28. For a similar use of tpdétoy cf. H. 1045% 26, De Jnt. 184 19. GANG phy ob8e Tov Kal’ aitd Neyoudvwy obS€ todro, Aristotle here anticipates an objection. Some one may say ‘it is no use asking what 70 ivariw etvaris. The thing denoted is not ka? atrd Xeydpevov—white is not xa aérd to. man—and therefore cannot be the essence of ‘white man’. The objection assumes, arbitrarily enough, that only what is internally xa@’ aird can be a xa aito predicate to something else. But Aristotle takes it seriously and shows that ‘white man’ may have something said of it which is not od xa atro in either of the senses in which a definition should not be od xa@ aire to its subject.
go. 7 8 ov. ‘ The other errs not by addition’, which is idiomati- cally equivalent to saying that it errs by omission; cf. Bz. /ndex 539% 14-47.
31-33. TO atts GANw TpocketcOat . . . TO GAO adt@, The anti- thesis is misleadingly stated. Really the error arises in one case because in the definition the definiendum is added to something else ; in the other because in the definiendum something else is conjoined with what is stated in the definition. apooxetofa: in 1], 31 refers, as mpoobécews does in |. 30, to the addition in ¢hought of a qualification ; the apooxetcOar which has to be understood in 1. 33 refers to the conjunction of a qualification in fact. Alexander’s understanding of detv mpookeicOa: in this line, which would remove the ambiguity, is indefensible.
34. Todefine is not dpiZew but dpiLerOar (Bz. Index 524? 8). JA?’s reading épiforto ipdriov must therefore be right. Alexander (470. 18) read either this or 6piZouro 76 iudrcov (so E?).
10302 1-2. It seems necessary to insert 76 before tt Hv etvar. ‘ But its essence is not to be white’. For the omission of 76 before Aevkd ewou cf. A, 101456 n., An. Pr. 67412, 13, Pl. Crat. 385 B2, 10, 408 A 5, Theaet. 176 B 3.
2. In pointing out the relevant senses of od xa airo Aristotle has
170 Commentary
The editions before Bz. place the full stop after the second instead of the first efvac in this line; the sense makes the change quite necessary.
3. Sep ydp ti, the reading of Ab yp. E, gives a good sense, and it is not necessary to read with Bz. drep yap (rdde) 7. For tu = 100e te cf. 10292 20, 24.
Otay 8 GAO kat GAou Adyntat. It might be said that a term like ‘man’, of which Aristotle thinks there is an essence, implies the predication of one term of another (‘rational’ of ‘animal’). Aristotle would reply that these are not dAda to one another since ‘rational’ exists only as-an attribute of ‘animal’ and has no separate existence. Cf. Z. 12, H. 6. On the other hand a man need not be white, nor a white thing a man.
6. dow 6 Aéeyos eottv dptopds. Any dvopa (like iparcov) can have a Néyos or combination of words, and any Adyos (like ‘white man’) can have a Adyos axpyBéorepos, which means the same as it; but that Adyos is not a definition, unless that which is thus explained is a mpOrov, a primary or simple real which is not the union of a subject with an irrelevant attribute.
g. A» and apparently Alexander (471. 28, 29) have Aéyo where our other authorities have A¢yw tai7év. The shorter reading is supported by An. Post. 92> 31 in yap dy dvopa O€cbar drowody Adyw.
date kat W IAuds dptopes eorat, i.e. the poem, being a combination of words, would be a definition of the word ‘Iliad’. The Iliad is a typical instance of the things that are only cwdécpue ev, Z. 1030? 9, H. 10452 13, An. Post. 93> 36, Poet. 14574 29.
11-14. Only yévous «dn have an essence, i.e. genuine species as opposed both to Platonic ei8y (cf. A. 9914 31 n.) and to collocations of substance + accident like ‘white man’. Platonic Forms xara petoynv Aéyerar—‘ participation ’” is one of Plato’s favourite ways of expressing the relation of particulars to the Form. Again, the Platonic Form need not express the innermost nature of its particulars, but is any universal under which they happen to fall, and may be a mere zaos or ovpBeBykds of them. In ‘white man’ the relation between the two elements is of the same external and accidental character. On the other hand, the genus which is implied in the name of a species does not ‘share’ in the differentia ; the differentia is not a mere ‘affection’ or ‘concomitant’ of it, but is its proper differentia. That xara peroyny is to be interpreted as above seems to be shown by 1037) 18, where
also Aristotle is speaking of the unity of a species as opposed to a term like ‘ white man’, and says évradda 8’ (in the species) od peréxet Oarépov Odrepov' 7d yap yevos ob SoKet peréxew Tov Suahopdv. No im- portant distinction seems to be intended here between xara petoxyv, kara. rdos, and ds cvpB_eBnkds.
22. 7TH pév, to substance.
25. oytkds, i.e. that which is not cannot be said to ‘be’ in the plain sense of that word, but speaking Aoyixés, with reference to linguistic usage (oi Adyor—cf. ras de? A€yew, 1. 27), we may say that it is, since we can say it zs non-existent. Cf. A. 1069%21. For Aoyuds cf. 1029) 13.n. So too a qualitative term is never the answer to the question ré éory in its primary sense, i.e. what is such and such a thing, but is an answer to the question what is such and such a quality. ‘A colour’ is the answer to the question ‘ what is white?’. Cf. 1028b 1, B. 996% 18-22, Zop. 103” 27 ff.
26. tues is perhaps a reference to Plato ; cf. Soph. 237, 256 ff.
27-28, Sel pev.. . exer, ie. such linguistic inquiries as Aristotle has been conducting since 1029) 13 (Aoyixds) and has referred to in ], 25 are important enough, but it is still more important to study the facts themselves. The mode of using the term ri jv eva, being now plain, Aristotle proceeds to draw the conclusion about the facts with regard to the ri jv etvas, viz. that non-substances have in a qualified sense a ri Hv etvat, Not an essence simply, but ‘the essence of a quality ’, &c. (I. 31).
29. dpotws is explained by domep Kal ro ti €or, 1. 30. The distinction between 7é éor: and ri Hv etvau is that while the former may stand for either a partial or a complete answer to the question ‘ what isso and ¢o?’, i.e, either for the genus or for the genus + the differentia, the latter always means the complete answer. Thus, while every ti jv evar iS a ri éott, the converse is not true. 7é ésru is sometimes distinguished from ré jv eivas (An. Post. 82» 31, 92% 7, De An. 430° 28), sometimes used in the same sense (Aw. Post, g1* 1, 93%29, A. 987% 20, 988b 29). Cf. Bz. Lndex 763° 47—764* 16.
32-34. Set yap... émorntdv. We must say that substance and non-substance ‘are’ in different senses of that term, or we must add a qualification in the latter case, saying that non-substance ‘is’ in a qualified sense (as we say the unknowable can be known—to be un- knowable, cf. “het. 1402°6), and remove the qualification in the former, saying that substance ‘is’ simply. So Alexander. But it is equally likely that zpooriévras and ddaipotvras both refer to the things which are not dvra in the full sense. If we say that they are évra we add a qualification to, and deduct from the full meaning of, dv.
35. doattws = cvvwvipus, ‘in the same sense’.
mpos TO adTd .. . kat éy, i.e. to the surgical art (mpds tarpuxyy, T. 1003) 1),
Terms which are neither épevvpa nor cvvdvupa are, as here, said to be zpos & in T. 1003* 33, K. 10618 11. In £. NV. 1096 27 Aristotle
172 Commentary
offers ad’ évds and xa7’ dévadoyiav as alternatives to mpds ev, and Seems to prefer, at least in the case of the various goods, kar dvadoyiav.
b 4. émorépws, i.e. whether we say that various categories ‘are’ in the same sense but with qualifications or deductions (@ 33) or that they ‘are ’,not in the same sense (xa év) but mpés & (*® 35-3). drrorépws can hardly refer to the alternatives stated in~* 32, 33, for Aristotle would not regard it as immaterial whether dv is a mere épdvupor, applied to the various ‘ beings’ by a mere play upon words.
8-10. toito 8. . . év, ‘i.e. with a Adyos which is the Adyos of some- thing that is one, not by continuity .. ., but in one of the senses of “one” which answer to the essential senses of “being” ’. écax@s héyerau 74 év is to be interpreted in the light of ll. ro-12. For the correspondence of the senses of ‘one’ to the categories cf. T. 1003 33, A. 1018 35, I. 1053 25, De An. 412» 8.
12. Since unity in some sense can be found in any of the categories, ‘white man’ (which is a union of terms from two categories) has a certain unity and can have a definition, though not in the same sense in which ‘white’ has one, while this again has a definition in a different sense from that in which a substance has one.
It is clear that the chapter does not do what it set out to do, viz. to discover whether essence is substance. It only tells us that it is substances alone that in the primary sense Aave essence, But this may be found to be a step towards the answering of the original question.
Flave coupled terms an essence or definition ? (ch. 5).
10g0> 14. With regard to (d) a coupled term like ‘ snubness ’, which per se belongs to the nose, not as ‘white’ belongs to Callias or to man but as ‘male’ belongs to animal, (i) how can such a term be defined, if definition must not involve an improper addition ?
23. Per se attributes are those in whose definition the account or the name of the subject must be present. If they have an essence, it must be in a sense different from the strict sense.
28. (ii) There is another difficulty about these. If snub nose = hollow nose, snub = hollow; but if this cannot be so because we cannot say ‘snub’ without implying the nose, ‘snub nose’ is tauto- logical, being = ‘hollow nose nose’. Thus, if such terms had an essence, an infinite regress would be involved.
10gI*z. Only substance, then, is definable, for definition of any- thing else would involve an improper addition; ‘odd’ cannot be defined apart from number.
5. Therefore (¢) if the terms are coupled, as in ‘odd number’,
such combinations cannot be defined, any more than terms like ‘odd’, or can be defined only in another sense of ‘ definition ’, 11, Substance, then, is the only or the primary subject of definition,