Theory and History of Ontology (www.ontology.co)by Raul Corazzon | e-mail: rc@ontology.co

Semantics and Ontological Analysis in Aristotle's Categories

Contents of this Section

Aristotle's Categories

Introduction

"The Categories, ascribed to Aristotle, has played a unique role in our tradition. It is the only philosophical treatise that has been the object of scholarly and philosophical attention continuously since the first century B.C., when people first began writing commentaries on classical philosophical texts. From early late antiquity until the early modern period, one would begin the study of Aristotle and the study of philosophy quite generally with the Categories and Porphyry's Isagoge. For several centuries, these two treaties, and the De Interpretatione, formed the core of the philosophical corpus which was still being seriously studied. Thus, it is hardly surprising that our received view of Aristotle -- whether we are aware of this in all its details or not -- was colored substantially by the Categories.

Already in late antiquity, however, doubts were raised about its authenticity, (1) though we know of no ancient scholar who, on the basis of such doubts, declared the treatise to be spurious. On the contrary, Ammonius claims that everyone agreed that it was authentic. (2) The writers of the Middle Ages and the scholastics of the early modern period seem to have had no doubt about the authenticity of the treatise; (3) presumably, they were relying mainly on the authority of Boethius. (4) It is tempting to suppose that this acceptance of the treatise by the scholastics is precisely what led Renaissance scholars like Luis Vives (5) and Francesco Patrizi (6) to raise doubts about this very foundation of both scholasticism and traditional logic, though they did not attempt to provide any detailed arguments for their conclusion. It remained for the nineteenth and twentieth centuries to examine the Categories critically with the aid of the new philology. And soon enough, there was an impressive roster of those staunchly maintaining that the treatise was not genuine. (7) Even H. Bonitz considered it to be of doubtful authenticity. (8) During the present century, opinion has again shifted in favor of the view that it is a genuine work of Aristotle's, though, to be sure, the doubts have not been entirely silenced. I. M. Bochenski, writing in 1947, thought the treatise of doubtful authenticity; (9) and in 1949, S. Mansion tried to argue against its authenticity. (10) Doubts especially about the second part, the so-called Postpraedicamenta, have never really ceased. (11)

Given the enormous influence this treatise has had on our view of Aristotle and on our interpretation of his writings, it seems extremely important to me to try, as far as possible, to lay these doubts to rest. Yet, I hope this investigation will also be of interest to those already firmly convinced that the Categories is a genuine work of Aristotle's; for it raises questions that interpreters of the treatise, in general, do not address and whose answers might well alter the standard view of this text.

The question of authenticity, however, turns out to be crucially linked to the question of unity. Given that it seems highly questionable whether the Postpraedicamenta were originally part of the treatise or were appended by a later editor, (12) it might seem as if the question regarding the authenticity of the treatise needs to be asked as two questions, viz., questions regarding the authenticity of the first and second part individually. Many authors have indeed taken this for granted and have thus assumed that the first part was authentic, the second either probably or certainly not. (13) Since, however, interest traditionally has focused almost exclusively on the first part of this treatise, we also find the tendency to regard the question of authenticity as primarily the question of the authenticity of the first part and so to leave the question of unity and the problem of the authenticity of the second part to more or less take care of themselves. Buhle already exhibits this tendency characteristic of many modern interpreters.(14) After having called attention to the apparent lack of connection between the Postpraedicamenta and the Praedicamenta and after briefly remarking (without providing any specifics) that some things in the Postpraedicamenta do not mesh well with other aspects of Aristotle's thought, he writes: "sed fac esse postpraedicamenta spuria, non idem tamen de Categoriis statuendum est." It is obvious -- as long as the authenticity of the first part is secure, it does not much matter to Buhle whether or not one considers the second part genuine. Such an approach, however, is methodologically highly suspect because the questions of unity and of authenticity cannot be separated without both prejudging the issue of unity and presupposing a certain interpretation of the Categories, especially since the lack of unity itself has been taken as providing strong prima fade grounds for judging the treatise to be spurious. (15) Therefore, in what follows, I will pay particular attention to the question of unity. The dangerous tendency to consider this treatise almost exclusively with reference to the first part and thus to jeopardize the status of the second part is, of course, reinforced considerably by the title. Hence. I will also discuss the title in connection with the question of unity." (pp. 11-12)

Notes

(1) Olymp., Prol. 22, 38ff.; Schol. 33a 28ff.; Brandis.

(2) Ammon., In Cat. 13, 25.

(3) The question of authenticity is either not discussed at all (cf. Ockham, Expositio aurea, Bologna 1469, f. gii) or discussed only very superficially and mechanically (cf. De Soto, Absolutissima commentaria, Venice 1574, 247ff.; Complutenses, Disputationes in Arist. dialecticam, Leiden 1668, 160; Gennadios, œuvres, VII, 119, 9, Paris 1936).

(4) Cf. Peter Abailard, Logica Ingredientibus, 116; Conimbricenses, In universam dialecticam, Cologne 1607, c. 297.

(5) De causis corrupt., art. 99 (according to Fabricius, Bibl. Gr., vol. II, 109).

(6) Discuss. Peripat., vol. I, Basle 1581, 20.

(7) C. Prantl, Geschichte der Logik, 1, 90; C. Prantl, in: Zeitschrift für die Altherthumswissenschaft IV (1846) 641-652; L. Spengel, in: Gelehrte Anzeigen (Munchen), 1845, c. 33-56; V. Rose, De Aristotelis librorum ord., 234ff.; A. Gercke, in: Arch. f G. d. Ph. 4 (1891) 424-441; E. Dupréel, in: Arch. f. G. d. Ph. 22 (1909) 230-251.

(8) Über die Kategorien des Aristoteles, in: Sitzungsber. Wien 1853, 593.

(9) La logique de Théophraste, 32.

(10) "La doctrine aristotélicienne de la substance et le Traité des Catégories" in: Proc. 10th Internat. Congr. of Philosophy, Amsterdam 1949, 1097-1100; cf. also her earlier paper, "La première doctrine de la substance: la substance selon Aristote" in: Rev. Philos. de Louvain 44 (1946) 349- 360.

(11) I. Düring says only that the authenticity of the Postpraedicamenta is likely (Aristoteles, 55); D. Ross thought that the Postpraedicamenta were generally regarded as spurious (Aristotle, 24 n. 2).

(12) See J. G. Buhle, Aristotelis Opera, vol. I, 1791, 436; Ch. A. Brandis in: Abh. Berlin 1811, 268ff.; E. Zeller. Philos. d. Gr., II 2 (4th edition), 1921, 67 n. 1; Th. Gomperz, Greek Thinkers, IV, 514; Überweg-Praechter, 379; D. Ross, Aristotle, 10; L. M. De Rijk, The Authenticity, in: Mnemosyne. 4 (1951), 159; I. Düring. R E Suppl, XI, s v. Aristoteles, 205, 61; J. L. Ackrill, 70; V. Sainati, Storia, 151ff. Some ancient authors took this line (Olymp., In cat. 133, 14), especially Andronicus (Simpl., In cat. 379, 8ff.).

(13) E.g. J. G. Buhle, 436; E. Zeller, II 2 (4 ed.) 24, 1921, 67; H. Maier, Die Syllogistik, II 2, 292 n. We hear of this view being taken by some in antiquity (Ammon., In cat. 14, 18ff.; Olymp., In cat. 133, 14ff.). Whether Andronicus was among these, as is often claimed, is doubtful; at any rate, we never hear that he argued against the authenticity of the Postpraedicamenta; we would assume, if this had been the case, that he would be referred to by name when their authenticity was being discussed.

(14) J. G. Buhle, 436.

(15) O. Hamelin, Le systeme d'Aristote, 27 and 131.

From: Michael Frede, The Title, Unity, and Authenticity of the Aristotelian Categories. In Essays in Ancient Philosophy. Minneapolis: University of Minnesota Press 1987, pp. 11-28.


"There is a theory called the theory of categories which in a more or less developed form, with minor or major modifications, made its appearance first in a large number of Aristotelian writings and then, under the influence of these writings, came to be a standard part of traditional logic, a place it maintained with more or less success into the early part of this century, when it met the same fate as certain other parts of traditional logic.

There are lots of questions one may ask about this theory. Presumably not the most interesting question, but certainly one for which one would want to have an answer if one took an interest in the theory at all, is the following: What are categories? It turns out that this is a rather large and difficult question. And hence I want to restrict myself to the narrower and more modest question, What are categories in Aristotle?, hoping that a clarification of this question ultimately will help to clarify the more general questions. But even this narrower question turns out to be so complicated and controversial that I will be content if I can shed some light on the simple questions: What does the word "category" mean in Aristotle? What does Aristotle have in mind when he talks of "categories"?

Presumably it is generally agreed that Aristotle's doctrine of categories involves the assumption that there is some scheme of classification such that all there is, all entities, can be divided into a limited number of ultimate classes. But there is no agreement as to the basis and nature of this classification, nor is there an agreement as to how the categories themselves are related to these classes of entities. There is a general tendency among commentators to talk as if the categories just were these classes, but there is also the view that, though for each category there is a corresponding ultimate class of entities, the categories themselves are not to be identified with these classes. And there are various ways in which it could be true that the categories only correspond to, but are not identical with, these classes of entities. It might, e.g., be the case that the categories are not classes of entities but rather classes of expressions of a certain kind, expressions which we--following tradition--may call "categorematic." On this interpretation these categorematic expressions signify the various entities we classify under such headings as "substance," "quality," or "quantity." And in this case we have to ask whether the entities are classified according to a classification of the categorematic expressions by which they are signified, or whether, the other way round, the expressions are classified according to the classification of the entities they signify. Or it might be thought that the categories are classes of only some categorematic expressions, namely, those which can occur as predicate-expressions. Or it might be the case that the categories themselves are not classes at all, neither of entities nor of expressions, but rather headings or labels or predicates which collect, or apply to, either entities or expressions, i.e., the category itself, strictly speaking would be a term like "substance" or "substance word." Or it might be the case that categories are neither classes nor terms but concepts. All these views have had their ardent supporters." (pp. 1-2)

From: Michael Frede, Categories in Aristotle. In Dominc O'Meara (ed.), Studies in Aristotle, Washington: Catholic University Press 1981, pp. 1-25. (Reprinted in: M. Frede, Essays in Ancient Philosophy, Minneapolis: University of Minnesota Press, pp. 29-48).


"The precise position to be assigned to the Categories in the Aristotelian system has always been somewhat of a puzzle. On the one hand, they seem to be worked into the warp of its texture, as in the classification of change, and Aristotle can argue from the premiss that they constitute an exhaustive division of the kinds of Being (A. Pst. I 22, 83b15). On the other hand, both in the completed scheme of his logic and in his constructive metaphysic they retire into the background, giving place to other notions, such as causation, change, actuality and potentiality. Investigation, has, moreover, been hampered, especially in Germany, by attempts to correlate them with the Kantian Categories, with which they have obvious points of contact. But Kant's formal a priori concepts by which the mind makes for itself a world, to use Mr Bosanquet's phrase, imply an attitude to knowledge and reality so utterly opposed to the Aristotelian that the comparison has tended to confusion rather than elucidation. Scholars now realise better that the Aristotelian Categories can only be understood in connexion with the problems of Aristotle's own age.

The best general account of the Categories known to me is that given by Maier, who accepts the interpretation of Apelt in its main lines, correcting it in some important points. (1) It is the great merit of Apelt to have firmly grasped the principle that, whatever the applications to which Aristotle put the scheme of the Categories, it is primarily connected with the use of linguistic thought to make assertions about reality and hence with the proposition, the judgment as expressed in language. In details, I think, he is misled by the associations of post-Kantian logic, which prevent him from entering fully into the attitude adopted by the early Greek logic towards the fact of assertion.

In view of the undoubted fact that the scheme of the Categories follows the lines of Socratic-Platonic thought, Gercke's suggestion (2) is tempting that it originated in the Academy. Gercke, whose own view of the Categories is strongly coloured by Kantianism, relies almost entirely on the greater point given to the arguments in the Ethics against the Idea of the Good if we suppose them to accuse Plato of inconsistency with his own doctrine of the Categories. Except as supplementing strong independent evidence an argument of this kind carries no weight. The case is certainly weakened if it can be shown that Aristotle uses the Categories to solve a philosophical problem in explicit opposition to the solution offered by the Academy. This can be done, I think. In Meta. XIV 2, 1088b18 he sets the Categories against Platonist doctrine. He is criticising the indefinite dyad, and traces the origin of this conception to 'their old-fashioned way of setting problems': the Platonists found it necessary to attack the Parmenidean dictum and establish the existence of 'what is not' (cf. Plato, Sophist 237A, 256D). But how will this account for the plurality of being (for being means sometimes substance, sometimes that it is of a certain quality, and at other times the other categories: 1089a7)? In the corresponding passage of the Physics (I 2, 184b15 sqq.) Aristotle solves the Parmenidean difficulty through the multiplicity of the Categories (186a25), and alludes to the inadequacy of the Academic solution (187a1).

The inference to be drawn from these passages, in conjunction with the chapter in the Ethics on which Gercke relies, is the negative one that Plato and his successors in the Academy did not apply the scheme of the Categories to the fundamental philosophical questions of Being and Good. Positive evidence must be sought in another aspect of the doctrine. Now the Topics exhibits the Categories in intimate association with dialectical logic. The work itself purports to codify methods in regular use but not hitherto systematically treated. That these methods were employed in the Academy is amply attested by the Platonic dialogues. (3) Further, as the Topics and particularly the Sophistici Elenchi show, they were developed in close connexion with the eristic logic of Antisthenes and the Megarians. This fact at once establishes a contact with the treatment of the problem 'one thing, many names' in Plato's Sophist (251A). This difficulty was removed by drawing a distinction between different kinds of being, and Aristotle himself regards it as finally disposed of by the doctrine of the Categories. That some of the kinds of being included in the scheme were already recognised in the Academy is plain. In the Topics relatives have a number of their own topoi and the varieties of relatives enumerated in the Categories follow closely on the lines of division in the Charmides. (4) Much of the matter of the Topics must have been common to Academy and Lyceum. But this is not to say that the Categories as a complete and exhaustive scheme belonged to the Academy. Eudemus tells us that Plato solved the difficulties of Lycophron and others by a dual distinction of being.

I shall accordingly assume in what follows that the scheme of the Categories was evolved in the course of efforts to establish a doctrine of judgment which should settle the difficulties raised by Megarian and other critics; that the application to the solution of the larger metaphysical problems was a later development; that the foundations of the scheme were laid in the Socratic tradition of the Academy; that the completed scheme is probably Aristotle's own; and that the original working out of the scheme did not contemplate extension beyond the metaphysics implied in predication to the more fundamental metaphysics of the First Philosophy. Hence we must look to the analysis of empirical propositions for the origin of the scheme.

Now if we examine the scheme itself, we find three aspects of it to have special significance:

(a) The first is the distinction between accidental predication (kata sumbebékos) and essential predication (kath' hauto). (5) What is musical may be literate, but only 'in virtue of something else' (kat' allo), viz.: qua Callias; Callias is literate essentially (kath' hauton). This distinction provides the first condition of scientific predication, and is regarded as of fundamental importance by Aristotle, who prefaces his accounts of such notions as unity and being with references to the accidental uses of these terms (Meta. V 6, 1015b16; 7, 1017a7).

(b) Closely connected with the previous distinction is the doctrine that all the Categories (including substance as predicate) imply a subject (hupokeimenon), which is the point of real connexion between the predicates, and provides the basis of their coexistence. The Categories classify the many 'names' which we apply to the individual (e.g. a man, Sophist251A), and give expression to the fact that he does not lose his unity in the process.

(c) Furthermore, all direct relations of implication and incompatibility lie within the Categories severally. They are, so to say, independent variables. The relation of genus to species is everywhere confined within the limits of a category and so is the relation of contrary opposition. This suggests a close connexion with the Platonic division, which, as we know from the Sophist and the Politicus and from Aristotle, was so prominent in the Platonic conception of scientific method." (pp. 75-77)

Notes

(1) Heinrich Maier, Die Syllogistik des Aristoteles, 3 volumes, Tübingen, 1896-1900, vol. II, pp. 277 ff.; Otto Apelt, Beitrage zur Geschichte der griechischen Philosophie, Leipzig, 1891, pp. 106 ff.

(2) Alfred Gercke, Ursprung der aristotelischen Kategorien, Archiv fur Geschichte der Philosophie 4, 1891, pp. 424 ff.

(3) Analysis of the arguments in the Charmides shows that nearly all make use of topoi dealt with by Aristotle in the Topics.

(4) Cf. with Cat. 6a36 sqq., Charmides, 168A. The list in Rep. 437B is the same and in the same order.

(5) See the distinction of 'being kata sumbébekos' and 'being kath' hauto' (Meta. V 7, 1017a7 ff.). Apelt's equation of 'being kath' hauto' with 'being said in virtue of no combination' (op. cit. 117) is manifestly wrong. Kath' ho or kath' hauto means that the determination attaches to the subject in respect of the subject itself and not in respect of the determination. See kath' ho and kath' hauto, Meta. V 18, 1022a14 ff.

Charles Melville Gillespie, The Aristotelian Categories, The Classical Quarterly 19, 1925, pp. 79-84. Reprinted in: J. Barnes, M. Schofield, R. Sorabji (eds.), Articles on Aristotle - Vol. 3, Metaphysics, London: Duckworth, 1979, pp. 1-12

The problem of the authenticity of the Categories

"The little treatise of Aristotle which stands at the head of the Organon has caused a great deal of difficulty to students, both ancient and modern. The bulk of the discussion has centred about the question of its place in the Organon and in Aristotle's system, and the character of the ten categories to which the greater part of the book is devoted. But there have been found also critics who expressed a doubt as to the authenticity of all or part of the treatise in question. To say nothing of the ancient commentators of Aristotle, the earliest attempt in modern times to cast a doubt on the genuineness of the work seems to be that of Spengel in Münchener Gelehrte Anzeigen (Vol. XX [1845], No. 5, pp. 41 sq.). He was followed by Prantl in Zeitschrift für Altertumswissenschaft (1846, p. 646), and in his Geschichte der Logik (I, p. go, n. 5), also by Valentinus Rose in De Aristotelis librorum ordine et auctoritate (p. 234 et seq.). Zeller, on the other hand (Philosophie den Griechen, 2nd ed., II, pt. a, p. 67, n. 1), decides in favour of the genuineness of the first part of the work, the Categories proper, and against the so--called Postpredicamenta from Chapter X to the end.

(...)

When we pass over to matters of doctrine, it si surprising how many points of contact there are between the two works [Categories and Topics]. I shall follow the Categories and point out the parallels in the Topics.

The homonyms, which are given a definition and an illustration in the beginning of the Categories, have a whole chapter devoted to them in the Topics, the fifteenth of the first book, where they are also called pollachos legomena. Of particular significance is 107a 18-20, for in 20 we seem to have a direct allusion to the definition in the Categories. We must see, Aristotle says, if the genera designated by the given name are different and not subordinate to one another, (...) (which is therefore a homonym), for the definition of these genera as connected by the name is different (...). The greater space given to homonyms in the Topics is not due so much to a development in doctrine as to the necessities of the subject. The object of the Topics is a purely practical one, to provide the disputant with ready arguments properly pigeon-holed, and a single general definition of homonyms is not adapted to such use. We must needs go farther and show in what different special ways homonyms can be detected. The Categories have more the appearance of materials gathered in the shape of preliminary definitions of necessary concepts.

Synonyms are referred to in the Topics 109b 7, 123a 27, 127b 5, 148a 24, and 162b 37. Of these, the first is the most important, since it states that the genera are predicated synonymously of their species; for the latter admit both the name and the definition of the former (...), assuming it as established that this condition constitutes synonymity. This is neither more nor less than a silent reference to the definition in the Categories (1a 6) [When things have the name in common and the definition of being which corresponds to the name is the same, they are called synonymous]. Moreover we have almost the very words of the Topics in another place in the Categories, 3b 2, [And the primary substances admit the definition of the species and of the genera, and the species admits that of the genus; for everything said of what is predicated will be said of the subject also.] 148a 24 also gives the same definition of synonyms merely in passing. Aristotle is dealing with the definition, and makes a statement that if the opponent makes use of one definition for homonyms it cannot be a correct definition, for it is synonyms and not homonyms that have one definition connoted by the name (...). He speaks of the definition as already known. (...).

Paronyms also are made use of in the Topics, 109b 3-12, in a way which shows the definition in the Categories is not purely grammatical, as it may seem at first sight, but has a logical significance quite as important as that of the former two. Paronymous predication is predication per accidens, as contrasted with synonymous, which may be per se (cf. also Trendelenburg, Geschichte der Kategorienlehre, p.27 et seq. and 30). Here also paronyms are not defined. It is assumed that the reader knows what they are. (...)

Categories 3, p. 1b 10-15 expresses very much the same thought as Topics IV, 1, p. 121a 20-6. The former states that whatever is true of the species is true of the individuals under the species (...), the latter that to whatever the species applies the genus does also (...). They both involve the logical hierarchy of genus, species and individual, and the two principles are: (1) The genus applies not only to the species but also to the individual ; (2) to the individual belongs not only the species but also the genus. What is especially important to notice is that, in the Topics, the principle is stated as already known and is applied to the particular case, thus assuming the existence of another treatise where these principles are stated and proved for the first time.

The treatment of the difference develops gradually in the Topics in the following passages: 107b 19 sq., 144b 12 sq., and 153b 6. The first of these is word for word the same with the statement in the Categories, 1b 16 sq., and they were both quoted above. Moreover the way in which the passage in the Topics is introduced, (...) makes it a direct reference to the Categories. Aristotle's doctrine concerning the difference so far is that of different genera which are not subordinated one to the other: the differences are different in species. In the second passage quoted above, 144b 12, Aristotle corrects this view by adding that the differences in the given case need not be different unless the different genera cannot be put under a common higher genus. In the third passage, 153b 6, Aristotle adds some more qualifications which make it clear that in the preceding statements the word etéron, in the phrase etéron ghenon, must be understood as including contrary genera (enantía). For there the case is different. If the contrary genera belong to higher contrary genera, their differences may be all the same.

The preceding examination seems to show very clearly that the Topics build upon the basis laid down in the Categories and carry the structure higher and broader. It would be a very absurd alternative to suppose that a later writer, making use of the Topics, found nothing else on the subject of logical difference than the first passage, which he copied verbatim in his treatise, where,' besides, it has no particular reason for existence. As a thought tentatively suggested, with the view of further elaboration and insertion as a proper link in a chain, the passage in the Categories assumes a different meaning, and its lack of connection with the preceding and following ceases to cause us serious difficulty.

If the view of the Categories taken here is justified by the preceding arguments and by what is still to come, it might even be a legitimate procedure to make use of the Topics in determining a disputed reading in the Categories. And we have one at hand in the passage quoted above on the difference." (pp. 97-103)

From: Isaac Husik, The Categories of Aristotle, in: Milton C. Nahm and Leo Strauss (eds.), Philosophical Essays, Ancient, Mediaeval, and Modern, Oxford: Blackwell, 1952, pp. 96-112. (Greek text of the citations omitted).

The recent critical edition by Richard Bodéüs (Aristote. [Catégories], Paris: Les Belles Lettres, 2001) contains a detailed status questionis on the problem of authenticity at pp. XC-CX.

The conclusion is: "Malgré ses doutes sur l'authenticité de l'ouvrage, l'éditeur, nous semble-t-il, reste donc autorisé à imprimer celui-ci sous l'autorité traditionnelle d'Aristote." p. CX. (Despite his doubts about the authenticity of the book, the editor, seems to us, remains authorized to print it under the traditional authority of Aristotle.)