An

Introduction

to the Aristotle's

First

Five

Chapters

of

Categories

BARRINGTON JONES 1. Introduction: account of paperl I have argued that a satisfactory both and nonof substantial individuals, postulation I Aristotle's achieved can be in the substantial, Categories by taking seriously of these individuals as things that are 'one in his characterization as 'a unit in a posthis characterization number' and by interpreting This approach to the Categories ha.s important sible act of enumeration'. of the work. of the remainder for the interpretation consequences In this essay I wish to present an account of the first five chapters based on the (bar chapter 4 which lays out the categories themselves) former paper. I wish to examine the four-fold division of 'the things In particular, that are' in chapter 2 and the two relations of 'being said of' and this 'being in' (or, rather, 'existing in') that are used to construct four-fold division, and the nature of 'primary substance' (or, rather, from 'secondary being') and the basis for its distinction 'primary substance' being'). The account that will be (or, rather, 'secondary different from any here is and substantially importantly developed other that I am aware of, and, even if it does not secure conviction, will hopefully make the dogma that the Categories is a its publication of work less readily tenable and force a re-thinking 'common-sensical' the usual account of the work. I wish to suggest that 'the things that are' are 'things that exist that the two relations ('being said of' and in existing are something', ouonx are that and 'being in') existential, 'primary substance' and that is the existentially primary mode of existence of individuals is due partly to a hypofrom 'secondary substances' their separation in language, partly to a recognition of count-nouns statization that, to the extent that counting forms a basis of the work, to that extent of 'the in the classification an ability to use language is presupposed cannot simply view things that are', and, finally, that Aristotle substances' as items of language because he recognizes 'secondary n an earlier

146

that being

we can have any one

a man as a unit of enumeration particular man that is that unit. and Synonymy :

without

there

2. Homonymy, As these

Paronymy

are introduced in the first chapter of the work would be all the individual men that there examples of synonyms a man and a man in a picture, of paionyms might be, of homonyms all the brave. All these are, so to speak, of the same ontological type. from this However, certain writers have wanted to exempt paronymy On their account paronyms are not, for example, type ot interpretation. all the brave, one to another, but a brave man and bravery. Thus, Hintikka writes: with synonymy " ... paronymy appears as a notion incomparable are defined to be and homonymy: Things synonyms and homonyms in so far as they share the same name, whereas two things are paronyms when they are called by different 'names' (terms) of which one is nevertheless derived (grammatically) from the other,"2 and, in more linguistic frame of mind, Owen claims that: " ... the definition of paronyms is merely grammatical. It shows... how adjectives can be manufactured from abstract nouns by the modifying word-ending."3 of paroNow, though there is no need to read the characterization in 1 in this way, whereby the paronyms chapter nymy as it is given are the item called by an adjective which is an inflected form of a and the property and hence as pro`name' for a property 'named', with an the characterizations of homonymy ducing asymmetry and synonymy in the first chapter, these writers have a point. For, is actually employed in the work, the relata are indeed as 'paronymy' what they claim (v. 6 b 11-14, 10 a 27-b 11, 2 a 27-29; v. too Phys. 245 b 11-12). This 'twist', however, is not restricted to paronymy. As homonymy and synonymy are employed in the work the relata are not items of the same ontological status. Thus, examples of synman and what he is, his eidos : onymy are a particular

notions

147

of substances and differentiae that all things "It is a characteristic called from them are so called synonymously. For all the predicates from them are predicated either of the individuals or of And the primary substances the species... admit the definition of the species and of the genera, and the species admits that of the genus... Similarly, both the species and the individuals admit the definition of the differentiae. But synonymous things were precisely those with both the name in common and the same Hence all the things called from substances definition. and difa trans. ferentiae are so called synonymously" 33-b 9, Ackrill; (3 v. too Top., 109 b 4-7). is employed in such a way that the colour white Equally, homonymy and a thing that is white are homonyms (2 a 29-32; v. too 3 a 15-17). between the three notions. As they are So there is no asymmetry introduced in the first chapter, the members of each of the three classes are of the same type; as they are used in the body of the work they are all given this 'twisted' application. In the earlier paper, I suggested that "the point of paronymy... is to license the inference from a certain number of literate individuals to that number of literacies.4 Miss J. Annas has suggested that "any would seem to go the other way. According inference-licence to Aristotle ... the literate man is so called from literacy."5 Thus, is from noun to adjective, "the direction of derivativeness not from to noun. "6 Now, calling paronymy an "inference-licence" adjective is not meant to explain his use of the word "from" in his characterizais not presented as a thesis about tion, and clearly the characterization expressed by "from" is meant counting. Probably, the 'derivativeness' of that type of explanation of the possession of to be a generalization individuals so common in the ethical properties by writings of Aristotle and the early dialogues of Plato. On this model, we explain why a person is brave by reference to his bravery, why Charmides is temin him, and more generally perate by the presence of temperance why a person is good by reference to his virtue. Such a model clearly the inference-licence in question: the good man is such presupposes because of his virtue, Charmides because of the presence tempeiate of temperance in him. Since the aim of the analysis of the Categories

148

and the present paper is not to claim that Aristotle out to construct an ontology on the explicit set self-consciously of but rather to endeavour to detera consideration basis of counting, of counting enable us to mine the extent to which considerations the work, all that is required is the presupposition of understand It is clear that counting does form part theses ielevant to counting. of the basis of the work, and so this enterprise is justified. what the appeal to counting does explain is the More importantly, of the three-fold division in chapter 1, and it is this that rationale most needs explanation. Why just these classes? To treat paronymy The interpretation as an exception is, as we have seen, unsatisfactory. does an and it has of in terms adequate explanation counting present in as much, indeed rather more, basis in the text as interpretations and the like.' 7 or 'universals' teims of 'grammar' Accordingly, it looks as though it will be profitable to approach the The measure of the profit of enumeration. work from the standpoint of to resolve or to illuminate difficulties. the is the ability approach So let us see how able it is. in "Individuals"

149

3. `Being In' and `Being Said Of' : distinct substances this is done on the basis When we discriminate of the appropriate and equally of the applicability sortal-expression, we can count and discriminate on in the case of non-substances sortal: two different colours are different the basis of the appropriate since no non-substantial colours and each is a colour. However, as an instance of the sortal in question, individual since such, is, white is not a distinct colour, such a discrimination Socrates' can, In the case of subin these cases, yield only a typical differentiation. stances, on the other hand, it is always possible for such a discriminadistinct instances of the sortal in question. tion to yield numerically claims of enumeration. are items which terminate We Substances count individual whites by counting white things and we count these simply by counting the things (men, chairs, tables) and not by counting or not, anything yet further. Now all the items, be they substances are in this way are synonyms. discriminated They synonyms in the introduced and in which that term is illustrated in chapter 1: sense Socrates and Plato, discriminated by means of the sortal "a man", definition applicable to them, are each men and have the appropriate and thus they are synonyms of one another; two colours will equally one to another. More importantly, stand in the relation of synonymy of what referred to by the sortal in question: are each is synonyms they Socrates is a man and can be called "a man" and will have the deffor white and a colour. to him, and similarly inition applicable item and the Since this is so, the relation between each discriminated referent of the sortal is that of 'being said of' (2 a 19-27, 3 a 17-20, 3 a 33-b 9). 'Being in' and 'being said of' are clearly distinct relations. This may In talking be satisfactorily brought out by the following consideration. of the thing that 'is in' something in the earlier paper I repeatedly and naturally used expressions of the form 'the o of a' - "Socrates' white". Now it seems to be distinctive of Aristotle's "Plato's literacy", that they cannot be said to be 'of' anything in a similar substances man of Socrates" does not make sense. One can use "of" "The way. with substance words but not in such a way as to suggest that the existence of that substance is somehow dependent on that of the thing of which it is. One can speak of the man of La Mancha, but Don Quixote would be who and what he is wherever he lived. Some philosto disguise this disparity by coining the ophers have attempted one can talk manhood of the manhood of of..."; expression "the 150

"the manhood of Socrates" refers most Unfortunately, readily to his virility and not to the fact that he is a human being. were used to refer to that in virtue of which Further, if "manhood" a human being is that, it must be objected that there are fundamental between "man" and "manhood". I can count under the disparities concept "man" ard achieve items such as Socrates and Plato, but the is to isolate Socrates' best I could do under the concept "manhood" manhood and Plato's and so on. Yet there is the need to recognize the without reference to someexistence of items which can be individuated thing other than what they themselves are - and manhood is not what Socrates is. This lack of an 'of' is not absolutely unique to substances. If we consider the relation between white and colour, we find the same Whereas we can refer to that in virtue of which Socrates impossibility. is white as "the white of Socrates", we cannot refer to that in virtue of which white is a colour as "the colour of white". White is a colour, it is not coloured (v. Top., 109 a 34-b 12). Equally the knowledge of literacy is different from the knowledge that is literacy. Accordingly, we must distinguish that in virtue of which Socrates is white from that in virtue of which he is a man and his white is a colour. (It should be noted that there is the same logical disparity between "white" and "whiteness" as that noted between "man" and "manhood". We can count using the sortal "white", but not using "whiteness". Indeed, it seems wrong to assert that whiteness is a colour.) Thus we can see that there is the same relation between Socrates and a man as between white and a colour, and that this relation is to from that between Socrates and his colour. be distinguished A consideration of counting, then, allows us to understand Aristotle's of to his two non-substantial individuals and postulation approach in relations of 'being present in' and 'being said of' such a way that Since a considerthey too are intelligible and readily distinguishable. ation of counting allows us to explain these two relations, it is clear that "the things that are" which are grouped by means of them cannot be such as to be incompatible with those conditions that have to obtain if an act of enumeration is to be possible. There are two that are especially germane to our present inquiry. conditions Firstly, whenever we count we have to count under some description or other, and the description must be such as to delimit precisely for any object whether or not it is an instance of the concept expressed That is, of each instance enumerable under the by the description." Socrates.

151

Each man must be a man and concept 'o' it must be true that it is a if we attempt to count, each colour a colour. The only alternative, is to miscount, and that is not a way of counting. Secondly, whenever I count under the concept 0, for me to be able to count a number of o's, there must exist 0's for me to count. As we have seen, to count no O's is not to count at all. Frege puts this tersely and well: "Affirmation of existence is in fact nothing but denial of the number nought".9 Sortal concepts are concepts which can have a number of distinct instances, and any instances that there are must a non-existent be existent: instance is self-contradictory. Therefore, whenever we have a case of counting it must be possible to frame the statement of the form result of the enumeration in an existential "There are n 0's". The 'things that are', then, must be things that exist. In that they exist and are countable, other than simply they must be something is not part of the subexistent. As Aristotle remarks elsewhere "existence the existent is not a genus" (Apo., 92 b 13-14). stance of anything; There are no such things as things which are simply what there are that exists must be something other than such things as. Everything merely existent, and, if it is to exist in such a way that it is countable or has countable instances, it must exist by being something, where the 'something' can be referred to by means of a sortal. Now, if this is so, we can see that 'being in a subject' must be 'exis7toxd[1.e:vov - is a general ting in a subject'. What is 'in' a subject it is 'in' (or 'of') was something property and it is only because what such as individual that there was a non-substantial individual, Socrates' colour. Now, it cannot be that what is 'in' the substantial and the non-substantial individual is both the general property the it is because individual; rather, general property is 'in' the sub'the property stantial individual that the non-substantial individual individual' exists. That is, if a property 0 in that substantial exists, there exists some subject and 0 exists in that it is 'in' at least one such subject. If a subject, s, exists and 0 is 'in' s, then 'the 0 in s' (e.g. the white of Socrates) exists. "is in" be taken existhat "is" in the expression This contention on the following textual grounds. Firstly, tentially can be supported the characterization of 'in a subject' (1 a 24-25) is patently existential.

152

v explains what it is that he is saying is 'in' a subject it as "what is in something not 8s Xlym 6... - and 7toxe:?[1.?vcp explains from what it is in". Secondly, as a part and cannot exist separately "for every colour exists in a 1 a 28 seems best taken existentially is of the claim that "the an elucidation and this example body" in a white subject". Finally, 2 a 27 speaks of "the is/exists particular things that are/exist in a subject". This description is clearly meant to embrace all the non-substantial types of existent given in chapter 2, and these, we have seen, as types of existent that are given there. Therefore, it is preferable to regard 2 a 27 as speaking of "those things that exist in a subject". The 'things that are', then, are things that exist and in existing are something, and, therefore, the four modes of being formed by the two relations 'existing in' and 'being said of' are modes of existing and we may present the divisions of 'the being something. Accordingly, things that are' in chapter 2 in the following form : and they are not (a) Certain items exist solely in having instances, such as to exist by being 'in' (or 'of') anything else (v. too 3 a 7-15), Aristotle e.g. a man ; (b) The instances of (a) must exist; like the items in (a) they are not which is an 'in' anything, but, unlike these, there is nothing instance of them; they do not themselves exist in having instances e.g. a particular man; Certain items exist in that they are 'in' some member of (b), and (c) they also have instances. These instances too must be in some member of (b). Thus, colour exists in that there are coloured bodies, and colour also has instances, the specifically different colours, each of which only exists, and hence can be an instance, if there are green and red and so forth bodies; (d) Certain items exist 'in' some member of (b) but do not themselves have instances, e.g. a particular literacy, Socrates' literacy. 4. Counting and Language :

It has been maintained very, very often that the Categories have an intimate connection with the analysis and explication of language, and this supposed has been most in analysis frequently presented terms of the sentence and the distinction between 'subject' and 'predicate'. It is worth pausing at this stage in our investigation to consider whether or not this is so. For to the extent that the activity of counting 153

is presupposed by what is said there, to that extent our ability to is use language presupposed. Showing this will also enable us to make a start in seeing what 'secondary sbustances' are. That is, we shall be able to see why Aristotle between and distinguished 'primary' and to see there are not substance', 'secondary why similarly 'primary' and 'secondary' and quantities and the like. Finally, we qualities shall be able to understand how Aristotle came to claim that "if the did not exist it would be impossible for any of primary substances the other things to exist" (2 b 5-6). This latter is problEmatic because we want to expostulate "But for a particular man to immediately: exist it must be the case that a man exists, and hence that the 'seconis one of the exists. After all, this 'secondary substance' dary substance' and if the appropriate things that exist in being said of something, the particular man, exists this must, ipso facto, exist." 'something', these in their turn, bring us closer to underwill, Discovering things what all notion of a 'primary substance' the is. standing important As an example of the construal of the Categories as offering a consider the following quotations from a paper theory of predication, Moravscik called on "Aristotle Predication".1 He claims to be by of one the nest of that Aristotle discusses in "selecting problems key these works", namely predication, and that Aristotle "discusses in the several interesting features of predication, and then Categories... between at least two different types of configuration distinguished that underlie predication,"11 'being in' and 'being said of', which Moravscik refers to as "inherence and predication" He respectively. and and the claim that terms begins by consideiing endorsing "general the correlated abstract singular terms, whether in subject or predthe same entity."12 Here "a general term" icate position, introduce is used as the equivalent of "the predicate in a sentence of subjectform".13 This claim is supported predicate by the following two we shall see that neither is independent of arguments (and argument the other) : in chapter four are "what 'expressions (i) The categories enumerated in no way combined' designate". "Given this fact, and the heterogeneous way in which the examples are specified, one must conclude that for Aristotle general terms

154

and the related tities. "14

abstract

singular

terms

introduce

the same en-

that underlie 'being in', is "one of the configurations (ii) 'Inherence', Aristotle says that: predication". " ... while the name of what inheres cannot be predicated of the to what term correlated inheres the can be so subject, general predmakes sense only if one assumes that icated. This characterization a sentence like 'Socrates is patient' is a configuration underlying of patience inherent in Socrates, and what inheres can be introor 'patience'. "15 duced by either 'patient' it seems as though Moravscik understands However, by "the related/ correlated abstract singular term" a word that can occupy the subject form. For he goes on to position in a sentence of subject-predicate remark: "It seems that Aristotle was strongly impressed by the fact that in one context is that what we use to describe and characterize which we specify as the subject in another context. Indeed, he that unless an expression introduces an might have thought also by a subject expression, it entity that can be designated cannot be a genuine predicate". 16 to deny that there is a theory of predicaNow, if we are to attempt in the Categories, we have to show that tion expounded or presupposed we shall try to show that it Specifically, (i) and (ii) are inadequate. is senseless to claim that the members of the categories are both items designated in no combined" and what are way by "expressions and correlated both by designated predicateby subject-expressions and that these two ways of referring to the members expressions, and further show that "predicate" of the categories are equivalent; observation that while the bears a different sense in the Aristotelian of the the term can be name cannot be general subject predicated from the sense appropriate to "predication" and "predicate" as these are used by Moravscik in expressing his own claims. expressions We shall argue these matters on independent grounds and then show how considerations of counting the possibility that the preclude be taken as expressing in the Categories can legitimately observations or expounding a theory of predication. At the beginning of chapter two Aristotle says:

155

"Of things that are said, some involve combination while others are said without combination. Examples of those involving combination are 'man runs', 'man wins'; and of those without combination 'man', 'ox', 'runs', 'wins"' (1 a 16-19 trans. Ackrill). Now, this could mean a number of things, depending on whether one "of things that are said", as takes the expression involving a reference to things that are spoken of or as involving a reference only to words that are spoken. That is, Aristotle could be saying : some are uttered alone and some are (a) When words are uttered, uttered in such a way as to form part of a sentence ('in combination') ; or else: (b) When things are spoken of, they are spoken of sometimes by means of single words and sometimes by means of sentences. Adoption of (b) here entails claiming that, once more, there is a conthat Aristotle means by "things spoken fusion of 'use' and 'mention'; of" "the words used to speak of things when the things are spoken of". However, whichever way we take this, it cannot be the case that the in chapter four are introitems in the various categories enumerated as the of duced by Aristotle referents subject expressions in subjectas the referents of words predicate sentences; they must be introduced uttered without being part of a sentence ; they are uttered or spoken of witlaout combination; if they are part of a sentence or what is referred Yet to by part of one, they are uttered or spoken of in combination. the two. it would seem, Moravscik, as we have seen, equates Rather, Aristotle must have in mind the fact that we do not always use senwe use single tences in speaking of or to signify things; sometimes words (and phrases: cf. "in the Lyceum," "in the market-place"; of this that come to mind are giving 'elliptical' 2 a 1-2). Examples and calling someone. Both of these examples answers to questions are given a predominant of place in the context of the enumeration the categories in the Topics, and this context also speaks in favour of adopting (b) above. Let us take the case of calling someone first since this is clear-cut. There comes a point in the Topics when Aristotle wishes to maintain that "the same" can be used in such a way as to suggest that an accident is the same as a substance: " ... a third use is found when it is rendered in reference to what or that-which-is-artistic is accidental, e.g. that-which-is-sitting 156

(sc. is the same as) Socrates; for all these mean to signify what is one in number. That what has just been said is true one would learn especially from cases where people change the calling for often, when we give an order to call by name (7tpocr-ryopocc;); one of the people who are sitting down, we change (sc. the calling) when the man to whom we gave the order happens not to underbetter from what is understand stand ; he will, we suppose, and we bid him call to us 'the one who is sitting' or accidental; 'the one who is talking' Tdv 81Xcy6ycvov. Clearly we suppose that we signify the same one according to the name and according to what is accidental" (103 a 29-39, author's translation). This speaks for itself and it shows that Aristotle was, at the least, aware of the activity of using an isolated expression ("the one sitting", "the one talking", "the artistic one") to call someone and to refer to someone. In Topics, I, 9 Aristotle enumerates the categories and then turns to make some observations about the use of the label "what it is" - Tt um - which can be used with reference to the members of each of the categories. He supports this by means of the following examples: "For when a man has been exposed one says that To XXd[1.EVO\lis - a man or an animal, what-has-been-exposed when and one says what it is and signifies the substance a white colour has been exposed one says that what-has-beenexposed is white or a colour, and one says what it is and signifies 'of what quality' Equally too if a six-foot length has been exposed is a six-foot length; exposed one says that what-has-beenone says what it is and signifies 'how much' -noa6v." (103 b 29-35; author's translation). Here one says of some object that has already been picked out, or and this is considered equivalent ex-posed, that it is such-and-such, to giving an answer to the question "What is that?" or "What is it?" (v. too 102 a 31-35). In answer one says "It is a as follows: This type of case is then characterized "Each of these signifies what it is if it is itself spoken of concerning itself and if the genus (sc. is spoken of) concerning this" (103 b 35-37; author's translation). To say of a o that it is a 0 is to say what it is and this is, in this type of case, to call it (or speak of it as) the thing itself. It seems clear that this type of situation is conceived by Aristotle as the uttering of a 157

word or phrase which refers to the thing itself, rather than the utterance of a sentence which makes an assertion about the thing. He concentrates on the word "man" and not on the sentence "It is a man". The use of the word that he has in mind is to refer to a particular man, the one who has been exposed, or picked out, and not to a species or the like. This seems sufficient to offer some support to the claim that "things spoken of without combination" (or "things uttered without combiare viewed as words or phrases used to refer to a isolated nation") object. This seems reasonable enough in itself. Otherwise particular we would (as Moravscik does) have to view Aristotle as considering words and phrases uttered as part of a sentence with disregard of the fact that they are part of a sentence, and this enterprise seems to me somewhat incoherent. a second on the last quotation given from It is worth concentrating the Topics. An expression signifies what a thing there before one is "if it itself is spoken concerning itself". What is spoken must be the expression, one would think, since this is also what signifies. But it is spoken "concerning itself" and this cannot be the expression but the thing referred to. This particular conflation of 'use' and 'mention' seems to support the suggestion that in the Categories Aristotle is not concerned of words but with the use of simply with the utterance words uttered to refer to things, or, better, to speak of things. Accordingly, it seems best to adopt something like (b) above. We have seen that it is a mistake to construe the items given as examples of the categories in chapter four as introduced as what the of subject-predicate sentences refer to. Rather, subject-expressions are the sorts of that can be to referred they thing by words that are in of not used the form sentences. This equation of "spoken without combination" and "subject-expression" was essential to the attempt to represent the Categories as dealing with predication. Let us turn to (ii), the argument denial that built on Aristotle's the name of what is present in something can be 'predicated' of what it is present in and assertion that "the general term correlated to what inheres" can be predicated. This claim and this assertion is explained in terms of 'predication'. However, it should now be obvious that when Aristotle speaks of 'predication' here he need only have in mind the use of an isolated expression to refer to something and not any kind of linguistic performance that we would describe philosophically as "predication" that it need not involve the utterand, particularly, 158

form. Thus, Aristotle does not ance of a sentence of subject-predicate illustrate his claim and denial by saying that "we utter the sentence 'A body is white"' or "we say 'A body is white"' but by remarking that "a body is called white" (2 a 32). to Thus, the Categories cannot be itself treated as a contribution of it takes its an account since predication, starting point, on the that are not the utterance of subjectlinguistic plane, from utterances it sentences. is tendentious to the word Equally, equate predicate with the concept of predication translated "predicate" that is employed in connection with the analysis of subject-predicate to deny that Aristotle sentences. This, however, is not necessarily does attempt to offer an account of what we would call "predication" in other works and that he uses our ability to use words in isolation in order to explain the nature of a sentential utterance. If he does do he is but this does not that he is mistaken, this, suggest equally mistaken in the Categories. He would only be mistaken there if he meant his account to explain facts about sentential utterances, or, more facts about As it his language. procedure is the happens, generally, reverse of this. He takes what he considers obvious facts about lan(v. 2 a 19-34). guage and uses them to support his ontological theses If counting forms one of the bases for the work this is only to be expected. We can only count if we possess a language that is richer than the mere seiies of numerals, or even than the series of numerals to"That is n", where "n" is any numeral. gether with the sentence-form Since what is one is specified with reference to its being something and say other than merely one, to be able to point to something "That is one" entails having the ability to apply some countnoun, or sortal, to it. If it is one, then it is one something, one man or horse or geranium. This is well put by Frege: "If I give someone a stone with the words: Find the weight of this, I have given him precisely the object he is to investigate. But if I place a pile of playing cards in his hands with the words: Find the number of these, this does not tell him whether I wish to know the number of cards, or of complete packs of cards, or even say of honour cards at skat. To have given him the pile in his hands is not yet to have given him completely the object he is I must add some further word - cards, or packs, or to investigate; honours. "1'

159

too is aware of this. "A unit is one such-and-such qua such" 1001 More he a in B, 26-27). specifically, argues Metaphysics, I (Meta., that "what is one in every genus is clearly some nature and this itself, the one, is not the nature of anything, but just as among colours one colour must be sought as what is actually what is one, so too one substance must be sought as what is among substance what is one" a (1054 9-13). actually for us to be able to count we must already be able to Furthermore, recognize any object as either being or not being a case in point. If we men are counting men in the room, we must be able to distinguish from objects other than men, like chairs, tables, packs of cards and pairs of shoes. This means that we must already be able to apply the word "man" to the men in the room correctly. As Aristotle supposes in the passage about the categories in the Topics, the objects in question are already 'exposed'. Finally, it is worth observing that the two types of sentence most with counting of are, neither of them, probably readily associated form. We with use sentences demonstratives as the subject-predicate grammatical subject (e.g. "That is one man") or existential sentences ("There are five of them"). Let us note an immediate consequence of this for the interpretation of the Categories. It has been suggested that the distinction between and 'secondary substance' 'primary' (2 a 11-19) is based on the supposed fact that a proper name cannot form a predicate expression whereas other expressions applicable to a substantial individual nameable by such a proper name (e.g. "a man") can be either the subjector the predicate-expressions.l8 Since a distinction between subjects and predicates is not, as such, to the point in considering the doctrines of the Categories, this cannot be the basis for the distinction between and substance'. 'secondary 'primary' Aristotle 5. The Priority of Primary Substances and the Ambiguity of Eidos :

of the preceding section we formulated In the first paragraph a over Aristotle's claim that "if the substances did not primary problem for any of the other things to exist" exist it would be impossible substances' exist in that they are said of (2 b 5-6). Since 'secondary

160

primary substances, if a primary substance exists then the appropriate secondary substance must ibso lacto exist. Therefore, there cannot be a question of any existential priority for primary substances. Yet this must surely be the thrust of Aristotle's remark. how Aristotle could have I shall make one attempt to understand come to make this claim which will prove not altogether satisfactory. will enable us to adopt a more satisSeeing how it is unsatisfactory factory approach. The word used for that secondary substance which "is more a substance than the genus" (2 b 7-8) is eidos. Traditionally, this has been "form" and "species". Now, a species, given two distinct translations More generally they are as we all know, is an aggregate of individuals. one of the things that something is, what it is - Ti !cr1'L.Thus the eidos of a particular man is... well, either a man (the form) or man (the we might say that Aristotle adopts two apspecies). Accordingly, proaches to what something is, which we may call a 'property'-view of what something is and an 'aggregate'-view. Thus, a property-view would concern itself with that in virtue of which each individual is would concern itself with an what it is, while the aggregate-view These approaches are distinct. For while aggregate of such individuals. an aggregate must have a plurality of members to exist, a property can exist with only one instance. Thus, in the De Caelo, Aristotle argues that even though it is possible for there to be a plurality of universes, if we regard 'a universe' simply as a form, there can only be one universe in fact, given that this form is that of 'a universe'. There could be no instance (277 b 27an additional surplus stuff (matter) to constitute 278 b 8). One might support the allegation of an aggregate-view by pointing to the fact that "the genus is always said of more than the eidos" (Top., 121 b 3-4) and that "the genus is said of most of all" a 30-31). It could then be suggested that in the Categories we find both a and that it is through a confusion of property- and an aggregate-view, these that Aristotle feels his claim of the priority of primary substance, to be justified. For Aristotle clearly holds that each man is in and of lq himself a man he is 'the certain man' (1 a 22, the role of the 2 a 13) and also 'this man' (3 b 10-23). Accordingly, However, by the individual. property of 'being a man' is pre-empted 'man' is the since existentially posterior to the existence of aggregate it must seem as as members of that aggregate, (some) individuals did not exist the aggregate would not though if the individuals 161

must be prior to the eidos. This and therefore the individuals constituent individual over the aggregate of constituted is priority also maintained by Frege: "A class, in the sense in which we have so far used the word, consists of objects' it is an aggregate, a collective unity of them; if so it must vanish when these objects vanish. If we burn down all the trees of a wood, we thereby burn down the wood."19 This approach would seem to explain how Aristotle could have held over aggregates. It would not, to the existential priority of individuals here show that these individuals are prior to the existence of however, the aggregate. Thus, it is not true that the wood in Frege's example must have ceased to exist when all the trees that compose it at any one time have ceased to exist. New trees could have grown during the of the old. Equally, and fortunately, it is course of the destruction of man that the existence the was not true species contingent solely upon the existence of Aristotle and his contemporaries. However, when we come to examine the text of the Categories for of eide, we find and the aggregate-treatments traces of the propertylike the propertythat matters are not this straightforward. Something view is to be found there: exist, seems to signify 'this something'. As regards "Every substance it is indisputably true that each of them the primary substances, for the thing revealed is an individual signifies 'this something'; and one thing in number. But as regards the secondary substances, though it appears from the form of calling TM a man or an animal - that a yop[ot4 - when one speaks of 'this this is not really substance something', secondary signifies true; rather it signifies what like something (poion ti) ; for the subject is not, as the primary substance is, one thing, but man The species and the genus and animal are said of many things... for they signify what mark off what like concerning substance b 10-16 and 20-21; author's like a substance 7tO?&.v Tcva (3 trans., after Ackrill). It might be suggested that what is of concern here is the fact that such as "man" and "animal", are grammatspecies- and genus-terms, ically singular. (This is suggested by Ackrill's rendering of rrp cx(y In this case the worry as "the form of the name"). 1tP<Tl1YOp(lXt; would be the same as that voiced by Russell:

162

"We may ask... what is to be said of the objects denoted by 'some man' and 'any man'. Grammar 'a man,' 'every man,' treats them all as one. But, to this view, the natural objection is, which one? Certainly not Socrates, not Plato, nor any particular persons is not all that is to However, I feel that this grammatical singularity to do with the point here. For what Aristotle stresses is something we call each individual man "a Now, certainly itpOG'Y)YOPLIX. 'calling' man". But what we do not do is call the 'collective unity' of all men "a man". The human species is not a massive sprawling man, a homo in Greek, you cannot use anthropos of the whole sapiens. Equally, of men assemblage (which is not to deny that you can make general statethe point ment about all men by saying "A man is...".) Accordingly, will be not that "a man" is used both of the individual man and of the species 'man', but that this man here before me is not only 'this man' but also a man. This, however, shows that we cannot find the existence in the text. 21 of the appropriate aggregate-view at first of an For the remarks that, sight, look like statements be more otherwise. view can now satisfactorily explained aggregative I mean the following remarks: for substance) signifies what like something; "(sc. secondary the subject is not, as the primary substance is, one thing, but the man (or "a man") and the animal (or "an animal") are said , of many things" (3 b 15-18), and "One marks off more with the genus than with the species for the one speaking of an animal takes in more than the one speaking of a man" (3 b 21). For, given that "a man" cannot be applied to the species 'man', it cannot be the case that "a man" is "said of many things" because it also refers to the species. Rather the point would seem to be that Each particular man is 'this made at the end of the last paragraph. is this man there are other man not necessarily man' but a men, each of which is equally (3 b 33-4 a 9) a man, and can just as well be called "a man." However, this man is himself a man and can be called "a man", and therefore, it might seem, says Aristotle, as though "a man" signifies a 'this'.

163

of the eide of the is a further reason against a treatment in terms. We have seen in connection with aggregative Categories trees that what a of the wood and the such treatment Frege's example would show is the existential priority of individuals in general over the aggregate they compose - individuals in general and not any particular the individuals since these latter could all cease to exist without to exist. Aristotle However, aggregate thereby ceasing appropriate must be making a stronger claim than this. If 'the certain man' it cannot is 'one in number' and so is a unit in a possible enumeration, man. be that 'the certain man' is any man - is simply an individual a number of men. We For then we would not be able to enumerate the point to one man and say "That's one man". But if, to enumerate to to one an individual we have man, man, number, point simply each time and correlate what we point to with a different numeral, I could point to the same man a number of times, and say on the first occasion "That's one man", on the second "That's two" and so on. But then we could never reach an end to counting any number of men, however small the number. There is no limit to the number of times I But then 'the certain man' would not be a. can point at anything. For no enumeration is possible. It possible unit in an enumeration. is obviously essential that I point to each man only the once. To point to the same man twice is to miscount. Therefore, if each man is 'one in number' this entails that he is 'this one and not that one'. It is 'this as opposed to the man and not that one' rather than 'this individual type of thing it is'. 22 reinforces our earlier suggestion about the role This consideration of the eidos-term "a man". In a room there is a man and a giraffe. How many animals are there there? Two - a man and a giraffe. But man for this to be the case there does not have to be this particular there; any man would do, so long as it is some man. Equally, there are two men in another room. But for this to be the case there do not have to be these men in the room - any two would do. Accordingly, "a man is said of many things". However, each 'thing' that it is said of must be this one and not that. Now it might look as though we have the makings of a defence for Aristotle. For there is no man other than 'these' men, the ones actually here. Any man there is to be found among them. But 'any man' is not another one beside them, hiding among them like a spy in a crowd. He is any of them, and so each of them, and only because they There

164

are there is any man there. Even if there were only one man there, then a man would be there: the man that is there is a man. But if no man were there, then there would not be a man not to be there. But, surely, one exclaims, if there was not a man there then that man would not be there. If there is not a man here, that man over there would not be here, nor would you be and neither would I. So it is just as much because there is a man here that any particular man is here as that it is only because the ones that are here are here that a man is here. And this is surely true. However, we have in the discussion up to now frequently spoken about "a man", - the word. "A man" is said of many things. We have had frequently to disentangle 'use' and 'mention' in our interpretations. Note especially our perplexities over whether "the things said / spoken are words or not, whether they are things of without combination" that are said and uttered or things spoken about. Accordingly, it would seem that these eide are the words we use in speaking of the things Aristotle does not simply we do speak about. With this difference: them as words. are characterized as things Although they present in that are said are said of of what they that exist (1 a 20-22 they et not as 'mere words', but as they are equally presented, indubitable realities along with such homely existents as 'the particular man' (1 a 20-b 9). In chapter five of the work, Aristotle suggests, and "signify as we have seen, that they "signify what like something" what like a substance" (3 b 10-21), where it seems they are treated as words in that they are said to signify23 but what they signify is presented as something other than what - T - the 'this something' roSe TL - is. If we adopt this suggestion, that Aristotle regards the words spoken of things as eide and 'hypostatizes' them, we can certainly explain how he could have come to do this. For we have seen in section (4) above that it is necessary if things that are 'one in number' are to there may be must be already far advanced exist that any enumerators in linguistic competence, if they have any such competence.24 Further-

165

one if more, it seems that things would only be, each one, enumerably there were enumerators In this Aiistotle presupposed. way argues in the Physics that there is such a thing as time only if there is such a thing as a mind, a Psuchi. He has argued that : " ... time is not movement, but only movement in so far as it admits of enumeration... Time then is a kind of number." (Phys., 219 b 2-5). On the basis of this he is able to make the following suggestion: if soul did not exist time would exist or not, is a question "Whether that may fairly be asked; for if there cannot be someone to count there cannot be anything that can be counted so that evidently there cannot be number; for number is either what has been or what can be counted. But if nothing but soul, or in soul reason, is qualified to count, there would not be time unless there were soul..." (223 a 21-26; trans. in such a way that they Hardie and Gaye). Similarly, things can exist and if there are enumerably distinct only if there exist enumerators, and these have any linguistic ability they must have exist enumerators sufficient ability to be able to speak of these objects in the appropriate distinct must exist in such a way. That is, things that are enumerably way that they can be spoken of. Aristotle then conceives this necessary condition of the existence of such items as the existence of a further class of objects which are related to them by the 'relation' of "speaking". It is as if he construed the verb "to say" by analogy with "to compare". Just as we can compare one thing with another (and they both have to exist to be comparable), so we can say one thing of another. This account is very like that of Miss Anscombe, with the difference observation but is embedded that it is not, unlike hers, an independent into a general pattern This gives the fmther adof interpretation. in her account can be vantage that what seems somewhat arbitrary seen as reflecting an essential truth about the situation presupposed Aristotle. She writes: by "Locke said that if you take a proper name, 'A', you can only discover whether A is, say, a man, or a cassiowary, by looking This preto see if A has the properties of man or a cassiowary. supposes that, having grasped the assignment of the proper name 'A', you can know when to use it again, without its being already determined whether 'A' is the proper name of, say, a man, or a as if there were such a thing as being the same without cassiowary: This is clearly false. Aristotle's being the same such-and-such. whatever it is, 'second substance' is indicated by the predicate, 166

say 'X', that is so associated with the proper name of an individual that the proper name has the same reference when it is used to refer to the same X. "25 The arbitrary factor in this account is the one that A. C. Lloyd to.26 points Why is there this demand for the possibility of 'referring to' ? How is it that the things existing brutishly there are so essentially intertwined with the referrings and identifications2' of human beings. the account of 'secondary in the substances' However, embedding context of enumerably discrete things forces on us considerations of our and activities. language linguistic It is worth noting, though, that it is not just a case of confusing the words we use to refer to the things with a special class of objects, eide. For we can construe a man as a possible unit in an enumeration without there being any particular man that he had to be - in the of the room the man and the giraffe, where we example containing just to how wanted discover many animals there were there. A difference between the Categories and the passage in the Topics dealing with the categories (A, 9) is worth noting. In the latter the that what is said of the man that is exposed possibility is entertained is the thing itself. A man is said of the man over there and the man that is said of him is the man over there, the same one. In the Categories,, by contrast, what is said of something is presented as always 'something else' (heteron) (1 b 10), and as a 'what like something' (poion ti) rather than the same thing that (ti) the thing exposed is, what (ti) it is. We have reached the following stage. Aristotle draws a distinction between 'primary' and 'secondary substances' because, while 'primary substances' are things that exist and not words, the latter are words as things. However, it is not open to Aristotle simply misrepresented to identify 'secondary substances' with words; for he recognizes that a man can exist without there being any one particular man that exists. However, this is not yet sufficient. For given that Aristotle cannot identify them with words spoken, why cannot he simply identify them with what a given substantial individual is? A 'secondary substance' then would be not the highly indirect 'what sort of what' (potion ti), that he but, simply, what the thing is. This, after all, is something comes to maintain in the Metaphysics.

167

To discover this we clearly have to look at the concept of a 'primary as that concept is employed in the Categoyies. substance' Substance:

(vi) Primary

is presented in the Categories not simply The particular individual as a 'this' but as a 'this something' 1'6e: rt, and this is what the 'primary for example, out (3 b 10-13). A 'primary substance' is, substance' marks 'the certain man' (2 a 13), and this is already something, a man. in a peculiar way. It is introsubstance' is characterized 'Primary duced as follows: and primarily and "O'Usia is what is spoken of most prominently most of all, viz the one which neither is said of a subject nor exists horse" man or the particular in a subject, e.g. the particular trans. after a Ackrill). (2 11-14, is purely negative. For we This is peculiar in that the characterization as the subject in which other things might expect it to be introduced are and of which they are said, especially since he does refer to it, in passing, as a subject (1 b 12-13, 3 b 16-17). Since he does not introduce it in this way, we might suppose that there is a distinction between the subject of which things are said and in which they are and the primary substance. Such a distinction would be that between the thing itself and its mode of being. We would then take the characof primary substance as giving what the primary mode of terization existence of such a subject is. characterization and fundamental We already know one important of such subjects. They are things that are enumerably one, and enumerably one in the sense that they can be included in any enumeration A unitary colour could be solely on the basis of being themselves. else 'in' which included only because there always exists something one that his colour is it exists. It is only because Socrates is thing one colour, but Socrates is one thing because he is in himself one thing. Any such object, then, must exist in such a way that it is, of What we must beware of at all individual. itself, a discriminable to Aristotle what we could call 'a thesis of costs here is attributing are a thesis to the effect that these individuals pure instantiation', the purely formal things which simply exist, which only instantiate he the sugWe have seen that o2csia. 'substance', repudiates concept are what there are such that there are which things simply gestion of what he a as. In the general specification Categories gives things 168

and kind of thing these must be - they must be 'one in number' of the kind of meant a man or horse. thing particular examples These things, then, cannot be pure occurrences, as it were; they must be things that exist in such a way that they are enumerably one, and, we have seen, enumerably one on the basis of their being what it is that they are: the certain man is one man and that is what he is, whereas the certain colour is, for instance, Socrates' colour but that is not what it is. In chapter two, these enumerably singular items are in terms of the modes of being 'being said of' and characterized substance' 'existing in'. Since it is in terms of these that 'primary is identified, we are justified in taking 'primary substance' to be the mode of being, or existence, of something that is enumerably singular, and enumerably singular in virtue of itself. the certain man the examples of 'primary substance', Accordingly, and the certain horse, will not be just particular objects that are of of modes of a certain class but viz. found, being objects found carried this ones. Before on line of getting away singular enumerably let us pause to demonstrate that ousia, 'substance', can thought, mean 'being' in the Categoyies. This, of course, is what we would expect from a linguistic and grammatical point of view. Ousia is the verbal it is obvious that noun of einai, "to be" or "to exist". Furthermore, Aristotle does recognize the common-place "existential" force of the word. Having introduced primary ousia by means of the characterization "the one which is neither said of a subject nor exists in a subject" (2 a 12-13), he argues, in the course of chapter 5, that: "It is reasonable that, after the primary ousiai, the species and genera alone of the other things are spoken of as (secondary) ousiai" (2 b 29-3 a 6). He also argues in favor of each of the marks of 'primary ousia' men5 (2 a 11-12). Its 'prominence' tioned at the start of chapter (2 b a 19-2 b 37-3 a 1), 'primacy' 6, v. esp. 2 b 5-6, with which cf. (2 Meta. A, 1019 a 1-14)28, and its 'most of all' (2 b 7-22). All this makes it overwhelmingly unlikely that ousia is used without any consideraforce and simply as a pure label for a tion of its common-place certain class of existents. There are, it seems to me, two indisputable occurrences of the word in this sense in the work, and the second one

169

shows that this is the sense it bears in the expressions "primary ousia" and "secondary ousia". the notions of homonymy When he introduces and synonymy in chapter one, great use is made of "the definition of being which corresponds to the name" (1 a 1-2, 4, 7, 9-10). It is agreed that this, ho logos... tis ousias, is indeed the definition of being. It is not simply the definition of being, being impossibly pure and simple, but the definition to the name. The man and the ox have corresponding in identical ways of being, where the being question is that corresponding to the name "an animal". The man and the man in the picture, on the other hand, have different ways of being corresponding to the name "an animal". One exists by being a picture or portrait, the other does not. Such a definition is correlated with "what being a o is for each of them" (1 a 5, 11). The relevant ways of existing are man exists by being 'whats'. A particular ways of being something, a man 73 b 83 a 5-8, being 24-32). (cf. In chapter remarks that ousia, Aristotle five, while discussing "it is a characteristic of... ousiai that all things that are called from them are so called synonymously" (3 a 33-34). Now, it hardly seems true that all things that are called the same name as any substance are so called synonymously. The man and the man in the picture are and not homonyms, synonyms, yet both are men. However, if we recall the 'twist' that is given to the notions introduced in chapter one, it will be seen that what Aristotle is saying here is that anything that is called after its own ousia is so called synonymously, and this is true. Thus, the man over there exists by being a certain man. Anything that is said of him, e.g. 'a man', is said of him synonymously, if what is said of him is his way of being, and his way of being is that of being a man. This remark about synonymy is made in connection with primary and secondary ousiai. Therefore, we may feel justified of ousia as 'being' or 'existence', an in retaining the interpretation on This reached hitherto grounds. interpretation purely general Acsomething. primary mode of being is that of being a particular cordingly, the role that we might expect the secondary modes of being to perform, that of providing the 'something' that the thing is to exist by being a particular one of, is already pre-empted by the primary mode of being of the thing. Thus, being a man, and the other secondary modes of being, are squeezed into an ontological limbo, and must remain so until either they are abandoned as fictions and figments or else the individual is no longer represented as existing by being 170

itself. (I speak of "existing by being itself" for the descriptions of the modes of being of these individuals is also a description of the individThe certain man exists by being the certain man.) uals themselves. mode of being of such a thing is that of being a man, The secondary a man as opposed to this man. There can, to repeat, for example man be a man there without there necessarily being this particular there. It seems mistaken, however, to hold that being a man is something different from being this man; for any man that is there will be one particular man and not another one. We have analysed the distinction between 'primary' and 'secondary into three factors. Firstly, the mode of being represented substance' the by 'secondary substances' is already pre-empted by the individuals. hover between being things spoken substances' Secondly, 'secondary is unwilling to assimilate of and things spoken. Thirdly, Aristotle substances' either to substances' or else to a 'primary 'secondary he that a man status because can exist recognizes purely linguistic man existing. It will readily be seen that without any one particular in producing a distinction between 'primary' the motives operative be operative in producing substances' cannot and 'secondary a between and distinction 'primary' 'secondary qualities', parallel or the like. Aristotle does not feel the temptation or 'quantities', to with our ability assert that the existence of qualities is concerned to use a certain class of words. Above all, whereas being this man is already being a man, being 'the certain colour' is not similarly being a colour is not, as such, a colour. Consequentcolour; for something's colour that the by being in something is that thing's colour can ly, be identified with the colour that it is; a move which was not readily available in the case of substances. (vii) Conclusion: of the Categories (other than the doctrines By following through on the basis of the characterization of the categories themselves) as what are 'one in number', we have been able to unite individuals these various doctrines in a coherent and intelligible way. The one area that has proved impossible to untangle in a defensible way is and 'secondary the distinction between 'primary' being'. We were in terms of a confusion of language able to explicate this distinction and things spoken of and the way in which what for Aristotle are as existing the fundamental class of 'things that are' are represented by being just exactly themselves. 171

Following this route has produced an interesting picture. We have seen how the existence of things is approached with a tacit framework which already presupposes the existence of human beings able to how this linguistic ability is worked into the speak a language, and existential fabric of the things themselves. The other interesting result is that this 'enumerative' approach to things has revealed Aristotle's with what we might call the 'radical' individuality of preoccupation the primary class of things that are. These are not simply individuals and members of classes; they are, each one, not another one; they are, each one, this one and not that. Of course questions now arise. In particular: What is so special that this can be assumed as a privileged means of about enumeration access to what there is?2a Princeton University

172