otTws aiTd adTrod ywpls av ely. Siebeck thinks this confirms his notion that the Parmenides is directed against the JZe/aphysics, but more probably Aristotle is pressing the difficulties raised by Parmenides in the dialogue,
33. dote kai td LGov might easily have been dispensed with, but is read by Asclepius as well as by the manuscripts, and need not be excised as Christ proposed.
ei péev obv 75 adtd. The other alternative comes in ? 7.
by-2. kal 8a ti ...todro; ‘If the Idea is present in separate things, does not this amount to separating it from itself?’
7-8. obxoiv ... &vOpwros. The argument, as ér in |. 9 indicates, is meant to be complete in itself. The conclusion dreipa as ézos eimely éorar dv 7 ovoia EGov is absurd, because things whose sub- stance is one are one (1038? 14).
Gmeipa Ws Eos eimety is an exaggeration, since for Aristotle the number of species in a genus is limited.
9-14. The argument is very obscure, but may perhaps be expanded as follows: ‘ Further, each of several things will be the Idea of animal. For (i) the “‘ animal” in each species of animal will be the substance of that species ; for it is after that and nothing else that the species is called (i.e. when we want to say what e.g. man essentially is, we say he is an animal, not an anything else); else that other would be the generic element in man. And (ii) each element in man, and therefore among others the element ‘‘ animal ”, is on the Platonic theory an Idea, We may infer that the “animal” in man is not the Zdea of one thing and the swdstance of another. Therefore the “ animal” in the various species of animal, which as we have seen is the suds/ance of each of these species, will be the dea of animal.’
For the force of kar’ dAAo déyerau 1. 10 cf. A. 987" 9.
14-16. ém. ... {dov; ‘Further, from what does this “‘ animal ” in each species spring? How is it derived from animal-itself? Or, if not derived from it, how can this ‘‘ animal”, whose very substance is animality, exist apart from animal-itself (the Idea of animal) ?’
15. Tas é§ adtod Laou; seems to be correctly explained by Alexander, TOS €k TOD TOLOVTOV adiToLwou oral TO év TG adToavOpaTw airoLdov ;
The reading and punctuation suggested by Bz. and given in the text is a great improvement on the traditional form ro fdov 6 ovicia, TovTo avTo wap avro, and derives some support from Al. 529. 17.
16, 17. The difficulties raised in *26->16 have referred to the relation of genus to species; Aristotle now says that even greater difficulties for the ideal theory attend the relation of species to sensible individuals.
Individuals, and therefore Ideas, are indefinable (ch. 15).
1039” 20. Concrete things are generable and therefore destructible ; forms are never in course of being destroyed any more than of being created; they are or are not, without generation or destruction.
27. This is why (1) particular sensible substances are not subjects of definition or of demonstration, viz. because they have matter capable of being and of not being. If knowledge can never become ignorance, there cannot be demonstration or definition of the contingent, but only opinion,
1040*2. For perishable things are no longer known when they have passed out of our perception, and, though the formula in the soul remains the same, there is then no longer definition or demonstration ; thus it is always possible to overthrow a definition of a particular, for it cannot really be defined.
214 Commentary
8. (2) Therefore an Idea, also, cannot be defined, for (a) it is said to be a separate particular. Its definition would have to consist of words, but new-coined words would not be understood, while old ones are common to all things of a class,
14. If it be said that, while the marks used in the definition separately belong to many things, together they may belong only to one, we answer (i) that they will also belong to both elements in the definition, e. g. ‘two-footed animal’ to ‘ animal ’ and to ‘ the two-footed ’” (where the elements are eternal, this mws¢ be so, since they are prior to the compound and, further, separately existent, because, if ‘man’ is to exist apart from its particulars, so must ‘animal’, and if ‘animal’, then also ‘ the two-footed’) ; further, (ii) that the elements are prior to the whole, and therefore are not removed when the whole is removed.
22. (2) Again, if the elements of Ideas are Ideas (as they must be, elements being simpler than the compounds), the elements, e.g. ‘animal’ and ‘two-footed’, will be predicable of many subjects. Otherwise how could they be recognized? All Ideas are thought to be shared in by a plurality of subjects.
27. In the case of eternal things, especially if they are unique, like the sun, people do not realize the impossibility of definition. Definition may err not only by adding irrelevant attributes such as ‘ going round the earth’ (the sun would still be a sun if it did not do this, for ‘sun’ means a certain substance) ;
83. but also by naming attributes which can belong to another subject. Such a definition will be common, while the sun was supposed to be an individual—Why does not some Platonist produce the definition of an Idea? The truth of our remarks would become apparent.
Bz. thinks that as in chs. 13, 14 Aristotle has proved that universals are not substances, he now proves that individuals are not substances. But (1) that is not Aristotle’s view, and (2) Bz. can arrive at it only by piecing together the conclusion of this chapter, that individuals are in- definable, with the conclusion of ch. 4, that substances are the subject of definition. If Aristotle had meant his readers to draw Bz.’s con- clusion, he could hardly have failed to call attention to it.
The interest of the chapter is simply in the problem of definition. Aristotle has asked in ch. 13 (1039 14-22) whether anything can be defined, and begins his answer here by pointing out that at any rate individuals cannot.
While chs. ro-14 have continued the line of thought of chs. 1-6 and contained no reference to chs, 7-9, ch. 15 has such a reference (1039? 26). But it also continues the line of thought of chs, 1-6, 10-
14, and in particular the problem of the possibility of definition, which was raised in ch. 13.
1039> 22. 6 dyos ddus, ‘the definition in its full extent’, not bound up with any particular matter. Bz.’s suspicion of éAws seems unjusti- fied; in the /zdex he quotes a phrase which is a good parallel, TO 6& dd.aiperov drAws (De Gen. ef Corr. 326% 28). Cf. also 1029" 6, 1033) 26.
24. ottws date POeipecOar, ‘in the sense that it is ever in course of being destroyed’. Cf. 1033> 5-6n., E. 10278 29n,
26, SéSetxrat, in ch. 8.
28. tav Kal” éxaota, cf. B. 999% 26n.
oUTe Gmrdderéis oti, ‘nor can there be demonstration adoué such substances ’, i. e. demonstration of their attributes.
30-31. 815 P0apra ... atTav. Aristotle shows no knowledge here of the distinction drawn in A. 1069? 30 between two kinds of sensible substance, the eternal (the heavenly bodies) and the perishable.
7a Kal Exacta, cf. B. 999% 26 n.
ZI. Kal 6 dpiopds emrornporiKdy, sc. ‘and therefore tOv dvayxaiwy’.
1.040% 2-4. Gdnhd te yap... . dméAOn, cf. 1036 6.
6. Bz. interprets tav mpds Opoy as meaning guod ad definitionem atlinet. Parallels to such a use of ra pds dpov may be found in Top. 102» 27, 120? 13 (éo7 d€ Tatra oro.xela THY mpds Tovs dpovs), but the genitive is very difficult. It seems better to suppose with Alexander that the phrase means ray pds dpov twos tpaypatevonevwv. But corruption may be suspected.
8. tov... Kal exacroy 7 idda, ‘the Idea is an individual’.
II, maow, ‘to all the members of the class denoted by the name’.
14-15. ci 8€ Tis... Umdpxew. The definition of an Idea might be defended against the attack just made, on the ground that, though each of the marks separately belongs to more than one subject, taken together they belong only to the Idea in question. Aristotle proceeds in 1. 15 to object to this defence, but, as Bz. observes, he himself gives a similar account of definition in An. Post. 96°33 éxacrov pev ei mAciov brdpéer, Gravta S& pH ert wAdov. The objection, then, must derive its force, if it has any, from the gecu/zar nature of the Ideas—from the fact that they are individuals and exist separately (I. 9).
The first objection is stated in I]. 15-17. ‘ Two-footed animal’ will belong both to ‘animal’ and to ‘ the two-footed’. Bz. thinks that dzdp- xew is here used (by an illegitimate change of meaning) not in the sense of ‘ be predicable of’, which it has in ], 15, but in the sense of ‘ be con- tained in the extension of’; but for this sense he offers no parallel. imdpxew has, in fact, its ordinary sense. Aristotle means that ‘two- footed animal’ is predicable (1) of ‘animal’ (not universally, but in certain cases), and (2) of ‘ the two-footed’ (universally, since two- footedness belongs only to animals, as a differentia should belong only to its genus, Zop. 143% 30). This objection would not apply to an ordinary definition, since that does not imply the existence of genus and differentia apart from the species. But according to the Platonists
216 Commentary
‘animal’, ‘the two-footed ’, and ‘man’ are separately existing Ideas ; and ‘two-footed animal’ is predicable of all three, and therefore not a proper definition of man.