Читать книгу Aristotle - Grote George - Страница 17

1 Scholia, p. 25, b. 37, seq. Br.; p. 321, b. 30; Diogen. L. iii. 62. The order in which the forty-six Aristotelian treatises stand printed in the Berlin edition, and in other preceding editions, corresponds to the tripartite division, set forth by Aristotle himself, of sciences or cognitions generally: 1. Theoretical; θεωρητικαί 2. Practical; πρακτικαί. 3. Constructive or Technical; ποιητικαί.

Patricius, in his Discussiones Peripateticć, published in 1581 (tom. i. lib. xiii. p. 173), proclaims himself to be the first author who will undertake to give an account of Aristotles philosophy from Aristotle himself (instead of taking it, as others before him had done, from the Aristotelian expositors, Andronikus, Alexander, Porphyry, or Averroes); likewise, to be the first author who will consult all the works of Aristotle, instead of confining himself, as his predecessors had done, to a select few of the works. Patricius then proceeds to enumerate those works upon which alone the professors in Italicis scholis lectured, and to which the attention of all readers was restricted. 1. The Predicabilia, or Eisagoge of Porphyry. 2. The Categorić. 3. The De Interpretatione. 4. The Analytica Priora; but only the four first chapters of the first book. 5. The Analytica Posteriora; but only a few chapters of the first book; nothing of the second. 6. The Physica; books first and second; then parts of the third and fourth; lastly, the eighth book. 7. The De Clo; books first and second. 8. The De Generatione et Corruptione; books first and second. 9. The De Animâ; all the three books. 10. The Metaphysica; books Alpha major, Alpha minor, third, sixth, and eleventh. Idque, quadriennio integro, quadruplicis ordinis Philosophi perlegunt auditoribus. De reliquis omnibus tot libris, mirum silentium.

Patricius expressly remarks that neither the Topica nor the De Sophisticis Elenchis was touched in this full course of four years. But he does not remark what to a modern reader will seem more surprising that neither the Ethica, nor the Politica, nor the Rhetorica, is included in the course.

2 Aristot. Topica, i. p. 104, b. 1, with the Scholia of Alexander, p. 259, a. 48 Br.; Scholia ad Analyt. Prior. p. 140, a. 47, p. 141, a. 25; also Schol. ad Categor. p. 36, a., p. 40, a., 8. This conception of the Organon is not explicitly announced by Aristotle, but seems quite in harmony with his views. The contemptuous terms in which Prantl speaks of it (Gesch. der Logik, i. 136), as a silly innovation of the Stoics, are unwarranted.

Aristotle (Metaph. E. i. p. 1025, b. 26) classifies the sciences as θεωρητικαί, πρακτικαί, ποιητικαί; next he subdivides the first of the three into φυσική, μαθηματική, πρώτη φιλοσοφία. Brentano, after remarking that no place in this distribution is expressly provided for Logic, explains the omission as follows: Diese auffallende Erscheinung erklärt sich daraus, dass diese [the three above-named theoretical sciences] allein das reelle Sein betrachten, und nach den drei Graden der Abstraktion in ihrer Betrachtungsweise verschieden, geschieden werden; während die Logik das bloss rationelle Sein, das ὃν ὡς ἀληθές, behandelt. (Ueber die Bedeutung des Seienden nach Aristoteles, p. 39.) Investigations περὶ τῆς ἀληθείας, ὃν τρόπον δεῖ ἀποδέχεσθαι are considered by Aristotle as belonging to τὰ Ἀναλυκτικά; enquiries into method in the first instance, and into doctrine chiefly with a view to method (Metaphys. Γ. p. 1005, b. 2. In Metaphys. Γ. 1005, b. 7, he declares that these enquiries into method, or analysis of the principia of syllogistic reasoning, belong to the Philosophia Prima (compare Metaphys. Z. 12, p. 1037, b. 8). Schwegler in his Commentary (p. 161) remarks that this is one of the few passages in which Aristotle indicates the relation in which Logic stands to Metaphysics, or First Philosophy. The question has been started among his Ἀπορίαι, Metaph. B. 2, p. 999, b. 30.

3 Respecting the title of Organon which was sometimes applied to the Analytica Posteriora only, see Waitz ad Organ. ii. p. 294.

These treatises are six in number: 1. Categorić;4 2. De Interpretatione, or De Enunciatione; 3. Analytica Priora; 4. Analytica Posteriora; 5. Topica; 6. De Sophisticis Elenchis. This last short treatise De Sophisticis Elenchis belongs naturally to the Topica which precedes it, and of which it ought to be ranked as the ninth or concluding book. Waitz has printed it as such in his edition of the Organon; but as it has been generally known with a separate place and title, I shall not depart from the received understanding.

4 Some eminent critics, Prantl and Bonitz among them, consider the treatise Categorić not to be the work of Aristotle. The arguments on which this opinion rests are not convincing to me; and even if they were, the treatise could not be left out of consideration, since the doctrine of the Ten Categories is indisputably Aristotelian. See Zeller, Die Phil. der Griech. ii. 2, pp. 50, 51, 2nd ed.

Aristotle himself does not announce these six treatises as forming a distinct aggregate, nor as belonging to one and the same department, nor as bearing one comprehensive name. We find indeed in the Topica references to the Analytica, and in the Analytica references to the Topica. In both of them, the ten Categories are assumed and presupposed, though the treatise describing them is not expressly mentioned: to both also, the contents of the treatise De Interpretatione or Enunciatione, though it is not named, are indispensable. The affinity and interdependence of the six is evident, and justifies the practice of the commentators in treating them as belonging to one and the same department. To that department there belonged also several other treatises of Aristotle, not now preserved, but specified in the catalogue of his lost works; and these his disciples Theophrastus, Eudemus, and Phanias, had before them. As all these three disciples composed treatises of their own on the same or similar topics,5 amplifying, elucidating, or controverting the views of their master, the Peripatetics immediately succeeding them must have possessed a copious logical literature, in which the six treatises now constituting the Organon appeared as portions, but not as a special aggregate in themselves.

5 Ammonius ap. Schol. p. 28, a. 41; p. 33, b. 27, Br.

Of the two treatises which stand first in the Aristotelian Organon the Categorić and the De Interpretatione each forms in a certain sense the complement of the other. The treatise De Interpretatione handles Propositions (combinations of terms in the way of Subject and Predicate), with prominent reference to the specific attribute of a Proposition the being true or false, the object of belief or disbelief; the treatise Categorić deals with these same Terms (to use Aristotles own phrase) pronounced without or apart from such combination. In his definition of the simple Term, the Proposition is at the same time assumed to be foreknown as the correlate or antithesis to it.6

6 Τὰ ἄνευ συμπλοκῆς λεγόμενα τῶν κατὰ μηδεμίαν συμπλοκὴν λεγομένωνα (Categ. p. 1, a. 16, b. 25). See Schol. ad Aristot. Physica, p. 323, b. 25, Br.; and Bonitz ad Aristotel. Metaph. (A. p. 987) p. 90.

The Categories of Aristotle appear to formed one of the most prominent topics of the teaching of Themistius: rebutting the charge, advanced both against himself, and, in earlier days, against Sokrates and the Sophists, of rendering his pupils presumptuous and conceited, he asks, ἠκούσατε δὲ αὖ τινος τῶν ἐμῶν ἐπιτηδείων ὑψηλογουμένου καὶ βρενθυομένου ἐπὶ τοῖς συνωνύμοις ἢ ὁμωνύμοῖς ἢ παρωνύμοις; (Orat. xxiii. p. 351.)

Reference is made (in the Scholia on the Categorić, p. 43, b. 19) to a classification of names made by Speusippus, which must have been at least as early as that of Aristotle; perhaps earlier, since Speusippus died in 339 B.C. We do not hear enough of this to understand clearly what it was. Boęthus remarked that Aristotle had omitted to notice some distinctions drawn by Speusippus on this matter, Schol. p. 43, a. 29. Compare a remark in Aristot. De Clo, i. p. 280, b. 2.

The first distinction pointed out by Aristotle among simple, uncombined Terms, or the things denoted thereby, is the Homonymous, the Synonymous, and the Paronymous. Homonymous are those which are called by the same name, used in a different sense or with a different definition or rational explanation. Synonymous are those called by the same name in the same sense. Paronymous are those called by two names, of which the one is derived from the other by varying the inflexion or termination.7

7 Aristot. Categor. p. 1, a. 1–15.

We can hardly doubt that it was Aristotle who first gave this peculiar distinctive meaning to the two words Homonymous and Synonymous, rendered in modern phraseology (through the Latin) Equivocal and Univocal. Before his time this important distinction between different terms had no technical name to designate it. The service rendered to Logic by introducing such a technical term, and by calling attention to the lax mode of speaking which it indicated, was great. In every branch of his writings Aristotle perpetually reverts to it, applying it to new cases, and especially to those familiar universal words uttered most freely and frequently, under the common persuasion that their meaning is not only thoroughly known but constant and uniform. As a general fact, students are now well acquainted with this source of error, though the stream of particular errors flowing from it is still abundant, ever renewed and diversified. But in the time of Aristotle the source itself had never yet been pointed out emphatically to notice, nor signalized by any characteristic term as by a beacon. The natural bias which leads us to suppose that one term always carries one and the same meaning, was not counteracted by any systematic warning or generalized expression. Sokrates and Plato did indeed expose many particular examples of undefined and equivocal phraseology. No part of the Platonic writings is more valuable than the dialogues in which this operation is performed, forcing the respondent to feel how imperfectly he understands the phrases constantly in use. But it is rarely Platos practice to furnish generalized positive warnings or systematic distinctions. He has no general term corresponding to homonymous or equivocal; and there are even passages where (under the name of Prodikus) he derides or disparages a careful distinctive analysis of different significations of the same name. To recognize a class of equivocal terms and assign thereto a special class-name, was an important step in logical procedure; and that step, among so many others, was made by Aristotle.8

8 In the instructive commentary of Dexippus on the Categorić (contained in a supposed dialogue between Dexippus and his pupil Seleukus, of which all that remains has been recently published by Spengel, Munich, 1859), that commentator defends Aristotle against some critics who wondered why he began with these Ante-predicaments (ὁμώνυμα, συνώνυμα, &c.), instead of proceeding at once to the Predicaments or Categories themselves. Dexippus remarks that without understanding this distinction between equivoca and univoca, the Categories themselves could not be properly appreciated; for Ens τὸ ὂν is homonymous in reference to all the Categories, and not a Summum Genus, comprehending the Categories as distinct species under it; while each Category is a Genus in reference to its particulars. Moreover, Dexippus observes that this distinction of homonyms and synonyms was altogether unknown and never self-suggested to the ordinary mind (ὅσων γὰρ ἔννοιαν οὐκ ἔχομεν, τούτων πρόληψιν οὐκ ἔχομεν, p. 20), and therefore required to be brought out first of all at the beginning; whereas the Post-predicaments (to which we shall come later on) were postponed to the end, because they were cases of familiar terms loosely employed. (See Spengel, Dexipp. pp. 19, 20, 21.)

Though Aristotle has professed to distinguish between terms implicated in predication, and terms not so implicated,9 yet when he comes to explain the functions of the latter class, he considers them in reference to their functions as constituent members of propositions. He immediately begins by distinguishing four sorts of matters (Entia): That which is affirmable of a Subject, but is not in a Subject; That which is in a Subject, but is not affirmable of a Subject; That which is both in a Subject, and affirmable of a Subject; That which is neither in a Subject, nor affirmable of a Subject.10

9 Aristot. Categor. p. 1, a. 16. τῶν λεγομένων τὰ μὲν κατὰ συμπλοκὴν λέγεται, τὰ δ ἄνευ συμπλοκῆςˇ τὰ μὲν οὖν κατὰ συμπλοκὴν οἷον ἄνθρωπος τρέχει, ἄνθρωπος νικᾷˇ τὰ δ ἄνευ συμπλοκῆς οἶον ἄνθρωπος, βοῦς, τρέχει, νικᾷ.

It will be seen that the meaning and function of the single word can only explained relatively to the complete proposition, which must be assumed as foreknown.

That which Aristotle discriminates in this treatise, in the phrases λέγεσθαι κατὰ συμπλοκὴν and λέγεσθαι ἄνευ συμπλοκῆς is equivalent to what we read in the De Interpretatione (p. 16, b. 27, p. 17, a. 17) differently expressed, φωνὴ σημαντικὴ ὡς κατάφασις and φωνὴ σημαντικὴ ὡς φάσις.

10 Aristot. Categor. p. 1, a. 20.

This fundamental quadruple distinction of Entia, which serves as an introduction to the ten Categories or Predicaments, belongs to words altogether according to their relative places or functions in the proposition; the meanings of the words being classified accordingly. That the learner may understand it, he ought properly to be master of the first part of the treatise De Interpretatione, wherein the constituent elements of a proposition are explained: so intimate is the connection between that treatise and this.

The classification applies to Entia (Things or Matters) universally, and is thus a first step in Ontology. He here looks at Ontology in one of its several diverse aspects as it enters into predication, and furnishes the material for Subjects and Predicates, the constituent members of a proposition.

Ontology, or the Science of Ens quatenus Ens, occupies an important place in Aristotles scientific programme; bearing usually the title of First Philosophy, sometimes Theology, though never (in his works) the more modern title of Metaphysica. He describes it as the universal and comprehensive Science, to which all other sciences are related as parts or fractions. Ontology deals with Ens in its widest sense, as an Unum not generic but analogical distinguishing the derivative varieties into which it may be distributed, and setting out the attributes and accompaniments of Essentia universally; while other sciences, such as Geometry, Astronomy, &c., confine themselves to distinct branches of that whole;11 each having its own separate class of Entia for special and exclusive study. This is the characteristic distinction of Ontology, as Aristotle conceives it; he does not set it in antithesis to Phenomenology, according to the distinction that has become current among modern metaphysicians.

11 Aristot. Metaphys. Γ. p. 1003, a. 21, 25–33, E. p. 1025, b. 8. ἔστιν ἐπιστήμη τις ἢ θεωρεῖ τὸ ὂν ᾗ ὂν καὶ τὰ τούτῳ ὑπάρχοντα καθ αὑτόˇ αὕτη δ ἐστὶν οὐδεμιᾷ τῶν ἄλλων ἐπισκοπεῖ καθόλου περὶ τοῦ ὄντος ᾗ ὅν, ἀλλὰ μέρος αὐτοῦ τι ἀποτεμόμεναι περὶ τούτου θεωροῦσι τὸ συμβεβηκός, &c. Compare p. 1005, a. 2–14.

Now Ens (or Entia), in the doctrine of Aristotle, is not a synonymous or univocal word, but an homonymous or equivocal word; or, rather, it is something between the two, being equivocal, with a certain qualification. Though not a Summum Genus, i.e. not manifesting throughout all its particulars generic unity, nor divisible into species by the addition of well-marked essential differentić, it is an analogical aggregate, or a Summum Analogon, comprehending under it many subordinates which bear the same name from being all related in some way or other to a common root or fundamentum, the relationship being both diverse in kind and nearer or more distant in degree. The word Ens is thus homonymous, yet in a qualified sense. While it is not univocal, it is at the same time not absolutely equivocal. It is multivocal (if we may coin such a word), having many meanings held together by a multifarious and graduated relationship to one common fundamentum.12 Ens (or Entia), in this widest sense, is the theme of Ontology or First Philosophy, and is looked at by Aristotle in four different principal aspects.13

12 Simplikius speaks of these Analoga as τὸ μέσον τῶν τε συνωνύμων καὶ τῶν ὁμωνύμων, τὸ ἀφ ἑνός, &c. Schol. ad Categor. p. 69, b. 29, Brand. See also Metaphys. Z. p. 1030, a. 34.

Dexippus does not recognize, formally and under a distinct title, this intermediate stage between συνώνυμα and ὁμώνυμα. He states that Aristotle considered Ens as ὁμώνυμον, while other philosophers considered it as συνώνυμον (Dexippus, p. 26, book i. sect. 19, ed. Spengel). But he intimates that the ten general heads called Categories have a certain continuity and interdependence (συνέχειαν καὶ ἀλληλουχίαν) each with the others, branching out from οὐσία in ramifications more or less straggling (p. 48, book ii. sects. 1, 2, Spengel). The list (he says, p. 47) does not depend upon διαίρεσις (generic division), nor yet is it simple enumeration (ἀπαρίθμησις) of incoherent items. In the Physica, vii. 4, p. 249, a. 23, Aristotle observes: εἰσὶ δὲ τῶν ὁμωνυμιῶν αἱ μὲν πολὺ ἀπέχουσι αἱ δὲ ἔχουσαί τινα ὁμοιότητα, αἱ δ ἐγγὺς ἢ γένει ἢ ἀναλογίᾳ, διὸ οὐ δοκοῦσιν ὁμωνυμίαι εἶναι οὖσαι.

13 Aristot. Metaphys. Δ. p. 1017, a. 7, E. p. 1025, a. 34, p. 1026, a. 33, b. 4; upon which last passage see the note of Bonitz.

1. Τὸ ὂν κατὰ συμβεβηκός Ens per Accidens Ens accidental, or rather concomitant, either as rare and exceptional attribute to a subject, or along with some other accident in the same common subject.

2. Τὸ ὂν ὡς ἀληθές, καὶ τὸ μὴ ὂν ὡς ψεῦδος Ens, in the sense /of Truth, Non-Ens, in the sense of Falsehood. This is the Ens of the Proposition; a true affirmation or denial falls under Ens in this mode, when the mental conjunction of terms agrees with reality; a false affirmation or denial, where no such agreement exists, falls under Non-Ens.14

14 Aristot. Metaph. E. 4, p. 1027, b. 18, p. 1028, a. 4. οὐ γὰρ ἐστι τὸ ψεῦδος καὶ τὸ ἀληθὲς ἐν τοῖς πράγμασιν ἀλλ ἐν διανοίᾳ οὐκ ἔξω δηλοῦσιν οὖσάν τινα φύσιν τοῦ ὄντος. Also Θ. 10, p. 1051, b. 1: τὸ κυριώτατα ὂν ἀληθες καὶ ψεῦδος. In a Scholion, Alexander remarks: τὸ δὲ ὡς ἀληθῶς ὂν πάθος ἐστὶ καὶ βούλημα διανοίας, τὸ δὲ ζητεῖν τὸ ἑκάστῳ δοκοῦν οὐ σφόδρα ἀναγκαῖον.

3. Τὸ ὂν δυνάμει καὶ τὸ ὂν ἐνεργείᾳ Ens, potential, actual.

4. Τὸ ὂν κατὰ τὰ σχήματα τῶν κατηγοριῶν Ens, according to the ten varieties of the Categories, to be presently explained.

These four are the principal aspects under which Aristotle looks at the aggregate comprised by the equivocal or multivocal word Entia. In all the four branches, the varieties comprised are not species under a common genus, correlating, either as co-ordinate or subordinate, one to the other; they are analoga, all having relationship with a common term, but having no other necessary relationship with each other. Aristotle does not mean that these four modes of distributing this vast aggregate, are the only modes possible; for he himself sometimes alludes to other modes of distributions.15 Nor would he maintain that the four distributions were completely distinguished from each other, so that the same subordinate fractions are not comprehended in any two; for on the contrary, the branches overlap each other and coincide to a great degree, especially the first and fourth. But he considers the four as discriminating certain distinct aspects of Entia or Entitas, more important than any other aspects thereof that could be pointed out, and as affording thus the best basis and commencement for the Science called Ontology.

15 Aristot. Metaph. Γ. p. 1003, a. 33, b. 10. Compare the able treatise of Brentano, Ueber die Bedeutung des Seienden nach Aristoteles, pp. 6, 7.

Of these four heads, however, the first and second are rapidly dismissed by Aristotle in the Metaphysica,16 being conceived as having little reference to real essence, and therefore belonging more to Logic than to Ontology; i.e. to the subjective processes of naming, predicating, believing, and inferring rather than to the objective world of Perceivables and Cogitables.17 It is the third and fourth that are treated in the Metaphysica; while it is the fourth only (Ens according to the ten figures of the Categories) which is set forth and elucidated in this first treatise of the Organon, where Aristotle appears to blend Logic and Ontology into one.

16 Aristot. Metaph. E. p. 1027, b. 16, p. 1028, a. 6.

17 Aristot. Metaph. Θ. 10, p. 1051, b. 2–15, with Schweglers Comment, p. 186. This is the distinction drawn by Simplikius (Schol. ad Categ. p. 76, b. 47) between the Organon and the Metaphysica: Αἱ γὰρ ἀρχαὶ κατὰ μὲν τήν σημαντικὴν αὐτῶν λέξιν ἐν τῇ λογικῇ πραγματείᾳ δηλοῦνται, κατὰ δὲ τὰ σημαινόμενα ἐν τῇ Μετὰ τὰ Φυσικὰ οἰκείως.

Τὰ ὄντα are equivalent to τὰ λεγόμενα, in this and the other logical treatises of Aristotle. Categ. p. 1, a. 16–20, b. 25; Analyt. Prior. i. p. 43, a. 25.

This is the logical aspect of Ontology; that is, Entia are considered as Objects to be named, and to serve as Subjects or Predicates for propositions: every such term having a fixed denotation, and (with the exception of proper names) a fixed connotation, known to speakers and hearers.

Τὰ λεγόμενα (or Entia considered in this aspect) are distinguished by Aristotle into two classes: 1. Τὰ λεγόμενα κατὰ συμπλοκήν, οἷον ἄνθρωπος τρέχει, ἄνθροπος νικᾷ. 2. Τὰ λεγόμενα ἄνευ συμπλοκῆς (or κατὰ μηδεμίαν συμπλοκήν) οἷον ἄνθρωπος, βοῦς, τρέχει, νικᾷ.

We are to observe here, that in Logic the Proposition or Enunciation is the Prius Naturâ, which must be presupposed as known before we can understand what the separate terms are (Analytic. Prior. i. p. 24, a. 16): just as the right angle must be understood before we can explain what is an acute or an obtuse angle (to use an illustration of Aristotle; see Metaphys. Ζ. p. 1035, b. 7). We must understand the entire logical act, called Affirming or Denying, before we can understand the functions of the two factors or correlates with which that act is performed. Aristotle defines the Term by means of the Proposition, ὅρον δὲ καλῶ εἰ ὂν διαλύεται ἡ πρότασις (Anal. Pr. i. 24, b. 16).

Τὰ λεγόμενα, as here used by Aristotle, coincides in meaning with what the Stoics afterwards called Τὰ λεκτά of two classes: 1. λεκτὰ αὐτοτελῆ, one branch of which, τὰ ἀξιώματα, are equivalent to the Aristotelian τὰ κατὰ συμπλοκὴν λεγόμενα. 2. λεκτὰ ἐλλιπῆ, equivalent to τὰ ἄνευ συμπλοκῆς λεγόμενα (Diogen. Laert. vii. 43, 44, 63, 64; Sext. Emp. adv. Mathemat. viii. 69, 70, 74): equivalent also, seemingly, to τὰ διανοητὰ in Aristotle: ὁ διανοητὸς Ἀριστομένης (Anal. Pr. I. p. 47, b. 22).

Hobbes observes (Computation or Logic, part i. 2, 5): Nor is it at all necessary that every name should be the name of something. For as these, a man, a tree, a stone, are the names of the things themselves, so the images of a man, of a tree, of a stone, which are represented to men sleeping, have their names also, though they be not things, but only fictions and phantasms of things. For we can remember these; and therefore it is no less necessary that they have names to mark and signify them, than the things themselves. Also this word future is a name; but no future thing has yet any being. Moreover, that which neither is, nor has been, nor ever shall or ever can be, has a name impossible. To conclude, this word nothing is a name, which yet cannot be name of any thing; for when we subtract two and three from five, and, so nothing remaining, we would call that subtraction to mind, this speech nothing remains, and in it the word nothing, is not unuseful. And for the same reason we say truly, less than nothing remains, when we subtract more from less; for the mind feigns such remains as these for doctrines sake, and desires, as often as is necessary, to call the same to memory. But seeing every name has some relation to that which is named, though that which we name be not always a thing that has a being in nature, yet it is lawful for doctrines sake to apply the word thing to whatsoever we name; as it were all one whether that thing truly existent, or be only feigned.

The Greek neuter gender (τὸ λεγόμενον or τὸ λεκτόν, τὰ λεγόμενα or τὰ λεκτά) covers all that Hobbes here includes under the word thing. Scholia ad Aristot. Physic. I. i. p. 323, a. 21, Brand.: ὀνομάζονται μὲν καὶ τὰ μὴ ὄντα, ὁρίζονται δὲ μόνα τὰ ὄντα.

Of this mixed character, partly logical, partly ontological, is the first distinction set forth in the Categorić the distinction between matters predicated of a Subject, and matters which are in a Subject the Subject itself being assumed as the fundamentum correlative to both of them. The definition given of that which is in a Subject is ontological: viz., In a Subject, I call that which is in anything, not as a part, yet so that it cannot exist separately from that in which it is.18 By these two negative characteristics, without any mark positive, does Aristotle define what is meant by being in a Subject. Modern logicians, and Hobbes among them, can find no better definition for an Accident; though Hobbes remarks truly, that Accident cannot be properly defined, but must be elucidated by examples.19

18 Aristot. Categ. p. 1, a. 24.

19 Hobbes, Computation or Logic, part i. 3, 3, i. 6, 2, ii. 8, 2–3.

The distinction here drawn by Aristotle between being predicated of a Subject, and being in a Subject, coincides with that between essential and non-essential predication: all the predicates (including the differentia) which belong to the essence, fall under the first division;20 all those which do not belong to the essence, under the latter. The Subjects what Aristotle calls the First Essences or Substances, those which are essences or substances in the fullest and strictest meaning of the word are concrete individual things or persons; such as Sokrates, this man, that horse or tree. These are never employed as predicates at all (except by a distorted and unnatural structure of the proposition, which Aristotle indicates as possible, but declines to take into account); they are always Subjects of different predicates, and are, in the last analysis, the Subjects of all predicates. But besides these First Essences, there are also Second Essences Species and Genus, which stand to the first Essence in the relation of predicates to a Subject, and to the other Categories in the relation of Subjects to predicates.21 These Second Essences are less of Essences than the First, which alone is an Essence in the fullest and most appropriate sense. Among the Second Essences, Species is more of an Essence than Genus, because it belongs more closely and specially to the First Essence; while Genus is farther removed from it. Aristotle thus recognizes a graduation of more or less in Essence; the individual is more Essence, or more complete as an Essence, than the Species, the Species more than the Genus. As he recognizes a First Essence, i.e. an individual object (such as Sokrates, this horse, &c.), so he also recognizes an individual accident (this particular white colour, that particular grammatical knowledge) which is in a Subject, but is not predicated of a Subject; this particular white colour exists in some given body, but is not predicable of any body.22

20 Aristot. Categ. p. 3, a. 20. It appears that Andronikus did not draw the line between these two classes of predicates in same manner as Aristotle: he included many non-essential predicates in τὰ καθ ὑποκειμένου. See Simplikius, ad Categorias, Basil. 1551, fol. 13, 21, B. Nor was either Alexander or Porphyry careful to observe the distinction between the two classes. See Schol. ad Metaphys. p. 701, b. 23, Br.; Schol. ad De Interpret. p. 106, a. 29, Br. And when Aristotle says, Analyt. Prior. i. p. 24, b. 26, τὸ δὲ ἐν ὅλῳ εἰναι ἕτερον ἑτέρῳ, καὶ τὸ κατὰ παντὸς κατηγορεῖσθαι θατέρου θάτερον, ταὐτόν ἐστιν, he seems himself to forget the distinction entirely.

21 Categor. p. 2, a. 15, seq. In Aristotle phraseology it is not said that Second Essences are contained in First Essences, but that First Essences are contained in Second Essences, i.e. in the species which Second Essences signify. See the Scholion to p. 3, a. 9, in Waitz, vol. i. p. 32.