Essentialism and Ontological Interdependence in Aristotle's "Categories" more

ESSENTIALISM AND ONTOLOGICAL INTERDEPENDENCE IN ARISTOTLE’S CATEGORIES Riin Sirkel The aim of this paper is to develop an essentialist interpretation of Aristotle’s Categories, according to which particulars (“primary substances”, prôtai ousiai) and universals (“secondary substances”, deuterai ousiai) are ontologically interdependent, i.e., their ontological dependence is not asymmetrical but mutual. This interpretation challenges three longstanding and deep-seated views about Aristotle’s Categories. Firstly, Aristotle believes that particulars enjoy ontological priority over universals in the sense that particulars can exist independently of universals, whereas universals depend on particulars for their existence. Indeed, according to what could plausibly be called a traditional interpretation, the relation of ontological dependence between particulars and universals has been seen as asymmetrical. Such asymmetry has been regarded as the “lynchpin of Aristotelian metaphysics”1. Secondly, in treating particulars as ontologically independent from universals, Aristotle commits himself to an ontological dualism of particulars and universals, which does not seem to be much different from Platonic dualism. The only difference, though of course not a small difference, is that Aristotle attributes ontological priority to particulars and denies that universals could exist without particulars instantiating them. Thirdly, in assigning ontological priority to particulars, rather than to Platonic Forms or anything resembling them, Aristotle turns the Platonic picture “upside down”. The Platonic dualism implies that universals – that is to say, Forms – are ontologically prior, while particulars have being only in so far as they participate in such Forms. My interpretation suggests that Aristotle does not attribute to particulars any unqualified ontological priority, and hence he is not committed to any                                                          I take this expression from Corkum’s (2008, 65) characterization of the traditional position. 1   1  robust dualism. I begin by explaining the notions of the particular and the universal, and examine the argument that is supposed to lead to the ontological priority of particulars. Then I develop the essentialist interpretation of the Categories and conclude by examining the consequences of my interpretation for Aristotle’s alleged anti-Platonism. PARTICULARS AND UNIVERSALS Aristotle does not use his standard terms for “universal” and “particular” (katholou and kath’ hekaston2) in the Categories. Instead, he relies on two phrases: “being said of a subject” (legetai kath’ hypokeimenou) and “being present in a subject” (en hypokeimenôi estin).3 Although the language Aristotle uses might suggest otherwise, these two phrases do not merely express linguistic relations, but above all they express ontological relations. More precisely, we will see that Aristotle construes them as relations of ontological dependence – both things that are “said of” and those that are “present in” a subject depend on the subject for their being. The traditional position4 holds, and I think correctly, that the “present in” relation (often called inherence) distinguishes substances from quantities, qualities, and other categories that the tradition has lumped together as “accidents”, and what we would today call “accidental properties”. Accidents are always present in a subject (they are always accidents of something), while substances (neither primary nor secondary ones) do not inhere in anything further (they cannot be said to be of anything in a similar manner). The “said of” relation (often called predication), on the other hand, is held to distinguish particulars from universals – universals are said or predicated of a subject, particulars are not. This position relies                                                          The term kath’ hekaston occurs once (at 2b3). Aristotle introduces these relations, somewhat abruptly, in chapter two of the Categories, and relies on them in his discussion of primary and secondary substances in chapter five. 4 See Ackrill (2002 [1963], 74). Since the publication of Ackrill’s translation and commentary on the Categories (1963), the traditional position is usually equated with his view. But see also Granger (1980), who offers a well-written overview and defense of the traditional position. 2 3   2  on Aristotle’s “standard” definition of the “particular” and “universal” in the De Interpretatione: “I call universal (katholou) that which is by nature predicated (katêgoreisthai) of many things, and particular (kath’ hekaston) that which is not” (17a38). Thus, these two relations distinguish substances from accidents, on the one hand, and particulars from universals, on the other. But Aristotle, at least in the Categories, does not see these distinctions as coinciding. Rather, they cut across each other, giving rise to the so-called fourfold division of “of things that are” (tôn ontôn, “of beings”) which is presented in chapter two of the Categories: Of things that are: (i) some are said of a subject but are not in any subject. For example, human being is said of a subject, this human being, but is not in any subject. (ii) Some are in a subject but are not said of any subject. (By “in a subject” I mean what is in something, not as a part, and cannot exist separately from what it is in.) For example, this knowledge of grammar (hê tis grammatikê) is in a subject, the soul, but is not said of any subject; and this white (to ti leukon) is in a subject, the body (for all colour is in a body), but is not said of any subject. (iii) Some are both said of a subject and in a subject. For example, knowledge is in a subject, the soul, and is also said of a subject, knowledge-of-grammar. (iv) Some are neither in a subject nor said of a subject, for example, this human being (ho tis anthrôpos) and this horse (ho tis hippos) – for nothing of this sort is either in a subject or said of a subject. (1a20-1b9).5 The combination of the “present in” and “said of” relations thus yields a distinction between two types of particulars (i.e., (ii) and (iv)) and two types of universals (i.e., (i) and (iii)). The former distinction has been the main focus of recent literature on the Categories. In particular, there has                                                          5 Here and in what follows I rely on Ackrill’s translation of the Categories (2002 [1963]). However, I do not follow Ackrill in all the details, and translate some of the expressions differently, e.g. while Ackrill translates ho tis anthrôpos as “the individual man”, I translate it as “this human being”.   3  been a fierce dispute over the precise nature of these things that are present in but not said of a subject (whether they are recurrent or nonrecurrent properties).6 It is not important for my present purposes to take sides in this issue, and I shall be content to agree with the traditional position that Aristotle is referring here to particulars in categories other than substance. I will focus on those things that are neither said of nor present in a subject – these are the ones Aristotle calls in chapter five of the Categories “primary substances”, such as “this human being” and “this horse” (2a11-14). It is generally agreed that when Aristotle speaks about primary substances, he has in mind concrete particulars, and, above all, naturally existing particulars (humans and horses). What is important for present purposes is the distinction between two types of universals: those that are both said of and present in a subject (e.g. knowledge, white), and those that are said of but not present in a subject (e.g. human being). The first type indicates universals in categories other than substance, whereas the second type corresponds to what Aristotle calls in chapter five “secondary substances” (2a14-19). Secondary substances include the species and genera under which primary                                                          6 According to one view, Aristotle’s reference to “this white” picks out a determinate property, e.g. a determinate shade of white, rather than a particular property unique to its possessor. The salient feature of this view is that, according to it, nothing prevents particular properties from being recurrent and repeatable. This view was originated by Owen (1965), and has been defended, most notably, by Frede (1987). But see also Furth (1988), Loux (2008 [1991]). According to another and more widespread view, Aristotle’s reference to “this white” picks out a non-recurrent property, i.e., a property that is peculiar to the particular to which it belongs. On this view, each white thing has its own, entirely distinct, property of whiteness (we might call such properties “tropes”). Owen calls this view “dogma” and equates it with Ackrill’s (2002 [1963]) view. See also (for a criticism of Owen) Moravcsik (1967), Granger (1980), Wedin (2005). According to a third view (which is a version of Ackrill’s view), defended by Matthews (2009), particular properties are non-repeatable instances of universal properties. This view emphasizes that although each white thing has its own property of whiteness, these particular properties themselves are instances of the universal property of whiteness (we might call this view “tropes plus universals”). As we will see, my interpretation offers indirect support to the latter type of view. I will argue that particular things (primary substances) are instances of universals, and hence it is reasonable to think that particular properties are likewise instances of universals.   4  substances fall, e.g. human being, animal, etc. This distinction is important because it can and has been interpreted as Aristotle’s (perhaps first) attempt to distinguish between essential and accidental predication: what is said of but not present in a subject is essential to its being what it is, whereas what is both said of and present in is accidental to its being what it is.7 In order to make referring to these types of universals easier, I will reserve the term “universal” for secondary substances, i.e., species and genera under which primary substances fall. When I want to refer to universals in other categories, I will use the term “accidents”. Now I will turn to Aristotle’s argument – sometimes called the “priority claim” – that is supposed to establish the ontological priority of particulars. PRIORITY CLAIM Aristotle begins his account of primary and secondary substances in chapter five of the Categories with the claim that “a substance (ousia) – that which is called a substance most strictly, primarily (prôtôs), and most of all – is that which is neither said of a subject nor present in a subject” (2a1113). Aristotle appeals to our linguistic intuitions about predication to tell us what the primary substances are. In a linguistic predication, there is a subject of predication and that which is predicated of a subject. But, to forestall a possible source of confusion: predication, as Aristotle conceives of it, is not merely a linguistic matter but also a matter of ontology. He thinks that when we predicate something of something (i.e., when we form subject-predicate sentences), we reveal (dêloi) something about something. More precisely, we reveal that something belongs (hyparchein)                                                          7 See Duerlinger (1970), who suggests that the predication-inherence distinction represents Aristotle’s “first attempt” to distinguish essential and accidental predication: “Aristotle does not exactly say that he is trying to distinguish what is essential from what is incidental to the nature of an individual thing, but his examples and statements strongly suggest that the attempt is being made” (p. 181). See also Loux (2008), who argues that all parties to the dispute over the precise nature of particular properties agree that “a distinction between two forms of metaphysical predication (what have been called strong or essential predication and weak or accidental predication) is operative in the Categories” (p. 21).   5  to something.8 So language is not a mere instrument for Aristotle – it reveals something about the way things are in the world. This suggests that there must be some sort of corresponding structure between the language we use and the way things are in the world.9 Now, in saying that primary substances are neither said of nor present in a subject, Aristotle identifies them with subjects themselves (hypokeimenon, lit. “that which stands under”). However, the fact that primary substances function as subjects does not yet distinguish them from universals because Aristotle claims that secondary substances can likewise function as subjects: “As the primary substances stand to everything else, so the species and genera of the primary substances stand to all the rest: all the rest are predicated of these” (3a1-6; 2b17-20). Indeed, if it is insisted, for example, that a species (e.g. human being) can function only as predicate, not as subject, then in the very act of insisting on this one makes the species a subject of predication. All Aristotelian commentators agree that the distinction between primary and secondary substances does not lie in the fact that primary substances are subjects simpliciter, but rather in the fact that they are primary or ultimate or basic subjects. The idea that primary substances are ultimate subjects is in scholarly literature often called the “subject criterion”.10 Aristotle himself does not express this point explicitly, but he does seem to think that primary substances are called substances most strictly because “all the other things” are predicated of them, while they are not predicated of anything more basic. So primary substances are the only subjects in the ontology of the Categories of which it is correct to assert that nothing                                                          See De Interpretatione 4 and 5 (esp. 17a15 ff.). Although it is reasonable to suggest that Aristotle (in so far as he insists that language or logos is revealing) is committed to the position that there exists an isomorphism between language and the world, it is important to notice that his commitment to such a position is not uncritical. As I will argue in the next section, Aristotle is well aware that the so-called subject criterion might not be the best guide for discovering primary substances. Thus, he seems to be aware that language is not a completely reliable guide to ontology. 10 See, e.g., Loux (2008, 23), Lewis (1991, chap. 2), Mann (2000, 24). 8 9   6  stands under them. This seems to be the core insight that leads Aristotle to his famous position that “all the other things” require primary substances as basic subjects for their existence: All the other things are either said of the primary substances as subjects or in them as subjects. This is clear from an examination of cases. For example, animal is predicated of human being and therefore also the particular human being (tinos anthrôpou); for were it predicated of none of the particular human beings it would not be predicated of human beings at all. Again, colour is in body and therefore also in a particular body; for were it not in any of the particulars (kath’ hekasta) it would not be in body at all. Thus all the other things are either said of the primary substances as subjects or in them as subjects. So if the primary substances did not exist it would be impossible for any of the other things to exist (mê ousôn oun tôn proton ousiôn adynaton tôn allôn ti einai). (2a34-2b6) Aristotle begins by claiming that everything which is not a primary substance is either said of and/or present in a primary substance. He tries to justify this claim on a case-by-case basis and argues that to predicate animal of human being is, ultimately, to predicate it of particular human beings for “were it predicated of none of the particular human beings it would not be predicated of human beings at all”. From this Aristotle draws a conclusion, the climax of the Categories, that if the primary substances did not exist, then neither would anything else. This is a strong conclusion. Aristotle is claiming that the existence of everything other than primary substances would be impossible were there no primary substances. From this it would follow that a universal conceived of as capable of existing on its own, independently of particulars, is an impossible entity – a fiction perhaps. This conclusion is evidently intended to show that universals are ontologically dependent on particulars, that they cannot exist independently of particulars. As such, this conclusion raises an important question, which is overlooked by most scholars. Namely, whether the   7  ontological dependence of universals on particulars is asymmetrical or mutual? Aristotle does not explicitly settle this issue, and those of his commentators who have addressed this question typically remain indecisive11. However, there is a tendency among commentators to regard the relation of ontological dependence as asymmetrical. It finds expression in the often-occurring talk about the ontological priority of particulars. When we ask what this ontological priority amounts to, then the most traditional (and perhaps natural answer) is that it is an asymmetrical relation of ontological dependence.12 The notion of ontological priority is traditionally identified with what Aristotle calls priority “by nature and substance”. In Metaphysics Δ 11, Aristotle states: “Some things then are called prior and posterior (protera kai hystera) ... by nature and substance, namely all things which can exist without (einai endechetai aneu) other things, whereas others cannot exist without them – a distinction which Plato used” (1019a1-4). There are at least three things worth noticing about this statement of priority. Firstly, ontological priority amounts to the capacity for independent existence, and is expressed in modal terms: A is ontologically prior to B if A can exist without B while B cannot exist without A. The conclusion of the above passage is presented in similar modal terms – universals cannot exist                                                          11 See, e.g., Ackrill (2002), who says that Aristotle’s conclusion is intended “to mark out primary substances as somehow basic (contra Plato). But the point is not very well expressed. For it may well be doubted whether (Aristotle thinks that) primary substances could exist if secondary substances and items in other categories did not do so (p. 83). See also Moravcsik (1976, 95), Shields (2007, 177-178). 12 Hence I think that Corkum (2008) is right to summarize the prevalent understanding of the Categories in the following way: “Aristotle holds that individual substances are ontologically independent from non-substances and universal substances but that nonsubstances and universal substances are ontologically dependent on substances. There is then an asymmetry between individual substances and other kinds of being with respect to ontological dependence. Under what could plausibly be called the standard interpretation, the ontological independence ascribed to individual substances and denied of non-substances and universal substances is a capacity for independent existence” (p. 65). See also Wedin (2005), who claims that “virtually all commentators assume that the Categories brand of ontological dependence is asymmetric” (p. 81).   8  without particulars.13 Secondly, ontological priority implies that the relation of ontological dependence is asymmetrical: A is ontologically prior to B just in case A can exist without, independently of, B but B cannot exist without A, i.e., the dependence is only one way. Thirdly, this notion of priority was, as Aristotle says, used by Plato. It is not overly surprising to find Aristotle endorsing what he takes to be a Platonic criterion for what is prior by nature and substance, given that Aristotle often uses the ideas of his predecessors for his own purposes. If the traditional interpretation is right, Aristotle uses this Platonic criterion of priority in a fairly radical manner, viz., he uses it to overturn the Platonic position. Nonetheless, it is important to notice that the above passage establishes only the ontological dependence of universals and accidents on particulars. It does not establish the ontological independence of particulars. In fact, nothing in the above passage excludes the possiblity that the relation between particulars and universals is one of mutual ontological dependence. Why, then, does the the traditional interpretation maintain that Aristotle is committed to the ontological priority of particulars? I believe that the traditional interpretation relies on two assumptions. The first assumption is that the relation of predication expresses ontological dependence – what is said of a subject is dependent on this subject for its being. The second and more important assumption is that the relationship between a subject of a predication and its predicate is irreducibly asymmetrical (i.e., predicates are predicated of subjects but not vice versa).14 Given these assumptions, it is clearly tempting to conclude that the relationship of ontological dependence between particulars and universals                                                          Aristotle discusses a similar form of priority also in chapter twelve of the Categories: “What does not reciprocate as to implication of being [is called prior]” (14a30). 14 This assumption has persisted successfully in the history of philosophy and has been challenged seriously only recently. One of the first authors to vigorously deny it is Ramsey (1925); he argues that there is no fundamental antithesis between subjects and predicates, and hence no irreducible asymmetry. 13   9  must likewise be asymmetrical. A further motivation behind the traditional interpretation might be Aristotle’s terminology. After all, Aristotle calls primary substances “primary” which might suggest that he attributes to them ontological primacy over universals.15 Finally, some16 contend that the conclusion of the above passage – “if the primary substances did not exist it would be impossible for any of the other things to exist” – would simply be pointless unless it is implied that the dependence is asymmetrical. However, there are reasons to doubt that Aristotle would be committed to this implication. The critical consideration here is the reason Aristotle gives for calling species and genera “substances”, which leads us to his essentialism. ESSENTIALISM It is well known that “secondary substances” are never mentioned, at least by name, in the Aristotelian corpus outside the Categories. So why does Aristotle call species and genera “substances”? There are authors, who think that genera and species of primary substances are substances because they can function as subjects, although not as primary subjects.17 But this is not the reason Aristotle himself gives for calling species and genera substances. Rather, he says the following: It is reasonable that, after the primary substances, their species and genera should be the only other things called substances. For only they, of things predicated (tôn kategoroumenôn), reveal (dêloi) primary substance. For if one is to say of the particular human being (tina anthrôpon) what he is (ti esti), he’ll do so appropriately (oikeiôs) by giving the species or the genus (though more informative is to give human being than animal); but to give any of the other things would be out of place – for example,                                                          See, e.g., Corkum (2008, 70). See Moravcsik (1967, 93). 17 See Kohl (2008), Moravcsik (1967), but also Lewis (1991) and Wedin (2005), who develop what Kohl (2008) calls “reductive accounts of subjecthood”. That is, they attempt to show that the subjecthood of secondary substances can be entirely reduced to that of primary substances. 15 16   10  to say “white” or “runs” or anything like that. So it is reasonable that these should be the only other things called substances. (2b30-37) The species and genera of primary substances are called substances, because only they, of the “things predicated”, answer the question “What is it?” It is Aristotle’s settled position that the answer to this question is a definition that reveals the essence (ti esti, to ti ên einai, ousia) of a thing.18 So we may contend that only species and genera reveal the essence of a primary substance.19 Aristotle thus draws a distinction between genera and species, on the one hand, and all the other “things predicated”, on the other. It is reasonably clear that this corresponds to the distinction between those universals that are said of but not present in a subject, e.g. human being, and those that are both said of and present in a subject, e.g. white (pale). These remain Aristotle’s stock examples. Practically all commentators agree that these distinctions correspond to those of essential and accidental predication.20 But even though it is generally acknowledged that Aristotle is, in the Categories, committed to such a distinction (to a position that can that can be labelled “essentialism”), it is not clear what this distinction amounts to.                                                          18 The definition (horismos) is described as “logos tês ousias” in the Topics (101b37, cf. 101b21, 103b9-10), and as “logos of what something is” (tou ti esti) in Posterior Analytics (93b30). Aristotle also links essence (to ti ên einai, lit. “the what it was to be” for a thing) with definition and a certain sort of essential predication also in Metaphysics Z 4, and argues that the essence of something is what the thing is “in virtue of itself” (kath” hauto), or, more precisely, “just what a this something is” (hoper tode ti). 19 I ignore here the difficulties surrounding differentia, and what category to put it in (whether it is substance, quality or something else). See, e.g., Irwin (2002 [1988], 64-66), who has a good discussion of the “anomaly of differentiae”. 20 That is, it is assumed that these distinctions correspond to the distinction which is most frequently marked by Aristotle by the terms “kath’ hauto” and “kata symbebêkos” (see, e.g., An. Post. A 4, 73a34-73b16, Met. Δ V 7, 18). The only author (to my knowledge) who denies that Aristotle is distinguishing in the Categories accidental from essential predication is Moravcsik (1976). His main argument is that if this distinction holds, then “the priority claim on behalf of substances becomes absurd” (p. 91). He does not argue for this position, and near the end of the article he contends that “we must set aside the thorny question whether Aristotle really means to draw an asymmetrical dependency claim between secondary and primary substances and to what extent” (p. 95).   11  Aristotle himself explains the distinction between secondary substances, e.g. human being, and accidents, e.g. white, in the following manner (2 a19-33; 3a10-20). He says that things are present in a subject when their definition is not predicated of the subject, although their name may be predicated of the subject. “For example, white, which is in a subject (the body), is predicated of the subject; for a body is called white. But the definition of white will never be predicated of the body” (2a31-33). Thus, accidents might share the same name with their subjects, but never the same definition and essence.21 Now, things are said of a subject when both their name and definition are predicated of it. “For example, human being is said of a subject, and the name is of course predicated (since you will be predicating human being of the particular human being), and also the definition of human being will be predicated of the particular human being (since the particular human being is also a human being)” (2a20-24). Thus, secondary substances share both their name and their definition with their subjects. On Aristotle’s account (3a35-b1), this makes this human being (e.g. Socrates) and the species human being “synonyms” (synonyma), i.e., things that share the same name (onoma) and the same “definition of essence” (logos tês ousias). This explanation suggests that although accidents are present in a subject, they are not part of the essence of the subject. When we predicate an accident of a primary substance like this human being, we are predicating a property of an independently existing thing. Accidents are properties that attach to the thing from outside, so to speak. But when we predicate a secondary substance, we are not just predicating a property; the predication reveals what the subject itself is.22 This implies that the subject                                                          21 This claim might be confusing, since we can evidently say that “white is a penetrative color” or “white is a color penetrative of sight” (this is how Aristotle defines white in the Topics, see 119a30, 158a38-b1). Aristotle’s point seems to be that in such predications the subject is something that is white. So to speak about “white” is a shorthand way of speaking about whatever happens to be white (e.g. Socrates, or, more precisely, his body). Evidently, we cannot say that “Socrates is a penetrative color”. 22 This raises the difficult issue of how we are supposed to understand the ontological character of secondary substances. Should we regard them as some sort of class,   12  of such a predication has to be that very thing that is predicated of it. For example, the subject of which human being is predicated has to be that very thing, a human being.23 The name “human being” or the definition of human being applies to Socrates not in virtue of something else that attaches to him from outside, but in virtue of what he is (to wit, a human being). If the essence of Socrates is being human, then Socrates cannot be without being human – Socrates will be human as long as he exists. So the anti-essentialist suggestion that Socrates could exist without (or independently of) being human would, on Aristotle’s view, take Socrates out of being. What, then, does this explanation say about Aristotle’s essentialism? According to anyone’s essentialism, some of the predicates or properties are essential to the thing, whereas others are not. Although this characterization is usually associated with Aristotle,24 it is potentially misleading. For it might suggest that there are some independently existing things that have some properties essentially (or “permanently”) attached to them, and some not. However, the picture that emerges from the Categories is that particulars are not things that can exist independently of their species and genera. Rather, particulars are things whose very being is determined by their species and genera.                                                          collections of particulars, or rather as some sort of property? For a further discussion of this issue, see Irwin (2002, 78-80, 264-265) and Code (1986). It is clear, however, that Aristotle does not regard secondary substances as properties like white, i.e., properties that attach to the thing from outside. 23 Aristotle makes this point very clear in the Posterior Analytics, where he claims that predicates which signify the substance (ta men ousian sêmainonta) signify the very things (hoper ekeino) they are predicated of, whereas accidental predicates are always predicated of something different (heteron ti), for example, white is predicated of a human being (see A 22, 83a24-35). 24 See Quine, who argues that “[Aristotelian essentialism] is the doctrine that some of the attributes of a thing (quite independently of the language in which the thing is referred to, if at all) may be essential to the thing and other accidental” (1966, 173-174; cf. 1960, 199-200). Quine is definitely responsible for the revival of interest in Aristotelian essentialism (despite, or perhaps because of, his rejection of it). For discussions of Aristotle’s views in relation to contemporary versions of essentialism, see Witt (1989, chap. 6), White (1972).   13  Further support for this interpretation can be found near the end of chapter five, where Aristotle introduces yet another important distinction between substances and predicates: It seems most distinctive of substance that being the same and one in number (tauton kai hen arithmôi), it can admit contraries. In no other case could one bring forward anything, numerically one, which is able to receive contraries… A substance, however, the same and one in number, is able to receive contraries. For example, this human being, being one and the same, becomes pale and dark, and hot and cold, and bad and good. Nothing like this is to be seen in any other case. (4a1021) Although Aristotle does not explicitly speak about change in the above passage (nor does he do so anywhere in the Categories), it is reasonably clear that what he takes to be the most distinctive feature of substances is their ability to survive accidental change. For Aristotle is obviously not claiming that a substance is both pale and dark, or good and bad, at the same time. The claim is rather that a substance can admit contraries at different times, while remaining the same substance. But that is just to say that substances can change while remaining the same. At first it may seem surprising that Aristotle makes the ability to survive accidental change distinctive of substances in general, while the most natural understanding of the above passage would suggest that it is a feature of primary substances alone. What, then, counts as a substance that persists through change? The answer seems to be that it is a primary substance that falls under species and genera. For it seems that a particular cannot receive accidental properties unless it is something essentially, unless it falls under species and genera. For example, this human being, say Socrates, may be pale at one time and tanned at another, but he remains a human being. The fact that he is, essentially, a human being is what makes Socrates the same over time. Thus, Socrates is not the kind of thing that can exist, or continue to exist, independently of species and genera that are predicated   14  of him. Rather, Socrates continues to be what he is as long as he belongs to species and genera, whatever other changes may come upon him. Aristotle’s essentialism thus appears to rely on his firm belief (at least at the time of writing the Categories) that every primary substance must belong to or exist in (hyparchein en25) a species and genera (2a14, 16). Aristotle does not justify this belief, but treats a particular’s belonging to species and genera as a primitive or unanalyzable fact.26 Montgomery Furth points out that even though in the Categories Aristotle never actually says that “a substantial individual has to belong to a substantial kind” (1978, 628, my emphasis), it is clearly intended. For otherwise it would follow that substantial individuals would be “capable of retaining their numerical identities through arbitrary migrations between substantial kinds or out of the kinds altogether” (Furth 1978, 628). In other words, it would be possible for one and the same thing to have at one time, one thing essentially predicated of it, and at another time, something else, or that nothing is predicated of it at all. But Aristotle would hardly accept that Socrates need not be a human being, or, even worse, that he could be human at one time and artichoke at another. So a thing’s species is, for it, a “migration barrier”. Aristotle does not explicitly express this view in the Categories, but in works from roughly the same time we find him denying that one and the same thing could migrate between species. In the Topics, for instance, he says that “it is impossible for the same thing to continue                                                          The word “hyparchein” is usually rendered “to belong to/in” or “to exist/be in”. Nonetheless, commentators tend to ignore the existential implication of the word in the Categories, saying that the primary substances are “contained” or “included” in their species (see, e.g., the Loeb translation). 26 See Loux (2008, 4; 34-36), who argues that in the Categories Aristotle is committed to the “Unanalyzability Thesis”, i.e., the idea that the primary substance’s belonging to a species is an unanalyzable fact. This thesis emerges when we consider that if the primary substance’s belonging to a kind (e.g. Socrates’ being the man he is) is analyzable in terms of some prior case of one thing’s being said of or present in another (these two relations exhaust the tools Aristotle has at his disposal in the Categories for ontological analysis or reduction), then Socrates cannot be a primary substance, but the more fundamental subject would have a better claim to this status. 25   15  yet entirely change its species; the same animal, for instance, cannot be a human being at one time but not another” (125b37-39, cf. 145a3-12).27 Although it is not unambiguously clear whether Aristotle is committed in the Categories to the view that denies the possibility of migration between species, he is definitely committed to the (weaker) view that a particular cannot migrate out of species altogether, to have nothing essentially predicated of it. There is every indication that Aristotle would not admit that the talk of primary substances as basic subjects leads us to what John Locke calls “something I know not what” or what contemporary philosophers call “bare particulars” or “bare substrata” (i.e., things that are essentially no kinds of things at all, that have nothing essentially predicated of them). A primary substance is presented in the Categories not simply as a “this” but as “this something”, tode ti (cf. 3b10-13), where “something” (ti) picks out a certain kind of thing.28 Accordingly, I sympathize with Michael Loux (2008) who argues that “we better capture the point of the “this something” epithet if we understand it as a kind of schema, where the term “something” functions as a placeholder for predicates expressing the species under which primary ousiai fall” (p. 29).                                                          27 This citation expresses the position that is traditionally attributed to Aristotle (esp. in his early writings). The traditional understanding of Aristotle’s position has been challenged in recent decades by scholars working on Aristotle’s biological writings, who have argued that Aristotle’s views on species (esp. in this biological writings) are more relaxed and compatible with evolutionary theories. See, e.g., Franklin (1986), Lennox (1987). The difficulty of determining whether or not Aristotle intends to rule out the possibility of arbitrary migration between species lies in the fact that in the Categories Aristotle does not seem to be concerned with, much less explicitly address, issues involving change and persistence. In fact, Aristotle discusses change only in so far as he claims in the passage cited above that substances are able to “admit contraries”. This seems to imply that in order to persist through accidental changes a substance must belong to the same species (and thus cannot change its species), but Aristotle does not state it expressis verbis. On the other hand, however, there is nothing in the Categories supporting the opposite suggestion that he does not intend to rule out the arbitrary migration. Hence, I am inclined to agree with Furth (and the traditional interpretation) that Aristotle is in the Categories committed to the same position as in the Topics and holds that particulars must belong to their proper species. 28 “Tode ti” is usually translated as a “particular” or an “individual” but its literal translation is “this something”. See Smith (1921).   16  Hence this epithet emphasizes that every particular must belong to a species. This does not seem to imply anything stronger (or more “technical”) than that every particular must be of a certain kind. Further, Aristotle’s only examples of primary substances are “this human being” and “this horse”. Verity Harte (2010) has pointed out that although Aristotle very often uses personal names in his examples, his talk of primary substances in the Categories is not such an occasion.29 Rather, he uses the grammatical formula which combines the article, indefinite pronoun and sortal term: ho tis anthrôpos, ho tis hippos. I agree with Harte that Aristotle uses such formulae to refer to any particular human or any particular horse when he does not mean to refer to any of them in particular.30 Again, the force of this formula is to emphasize that particulars are instances of universals under which they fall. Accordingly, to identify a primary substance is to identify things that belong to certain species, that are of a certain kind (e.g. particulars-ashumans or particulars-as-horses). A primary substance, e.g. this human being, is already something, a human being – the secondary substances provide the “something” that the thing is. Thus the assumption underlying Aristotle’s discussions in the Categories seems to be that for a particular thing to be it has to be something, and what a thing is, is determined by genera and species.31 Since the existence of a particular                                                          29 It is very common among scholars to give names to one’s examples of primary substances. Although I have followed this common practice (and sometimes used Socrates as an example), it is important to notice that this is not Aristotle’s own practice in the Categories. 30 Harte (2010) argues that the difference that is made by the use of the expression ho tis anthrôpos rather than a proper name is its deliberate indefiniteness: the expression ignores what may be distinctive or special about any one particular individual. She suggests that such an expression is used in a similar manner by Sophocles and by Aristotle in the De Interpretatione 7 (17b37-18a7). 31 See Loux (2008, 26-33) and Jones (1972, 1975), who both locate the essentialist reading of “to be” as early as the Categories. One consideration that supports this reading is Aristotle’s claim that each primary substance is “one in number” (Cat. 4a10-13). Both Loux and Jones argue that Aristotle denies that there is anything that counts as just being one thing (Met I, 1-2). Rather, to say that the primary substances are one in number is to say that they fall under universals that provide us with measures for counting their   17  thing is dependent on its being something, i.e., dependent on a universal’s being said of it, it follows that particulars cannot exist without their universals. In other words, Aristotle’s commitment to essentialism in the Categories, and his rejection of the notion of the bare particular, entails that particulars are ontologically dependent on universals. Before drawing out the consequences of my interpretation for Aristotle’s anti-Platonism and for his “problem of katholou”, I would like to spend more time analyzing the bare particular view as an alternative to essentialism. In fact, it seems to be the only feasible alternative available to those who believe (as Aristotle does) that ontologically privileged entities must function as subjects. Why does Aristotle find the notion of a bare particular philosophically repugnant? We cannot simply say that he fails to see as far as bare particulars: he occasionally flirts with the idea of a bare substratum in his Metaphysics. In Metaphysics Z 3, he includes basic subject (hypokeimenon eschaton) on his list of possible candidates for the title of substance, and uses a thought experiment to discover what sort of thing the basic subject is. Aristotle argues that when we abstract from a concrete substance everything that could be predicated of it, we are left with matter “which in itself is neither a something (ti) nor a quantity nor otherwise determined” (1029a25). Although it is not obvious what                                                          instances. As Jones (1972) puts it: “Since what is one is specified with reference to its being something other than merely one, to be able to point to something and say “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” (p. 159). Another, related consideration is that Aristotle denies that there is anything that counts as just being (to on, “existent”). Aristotle does not express this point explicitly in the Categories, but in works from roughly the same time, we find him denying that “being is a genus”, i.e., everything that is/exists must be something other than merely being/existent (e.g. De Soph. El. 172a37-39, Top. 127a26-35, An. Post. 92b13-14). Loux argues that there is strong evidence that, even in the Categories, Aristotle took the categories “to provide the parameters for disambiguating the term” (p. 27-28). See also Matthews (2009), who argues that Aristotle is committed to a principle (he calls it “Aristotle’s Principle”) that everything that exists is a something or other. The essentialist interpretation of “to be” implies that we cannot, in case of primary substances, distinguish between essential and existential dependence (we cannot argue, for instance, that primary substances depend on secondary ones for their essence but not for their being) – their essential dependence implies existential dependence.   18  conclusion Aristotle wants us to draw from this thought experiment, it is clear that he does not accept the conclusion that bare matter, as basic subject, is a substance. He rejects this as an impossible result on the grounds that matter is nothing determinate, but a substance must be something determinate – a substance, he claims, must be a “this something” (tode ti, 1029a28). So Aristotle seems to be aware that the “subject criterion” (when pushed to its limits) leads us to bare substrata, and remains anxious to hold on to the idea that something of which nothing is essentially predicated cannot be a substance; it cannot be a “this something”. Paul Vincent Spade (1999) calls the bare particular view a “Platonic view of things”. On this view, we have an object or thing, which has certain properties or features that are somehow (by participation, say) attached to it, but by itself, all on its own, the object has no properties. In light of Spade’s interpretation we might regard Aristotle’s reluctance to adopt the notion of a bare particular as a reaction to the Platonic approach.32 But what precisely is wrong with the bare particular view? Spade argues that on the bare particular view there are pressures from two different directions. On the one hand, the bare particular view runs into trouble with the principle of the Identity of Indiscernibles.33 Since the bare particulars do not have any features of their own, there seems to be nothing to distinguish them. So we are pushed toward the view that there is really just one substratum. On the other hand, if there really is just one                                                          Similar suggestion is made by Mann (2000), who agues that Aristotle challenges the Platonic view which treats all predication in the sensible world as accidental predication. I shall not consider the question whether such an understanding of Plato is justified, but see Code (1986), who discusses Plato’s position in the Phaedo, although he also emphasizes that he examines Plato “through Aristotle’s eyes”. Code argues (and this, I believe, is a widely accepted view) that Aristotle holds that sensible particulars are endowed with essential natures, whereas Plato holds that “the inhabitants of the sensible realm Are not the natures in which they participate”, e.g. “Socrates merely Has, or participates in, man, without Being man” (p. 428). 33 This principle states that if two or more things share all their properties/features, they are the same thing. That is, things are really the same thing unless there is some feature to distinguish them. 32   19  substratum underlying bare particulars, then we seem to run into trouble with the Law of Non-Contradiction, because this one underlying something will have contrary properties at the same time. So Spade contends that “the Identity of Indiscernibles would lead us to say that there is only one such object. On the other hand, the Law of NonContradiction would lead us to say there are several “bare particulars” that play this role. Neither alternative solves the problem once and for all” (1999, 5). On Aristotle’s view, primary substances have – in virtue of belonging to their species and genera – internal features of their own. A primary substance is something (human being, say) all by itself or in its own right. And since primary substances can be distinguished from one another by their own internal features, the problem with the Identity of Indiscernibles disappears. Also the worry about the Law of NonContradiction is lessened – if there are many subjects then we do not need to worry about one and the same thing having contrary properties at the same time. Further, the fact that primary substances belong to their species and genera also explains why they can have contrary properties at different times. This comparison with the bare particular view suggests that secondary substances make the primary substance something determinate, something that can be distinguished from other substances, something that can underlie accidental changes – this might be the reason why Aristotle believes that primary substances must exist in, or fall under, secondary substances.34                                                          34 This interpretation suggests that Aristotle of the Categories (which is innocent of the matter-form distinction) has a hold of the phenomenon we would nowadays call individuative universals or sortal universals, etc. So, Aristotle of the Categories might agree with Strawson, who argues that “a sortal universal supplies a principle for distinguishing and counting individual particulars which it collects” (1964, 168). Loux (2006) argues that on such a view there is no “special problem of explaining the particularity of concrete objects. Just in virtue of instantiating its proper kind a concerete object is marked out as a particular… Furthermore, the multiple instantiation of the kind is, by itself, sufficient to secure the existence of numerically different particulars” (p.112). The idea that   20  By now it should be obvious that there is a kind of essentialism at work in the Categories and that this essentialism is not merely epistemological. That is, secondary substances do not merely give us knowledge of primary substances but determine the very being of those substances.35 Aristotle’s commitment to essentialism has major consequences for the traditional view which attributes to Aristotle the position that particulars are ontologically prior to universals, i.e., the relation of ontological dependence between them is asymmetrical. I believe that the required asymmetry can be defended with respect to the relationship between primary substances and accidents. Aristotle might be able to show that things present in a subject are ontologically dependent on primary substances as their subjects, whereas primary substances do not in the same way depend on them. The reason for this is that accidents are not essential to primary substances – a particular human being would still be what she is, regardless of whether or not she is pale. Nonetheless, the details of this account are messy and complicated. For it seems that if a particular human being is to exist she must be of some colour, of some weight, and so on (she cannot exist without all the so-called determinable accidents, like being coloured). But I believe that it is possible to develop an account which makes plausible the idea that the relationship here is asymmetrical. One possible account of asymmetry, in rough terms, is the following. The claim that a primary substance can exist without those accidents that inhere in it does not imply that the substance might be lacking in accidents altogether, but only that it is capable of possessing different accidents from the ones it actually has. The accidents, on the other hand, cannot exist unless substances also exist, i.e., unless they are                                                          Aristotle of the Categories has a hold on the notion of individuative predication is developed also by Furth (1988). 35 The distinction between essential and accidental predication might well be based on epistemological considerations. But I do not think that secondary substances merely give us knowledge of primary substances. Rather, they determine, so to speak, the conditions for the existence of primary substances. They are, as Furth puts it, “something constitutive or even causal, required for the continuity of the substantial individuals across time” (1978, 631).   21  instantiated by substances. So humans, for instance, can exist without exhibiting a given quality, such as being pale or being bald, but pallor or baldness cannot exist without some particular human having it.36 However, the relevant sort of asymmetry does not obtain in the relationship between primary and secondary substances. It is Aristotle’s view that there would not be a species such as human being without there being particular human beings, but it is not the case that a particular human being could exist without belonging to the species human being, that is, without being human. Thus, the essentialist interpretation suggests that in addition to the dependence of “all other things” on primary substances, there is also a reverse dependence that primary substances have on secondary substances. Stated otherwise, to be a particular is to be a particular of a certain sort, i.e., to be an instance of a universal under which a given particular falls, while to be a universal is to be instantiated by particulars that fall under it. In what follows, I consider the consequences that my interpretation has for Aristotle’s alleged antiPlatonism.                                                          36 We are now in a position to discuss briefly my earlier remark (footnote 6) that my interpretation offers indirect support to the view that Aristotle’s reference to “this white” picks out an unrepeatable instance of a universal accident white. On my interpretation, primary substances (e.g. this human being) are instances of secondary substance (e.g. human being). Hence we may suppose that particular accidents (e.g. this white) are likewise instances of universal accidents (e.g. white) – both this human being and this white are “this somethings”. What is the role of particular accidents? I believe that they mediate the dependence that universal accidents have on primary substances. That is to say, for the universal, human being, to exist is merely for there to be particular humans. But for the accident, white, to exist is for there to be instances of whiteness in particular humans. (A convincing account of this kind of two-step dependence is developed and defended by Deurlinger (1970).) So we can say that white depends on this human being because this white depends on the particular human being (Socrates say) in which it exists. This raises the question of why universal, human being, is dependent on this human being (Socrates say, rather than Callias). Aristotle avoids this question because he does not use proper names but indefinite formulas such as “this human being”. I suggested earlier that Aristotle uses this formula to refer to any particular humans when he does not mean to refer to any of them in particular. Hence it is safe to say that human being is dependent on this human being, for it cannot exist without some human beings, even though it is not dependent on any of them in particular. For a further discussion of these complications, see Corkum (2008).   22  CONSEQUENCES The possibility of interpreting Aristotle’s position in a way that makes particulars ontologically dependent on universals has not gone completely unnoticed by scholars. Martin Tweedale (1993), Montgomery Furth (1978, 1988), and Michael Loux (2008 [1991]), in particular, have taken this interpretation seriously. Tweedale thinks that the ontological interdependence, as I have called it, between universals and particulars indicates a “definite tension, if not outright contradiction” (1993, 78) right at the heart of Aristotle’s ontology. Loux and Furth, who focus more specifically on the Categories, see the dependence that particulars have on species as a problem that motivates or inspires Aristotle’s discussion of substance in his later writings. Furth thinks that Aristotle is well aware that this dependence “must immediately begin to erode their ultimacy as subjects” (1978, 631). This, he believes, is the reason why Aristotle did not definitely settle the matter by making their dependence explicit. So the omission of the clear statement of this position is not an oversight on Aristotle’s part, but “the author, seeing plainly the edge of abyss and knowing that it could not possibly be plumbed within the scope of the work in hand, deliberately, silently drew back. He hoped, perhaps, that no one would notice, and hardly anyone did” (p. 631). But why should we see the ontological interdependence as a problem in the first place? It seems that it constitutes a problem on the assumption that Aristotle intends the subject criterion to establish an unqualified ontological priority of particulars (i.e., a priority over universals and accidents), which, then, enables him to turn the Platonic picture “upside down”. But there does not seem to be enough evidence to show that Aristotle wants to overturn the Platonic position rather than change it. This leads us to the question of what precisely are the consequences of the essentialist interpretation for Aristotle’s alleged anti-Platonism.   23  According to the traditional interpretation, Aristotle’s famous conclusion in chapter five that “if the primary substances did not exist, it would be impossible for any of the other things to exist” commits Aristotle to an unqualified ontological priority of particulars. This position may be regarded as the prime example of Aristotle’s anti-Platonism, for in assigning ontological priority to particulars, rather than to universals, he turns the Platonic picture upside down. According to the essentialist interpretation, his position is more nuanced. On this interpretation, Aristotle is not committed to the ontological priority of particulars over secondary substances (or else he needs to give up the assumption that particulars are “this somethings” and embrace the notion of a bare particular). Rather, we could say that Aristotle equalizes the status of primary and secondary substances. It is no more possible for universals to exist without particulars than it is for particulars to exist without universals. But even on the essentialist interpretation there is a sense in which particulars as “this somethings” remain ontologically prior, viz., they are prior to accidents that belong to them. Accordingly, the answer to the question of whether particulars are ontologically prior is in a sense “yes”, and in a sense “no”. It is “yes” if by universals we mean accidents, and it is “no” if by universals we mean secondary substances. The interpretation according to which particulars as “this somethings” are ontologically prior to accidents gives us also a possible explanation for why Aristotle subsumes both particulars and universals under the label “substances”, viz., he does so to distinguish them from accidents which lack the independence that “this somethings” have.37                                                          37 Further support for the view that Aristotle is committed to the ontological priority of substances over accidents is found in the Metaphysics. Aristotle often uses separation terminology (chôris and its cognates) to refer to ontological priority. In the opening chapter of Metaphysics Z, he uses the term “separate” to express the ontological asymmetry between substances and accidents. This chapter clearly suggests that the relevant “somethings” from which substances are separate are accidents (while the relevant “somethings” from which the Platonic Forms are separate are sensible particulars). Since Aristotle seems to be committed to the ontological priority of   24  Nonetheless, even if we agree that ontological priority distinguishes substances from particulars, there still remains the question of why Aristotle uses the labels “primary substance” and “secondary substance”. It might seem that my essentialist interpretation, according to which primary and secondary substances are ontologically interdependent, leaves mysterious the point of using such labels. Loux, for instance, worries that the essentialist interpretation “leads us to wonder just why should we say that substance-species or whatever it is that makes basic subjects be what they are are only derivatively or secondarily ousiai. Do they not have as much right to the title ‘primary ousia’ as the basic subjects themselves?” (2008, 48). This worry seems to rest on the assumption that the criterion for being a substance is ontological priority. This assumption is adopted by the traditional interpretation which maintains that the subject criterion is used to establish not only the ontological dependence of secondary substances on primary ones, but – by implication – the ontological independence of primary substances. My interpretation challenges this implication, but there is a sense in which it does not challenge the subject criterion itself, i.e., the idea that primary substances are ultimate subjects of which everything else is predicated but which themselves are not predicated of anything more basic. On my interpretation, primary substances cannot exist independently of universals predicated of them. Nevertheless, they remain primary in the sense of not being ontologically dependent on any further subjects. Stated otherwise, particulars are instances of universals, but particulars themselves do not have instances. The essentialist interpretation is thus compatible with what could plausibly be called the weak reading of the subject criterion. This reading agrees with the traditional view that secondary substances depend for their existence on primary substances as their subjects. But it differs from the traditional reading in that it does not draw the conclusion that primary substances can therefore exist independently of secondary substances.                                                          substances over accidents in the Metaphysics, it is not unreasonable to suggest that he is committed to this position already in the Categories.   25  The minimum reading only draws the conclusion that primary substances exist independently of further subjects; they do not need to be instantiated in order to exist. Thus primary substances retain a sort of primacy and independence, but their primacy is rather weak one – it does not imply the capacity to exist independently from universals predicated of them.38 In so far as this reading of the subject criterion does not establish any strong independence of particulars, it might not appear to be wholly satisfactory. But it provides a possible answer to the question of why Aristotle uses the labels “primary” and “secondary”. Namely, secondary substances are “secondary” because they are always instantiated by primary substances, whereas primary substances remain “primary” in the sense of not having instances.39 So, the essentialist interpretation suggests that Aristotle does not require primary substances to be capable of existing independently of universals, and hence he does not overturn the Platonic position. But Aristotle definitely changes it, and he changes it in a more radical manner than is usually thought. The traditional interpretation holds that particulars can exist independently from universals, thereby committing Aristotle to an                                                          38 A somewhat similar suggestion is advanced by Corkum (2008), who challenges the traditional understanding of the notion of ontological independence, and proposes to weaken the notion of ontological independence from a capacity for independent existence to the possession of certain ontological status. He leaves open the question of what precisely this ontological status amounts to as a subject for further research. My interpretation differs from Corkum’s in that I do not challenge the traditional understanding of the notion of ontological independence as a capacity for independent existence. I only challenge the idea that primary substances can exist independently of universals. On my interpretation, primary substances are primary because they can exist independently of further subjects (because they are not predicated of anything further), not because they can exist independently of universals predicated of them. 39 The fact that this criterion admits of different readings (a weaker and a stronger one) might well be the reason why Aristotle complains in Metaphysics Z 3 that this criterion is “not only obscure (adêlon), but makes matter substance” (1029a11). The weaker reading gives us only a weak sort of independence. The stronger reading establishes an ontological priority over universals. This, however, is too strong, since particulars cannot exist without universals predicated of them. As Aristotle argues in Metaphysics Z, when we eliminate everything that that is predicated of a subject, we do not end up with a substance and “this something” but with an indeterminate matter.   26  ontological dualism of particulars and universals. The essentialist interpretation, on the other hand, holds that particulars and universals are ontologically interdependent, and their ontological interdependence implies that “primary” and “secondary” cannot be labels for irreducibly distinct types of entity. Hence, this interpretation does not commit Aristotle to any robust dualism. Nonetheless, the question of what precisely is the positive view that emerges out of this interpretation is not an easy one to answer. On the one hand, the notion of ontological interdependence does not imply that particulars and universals are irreducibly distinct types of things, complete entities in their own right, that are somehow dependent on each other. On the other hand, to say that their ontological interdependence amounts to identity seems also to go beyond what Aristotle has in mind in the Categories. The notion of identity is nowadays usually equated with the principle which can here be loosely expressed by saying that if A and B are identical, then whatever is true of the one is true of another.40 But it does not seem to be the case that whatever is true of a primary substance is true of a secondary substance, and vice versa. Firstly, secondary substances are predicated of a subject, whereas primary substances are not. Secondly, in addition to essential predications, we can also make accidental predications of a primary substance: for example, “Socrates is white”. This predication is true of Socrates but is no part of his essence. So it seems that there are more things true of Socrates than are stated in a definition expressing his essence. Hence, it does not seem to be the case that primary and secondary substances are strictly identical. These considerations suggest that Aristotle of the Categories is interested in some sort of a middle position between robust dualism and strict identity. And                                                          40 Aristotle seems to formulate something that looks like this principle (“Leibniz’s Law”, as it is usually called) in Topics H (152b25-29). For a further discussion, see White (1971), who argues that the Topics is the only work of Aristotle which reveals a “relatively firm grasp of something like the notion of identity” (p. 177).   27  here we are at the beginning of a very long and complicated story, which, to borrow from Porphyry, is too deep for the present investigation.41 To sum up, I have shown that there is a kind of essentialism at work in the Categories which undermines the traditional view according to which Aristotle assigns an unqualified ontological priority to primary substances over universals, thus overturning the Platonic picture which assigns ontological priority to universal Forms. I have argued that Aristotle distinguishes between two types of universals, and while primary substances can be said to be ontologically prior to accidents, they cannot be prior to secondary substances. This is because primary substances cannot exist independently of secondary substance – to be a particular is to be a “this something”. On this interpretation, Aristotle does not turn the Platonic picture “upside down” but rather equalizes the status of universals and particulars.                                                          Porphyry famously formulates three questions concerning the ontological status of species and genera but refuses to answer them, saying they are too “deep” (Isagoge, 1, 1014). Although Aristotle does not treat universals as complete entities in their own right, he seems to want to give them some sort of (weak) ontological status. Nonetheless, he does not seem to think that their ontological status is particularly worrisome as long as we do not separate them from particulars and regard them as embedded in the very being of particulars. Indeed, he might even think that asking about their ontological status implies treating them as complete entities in their own right, separating them from particulars. However, even if we avoid questions concerning the ontological status of universals, it is hard to avoid questions about the distinction between particulars and universals. The questions concerning distinctions between different entities were intensely discussed by medieval authors, especially Duns Scotus. Scotus’ notion of formal distinction, in particular, seems to be a good (perhaps the best) attempt to formulate such middle position Aristotle is interested in the Categories. Formal distinction is a real (and not simply mind-dependent), but it is not a distinction between two things, fullfledged entities. Rather, it is a distinction between two items (or aspects?) that are really the same but definitionally independent from one another. So items really the same but formally distinct do not need to be such that what is true of one must be true of another. Although the basic idea of formal distinction is simple enough, it is difficult to establish this position. Ockham has famously argued that Scotus cannot maintain this distinction, but it is reduced to a (unqualifiedly) real distinction, i.e., a distinction between two things or entities. For a further discussion of Scotus’ formal distinction and Ockham’s criticism, see Tweedale (1999). 41   28  BIBLIOGRAPHY Aristotle. 1831-70. Aristotelis Opera, 5 vols. I. Bekker (ed.). Berlin: G. Reimer. Aristotle. 1995. The Complete Works of Aristotle, 2 vols. J. Barnes (ed.). Princeton: Princeton University Press. Porphyry. 2003. Introduction. Translated with an Introduction and Commentary by J. Barnes. Oxford: Clarendon Press. Ackrill, J. L. 2002 [1963]. Aristotle: Categories and De Interpretatione. Oxford: Clarendon Press. Code, A. 1986. “Aristotle: Essence and Accident.” In: Philosophical Grounds of Rationality, R. E. Grandy, R. Warner (eds.). Oxford: Clarendon Press, pp. 411-440. Corkum, P. 2008. “Aristotle on Ontological Dependence.” Phronesis 54, pp. 65-92. Duerlinger, J. 1970. “Predication and Inherence in Aristotle’s Categories.” Phronesis 15, pp. 179-203. Franklin, J. 1986. “Aristotle on Species Variation.” Philosophy 61, pp. 245252. Frede, M. 1987. Essays in Ancient Philosophy. Minneapolis: University of Minnesota Press. Furth, M. 1978. “Transtemporal Stability in Aristotelian Substances.” Journal of Philosophy 75, pp. 624-646. Furth, M. 1988. Substance, Form and Psyche: An Aristotelian Metaphysics. Cambridge: Cambridge University Press. Granger, H. 1980. “A Defense of the Traditional Position Concerning Aristotle’s Non-substantial Particulars.” Canadian Journal of Philosophy 10, pp. 593-606. Harte, V. 2010. “What’s a Particular and what makes it so? Some thoughts, mainly about Aristotle.” In: Particulars in Greek Philosophy: The Seventh S.V. Keeling Colloquium in Ancient Philosophy, R. W. Sharples (ed.). Leiden: Brill, pp. 97-125. Irwin, T. H. 2002 [1988]. Aristotle’s First Principles. Oxford: Clarendon Press. Jones, B. 1972. “Individuals in Aristotle’s Categories.” Phronesis 17, pp. 107123.   29  Jones, B. 1975. “An Introduction to the First Five Chapters of Aristotle’s Categories.” Phronesis 20, pp. 146-172. Kohl, M. 2008. “Substancehood and Subjecthood in Aristotle’s Categories.” Phronesis 53, pp. 152-179. Lennox, J. G. 1987. “Kinds, Forms of Kinds, and the More an the Less in Aristotle’s Biology.” In: Philosophical Issues in Aristotle’s Biology, A. Gotthelf, J. G. Lennox (eds.). Cambridge: Cambridge University Press, pp. 85-95. Lewis, F. A. 1991. Substance and Predication in Aristotle. Cambridge: Cambridge University Press. Loux, M. J. 2006. Metaphysics: A Contemporary Introduction. New York and London: Routledge. Loux, M. J. 2008 [1991]. Primary Ousia: An Essay on Aristotle’s Metaphysics Ζ and Η. Ithaca and London: Cornell University Press. Mann, W.-R. 2000. The Discovery of Things: Aristotle’s Categories and Their Context. Princeton: Princeton University Press. Matthews, G. B. 2009. “Aristotelian Categories.” In: A Companion to Aristotle, G. Anagnostopoulos (ed.). Oxford: Wiley-Blackwell. Moravcsik, J. M. E. 1967. “Aristotle on Predication.” Philosophical Review 76, pp. 80-96. Owen, G. E. L. 1965. “Inherence.” Phronesis 10, pp. 97-105. Quine, W. V. 1960. Word and Object. Cambridge: MIT Press. Quine, W. V. 1966. The Ways of Paradox and Other Essays. New York: Random House. Ramsey, F. P. 1925. “Universals.” Mind 34, pp. 401-417. Shields, C. 2007. Aristotle. London and New York: Routledge. Smith, J. A. 1921. “Tode ti in Aristotle.” Classical Review 35, p. 19. Spade, P. V. 1999. The Warp and Woof of Metaphysics: How to Get Started on Some Big Themes – http://pvspade.com/Logic/docs/WarpWoo1.pdf Strawson, P. F. 1964. Individuals: An Essay in Descriptive Metaphysics. London: Methuen & Co Ltd. Tweedale, M. M. 1993. “Duns Scotus’s Doctrine on Universals and the Aphrodisian Tradition.” The American Catholic Philosophical Quarterly LXVII, pp. 77–93.   30  Tweedale, M. M. 1999. Scotus vs. Ockham – A Medieval Dispute Over Universals. Volume II: Commentary. Lewiston, Queenston, Lampeter: The Edwin Mellen Press. Wedin, M. V. 2005. Aristotle’s Theory of Substance: the Categories and Metaphysics Zeta. Oxford: Oxford University Press. White, N. P. 1971. “Aristotle on Sameness and Oneness.” The Philosophical Review 80, pp. 177-197. White, N. P. 1972. “Origins of Aristotle’s Essentialism.” The Review of Metaphysics 26, pp. 57-85. Witt, C. 1989. Substance and Essence in Aristotle: An Interpretation of Metaphysics VII–IX. Ithaca and London: Cornell University Press.   31