CATEGORY 223 ject and predicate. He draws attention to the fact that things are spoken of either in the connection known as the proposition, e.g., "a man runs," or apart from such con nection, e.g., "man " and "runs." He then proceeds, " Of things spoken of apart from their connection in a proposi tion (TOJV KO.TO. fj.rj8efj.iav o~vfj.irXoi<r)v Aeyo/xeVwi/), each signi fies either Substance (oicrt a), or Quantity (7rocroV),or Quality (TTOLOV), or Relation (irpos TI), or Where (i.e. Place, TTOV), or When (i.e. Time, Trore), or Position (KelaBaC), or Possession (e^av), or Action (Trotetv), or Passion (Trao-^etv). Oucrt a, the first category, is subdivided into rcpur-q ova-la, or primary substance, which is defined to be ro Se TI, the singular thing in which properties inhere, and to which predicates are at tached, and SeuTepat oixriai, genera or species which can be predicated of primary substances, and are therefore ouat a only in a secondary sense. Nevertheless, they too, after a certain fashion, signify the singular thing, ro Se TI (K., p. 3b, 12, 13). It is this doctrine of irpurr) ova-La that has raised doubts with regard to the authenticity of the KO/TT;- yopt at. But the tenfold classification, which has also been captiously objected to, is given in an acknowledged writing of Aristotle s (see Topica, i. 9, p. 103b, 20). 1 At the same time it is at least remarkable that in two places where the enumeration seems intended to be complete (Met., p. 101 7a, 25; An. Pos., i. 22, p. 83a, 21), only eight are men tioned, ex tv an d Kfltr&aL being omitted. In other passages 2 six, five, four, and three are given, frequently with some addition, such as /cat at dAAcu K. It is also to be observed that, despite of this wavering, distinct intimations are given by Aristotle that he regarded his list as complete, and he uses phrases which would seem to indicate that the divi sion had been exhaustively carried out. He admits cer tainly that some predicates which come under one category might be referred to another, but he declines to deduce all from one highest class, or to recognize any relation of sub ordination among the several classes. The full import of the categories will never be ade quately reached from the point of view taken up in the Ka-n/yopiat, which bears all the marks of an early and pre liminary study. For true understanding we must turn to the Metaphysics, where the doctrine is handled at large. The discussion of Being in that work starts with a distinc tion that at once gives us a clue. To ov is spoken of in many ways ; of these four are classified TO ov Kara o-v/x/?e/??7/cd?, TO ov <Ls aXrjOes, TO ov owa.fj.et /cat evepyeia, and TO ov Kara TO. o~^/xaTa TOJV KarrjyopLwv. It is evident from this that the categories can be regarded neither as purely logical nor as purely metaphysical elements. They indicate the general forms or ways in which Being can be predicated ; they are determinations of Being regarded as an object of thought, and consequently as matter of speech. It be comes apparent also why the analysis of the categories starts from the singular thing, for it is the primary form under which all that is becomes object of knowledge, and the other categories modify or qualify this real individual. Ilcu Ta oe TO. yiyvo/xei a VTTO re Ttvo9 yiyverai Kat CK TIVO? /cat TI. To oe ri Ae yco xaO e/cao-r^v Kan/yoptar* 77 yap ToSe r/ TTOO-OV 17 TTOIOI/ r) 7rov (Met., p. 1032a, 13-15). . . . The categories, therefore, are not logical forms but real predi cates ; they are the general modes in which Being may be expressed. The definite thing, that which comes for ward in the process from potentiality to full actuality, can only appear and be spoken of under forms of individual ity, quality, quantity, and so on. The nine later categories all denote entity in a certain imperfect fashion. The categories then are not to be regarded as heads of predicates, the framework into which predicates can be 1 Against this passage even the cross-grained Prantl can raise no objection of any moment ; see Ges. der Logtk, i. 206, n. See Bonitz, Index A ristotelicus, s.v. , and Prantl, Ges. d. Log., i. 207. thrown. They are real determinations of Being allgemevie Bestimmtheiten, as Hegel calls them. They are not summa genera of existences, still less are they to be explained as a classification of nameable things in general. The objections Mill has taken to the list are entirely irrelevant, and would only have significance if the categories were really what they are not an exhaustive division of concrete existences. Grote s view (Aristotle, i. 108) that Aristotle drew up his list by examining various popular propositions, and throw ing the different predicates into genera, " according as they stood in different logical relation to the subject," has no foundation. The relation of the predicate category to the subject is not entirely a logical one ; it is a relation of real existence, and wants the essential marks of the preposi tional form. The logical relations of TO ov are provided for otherwise than by the categories. Aristotle has given no intimation of the course of thought by which he was led to his tenfold arrangement, and it seems hopeless to discover it. Trendelenburg in various essays has worked out the idea that the root of the matter is to be found in grammatical considerations, that the categories originated from investigations into grammatical functions, and that a correspondence wall be found to obtain between categories and parts of speech. Thus, Substance corre sponds to noun substantive, Quantity and Quality to the adjective, Relation partly to the comparative degree and per haps to the preposition, When and Where to the adverbs of time and place, Action to the active, Passion tj the passive of the verb, Position (/ceicrtfat) to the intransitive verb, e^eu/ to the peculiar Greek perfect. That there should be a very close correspondence between the categories and grammatical elements is by no means surprising; that the one w r ere deduced from the other is both philosophically and historically improbable. Reference to the detailed criticisms of Trendelenburg by Ritter, Bonitz, and Zeller will be sufficient. Aristotle has also left us in doubt on another point. Why should there be only ten categories] and why should these be the ten] Kant and Hegel, it is well known, signalize as the great defect in the Aristotelian categories the want of a principle, and yet some of Aristotle s expressions would warrant the inference that he had a principle, and that he thought his arrangement exhaustive. The leading idea of all later attempts at reduction to unity of principle, the division into substance and accident, was undoubtedly not overlooked by Aristotle, and Brentano 3 has collected with great diligence passages which indicate how the complete list might have been deduced from this primary distinction. His tabular arrangements (pp. 175, 177) are particularly deserving of attention. The results, however, are hardly beyond the reach of doubt. There was no fundamental change in the doctrine of the categories from the time of Aristotle to that of Kant, and only two proposed re-classifications are of such import ance as to require notice. The Stoics adopted a fivefold arrangement of highest classes, yevtKwraTa. To ov or ri, Being, or somewhat in general, was subdivided into wroicet- fj-eva or subjects, Troia or qualities in general, which give defiuiteness to the blank subject, TTO>S l^ovra, modes which further determine the subject, and. Trpo s Tt TTWS exovra, definite relative modes. These categories are so related that each involves the existence of one higher than itself, thus there cannot be a Trpo s TI TTWS IXDV "which does not rest upon or imply a TTOJS exov, but TTWS exov is impossible without TTOIO V, which only exists in vtroKct/uiw, a form or phase of TO ov. 4 3 Brentano, Bedeviling den Seienden nach A., pp. 148-178. 4 For detailed examination of the Stoic categories, see Prantl, Ges. iL Logik, i. 428, sqq.; Zeller, PA. d. Gri., iii. 1, 82, sqq.; Trendelen-
burg, Kateg., p. 217.