In this chapter Aristotle raises two problems about ra ody dr\G GAA ovvdedvacpéva, terms which stand for an attribute which is coupled with a subject (réde év rGde J. 18) not accidentally but in virtue of the very nature of the attribute; the attribute being xa6’ avro to the subject in the sense that it cannot be defined without either naming or defining the subject (I. 23), i.e. in the second of the senses recognized in the Postertor Analytics (73° 37-> 3). Ta cvvde- dvacpeva are in fact propria (e.g. snub) or unions of subject and proprium (e.g. snub nose), which are initially distinguished from (a) substances, (4) terms in other categories, and (c) combinations of substance and accident (e.g. white man), all of which have been shown in ch. 4 to be definable, though (a) are definable in a more proper sense than (4), and (6) than (c). But ultimately all terms in categories other than substance are shown to be in principle of the same type as ‘snub’, in that the ‘definition’ of them must be €x mpoobécews, must involve a reference to the substance to which they belong (10314 2-5). The question whether ‘ the snub’ has a definition is significant for Aristotle because mévra ra gvoixd duoiws TO oid déyovrar (E. 102534). But there is an important difference between 70 oipov and natural substances; cf.n.ad loc. The conclusion drawn from the first problem (1030 14-28) is that such terms cannot be defined, or can be defined only in a secondary sense, kadaep eipjxapev. The reference here is to # 17— 13, but Aristotle does not mean that he has already mentioned this particular sense (for the cvvdedvacpéva are a different class of terms—cf. > 20—from terms like ‘white man’, which he was there referring to), but that he has said there are secondary kinds of definition.
There cannot be a proper definition of 7a cvvdedvacpéva because the account of them must be é« mpoo@écews, and this prevents it from being a proper . definition (1029> 30). The ovrdcdvacpévov is a quality ; yet you cannot give an account of it without mentioning its subject as well as it. ‘The equal’ is ‘a quantity which is equal’; ‘the male’ is ‘an animal which is male’, Thus you define X as XY and break the rule against zpdceors.
The second problem of the chapter is stated in 1030> 28— TOSIA I:
If ‘snub nose’ = ‘hollow nose’, ‘snub’ = ‘ hollow’. But ‘snub’ is not = ‘hollow’, since ‘snub’ implies a reference to the nose while ‘ hollow’ does not. .*. ‘Snub nose’ is not = ‘hollow nose’, ,*. If we say ‘snub nose’ at all, we are saying what is = not to ‘hollow nose’ but rather to ‘ nose which is a nose-which-is-hollow’. Such terms cannot have an essence, since this would involve an in-
174 Commentary
finite regress; ‘ nose which is a nose-which-is-snub ’ will involve ‘ nose’ once more.
By ‘such terms’ (1030) 34) Aristotle seems to mean such terms as “snub’, since this is the class of terms which he is throughout the earlier part of the chapter trying to prove_indefinable (he advances to ‘snub nose’ only in 10315), But he is evidently assuming that ‘snub’ is equivalent to ‘snub nose’ (which is what is implied in saying that the account of it must be ék zpocOécews) ; for it is only by applying this equation that he reaches the term ‘ snub nose nose’ in 1. 35. He had previously (I. 33) reduced ‘snub nose’ to ‘hollow nose nose’, which leads to no infinite regress, but now, substituting for ‘snub ’ its equivalent ‘snub nose’, he gets the form ‘ sud nose nose’, and has no difficulty in showing that if the same substitution be re- peated we are landed in an infinite regress.
I take ci 8@ py l. 35 as = ef d€ Tis Adyer Oru ora Kal ev TovTOLS TO TL hv evar kat opicpos (Al. 478.15). Alternatively, we might treat dud... eivat ll. 34, 35 as parenthetical, and interpret ‘ otherwise—i.e. if it be denied that the snub-nose is a hollow-nose-nose—we shall be com- mitted to a process ad ¢nfinitum. For (since the saud is the nose that zs snub,the snub nose will be the snub-nose nose: and) the snub-nose nose will contain yet another zose (and so on ad zfinitum)’. ‘This may well be right. In any case Bz. is wrong in supposing that the introduction of fils pis oxy in 1. 35 is a mere slip.
To this ‘infinite regress’ argument for the indefinability of ‘snub’ Aristotle himself in effect supplies the answer in Soph. 7. 182° 4. ‘The snub’ = ‘a snub nose’, but it does not follow that- in ‘snub nose’ we can substitute ‘snub nose’ for ‘snub’, and so ad infinitum. For in ‘snub nose’ ‘snub’ does not mean ‘the snub’, i.e. ‘that which is snub’, but a quality of the nose (puds rodi, otov wafos), so that ‘snub nose’ is analysed not into ‘snub nose nose’ but into ‘ nose having the kind of hollowness proper to a nose ’, in which no infinite regress is involved (jor otdev drorov, <i } pis } oyun pis éorw exoved kowWdryta fuvos). Le. the Sophistice Elenchi draws the distinction which Aristotle fails to draw here between 76 cysov in the sense of ‘snub’ and 76 ody in the sense of ‘that which is snub’. The corresponding distinction between the two senses of 76 Aevkoy is drawn in 1031 23. The main upshot of the discussion of both problems is that 7d cysdy is not strictly definable because the account of it must be éx mpoabecews. The two problems bring out the two aspects in the notion of ék zpocécews. The first shows that the account of such a term must introduce something other than the term, viz. the under- lying substance (1030 23-26); the second, that it must involve a tautology (1031 4).
1030) 24. of = rovrov ov.
27. xabdmep cipyxapev, © 17-13.
10gi*1. Aristotle now concludes that in general only substance is definable. If there be a non-substance X, there is always a subject Y which it presupposes, and it could only be defined as XY, i.e. &
apoobécews. Therefore it cannot be defined at all. To this also the Sophistict Elenchi evidently provides the answer.
3. mood can hardly be right, but neither Bz.’s dpréov nor Goebel’s moAXod is convincing. ,
8-9. Aro... 7d Ti Hy etvar. I.e. either such terms can be properly defined in some way which does not involve the addition of the subject (which Aristotle does not admit that they can), or we must distinguish improper from proper definition and say they can only be defined in an improper sense of ‘ define ’.
8-9. kaddmep...7d Ti Hv etvar. The construction is due to a fusion of two possible constructions, Kafdmep eA€x On, Nexréov éort ToANaxOs AeyerOar Tov Spicpdv, &c., and KaOdmrep édéyOn modrdaxGs héyeoOar 6 opicpos, &c., ovrw viv Aexréov. For similar fusions cf. De An. 414° 22, H.A. 49817, 60025, P.A. 656° 15, L. NV. 1135> 29, Rhee. 1356% 10, 14059. Cf. also. B. 100081 n.
KaQdtep éXéx0n, 1030% 17—) 13,
Ts a thing the same as tts essence ? (ch. 6).
1031*15. Is a thing the same as its essence? This bears on the study of substance, for a thing seems to be = its substance, and its substance to be = its essence.
1g. (1) An accidental unity like ‘white man’ would seem not to be = its own essence. For else essence of man = essence of white man, since man = white man.
24. But perhaps it does not follow from white man being = essence of white man that the essence of accidental unities is the same as that of the simple terms (essence of white man = essence of man); for the extreme terms of the syllogism are not identical in the same way with the middle term. _ It might, however, seem at least to follow that the accidental extremes (e.g. essence of white and essence of musical) are the same ; but they are not.
28. (2) Is a per se term necessarily the same as its essence? Take (a) primary terms like the Ideas? If the Good Itself is to be different from the essence of good, (i) there will be substances prior to the Ideas, if essence is substance.
bg, (ii) If the Ideas are separated from their essences, (a) they will not be known, and (@) the essences will not be existent. For (a) we know a thing only when we know its essence, and
7. (f) if the essence of good is not good, the essence of being will not be, arid since all essences are on the same footing, no essence will be.
176 Commentary
11, (iii) That to which ‘being good’ will not attach (sc, the Good Itself) will not be good.
Therefore all terms (whether Ideas or not) which are self-subsistent must be the same as their essences,
(15. If the Ideas are such as the Platonists suppose, it will not be substratum that is substance, since they are substances which do not imply a substratum.)
18. (2) That a thing is the same as its essence is also clear from this, that to know a thing is to know its essence.
(22. If we consider an accidental term like ‘ the white’, the essence of white will not be the same as that which is white (the white man), but it will be the same as the quality white.)
28. (c) The absurdity of separating a thing and its essence is further seen if we put a name to each essence, for then it will have another essence of its own; it is better to recognize at once that some things = their essences.
32. The definition of a thing, also, = the definition of its essence ; for it is not per accidens that e.g. unity and its essence are one. To separate them would produce an infinite regress,
b 4. Each simple fer se term, then, is the same as its essence.
6. The sophistical objections are to be met in the same way as the question whether Socrates = being Socrates.
Aristotle’s doctrine in this chapter is that ra Aeydueva Kata ocvpPe- Byxés (i.e. terms denoting a union of a subject with an accident) are not, and ra xa attra Aeydueva (terms denoting a self-subsistent unity, i.e. either a summum genus or a species, either in the category of substance or in some other, 1031) 27, 28) are, identical with their essence. E,g. ‘to be a man’ sums up the whole substantial, per- manent nature of each individual man and is identical with each and every man; ‘to be a sitting man’ does not express the permanent nature of any man, and ‘ to be a white man’ expresses the permanent nature of some men but not of others.
Aristotle first discusses accidental terms (1031 19-28) and puts for- ward a proof of their non-identity with their essences, If
(x) a white man = the essence of white man,
then, since (2) a man = a white man,
.*. a man = the essence of white man.
Now, if (1) is true, similarly (3) the essence of man = a man,
.*. the essence of man = the essence of white man.
But this is evidently not true. Therefore a white man is not = the essence of white man.
This reductto ad absurdum, however, Aristotle points out in 1, 24, fails. It does not follow that the essence of accidental combinations.
is identical, sc. with that of the corresponding simple terms. For the extremes are not identical in the same way, sc. with the middle term. In the first syllogism the major term is absolutely identified with the middle, while the minor is identical with the middle only per accidens ; in the second syllogism the converse is true.
Aristotle next puts forward (I. 25 éAX’ tcws xrd.) an alternative reductio ad absurdum. If the fallacy of the above reductio be detected, it might at any rate seem to follow from the identification of an acci- dental term with its essence that the accidental extremes, essence of white and essence of musical, are identical. But evidently they are not. Therefore accidental terms are not identical with their essences. The argument here implied is:
The musical man = the essence of musical man. The man = the musical man. ’ The white man = the man. The essence of white man = the white man. .*. The essence of white man = the essence of musical man. .*. The essence of white = the essence of musical.
This conclusion might seem to follow, because here musical man and white man are both identical with the middle term man in the same way, i.e. per accidens. The argument is, of course, unsound ; but Aristotle does not commit himself to its accuracy—he merely says ddéevev dv cvpBaivew.