Aristotle's Subject Matter Dissertation Presented in Partial Fulfillment Of

Home , Schema (Kant), Subject (philosophy)

Aristotle’s Subject Matter

Dissertation

Presented in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy in the Graduate School of The Ohio State University

Keren Wilson Shatalov, MLitt Graduate Program in Philosophy

The Ohio State University 2019

Dissertation Committee: Allan Silverman, Adviser Tamar Rudavsky Lisa Downing

iii Abstract

iii

Dedication

To my father, whose conversations were my first lessons in philosophy.

iv Vita

May 2004 ...... Logan County Home Educators May 2008 ...... B.A.Liberal Arts, Thomas Aquinas College November 2009 ...... MLitt, University of St. Andrews 2010-2017 ...... Graduate Teaching Associate, Department of Philoosophy, the Ohio State University 2017 to the present ...... Sawyier Fellow, Department of Humanities, Illinois Institute of Technology

Fields of Study

Major Field: Philosophy Specialization: Ancient Greek Philosophy

v List of Figures

Fig. 1: The number and nature of the principles ...... 64 Fig. 2: The simples and complexes in change ...... 103

vi Table of Contents

Abstract………… ...... ii Vita………………………...... iv List of Figures...... v Chapter 1: Background and Overview ...... 1 Chapter 2: Ὑποκείμενον in the Logical Works ...... 27 Chapter 3: Physics A.1-6 ...... 63 Chapter 4: Matter as Ἡυποκείμενον in Physics A.7 ...... 98 Chapter 5: Matter as Analogous ...... 134 Bibliography…… ...... 151

vii Chapter 1 Background and Overview

§.1 Background:

In my dissertation I examine Aristotle’s concept of matter by highlighting the tools from his Organon which he uses to introduce matter in his Physics. I make use of logical concepts Aristotle develops in his work on explanation in Posterior Analytics, especially his concept of subject or ὑποκείμενον, to argue that matter, for Aristotle, must be understood not as a distinct ontological category but as a term of art denoting a part of an explanation in natural philosophy. By presenting an analysis of Aristotle’s concept of ὑποκείμενον from his logical works, I show how Aristotle uses it to spell out just what explanatory role matter plays, and what this means for what it is to be matter. My understanding of Aristotle’s concept of matter depends on how I understand Aristotle’s notion of ὑποκείμενον in his logical works, since I claim that it is this notion he is employing when he introduces ὑποκείμενον as a principle of change in Physics A. I find that extant views in the literature on Aristotle’s concept of matter are also dependent on whether scholars see Aristotle as using ὑποκείμενον synonymously between his Physics and his logical works and, if, they do, on how they understand Aristotle’s notion of ὑποκείμενον in his logical works. In what follows I aim to give a sense of how important these two points are for understanding Aristotle’s matter by tracing their influence on some of the prominent views concerning matter in Physics. Seeing the range of readings already offered will, I hope, help to show what the reading I will give in this dissertation can add to the conversation about what matter is for Aristotle. Rather than attempt to capture all of the variety of proposals which have been made, I will take David Bostock, William Charlton, and Sarah Waterlow as my primary interlocutors. These three have all written works which are now classics on Aristotle’s Physics: Waterlow’s Nature, Change, and Agency in Aristotle’s Physics, Charlton’s translation and commentary in his Aristotle’s Physics, I & II for the Clarendon Series, and Bostock’s essays in his Space, Time, Matter, and Form. Sir David Ross’s tome Aristotle’s Physics makes him a fourth important voice, standing in a unique position as one whose scholarship both made possible and influenced, in some way or other, all the 1 rest, including my own. I will take on Bostock, Waterlow, and Charlton individually; Ross will be consulted throughout as a source of insight. Though they differ among themselves in many ways, one thing they all take in common is reading Aristotle’s notion of ὑποκείμενον or subject from his logical works as being coextensive with his notion of οὐσία or substance, at least for the purposes of their commentary on Aristotle’s Physics. As I will show, this has great bearing on whether they read ὑποκείμενον in Physics as being synonymous with the subject from the logical works, as well as what this means for matter, whether they do or do not take it as synonymous. First, some things held in common. On my view, in Physics A Aristotle aims to provide a model or schema for explaining any change. Physics A consists of a search for the principles of natural science, and by Aristotle’s lights the consideration of the principles of a science is not itself a part of the science, but a necessary complement of and foundation for it. David Bostock puts the same thought this way: “Roughly, what is introduced as if it were a continuation of the physicists’ investigation of nature has instead become a meta-investigation of the general form which any account of change must take, whether it is an account of natural change or not.” (“The Principles of Change in Physics I”, pg.4) In the case of natural science, since natural science is the science of natural, changeable things, the principles required are the principles of change and of changeable things. The principles of change are the nodes or elements of an explanation for change, which together form a schema to be filled in when explaining any particular change. If the principles are three in number, say, α, β, and γ, then in every change there must be something which plays the role of and thus is appealed to as the α, the β, and the γ in each particular change. Charlton, like Bostock, emphasizes that in Physics A Aristotle “writes as a philosopher, not as a scientist” (pg. ix); his topic there “may, perhaps, most aptly be classified as philosophy of science (ibid). Waterlow and Ross agree about this. Going beyond this consensus, I maintain that Aristotle’s goal in Physics A is to characterize the roles of α, β, and γ, so that scientists doing research in the field, as it were, might figure out which things play them in particular instances of change. Having provided a framework for the explanation of change, he has thus also provided guidance to scientists undertaking to explain, predict, or extrapolate from particular instances or 2 kinds of change. Thus in Physics A Aristotle is engaged in the same kind of project as he is in the beginning of Physics B when he distinguishes among the different kinds of causes. There, for example, he defines the efficient cause as “the primary source of the change or the staying unchanged” (194b29-30). This definition tells what it is to be an efficient cause; it characterizes what sort of explainer the efficient cause is. But which thing or kind of thing is the efficient cause in a particular case must be filled in in part empirically, by going into the world and seeing what makes what happen. A good example of this complementarity between the kind of view Aristotle aims to provide in the first book of his Physics and the kind of empirical investigation that a modern reader might be more inclined to call “scientific” is the way in which Aristotle puts to use the account of change from Physics A, plus the remaining causes he posits in Physics B, in his On Generation and Corruption (henceforth, DGC.) In that work Aristotle employs the schema of explanation from Physics A, and the full set of causes from Physics B, to organize and make sense of the data he has from his empirical research. For example, Aristotle points out that among animals the parents are the efficient causes of their offspring. Then, becoming more fine-grained in his causal analysis, he goes on to propose, as a result of the empirical investigations he has undertaken, that the father causes by providing his sperm, the mother by providing her menstrual blood. This raises the question of cause at an even finer level of granularity. Aristotle continues his investigation as a natural scientist, still guided by his understanding of causes; the sperm, Aristotle concludes, acts as efficient cause by imparting to the menses a certain motion and heat. Bostock, Waterlow, Charlton, and Ross generally agree that this is the kind of project in which Aristotle is engaged in his Physics A, though they may disagree about the role which empirical research is meant to play in relation to this. More to the point for this dissertation, they differ in some significant ways regarding the framework of explanation Aristotle proposes, and the role of ὑποκείμενον in it. There are many subtle ways in which their readings differ from each other, and from the one which I will give in this dissertation. However, since my aim in this dissertation is to show how Aristotle’s notion of ὑποκείμενον informs his notion of matter through his discussion in Physics A, in what follows I will focus on how each commentator understands Aristotle’s notion of 3 ὑποκείμενον from his Organon, and what role they think it plays in his introduction of matter in Physics A. I will aim to summarize how this plays into their understanding of Aristotle’s notion of matter. Regarding Aristotle’s notion of ὑποκείμενον or logical subject from the Organon, David Bostock says the following: “to say that X underlies Y is just to say that X is the subject of which Y is predicated” (“Aristotle’s Theory of Matter”, pg. 31). He does not, however, elucidate further what it is to be the subject of which something is predicated, except to suggest that the subject always is a substance. This is significant, for it affects how Bostock reads the way in which Aristotle argues that there must be a ὑποκείμενον in every change. In Physics A, on Bostock’s reading, Aristotle first introduces the term “ὑποκείμενον” into his discussion of change by proposing that there is always a substance involved in change. In considering this, Aristotle, on Bostock’s view, comes to realize that something always persists through change, and that this is key to understanding change. Bostock says that in A.7 “the expression which Aristotle here uses for what persists is ‘what underlies,’” or ὑποκείμενον. (“Principles of Change”, pg. 7) Thus, according to Bostock, Aristotle uses the term with two meanings within Physics A. Its initial use there imports the technical sense from his logical works by referring to or implying the presence of substance. Considering the need for a substance in change, however, leads Aristotle to begin to use a distinct meaning for the term, such that, as a principle of change, it means only “what persists.” Bostock argues that when Aristotle finally posits the ὑποκείμενον as a principle of change at the end of Physics A.7 he has in mind this second meaning, rather than the first. He is not positing substance as a principle of change; rather, he is claiming merely that there is always something which persists through change. Because Bostock seems to take ὑποκείμενον from the logical works as being coextensive with οὐσία or substance, he argues that if it were the ὑποκείμενον from the logical works which Aristotle was claiming in Physics is a principle of change, then it would follow immediately from the Physics view of the principles that matter is substance. Yet Aristotle implicitly denies that this follows when he claims that it is an open question whether the matter or the form is the substance (“Principles of Change,” pg. 14, cf. 191a19-20). From this follows a dilemma for interpreters, which Bostock 4 solves by concluding that ὑποκείμενον in Physics A, at least when it is being used to name a principle of change, is not synonymous with ὑποκείμενον as subject of predication, the term from the logical works. This also solves another difficulty Bostock sees for Aristotle. Since on Bostock’s reading only substances can be ὑποκείμενα according to the treatment in the logical works, then in Physics there would be a contradiction when he seems to allow that form is predicated of matter. Since form and matter together constitute a substance, this would be to allow that something which is not a substance can be the subject of predication:

I see two possible ways for avoiding it while still preserving Aristotle’s main doctrine of becoming. One possibility would be to abandon the doctrine of the Categories on predication, and to say that there are subjects of (accidental) predication which are not substances; in particular, a form may be predicated of matter, and matter is not substance. (…) But a more appealing possibility would be to retain the view that forms are always predicated of substances but to deny that that makes substance into the third principle. For according to the most plausible version of the doctrine of chapter 7 the third principle is not what underlies in the sense of what is a subject of predicates…, but rather what underlies in the sense of what persists through change. (“Principles of Change,”pg. 14)

Ὑποκείμενον as a principle is thus according to Bostock, a sort of black box for whatever persists through the change. Bostock ultimately characterizes this sense of ὑποκείμενον as meaning “either what Y is made from or what Y is made of.” (“Theory of Matter,” pg. 31) On Bostock’s reading Aristotle proposes a single model for all change—a replacement model. On this model, change consists in one thing persisting through a change (the ὑποκείμενον), while one of its properties or characteristics is replaced by another. In change of accidental or non-essential properties, change consists in one property belonging to a substance being replaced by another. In substantial coming-to-be or generation, the change consists in some material persisting while the essence or form it 5 had at the beginning is replaced by another. In both change of accidental properties and coming to be or generation, the actual mechanism of the change is left for another conversation, nor does the general framework proposed in A7 provide any guidance for how one might investigate or understand this. In Physics A Aristotle points out only that all change consists in replacement. Bostock argues further that Aristotle eventually uses the term ὕλη, usually translated as “matter,” to rename the persistent principle in change. (e.g., “Principles of Change,” pg. 7) Thus matter, for Bostock’s Aristotle, is whatever underlies the exchanged properties (or substantial forms) insofar as it persists through the change, for this is all the elucidation he can find in Physics A for the idea that matter is “what Y is made from or what Y is made of” (ibid). Given the significance of the notion of ὑποκείμενον for Aristotle in his Categories, An. Po., and even his Metaphysics, I think it would be surprising for him to use the term ambiguously in the way Bostock suggests, with one sense in his logical works and a different sense in his Physics. It would be especially surprising if Aristotle used a new signification for the term in his Physics within the same discussions in which he employs the sense from the logical works, and without flagging the difference. Further, Bostock’s proposal for why Aristotle would abandon the technical meaning from the logical works for the “black box” signification in his final enunciation of the principles of change is insufficiently motivated, as I will argue in my chapter 4. But the interesting thing for my purposes is what this means for Aristotle’s notion of matter. On Bostock’s reading, matter is the immediate underlier in the sense of being what something is immediately made out of or from. If X is the matter of Y and Y the matter of Z, X is not thereby the matter of Z (“Theory of Matter,” pg. 33). Consequently, matter is relative. Since Aristotle holds that ultimately one can get any stuff from any stuff, provided there are the appropriate intermediate changes, this requires that Aristotle posit some very basic kind of stuff which is such that it can take on the appropriate forms to act as underlier for substantial change. This is what prime matter is: a basic, featureless matter, the only stuff which truly persists through all change, and which underlies the becoming of anything. Bostock proposes a significant objection to this view of matter: it provides no criterion of identity for the basic stuff. If there is no distinctive identity for matter, this challenges the idea that it is what persists through 6 change. For it is not clear how one could pick out any “what” which so persists. Thus Aristotle’s theory of matter, as reconstructed by Bostock, faces this serious difficulty. However, this difficulty can be avoided if one sees Aristotle’s appeal to ὑποκείμενον in the Physics A discussion as a consistent invocation of the concept of ὑποκείμενον developed in his logical works, and further if one sees that in the logical works Aristotle does not hold that only substances can be ὑποκείμενα. Doing so has the capacity to recast the discussion so that understanding Aristotelian matter is not about discovering some basic kind or kinds of stuff, but is rather concerned with explicating an explanatory principle such that a variety of kinds of stuff could play that role. William Charlton, in his 1970 translation and commentary on the first two books of Physics, takes a different approach than does Bostock, inasmuch as he reads Aristotle’s use of the term ὑποκείμενον in the Physics A discussion as being univocal throughout with the usage that appears in Categories and An. Po. However, like Bostock, Charlton takes it that a ὑποκείμενον in this sense is always a substance. This marks an important difference between his reading and mine, and one which has significance for Aristotle’s understanding of matter. Against Bostock, Charlton reads Aristotle as positing the logical ὑποκείμενον as a principle of change; on his reading that Aristotle’s ὑποκείμενον is a principle ofchange means, roughly, that in every change there must be a substance, a thing, which undergoes the change. In part because of this, Charlton also disagrees with Bostock about whether Aristotle holds that something always persists through change. If Aristotle did, this would undermine his claim to have provided an explanation of substantial coming to be and destruction, for intuitively no substance persists in such cases of change. Since no substance persists through change, one might worry that Charlton’s reading requires that Aristotle hold that substantial generation and substantial destruction are creation ex nihilo and annihilation, respectively. Charlton denies this, however (pg. 140).1 One reason he has for thinking that, on Aristotle’s view, it is not necessary that something persists

1 He seems to suggest that the reason is just that in the natural world, there are predictable pairs of what is present before the coming-to-be and what is present after it, and to grasp this is to see that there is no need to posit creation ex nihilo or annihilation. See Charlton, pg. 140, quoted on my page 8. 7 through change in every case is a response to a problem concerning identity criteria for matter, similar to the problem Bostock saw:

The idea, however, that if there is nothing which remains throughout a change, then things come to be out of or pass away into nothing, is mistaken. Between alteration on the one hand, and creation and annihilation on the other, there is a third possibility. If you have a glass jar from within which you have removed the air and everything else you can find; and you see a frog suddenly appear in it; then you might call that coming to be out of nothing. (…) But when the passing away of one thing is always and intelligibly attended by the coming to be of another, for instance when wood passes away in smoke and flames, or a saucer of water passes away and the air is refreshed, then we do not say that the first thing has passed away into nothing, but into the second, and we say that the second has come into being, not out of nothing, but out of the first. Yet we cannot say that there is something which remained throughout and underwent these transformations, unless we can find some description under which this thing can be identified throughout. (pg. 140)

On this reading, it would be a mistake to think of change as the replacement of characteristics or properties within something persistent. Instead, Charlton approaches it by thinking about one thing coming to constitute, or developing into, another. The proper way to think about what has happened in change, on Charlton’s reading, is not that some property in a subject has been replaced, but rather that the terminus a quo has come to be or developed to be a certain way. Thus matter is not what underlies as what persists through the change. Rather, matter for Aristotle, according to Charlton, is the “constituter” or what constitutes the outcome of the chage. It is revealed by the terminus a quo of the change. Form is the thing constituted, the terminus ad quem. However, even if Charlton is correct that we should think of matter’s relation to form, or to the substance which is the matter-form composite, as being that of constituter to thing constituted, it is not clear that he has in fact avoided the criterion of identity problem Bostock points to, or 8 the problem of coming to be ex nihilo. If nothing persists through change, how can something which was present before the change constitute what is present at the end? Or if there is no account under which what was present before the change persists through it, on what grounds can we say that what has come to be has not done so from nothing? Because Charlton holds that a ὑποκείμενον is always a substance, he seems stuck with a dilemma: either Aristotle has not provided an explanation of substantial change as he has claimed, or he has failed to avoid positing change as consisting in something coming from, and something else passing away into, nothing. In his commentary on Physics A.9 Charlton discusses Aristotle’s attempt to avoid these problems by distinguishing among the ways in which something might be said to be or not to be. After remarking that this is “a little difficult” he says the following: “and his reasons seem fair: if you are making an ivory billiard ball, you are not making either the ivory or what the ivory will constitute, a sphere: you are making the ivory constitute a sphere.” (pg. 83) I worry that this does little more than re-describe the change and restate the problem: what is it for the ivory to do so or to come to do so? Charlton is surely right that Aristotle’s solution has something to do with his understanding of being potentially vs. actually. Yet to say this is not to provide that solution. For one thing, at the end of A.7, in the course of which he has not appealed to actuality and potentiality, Aristotle says that he has said the same thing elsewhere in terms of actuality and potentiality. What is this same thing he has said, and how does he say it in terms of ὑποκείμενον, form, and lack in Physics A.7? These are questions Charlton’s commentary sets up but leaves unanswered. In the first part of her Nature, Change, and Agency, Sarah Waterlow provides a reading of Aristotle’s Physics A analysis of change which helps to answer these questions. On her reading, Aristotle’s account of change is motivated to avoid a certain dilemma: positing coming to be ex nihilo, on the one hand, and positing something persistent, on the other hand, in terms such that there is finally no change. In Physics A Aristotle shows that “the distinction between things and properties furnishes us with an answer against those who accuse becoming of entailing ‘ex nihilo’ (pg. 22, para. 26), while the distinction between “substance-constitutive characteristics and others” (pg. 22,

9 para. 27) which underscores this distinction also helps to avoid the other horn of the dilemma. On her reading Aristotle’s use of the term ὑποκείμενον in Physics A adumbrates the distinction between things and properties which she presents as solving the problem. For Waterlow, Aristotle and his predecessors agree that change only makes sense if something persists through change (in order to avoid the ex nihilo horn of the dilemma), and this “vague notion was articulated by means of the concept of an ‘underlying subject’” (pg. 27, para. 33). This concept of the underlying subject is shown in the course of the argumentation of Physics A to “depend on the distinction between substances and attributes, and therefore on the distinction between substance-constitutive characteristics and others.” (pg. 27, para. 33) On Waterlow’s reading Aristotle takes the reverse of the tactic Bostock ascribes to Aristotle; he introduces ὑποκείμενον as a “vague notion” which points to whatever persists through change, rather than as the term of art he had employed in his Categories. However, in the course of his analysis of change he comes to hold that there can only be an underlying subject in the sense of a persistent through change if there is a substance, a thing. In this way Aristotle sharpens the intuition behind the vague concept of ὑποκείμενον to posit as a principle of change that ὑποκείμενον which, in the logical works, was substance. Aristotle further sharpens this understanding of ὑποκείμενον as a principle by clarifying that there can only be a persistent substance if we can differentiate between substance-constitutive characteristics and other attributes. That is, there can only be a ὑποκείμενον or substance which persists through change if there is something for which an essence can be picked out. Thus, for Waterlow, Aristotle’s notion of ὑποκείμενον is such that only substances can be ὑποκείμενα, for only substances have essences, and this is what it takes to persist through change. On Waterlow’s view, the right way to think about natural change is as a sort of compromise between a change whose outcome is wholly determined by something external to the changing thing (such as might occur in creation ex nihilo), and change whose outcome is determined wholly by the changing thing and is independent of external causes, conditions, or influence. In natural change, the nature of the changing thing limits the possibilities of the outcome of the change and directs the change toward some terminus, even while external conditions can affect its success in reaching this 10 outcome, or dictate some of the characteristics it will have (within the bounds set by the nature). (pg. 5, para. 7) Aristotle holds that one can only make sense of the kind of orderly, predictable change which occurs in the natural world if it is change of this kind. If the change were entirely dependent on external conditions, there would be no nature or characteristic under which the thing might be said to persist through the change. If the change were entirely independent of external conditions, this points toward the other horn of the dilemma. On Waterlow’s reading change of accidental properties follows the replacement model, but substantial generation and destruction do not, or not quite. In both kinds of change what persists is the substance. This requires that substantial change have a somewhat different structure than change of accidents, since in substantial change it is the substance which is coming to be or being destroyed. Waterlow suggests that in substantial generation the substantial form of the outcome of the change governs substantial coming to be, and thus is what persists through the change. In substantial coming-to-be the substance is present in the terminus a quo in potency through the substantial form. Note that it would be misleading to try to put the same thing by saying that the substance is present at the beginning potentially, and at the end of the change actually. To say that something is present potentially seems to suggest that it is not present at all, but it could be. Instead, the idea is that the substance is in fact present at the terminus a quo, and throughout the change to the terminus ad quem in virtue of the substantial form which governs or directs the change. In the beginning, however, its way of being is being potentially, while at the end it is being actually. The substance which persists is the matter in both the case of change of accidents and substantial change, for matter is just what persists through the change, as what develops into the terminus ad quem. The difference between the two kinds of change is that in substantial generation the matter which develops does so out of its own nature, which cannot be defined without reference to the endpoint in the change (cf. pg. 47), while in change of non-essential properties what develops does so in such a way that the outcome is constrained by the nature of what is present at the beginning of the change, while the nature of that substance can be defined apart from the outcome of the change.

11 In this way, on Waterlow’s reading Aristotle’s use of the term ὑποκείμενον is part of a rough first attempt to find a solution to what Charlton calls “the more fundamental problem which influenced all the ‘thinkers of earlier times’” ( pg. 80): the dilemma between change ex nihilo and persistence which undermines the possibility of change. But it is not finally the solution. The solution comes by making the proper distinction between thing and property, and by recognizing that this requires positing essences. A difficulty with the solution as Waterlow puts it is that, in the end, it is not clear that her interpretion ascribes to Aristotle an explanation of substantial generation so much as substantial development. The view which Aristotle arrives at explains how an oak tree comes to be from an acorn by seeing that the acorn is really the same substance as the oak tree (it has the same substantial form), but not the mature form. The fact that it goes by a different name is a quirk of language, and not indicative of a substantial difference. If this is the case, then it is not quite clear why the change should count as the coming to be of the substance; it leaves open the question of how that substance, of which the acorn is the immature form, could come to be simpliciter. Nevertheless, Waterlow’s central insight that for Aristotle the explanation of change depends on the distinction between thing and property seems right, as does the idea that the notion of a ὑποκείμενον Aristotle employs has something to do with this. I will argue, however, that the concept of the ὑποκείμενον is not a rough concept which must be supported by the metaphysical distinction between thing and property, but rather a very precise concept which has ramifications for the distinction between thing and property. Further, Waterlow’s reading seems to make it so that only substances, qua items having essences, can be matter, for only things with essences can be ὑποκείμενα.2 By contrast, I argue that the constraint that Aristotle puts on something being a ὑποκείμενον, and thus the matter in an instance of change, is weaker than this. All that is needed to allow that something persists is that what persists does so under an appropriate description, not that what does so persists qua substance or under an account for which the ontological correlate is an essence. All that is necessary is that there be something

2 Nevertheless, in the Appendix to Chapter 1 of Nature, Change, and Agency,Waterlow suggests that Aristotle might posit prime matter as the proximate matter of the elements. (pg. 46) 12 which is the precondition for the outcome of the change which persists through the change. This will allow Aristotle to explain not only substantial development and accidental change, but also substantial generation. What is required for something to be a ὑποκείμενον is not that it has an essence or is a being per se, but rather that it, under the description according to which it persists, is the precondition for the outcome of change. This, I will argue, fits better with Aristotle’s concept of ὑποκείμενον from his logical works and his claim that talking about change in terms of ὑποκείμενον is another way of getting at the same truth that can be characterized in terms of potentiality. Like Waterlow and Charlton, I think that in Physics A Aristotle holds that explaining change means rooting it in what came before, such that the outcome can be grasped in the light of a proper understanding of the terminus a quo as a principle of the change. Similarly, I think that Aristotle’s account of change is closely tied to the distinction between substance and property which Aristotle motivates elsewhere in his corpus. However, I think the way that it does so is by relying on his technical notion of ὑποκείμενον, a notion developed in his logical works, rather than by relying on his notion of substance. Substances are ultimately the only things in Aristotle’s ontology; but to be a ὑποκείμενον is not to be a substance. Consequently, the distinction between a ὑποκείμενον and what it is ὑποκείμενον of is not the same as the distinction between thing and property, as I argue in my chapter 2. Nevetheless, the distinction between ὑποκείμενον and what it is ὑποκείμενον of does provide a metaphysical schema in the context of which Aristotle’s distinction between thing and property can be understood. In this way, I think that Charlton (and Bostock, in part) is right to see Aristotle’s use of that term in Physics A as an invocation of the logical and conceptual apparatus developed in the logical works. There are several reasons for this. One is that the term was, in those logical works, developed as a key term in Aristotle’s theory of scientific explanation or knowledge, and as a key concept in his metaphysics. It is a loaded term for Aristotle. In Physics A Aristotle’s discussion seems to assume and to contain references to both the discussions of scientific knowledge in which the term ὑποκείμενον appears, and to the metaphysical themes Aristotle employs the term to elucidate. Because of this, I think that the best interpretive strategy is to read it as bringing into the Physics discussion the

13 conceptual machinery of the logical works, and the metaphysics which comes along with that. This helps to introduce a further point of difference between the interpretation I will develop here and those given by Bostock, Waterlow, and Charlton. I argue, contra Waterlow, Bostock, and Charlton, that ὑποκείμενον should not be understood as synonymous or coextensive with substance. I argue that to be a ὑποκείμενον is to be the precondition for, or what is logically presupposed by, something else. To be a ὑποκείμενον is to be something which, just by being what it is, is suited also to be something else, or has the characteristic which is presupposed by some other characteristic. For example, to have something which is in tune one must have a musical instrument, something which is tunable; a musical instrument is the ὑποκείμενον for a tuned thing. Similarly, to have an educated thing one must have an intelligent thing, since intelligence makes something educable; a human person is the ὑποκείμενον for someone or something which is educated. By being a musical instrument something is suited to be tuned, lacking only the action of some agent in tuning it. By being an intelligent creature a person is something suited to be an educated thing, lacking only a teacher or the right opportunities to become educated. A characteristic in virtue of which something is a ὑποκείμενον need not determine or affect every other characteristic; it need only be the grounds for a single other characteristic. Again, it need not be that in virtue of which the thing is, fundamentally, it need only be what is presupposed by some other characteristic which it has Though I do not agree with Bostock that ὑπόμενον is a synonym for ὑποκείμενον in Physics A.7, such that to be a ὑποκείμενον is just to be what persists through change, I think Waterlow, Bostock, and Ross are correct that, for Aristotle, something must persist through every change. I can see no other way of understanding Aristotle’s repeated claim (190a10, 190a17-20) that when something changes, something persists [ὑπομένει, a19] and something else does not [οὒχ ὑπομένει, ibid]. The word ὑπομένω means “to remain,” a natural translation of which in this context is “to persist.” Charlton never quite explains what he thinks licenses him to read it otherwise. But if something persists, how could it be other than a substance? By Aristotle’s lights there are no free-floating non-substances, no non-substances which exist without being in or in some other way dependent on 14 substances. In this light it is clear why Waterlow, Bostock, and Ross should see Aristotle as being compelled to say that all ὑποκείμενα are substances. Aristotle argues that a ὑποκείμενον is present in every change through the same considerations by which he argues that something persists through every change, so it seems as though the same reasons compel him to posit a ὑποκείμενον as a principle in change also compel him to posit substance as a principle. However, I argue that on the account given in his Physics one does not have to appeal to the thing’s being a substance to understand the instances of change in which it is present as ὑποκείμενον. That is to say, one does not have to appeal to its being a per se or essential being to understand the change. Rather, one must appeal to the sense in which it is the the ὑποκείμενον for some further characteristic. It might be a ὑποκείμενον under the same account by which it is a substance. Nevertheless, once we have appropriately divorced Aristotle’s notion of substance from his notion of ὑποκείμενον, my reading of Physics A.7 will show that that something’s having an essence is not the relevant fact about it for understanding the changes in which it is involved as ὑποκείμενον.

§.2 Abstracts

Chapter 2 Abstract:

The goal of this chapter is to provide an interpretation of Aristotle’s logical notion of a ὑποκείμενον, as it appears in his Categories and Posterior Analytics. This interpretation will distinguish being a hypokeimnon from being a substance, and will lay the groundwork for my interpretation of Aristotle’s notion of matter in the chapters of my dissertation which follow. In his Categories Aristotle famously discusses what it is to be a substance. That he does not give a full exposition of what it is to be substance in Categories is also well known, for in discussing substance Aristotle is participating in a debate in which the stipulated meaning of substance or οὐσία is already established. In general, to be a substance is to be the sort of thing which is such that everything else in one’s ontology 15 exists by depending on or being in some way related to a substance. Even with this established, a theory of substance is needed, one which will give criteria for being a substance that will explain how something which fulfills these criteria can do the work of being a substance. Scholars generally agree that in Categories Aristotle presents the criterion for something being a substance as its being a ὑποκείμενον, thus laying great importance on Aristotle’s notion of ὑποκείμενον. Yet the question of what it is to be a ὑποκείμενον for Aristotle tends to be under-studied. There is not, after all, a similar background of philosophical debate for Aristotle’s notion of the term ὑποκείμενον as for the concept of substance. ὑποκείμενον is a term whose use is developed and stipulated by Aristotle. In Categories he gives some discussion of the theoretical framework of which it is a part, detailing some of the rules of the predication relations it underlies. Yet he provides no theory of or criteria for being a ὑποκείμενον. He does present examples of things he claims are and others which are not, subjects in the relevant sense. He also suggests that there is a distinctive relationship between being a substance and being a ὑποκείμενον. From this it is often inferred that the latter is the criterion for the former. In this chapter I argue that ὑποκείμενον is not presented as the criterion for substancehood in Categories, for it is not the case that in Categories Aristotle proposes that all and only substances are subjects. As I will show, Aristotle presents many examples of non-substances which are ὑποκείμενα. He also never claims that all substances are subjects, and, further, it does not follow from anything he says that all substances are subjects. I also present some mixed evidence from elsewhere in the Aristotelian corpus that he is willing to admit as a substance something which is not a subject. This compels us to ask anew what it is to be a ὑποκείμενον, and what being a ὑποκείμενον is meant to tell us about being a substance. In so doing it calls into question some entrenched views both about what it is to be substance and what it is to be a ὑποκείμενον for Aristotle. It also puts into sharper relief the question of the criteria for being a ὑποκείμενον. I argue that in Categories Aristotle provides few clues about his theory of subjecthood, beyond generally laying out the framework of ontological predication in which it functions. But Categories does contain an overlooked pointer to the discussion of strict 16 and accidental predication in Posterior Analytics A.22. I argue that we can extract from this discussion Aristotle’s criterion for being a ὑποκείμενον. On my reading, the text in An. Po. A.22 shows that Aristotle’s criterion for being a ὑποκείμενον is for something to be the grounds for an ontological predicate by being ὅπερ τι what it is. That is, for something x to count as a subject for something else y, in the relevant sense, is for the former to be, qua x (or ὅπερ x τι), a presupposition or precondition for y. I argue that being ὅπερ x τι in the sense Aristotle has in mind does not require being an essential or καθ’αὑτό being. In other words, the criterion of being ὅπερ x τι for subjecthood does not require something to be a substance in order for it to be a subject. I argue that it also provides a new perspective on why Aristotle thought it was informative to characterize substance in terms of ὑποκείμενον in Categories. This understanding of what it is to be a ὑποκείμενον, moreover, serves as a pillar of my interpretation of Aristotle’s account of change and introduction of matter in his Physics A. As I argue in the following chapters, Physics A contains many references to the discussions and theses detailed in Aristotle’s logical works, including his Categories and Posterior Analytics. Aristotle’s appeal to the notion of a ὑποκείμενον is just one of these references. Since Aristotle eventually posits ὑποκείμενον as one of three principles of change, what this means for his account of change depends significantly on how one understands the notion of being a ὑποκείμενον to which he is appealing. Consequently, my understanding of Aristotle’s notion of ὑποκείμενον has ramifications for how to read Physics A.

Chapter 3 Abstract:

Having drawn from Categories and Posterior Analytics what I take to be Aristotle’s logical-metaphysical use of the term ὑποκείμενον, in this chapter I turn to the first book of his Physics. The primary aim of this chapter is to show how Aristotle’s notion of ὑποκείμενον enters his dialectical search for the principles of change in Physics A. In order to do so, I begin by pointing out the several references to themes from Aristotle’s logical works which can be found in A.1-3. In doing so my goal is two-fold: (a) tracing out the many references to Aristotle’s logical works which can be found in 17 these chapters in order to clarify what question Aristotle is trying to answer in Physics A, and (b) by emphasizing the continuity between Physics and the logical works to provide support for my decision to read the term “ὑποκείμενον” as being synonymous between the discussions contained therein. For in Physics A.1-3 explains what Aristotle is trying to accomplish in that work by appealing to notions of scientific explanation which he investigates and explains more fully in his Posterior Analytics. Further, in laying out the possibilities for principles and in attacking the monists, Aristotle appeals to a metaphysical schema of substance and properties such as he lays out in his Categories and is supported in his Posterior Analytics. Physics A.1 begins with Aristotle rehearsing his claim that scientific knowledge arises from knowledge of the principles, causes, and elements of the subject of the science; this echoes, if in a compressed way, the conclusion of the discussion of principles with which Aristotle opens Posterior Analytics A. It continues by appealing to the familiar Aristotelian distinction (familiar from, e.g., An. Po. A.2 71b29-72a5) between what is “more knowable and clear to us” vs. “those which are clearer and more knowable by nature” for advice about how the one investigating the natural world should proceed. The references to the role of principles in scientific knowledge and to the distinction between what is more knowable to us and more knowable by nature are appeals to notions developed by Aristotle, and not to a common background to which any philosopher of his time or milieu would assent. Yet the advice Aristotle gives at 184a18- 25 for how the scientist should proceed seems to be in tension with the referenced passages of An. Po., for he suggests moving from what is more knowable to us toward what is more knowable by nature, the opposite of the direction he had proposed in An. Po. Further, in Phys. A.1 Aristotle seems to suggest that to move in this direction we start from what is universal [καθόλου] and move toward what is particular [καθ’ἕκαστα]. This appears to be in conflict with Aristotle’s claim in An. Po. that what is more familiar to us are the particulars, which are closer to sensation, while what is less familiar to us and more knowable by nature are what is further from sensation. Both of these tensions seem to undermine the claim that in Physics A Aristotle is relying on views the locus of discussion for which is An. Po.

18 I argue that both these apparent tensions can be resolved, and that their resolution supports reading Physics A as taking for granted discussions from An. Po. Ross, in his edition and commentary on Physics, provides a satisfactory solution to the first. He points out that the first book of Physics involves the search for principles, a project Aristotle only describes in the last chapter of An. Po., and not the project he has in mind in the text from earlier in An. Po. from which the tension arises. The second, I argue, can be resolved in part by relying on Jonathan Barnes’ contention in his translation and commentary on Posterior Analytics that the question of what is more familiar to us is an empirical question (pg. 97), such that there is no one answer that will hold in all cases. Still, this alone does not clear Aristotle of contradicting himself between Phys. A.1 & An. Po. A.4. I argue that when in the Physics passage Aristotle suggests that we should move from what is καθόλου to what is καθ’ἕκαστα, καθ’ἕκαστα does not there denote the particulars which are the logical opposites of universals. Barnes’ contention helps to show that it is not inconsistent for Aristotle to set a universal as a starting point in the search for principles, and my interpretation of the use of καθ’ἕκαστα shows that Aristotle is not contradicting himself by suggesting that particulars are more knowable by nature than the universals they fall under. Aristotle can thus be accused of infelicity and lack of clarity in expression here, but not of inconsistency in his use of a key notion from his epistemology and philosophy of science. Further support for the consistency and continuity between Physics A and Posterior Analytics can be found by reflecting upon the plan of investigation Aristotle lays out in the opening lines of A.2. In A.1 Aristotle introduced his goal in Physics A as being the search for the principles of a science. In A.2 Aristotle clarifies that the science for whose principles he is searching is the science of nature. He also outlines the possible views about the principles by dividing them up according to number and kind of proposed principles. This outline provides a plan of investigation which he follows from A.2-A.7. Since his reasoning in these chapters forms a continuous argument, if Aristotle is relying on An. Po. discussions in Physics A.1, then it is reasonable to take him as doing so in A.2-7, too. This lends support for the claim that when Aristotle introduces terms such as ὑποκείμενον in A.2-7, a term which had played such an important technical role

19 in his logical works, he is introducing into the Physics discussion the same notion which he had employed and reflected on in An. Po. I argue that Aristotle’s response to Parmenides and Melissus in A.2 & 3 gives more positive support that the notion of ὑποκείμενον employed in Categories and discussed in Posterior Analytics is important to Aristotle’s search for principles in Physics A. In Phys. A.2 Aristotle argues that Parmenides and Melissus make no contribution to natural science when they propose that all is one. Rather than leaving it at this, however, Aristotle offers a rebuttal to Eleatic monism. He does so, he says, because the examination of their view “is still not without scientific interest.” (185a20) I argue that the “scientific interest” lies in the way in which an examination of their view highlights something that they missed that is key to natural science, according to Aristotle: that being is said in many ways. The ways that being is said which Aristotle claims the Eleatic monists overlooked can be divided along three different lines: the categories, actuality and potentiality, and ontological subject (or ὑποκείμενον) and predicate. All three of these will be important in subsequent discussions in Aristotle’s Physics. I argue that the last, the distinction between ontological subject and ontological predicate, is key to the search for the principles of change. I defend this by showing how Aristotle highlights the way this distinction impedes Parmenides from making any contribution to natural philosophy. I argue also that the text here is also consistent with the understanding of ὑποκείμενον I defended in my chapter 2. Finally, I set the stage for my reading of Physics A.7 by showing how Aristotle brings the need for a ὑποκείμενον into his discussion of principles in A.4-6. In A.4 Aristotle derives from Anaxagoras’s proposal for the principles of natural philosophy this useful conclusion: “in nature nothing acts on, or is acted on by, any other thing at random, nor may anything come from anything else” (188a31-34, emphasis mine). Anaxagoras, according to Aristotle, had thought that anything could come from anything on the grounds that everything is present in everything. While disagreeing with Anaxagoras about whether everything is in everything and anything can come from anything, Aristotle draws a general lesson from Anaxagoras: that there is a necessary precondition for every change.

20 Aristotle uses this in A.5 to argue that the precondition for any change is the opposite or contrary of its outcome. Consequently, he proposes that contraries are principles of change. Bostock has proposed that this is a superficial conclusion, for Aristotle has said no more than that “a thing and its opposite” are principles. I argue that the contraries Aristotle has in mind are not merely contradictories; they are not merely x and not-x. Things come to be not merely from their contradictories (what is not-x), but from something having a contrary property which is also the presupposition of the characteristic they will have at the outcome of the change. For example, the precondition for something being or becoming tuned is not merely that it be not-tuned, but that it be out-of-tune. A piano or harp can be out-of-tune, and must be if it is to become tuned. A fish, on the other hand, can only be not-tuned. It is never out-of-tune; thus it cannot be tuned. Aristotle argues that a particular pair of contraries, form and lack, are the principles and termini of every change, and uses examples like my tunedness example to fill out what he means by form and lack. In A.6 Aristotle presents other considerations which suggest that one must also posit a ὑποκείμενον in change. These considerations are drawn from his reflections upon some views by his predecessors, as well as from his own metaphysical commitments. I argue that the philosophical considerations according to which Aristotle motivates positing a ὑποκείμενον in change suggest that the best way to read Aristotle’s appeal to ὑποκείμενον in this chapter is as synonymous with ὑποκείμενον in his logical works. This, in addition to the many other references to the themes and conclusions from the logical works, underscores the synonymity between ὑποκείμενον in Aristotle’s final account of the principles of change in A.7 and his discussion of scientific explanation and metaphysics in his logical works. However, since the reasoning by which he proposes that there must be a ὑποκείμενον is orthogonal to the reasoning by which he proposes that form and lack are principles, Physics A.6 presents Aristotle with the difficulty of putting together a final, coherent account of the principles of change. This is the task he takes up in A.7.

Chapter 4 Abstract:

21 In this chapter I provide a textual analysis of Aristotle’s reasoning in Physics A.7, whereby he proposes ὑποκείμενον as a principle in change. I focus on the framework of complexes and simples which he proposes at 189b32-190a5. Aristotle proposes this framework as a way of thinking about change which fits with both some common ways of talking about or describing change and with his conclusions from A.5 & 6, and which he can use to illustrate the logical analysis by which he arrives at his account of the principles. I argue that Aristotle applies to this framework a logical analysis which depends centrally on the notion of a ὑποκείμενον from An. Po. This analysis, he proposes, makes change intelligible and provides guidelines for the natural scientist. As a result of this analysis Aristotle posits two different but consistent ways of understanding the principles of change. Aristotle concludes that in a way there are three principles—a ὑποκείμενον, εἶδος (form), and στέρησις (lack)—,and in a way there are two principles—ὑποκείμενον and εἶδος. Understanding either set of principles as principles requires understanding the way in which a ὑποκείμενον stands as the precondition for what it is ὑποκείμενον of, by being just what it is (ὅπερ x τι). The framework Aristotle proposes is constructed as follows: in every change there is a logically complex item at both the terminus a quo and the terminus ad quem, each complex item being composed of a pair of logically simple items, such that each has as a part one member of a pair of contraries. In Aristotle’s favored example, the case of an unmusical man becoming a musical man, the complexes are the unmusical man and the musical man, while the simples are the unmusical and the man, on the one side, and the man and the musical on the other side. The complexes are logically complex in that they are one in number—in a fundamental sense there is really only one thing there—while they are two in form or account [λόγος]. They are two in account in that the terminus answers to two different definitions [λόγοι], corresponding to its two names, e.g., unmusical and its definition, man and its definition. Because of this, there is another sense in which each of the termini is two: e.g., an unmusical thing and a man. Aristotle calls the terminus a quo the ὑποκείμενον of the change. Further, the complexes are constructed such that each has one of a pair of contraries, e.g., musical and unmusical, plus something which is not a contrary to anything else involved in the change, e.g., man.

22 The ὑποκείμενον persists through the change under the account according to which it is not a contrary, while it does not persist according the account by which it is. That the ὑποκείμενον persists under one of its two λόγοι captures the insight which was the concern of the discussion in A.6: in every change there must be some one thing which undergoes the change, and which is such as to be affected by something else so as to change. That the ὑποκείμενον under the other of its λόγοι is a contrary, and contrary to one of the λόγοι of the result of the change, captures the insight from A.5. Further, Aristotle’s insistence that the ὑποκείμενον is complex in account shows how the terminus a quo he has pointed to accurately picks out the precondition for the change. As Aristotle had argued in A.5, what it means for the contrary at the terminus a quo to be or to have a “lack” is not that it is or has the absence of the characteristic which will be present at the outcome of the change. Rather, this terminus is positively characterized, and the characteristic which is relevant to understanding the change is the one which will be presupposed by the characteristic which will be present at the outcome, or, in other words, the one which will be the grounds for the outcome characteristic. Aristotle plays upon an ambiguity in the meaning of τὸ + neuter expressions in Greek to give the two sets of principles. In the enumeration of the principles of change as three, ὑποκείμενον, lack [στέρησις], and form [εἶδος], Aristotle takes το + μουσικόν to refer to an abstract property: musicality (or unmusicality). This enables him to clarify the understanding of the precondition for change from A.5. On this enumeration the ὑποκείμενον, under the account according to which it persists, picks out something as having the foundational characteristic which will be presupposed in the outcome. For example, if the outcome is something musical, the ὑποκείμενον of the change in the 3- principle enumeration will be something rational, the man, since the outcome’s musicality is based on its being rational. But to see this item, the person, as the ὑποκείμενον of the change requires seeing it not just as something having a certain characteristic statically, but to see it as not yet having another. Thus Aristotle posits the lack, e.g. unmusicality, as a second principle. In this enumeration, Aristotle uses the τὸ + neuter construction to refer to an abstract characteristic, such as unmusicality. In the two- principle enumeration, where Aristotle names the principles ὑποκείμενον and form, he takes the principles to be a pair of contraries. Here he treats the τὸ + neuter construction 23 as picking out a thing which has some characteristic, qua having that characteristic. To use his example, τὸ μὴ μουσικόν refers to an unmusical thing, qua unmusical, and it is this which is the ὑποκείμενον. The two accounts belonging to the ὑποκείμενον reveal the structure of the precondition for the change just as the two principles present at the terminus a quo did in the 3-principle enumeration. By logically distinguishing the elements involved in the change Aristotle not only avoids the dilemma which had faced his predecessors, but also accomplishes what he set out to do, namely, to discover the explanatory structure according to which any change can actually be explained. The appeal to ὑποκείμενον is key to accomplishing this. In positing a ὑποκείμενον Aristotle points out that there is a discoverable precondition for every change. He clarifies what it is to be a ὑποκείμενον, and in so doing explains what it is to be the precondition for change: having the two λόγοι in the way described. These λόγοι indicate what a scientist must discover about something to accurately pick out the precondition for any change. One has not understood what something has come to be from until one understands the outcome well enough to see what characteristics it is built upon. To explain the change, one must pick out the relevant state of affairs at the initial terminus of the change, by pointing to something which has a complex λόγος as described above. The item which is the terminus a quo understood in this way is the matter.

Chapter 5 Abstract:

In this chapter I explain what the textual analysis from my chapter 4 means for Aristotle’s understanding of matter. I make use of Mary Hesse’s analysis of analogy and analogous reasoning in Aristotle to argue that Physics A.7 reveals that matter, for Aristotle, is analogous. That is to say, matter does not form a distinct category of being, except by analogy, such that it does not form a rival to substance, and things entering into any of the non-analogous categories Aristotle posits might serve as matter in the case of some change. Ultimately, to be matter is to play a certain role in the explanation of change by being the ὑποκείμενον of that change. One significant upshot of this is that scholastic prime matter is inconsistent with Aristotle’s concept of matter. 24 Because to be a ὑποκείμενον is just to be what has the characteristics upon which are founded, or which are presupposed by, the the thing which is present at the outcome of the change, finding out what the matter of some change or substance is will always be an empirical question. Since the ὑποκείμενον in an instance of change is the precondition for the outcome of that change, its ὑποκείμενον will also be matter insofar as it is the precondition for the outcome of that change. Matter is always some other natural thing, which has the characteristics which are the foundation for further ones. If it is a natural thing, however, it is itself in principle changeable; the world Aristotle envisions is one in which everything which is in it is subject to change. Because of this, there is nothing which will count as matter for Aristotle which is not itself subject to change. But if the thing which counts as matter for one change is itself subject to change, it will itself have some matter. Thus there will be no matter which is fundamental in the sense that it is not itself from anything else as matter. This helps to make sense of why the basic material elements Aristotle posits can change into one another, unlike his predecessors’ If there is no fundamental matter in the sense of something which itself does not come to be, then a fortiori there is no prime matter in the scholastic sense. Charlton provides evidence that can be taken to show that Aristotle never uses the expression πρῶτη ὕλη anywhere in his corpus to refer to some matter which is prime by being fundamental in this way. Further, given that to be matter is to be a ὑποκείμενον, and to be a ὑποκείμενον is to be something positively characterized, for Aristotle there can be no matter which is bare or featureless. This is a further stroke against scholastic materia prima. On the basis of my exegesis of Aristotle’s account of change in Physics A, I argue that something like scholastic materia prima could not do the explanatory work required of it qua matter, in Aristotle’s view. My understanding of Aristotle’s matter as analogous and as ὑποκείμενον also has ramifications for further work on Aristotle’s theory of substance. My argument in my chapter 2 that Aristotle never introduces being a ὑποκείμενον as a criterion for substance opens up the possibility that there is no tension between Aristotle’s discussion of substance in his logical works and his discussion of substance in his Physics and his Metaphysics. Many scholars have argued that there is an incompatibility between these approaches, seeing tension between Aristotle’s depiction of substance as ὑποκείμενον in 25 Categories and his introduction of matter as ὑποκείμενον in Physics. On this line of thinking, Aristotle is compelled by his reasoning in Physics and Categories to conclude that matter, or prime matter, is primary substance. This involves his metaphysics in difficulty, for on independent grounds matter (or prime matter, for those finding the doctrine of prime matter in Aristotle) is disqualified as a candidate for being a substance. This prompts Aristotle to revise his theory of substance, a task which he undertakes in the central books of his Metaphysics, especially Metaphysics Z. Metaphysics Z.3 is a key text for interpreting Aristotle’s solution to this difficulty, for in it Aristotle seems (to some) to set aside ὑποκείμενον as a criterion for substancehood on the grounds that it would make matter to be primary substance. My reading of ὑποκείμενον, and matter as ὑποκείμενον, helps to dissolve this one possible source of tension between Aristotle’s later theory of substance and his discussion of substance in his logical works. Further, it lays constraints on how one can interpret Metaphysics Z.3. In Z.3 Aristotle presents what has been called a “stripping argument,” a sort of thought experiment in which he imagines “stripping” each of the predicates or properties from a substance to reveal the ultimate ὑποκείμενον . What he arrives at is something bare, which, on the grounds that it is bare or featureless in itself, cannot stand as primary substance. From this it appears that Aristotle retracts his earlier proposal that ὑποκείμενον is the criterion for substancehood, and so embarks on a search for a new criterion. On my understanding of ὑποκείμενον in the logical works and matter in Physics A, in Metaphysics Z.3 Aristotle is criticizing a view which is not his own, or, rather, a mistaken understanding of his own. For neither does his concept of ὑποκείμενον or of the predication relations which it underlies allow for a bare ὑποκείμενον, nor is matter bare, on his view. In other words, the “stripping” which is imagined in Z.3 is impossible on his own view. It does not provide him with occasion to retract his claim from Categories that those things which are the ultimate subjects of predication (ὑποκείμενα which are not themselves predicates) are primary substances, and it does not show that he is compelled to take matter as primary substance if he does not reject the Categories claim.

26 Chapter 2: Ὑποκείμενον in the Logical Works

§.1 Introduction

Since in this dissertation I wish to argue that conceiving of matter as ὑποκείμενον is important to understanding what matter is for Aristotle, I must be clear about how I think Aristotle understands ὑποκείμενον in this context. I think that when Aristotle introduces matter as ὑποκείμενον in Physics A, he is appealing to a technical notion he develops in his logical works, specifically in his Categories and Posterior Analytics. My task in this chapter is to spell out what it is to be ὑποκείμενον in this sense; in the following chapter, chapter 3, I will argue that this is the sense Aristotle employs in his Physics. To suggest why it matters to give a treatment of what it is to be ὑποκείμενον in its own right, let me briefly point to a puzzle in Aristotle interpretation which this sort of investigation might help to solve. For one line of interpretation, e.g., Loux (1991), Lewis (1992), and Gill (1989 & 1996), the claim that Aristotle uses ὑποκείμενον synonymously between Categories and Physics will come as a friendly remark. Loux, Lewis, and Gill, holding that matter is introduced as ὑποκείμενον in the sense from Categories are concerned with the apparent tension between Aristotle’s suggestion in Categories that substances such as Socrates and Bucephalus are ὑποκείμενα in an important way, and his treatment of matter as ὑποκείμενον in Physics. These scholars hold that in Categories being a ὑποκείμενον is presented as a criterion for being a substance. If matter is ὑποκείμενον in the same sense, they worry, it will also be substance, and, indeed, the primary or ultimate matter will be primary substances. This would constitute a serious problem for Aristotle, since in Physics he also finds matter to be one of the compositional parts of things such as Socrates and Bucephalus, which he treats as substances both there and in Categories. If primary substances are the fundamental beings in Aristotle’s ontology, and their being ὑποκείμενα is what enables them to be so, it seems that it cannot be that both Socrates and his matter are substances. For then there would be a substance in a substance, a substance which depends on some other substance just the way everything else is supposed to depend on substances. On the one hand, this might 27 simply be a problem in Aristotle’s view. But scholars are typically, and rightly, motivated to avoid positing this. Thus Gill, Loux, and Lewis argue that when Aristotle posits matter in his analysis of natural substances, this prompts him to re-examine and consequently to reformulate his metaphysics of substance, a task which he undertakes in his Metaphysics, especially Metaphysics Z. In other words, when in Physics Aristotle introduces matter as ὑποκείμενον, and then as one of the two parts of hylomorphic composites, he finds he must rethink his criteria for being a substance, and a primary substance. Loux and Lewis argue that in M.Z Aristotle rejects his Categories characterization of substance as the ultimate ὑποκείμενον, though they disagree about what view Aristotle offers in its place. Gill, on the other hand, argues that Aristotle reexamines his notion of ὑποκείμενον so as to preserve it as a criterion for substance. Thus Gill, too, suggests that Aristotle must modify his view of substance after Physics, but he does so at least in part by reevaluating his view of ὑποκείμενον. Sean Kelsey, on this other hand, in his (2008), offers an evocative interpretation of the role of ὑποκείμενον in the argument of Physics A.5-7 which suggests that the view of matter which emerges is not such that it challenges the claim of ὑποκείμενον as a criterion for substancehood. On the other hand, another line of interpretation holds that in Physics Aristotle is using the term ὑποκείμενον homonymously with his use in Categories; he means something different by this term than he had meant there. One way of characterizing what ὑποκείμενον in Physics A is on this view is “whatever persists”. Morison (forthcoming) holds this view. David Bostock, on the other hand, suggests that it means “what Y is ‘out of’ ( ἐκ), which means either what Y is made from or what Y is made of.” (Space, Time, Matter, Form, pg. 31) Whatever happens between Categories and Metaphysics Z, on this sort of line Aristotle does not have to reconsider whether ὑποκείμενον forms a criterion for substance, at least on the grounds of matter being ὑποκείμενον. This disagreement between these two lines of interpretation depends a great deal on what it is to be a ὑποκείμενον in Categories. But there is a curious lacuna in scholarship on Aristotle’s logic and metaphysics, in that few authors investigate Aristotle’s notion of ὑποκείμενον in its own right, even by confining this investigation to Aristotle’s logical works, as I will in this chapter. Though 28 they tend to agree that it is offered as a criterion for substancehood in Categories, discussion of what it is to be ὑποκείμενον is offered only in passing, and generally only with an eye to what is relevant to the discussion of substance. There is a reason for this: substance seems the more compelling topic, since it is about this that Aristotle is in disagreement with Plato, and it is in the interest of this disagreement that Aristotle introduces the notion of ὑποκείμενον. But if being a ὑποκείμενον is so key to Aristotle’s exposition of his ant-Platonic view of substance, at least in his logical works, to the extent that we do not understand being a ὑποκείμενον we cannot understand what Aristotle is trying to tell us about what it is to be a substance. In Categories Aristotle gives neither a full exposition of what it is to be substance nor, I argue, of what it is to be a ὑποκείμενον. That Aristotle does not give a full exposition of what it is to be substance in Categories is well known. Loux, for example, argues that in Categories substance is taken as an “antecedently understood term” (Primary Ousia, pg. 15), such that Aristotle cannot merely stipulate the meaning, but is participating in a debate in which the stipulated meaning is already established. In general, to be a substance for Aristotle or Plato is to be the sort of thing which is such that everything else in one’s ontology exists by depending on or being in some way related to a substance; it is what grounds everything else in one’s ontology. Alan Code (1986) and Russel Dancy (1975) similarly point out that although Aristotle and Plato disagree about what the criteria for being a substance is and which things are substances, they are still arguing about how to fill out a term of art on whose general use they agree. What each must provide is a theory of substance, one which will give criteria for being a substance that will explain how something which fulfills these criteria can do the work of being a substance. There is not, after all, a similar background of philosophical debate for Aristotle’s notion of the term ὑποκείμενον as for the concept of substance. It is a term whose use, like that of ἐντελεχεία, is developed and stipulated by Aristotle. In Categories he gives some discussion of the theoretical framework of which it is a part, detailing some of the rules of the predication relations it underlies. Yet he provides no theory of or criteria for being a ὑποκείμενον. He does present examples of things he claims are and others which are not, subjects in the relevant sense. He also suggests that there is a distinctive 29 relationship between being a substance and being a ὑποκείμενον. From this it is inferred that the latter is the criterion for the former. This has contributed to how brief and gloss-like are most explicit attempts to characterize Aristotle’s logical notion of ὑποκείμενον. For example, David Bostock characterizes the sense of ὑποκείμενον from Categories as follows: “In the Logical Works, to say that X underlies Y is just to say that X is the subject of which Y is predicated.” (Space, Time, Matter, Form, pg. 33). This is helpful as an initial gloss, but not very informative, finally, about what it is to be a ὑποκείμενον. Loux, similarly: “Let us say that a thing x is a subject for a thing y just in case y is either said of or present in x, and that a thing is a basic subject just in case it is the subject for something else but nothing is its subject.” (Primary Ousia, pg. 23) Russell Dancy (1975) argues that being a subject means being a subject only of the said-of predication relation outlined in Categories, and not also of the relation of being in-a-subject. Bostock’s gloss seems to be defining the relationship of being a subject (or “underlying” in the relevant logical sense”), but uses the term “subject” in the definition. Loux’s definition tells us which relata subjects are in a relation with, and Dancy’s definition seems to do the same; this is something like defining “father” by pointing out that a father is a father of a child. Much is left out. In the case of ὑποκείμενον, it is still unclear what sort of thing underlies in this way, what sort of characteristics it must have to be a ὑποκείμενον, or what it is to underlie whatever in fact it does underlie. In other words, we can ask similar questions of ὑποκείμενον as those concerning substance to which many authors think Aristotle is proposing answers in Categories. I begin with a brief survey of where, and how, the word ὑποκείμενον is used in Aristotle’s corpus, in order to narrow my focus to the use, and the passages, which are salient to my particular project. After this I will distinguish several different senses of “subject” to shed light on how Aristotle is using ὑποκείμενον in the relevant passages. I argue that Aristotle’s use of ὑποκείμενον strictly speaking (κυριώς) maps onto what I am calling an “ontological subject”, and his use of ὑποκείμενον in the context of accidental predication (in a sense to be distinguished), maps onto what I am calling “sentence- subject”.

30 While drawing from Categories some basic features of being a ὑποκείμενον, I will argue that in Categories Aristotle does not present a full picture of what it is to be a ὑποκείμενον. Further, I argue that in Categories being a ὑποκείμενον is not presented as the criterion for being a substance, for it is not the case that all and only substances are subjects in that work. As I will show, Aristotle presents many examples of non- substances which are ὑποκείμενα. He also never claims that all substances are subjects, and, further, it does not follow from anything he says that all substance are subjects. I also present some mixed evidence from elsewhere in his corpus that he is willing to admit as a substance something which is not a subject. This compels us to ask anew what being a ὑποκείμενον is meant to tell us about being a substance, thus calling into question some entrenched views both about what it is to be substance and what it is to be a ὑποκείμενον for Aristotle. It also puts into sharper relief the question of the criteria for being a ὑποκείμενον. I argue that in Categories Aristotle provides few clues about his theory of subjecthood, beyond generally laying out the framework of ontological predication in which it functions. But Categories does contain an overlooked pointer to the discussion of strict and accidental predication in Posterior Analytics A.22. I argue that we can extract from this discussion Aristotle’s criterion for being a ὑποκείμενον. On my reading, the text in An. Po. A.22 shows that Aristotle’s criterion for being a ὑποκείμενον is for something to be the grounds for an ontological predicate by being ὅπερ τι what it is. I argue that being ὅπερ x τι in the sense in Aristotle has in mind does not require being an essential or καθ’αὑτό being. In other words, the criterion of being ὅπερ x τι for subjecthood does not require something to be a substance in order for it to be a subject. This supports my reading of Categories. I argue that it also provides a new perspective on why Aristotle thought it was informative to characterize substance in terms of ὑποκείμενον in Categories.

§.2 Uses of ὑποκείμενον

The word ὑποκείμενον is a present participle from ὑποκείμαι, whose basic meaning is “lying under”; it comes from ὑπο + κείμαι, and literally concerns position. It 31 also often functions as the passive form of ὑποτίθημι (“place under”). The many metaphorical uses of the term (both in Aristotle’s work and in Attic Greek more widely) stem from this basic sense and its use as a passive for ὑποτίθημι. Aristotle does use the term in its basic or literal sense; Bonitz calls this sensu locali (797b23) and supplies a number of instances. Other instances beyond those Bonitz points out may be found easily, e.g., De Caelo B.13.249b19, where Aristotle points out that Anaximenes and Anaxagoras held that the earth is flat and rests on underlying (ὑποκείμενον) air. The first of the metaphorical significations the LSJ lists is “to be established, set before one (by oneself or another) as an aim or principle3”, and it includes among other instances from ancient authors (such as Plato, Demosthenes, and Polybius) Aristotle’s Politics 1275a35 as an instance. The second metaphorical signification given is “to be assumed as a hypothesis”4, or, in an 1846 American edition of the dictionary “to be laid down, assumed as a ground of argument”. (pg. 1568) The LSJ offers several instances from Plato, as well as from Aristotle. To take an example not cited there, Aristotle uses the term in this way in, for example, An. Po. B.12.62a27: “consequently the supposition [ὑποκείμενον] is false.” Bonitz also gives this sense of ὑποκείμενον in his Index, characterizing it as positum aliquid est (sive sumptum modo et concessum sive demonstration firmatum) tamquam fundamentum ex quo alia concludantur (797b33-35). To be a ὑποκείμενον in this sense is to play a certain role in argument or demonstration; a statement or proposition will be a ὑποκείμενον of this kind. Bonitz gives many more examples, but the An. Po. instance is apposite here. For while use of ὑποκείμενον and ὑποκείσθαι in this sense is sprinkled throughout Aristotle’s corpus, it is in his Analytics that Aristotle examines what makes something to be an appropriate supposition, principle, or premise. The LSJ lists other metaphorical uses (e.g., “to be suggested”, “to be in prospect”, “to be subject to”), many of which one can also find Aristotle, among other Greek writers, employing. More importantly, it also lists uses “in Philosophy.” The focus in the LSJ entry is on certain Aristotelian and Platonic instances. The general sense it gives is

3 http://www.perseus.tufts.edu/hopper/morph?l=u(pokei/mai&la=greek#lexicon 4 ibid 32 as follows: “in Philosophy, to underlie, as the foundation in which something else inheres, to be implied or presupposed by something else.”5 The first three instances it gives are Platonic: Protagoras 349b, Cratylus 422d, and Republic 581c.6 In the first Socrates asks Protagoras whether the names “wisdom,” “temperance,” “courage,” and “justice” all point to one virtue, or whether underlying [ὑπόκειταί] each of these is something different. Here to “underlie” appears to mean being what some name refers to. In Cratylus 422d, Socratres asks Hermogenes how the first names could reveal the natures of the things they name when they are not based on any prior names [οἷς οὔπω ἕτερα ὑπόκειται]. Here the verb points out the inverse of “being derived from”, where the derivation is etymological. In Republic 581c Socrates claims that there are three kinds of pleasure underlying [ὑποκείμενον] the three kinds of men (the lover of wisdom, the lover of victory, and the lover of gain). Here underlying seems to mean being that upon which a distinction or categorization is based. Each of these ways Plato uses the term might be ways Aristotle does at some point in his corpus, but none of them as yet captures the sense in which ὑποκείμενον and its verbal forms are used predominantly in Categories and Posterior Analytics. A sign of this is that, with the exception of the Timaeus Locrus, the LSJ immediately turns to Aristotelian examples. The Timaeus Locrus instance is a curious one, since in that work the term ὑποκείμενον is used to refer to the “receptacle” or “omnireceptacle” (πάσης γενέσεως ὑποδοχἠν, 49a5-6) of Plato’s Timaeus, and to characterize it as matter [ὕλη]. Plato himself uses neither ὕλη nor ὑποκείμενον to refer to his receptacle, and indeed these are both terms which, if used in that way, would be typically Aristotelian uses. This is one reason scholars have thought the Timaeus Locrus was a much later work, written perhaps in the 2nd century C.E., though a reason Ryle offers for dating it as 4th century B.C.E., during Aristotle’s lifetime. Ryle, in his 1965 article on the Timaeus Locrus points out that the use of these terms, with some others, is uniquely Aristotelian, in that apart from their appearance in Timaeus Locrus they appear

5 http://www.perseus.tufts.edu/hopper/morph?l=u(pokei/mai&la=greek#lexicon 6 Unless otherwise indicated, all quotations from Plato are taken from the translations included in Plato: Complete Works, Ed. John M. Cooper. 33 nowhere else until the writings of some of Aristotle’s students and followers. Thus whether we take the Rylean view that the work is by a contemporary of Aristotle’s or the Taylor view that it is by a later forger, it remains that the use of the term ὑποκείμενον in that work seems to be derived from Aristotle’s own usage.7 Because of this it does not seem to stand as an exception to the claim that ὑποκείμενον in Aristotle’s logical works is a technical term developed uniquely by Aristotle. The LSJ entry divides up the Aristotelian uses of ὑποκείμενον as follows: “(1) to the matter which underlies the form, opp. εἶδος, ἐντελέχεια …; (2) to the substance (matter + form) which underlies the accidents, opp. πάθη, συμβεβηκότα …; (3) to the logical subject to which attributes are ascribed, opp. τὸ κατηγορούμενον.” It is, roughly, these with which I am concerned. However, I do not wish to rely on the LSJ’s three-fold division of the signification of ὑποκείμενον, since it presumes answers to questions I am asking. In the same way, I wish to focus on the instances in Aristotle which Bonitz includes under the 3rd signification he points out: positum esse (τὸ ὑποκείμενον id quod positum est) tamquam fundamentum cui alia inhaerant. (798a11-13) I am especially concerned with two of the significations Bonitz lists under this: τῶ ὑποκειμένω, τῆ οὐσία, qae est τόδε τι, καθ’ αὑτό (798b20-21), and subjectum logicum (798b32), which he lists as separate, though related senses. I will argue that “that in which other things inhere” is the logical subject, and further that, at least in Categories and An.Po., Aristotle never uses ὑποκείμενον as synonymous or coextensive with οὐσία, and never sets as a mark or criterion for being ὑποκείμενον that something be καθ’ ἁυτό. This I will argue in this chapter. More to the point, I argue that for Aristotle there are not distinct metaphysical and logical senses of the term ὑποκείμενον. Now, this is not necessarily to say that Bonitz has made a mistake in his index. Rather, I think that Aristotle’s logical sense of ὑποκείμενον has metaphysical ramifications. Sometimes he discusses ὑποκείμενον with an eye to its content as a logical notion, and sometimes more with an eye on the

7 Ryle also suggests that the use of the term in this work has more in common with the use in Physics than in Categories or Metaphysics. This would strengthen my case even further, were it not that I intend to argue that ὑποκείμενον is used synonymously in Categories and in the relevant Physics passages where it picks out matter. 34 metaphysical ramifications. Bonitz’s division may appropriately reflect Aristotle’s different interest in the different contexts in which he uses this term. Again, Bonitz does divide Aristotle’s use of ὑποκείμενον as “what is posited in which others inhere” into three distinct categories: matter, οὐσία, and logical subject (798a25-30). This seems to me to be an error. My task here is to draw these three together, to show that they are all rooted in the sense of ὑποκείμενον as logical subject. Now, showing that these three are linked throughout Aristotle’s corpus is too big a project for the present, not the least because of the additional textual issues that would be involved. My goal in this dissertation is to focus on ὑποκείμενον in Categories, Posterior Analytics, and Physics. My goal in this chapter is to begin in Categories and Posterior Analytics.

§.3 Grammatical and sentence-subjects

I think that the relevant notion of ὑποκείμενον in Aristotle is born out of a commitment which he expresses at the opening of On Interpretation when he says

Now spoken sounds are symbols of affections in the soul, and written marks symbols of spoken sounds. And just as written marks are not the same for all men, neither are spoken sounds. But what these are in the first place signs of—affections in the soul—are the same for all; and what these affections are likenesses [ὁμοιώματα] of—actual things—are also the same. (16a3-8)

The idea is that the written sentences are signs or symbols of spoken sentences, which are themselves signs or symbols of what is in our minds. What is in our minds, on the other hand, is a likenesses rather than a sign or symbol of what is in the world. The first part seems familiar and unsurprising; of course our words are symbols of what is in our minds, and of course what we write symbolizes what we say. Other scholars have puzzled over what it is for words and other linguistic items to be symbols, or, in the modern parlance, representations; this may not be straightforward. For my purposes it is 35 also striking that Aristotle treats what is in our minds not as merely artifacts, but as “affections,” or παθήματα. They are not like our spoken or written words, products of our minds; they are in some way produced by things in the world on our minds. Further, and still striking, these παθήματα are not symbols, or signs or representations of things in the world; instead, they are likenesses of them. Their being likenesses is in part because they are παθήματα.8 Because the things in the soul or mind are παθήματα, the relation between our words and things in the world is not a straightforward case of signification or symbolization. Aristotle seems to suggest, both here and in other difficult passages9, that our minds are to some degree informed by their contact, mediated by our senses, with the world external to them. Because of this, I suggest, Aristotle thinks we can learn something about the way things in the world hold together by thinking about how we have “likenesses” of them in thought and symbols of these in speech. In Categories he does this by reflecting on how things in the world must be related if sentences about them are to be true. In Categories Aristotle limits the domain of his discussion to simple, true sentences. He assumes that these sentences are symbols or signs of παθήματα which are likenesses of things in the world, and emphasizes that the names we use are also symbols of likeness of things. This suggests that something about the structure of a sentence is important, too; different ways of putting together παθήματα in thought and symbolizing this in speech will be likenesses of, or limn, different ways the things in the world might be together. In An. Po. Aristotle reflects on this further by examining how things in the world must be related to each other if they are knowable, and not just if we are to successfully say true things about them. In On Interpretation (henceforth, DI) he lays some groundwork for both of these discussions by distinguishing different kinds of words, sentences, and expressions, and by discussing the functional parts of sentences. In doing so he builds on his predecessors’ and contemporaries’ discussions of grammar in the context of rhetoric, but with the payoff of contributing a substantial metaphysical theory.

8 Aristotle’s De Anima gives a more detailed account of how, on his view, the human intellect is both passive and active. 9 E.g. Posterior Analytics B. 19. 36 Grammatical and linguistic study among the ancient Greeks began with the 5th century sophists.10 Mixing in the milieu of the sophists of the 4th century as they did, Plato, and Aristotle, too, had an interest in grammatical science (and even in phonics), as is evident in their written works. In Plato’s Sophist, for example, the Visitor conducts an examination of the different kinds of names or words.11 In one part of this examination he aims to discover which different kinds of words are such that they “indicate something when you say them one after another” and so “fit together” to form a sentence (261d). In this context he distinguished two different kinds of words. These different kinds of words are such that when a word of one kind is said with a word of the other a sentence is formed and an assertion (or denial) is made. Of these two kinds, one includes words which have to do with action, words which he calls ῥήματα. The other includes words which designate something on their own, which he calls ὀνόματα. This distinction between ῥήματα and ὀνόματα is consistent with a basic understanding of the distinction between verbs and nouns. In the simplest sentence a noun names a thing while a verb indicates an action. Further, minimally, a noun and a verb are required for the formation of a complete sentence. In the opening line of DI 2 Aristotle defines an ὄνομα, a noun, as “a spoken sound significant by convention, without time, none of whose parts is significant in separation.” (16a19) Note that Aristotle thinks of an individual noun, or ὄνομα, as having different forms without being a different ὄνομα altogether. To use his example, Φίλωνος and Φίλωνι are not distinct nouns, according to Aristotle, but simply forms of Φίλων (16a32-16b5). In DI 3 Aristotle goes on to define what he calls a ῥῆμα as “what additionally signifies time, no part of it being significant separately; and it is a sign of things said of something else.” (16b6) Aristotle thus characterizes the difference between ὀνόματα and ῥήματα as being between words

10 Pernot, Laurent. Rhetoric in Antiquity. Washington, DC: Catholic Univ. of America Press. (20050), pp. 14 11 Here the Visitor is discussion words, not, as Plato’s characters do elsewhere in his writing, non-linguistic items according to the way they are represented in speech. The Visitor and his interlocutors are trying to uncover “that which is not”, and have agreed that it is “one kind among others, but scattered over all those which are.” (260b) Since it is so scattered among the kinds of things which are, they decide to examine whether “it blends with belief and speech” (ibid) by first thinking about speech. 37 which do not and not signify time and imply something said of something else, rather than as being between words which do not and do signify an action, as the Visitor had suggested in Plato’s Sophist. This allows Aristotle to explain why ὑγιαίνει (“is healthy”) is a verb or ῥῆμα even though ὑγιεία (“health”) is a noun and seems to refer to the same thing. The difference is that ὑγιεία does not also signify time. Furthermore, it is not the sort of word which always signifies something predicated of something else [ἀεί τῶν καθ’ἑτέρου λεγομένων]; ὑγιεία is significant on its own, apart from a predication. A word such as ὑγιαίνει is not significant on its own; it suggests something else involved, a he or she or it, or we or you or I, which must be supplied in translation. They are incomplete on their own, and must be partnered with something else. 12 Because of this, Aristotle’s distinction between ὄνομα and ῥῆμα is not exactly the same as the elementary-school distinction between nouns and verbs. It has something in common with the distinction between grammatical subjects and predicates. Grammatical subjects and predicates, though linguistic items themselves, are not different kinds of words, even though usually one finds nouns in the subject of a sentence, and verbs in the predicate. Grammatical subjects and predicates are not words but different integral parts in a complete sentence; sentences are composed of subjects and predicates. Grammatical subjects and predicates are distinguished by their roles. Roughly, subjects pick out or point to what one is making an assertion or denial about, while grammatical predicates pick out or point to what one is asserting or denying of that. For Aristotle, of course, this “picking out” or “pointing to” will not be merely mediated signification, as I said above. For Aristotle words are not straightforwardly symbols of things in the world. Rather, subjects signify παθήματα in the soul, which are likenesses of things in the world. Grammatical predicates also involve ῥήματα, which imply a relation holding between the παθήματα in the subject and in the predicate. Both words and lengthier expressions can be grammatical subjects or predicates. For example, when one says the sentence “Jane runs,” the name “Jane” is the grammatical subject. But the subject of a sentence can also be several words together,

12 In this respect verbs, for Aristotle, are incomplete in a similar sense as Frege’s functions. They need or imply something else. 38 even including a finite verb. For example, in the sentence “The boy who took Jane’s hat is running away,” the grammatical subject is the whole expression “the boy who took Jane’s hat.” Now, this is not to say that the whole expression is an ὄνομα, according to Aristotle, or a similarly complex expression constituting the grammatical predicate of a sentence is a ῥῆμα. Aristotle’s discussion falls short of fully distinguishing nouns and verbs from grammatical subjects and predicates. Nevertheless, that he is thinking of nouns and verbs as integral parts of sentences suggests this further distinction. Of course, the term “subject” is ambiguous in English. One could also say that Jane herself was the subject of the first sentence, while the boy who took Jane’s hat is himself the subject of the second. But this is to use the term “subject” in a different way, to name not a part of a sentence but instead the thing in the world which is picked out by a part of the sentence, and this insofar as or in the respect according to which it is picked out. That is, Jane gets to be the subject in this sense because Jane is picked out by the grammatical subject, the name “Jane.” Similarly, running is the predicate of the sentence “Jane runs,” since running is the action picked out by the grammatical predicate. I will call the item in the world which is picked out by the grammatical subject the sentence- subject, and the item in the world picked out by the predicate the sentence-predicate. The term ὑποκείμενον, usually translated as “subject”, arises in DI for the first time in the context of Aristotle’s discussion of sentences and predication in chapter three, when Aristotle says that a verb “is always a sign of what holds, that is, holds of a subject” (καὶ ἀεὶ τῶν ὑπαρχόντων σημεῖόν ἐστιν, οἷον τῶν καθ’ὑποκειμένου 13). (16b10-11) Given that the English term “subject” can mean what I am calling the grammatical subject or what I am calling the sentence-subject, it is fair to ask which one best fits what Aristotle has in mind here. By ὑποκείμενον Aristotle certainly does not mean here the grammatical subject, for a grammatical subject is a linguistic item—a word or set of words—and in 16b11 the ὑποκείμενον is introduced as what some word or words are a sign of. This usage, then, agrees with Plato’s in Protagoras 349b. Aristotle neither offers a definition of σύμβολον nor of πρᾶγμα, but from 16a5-8 it appears that words are σύμβολα, since they signify affections in the soul, and through these, πράγματα, or things

13 Alternatively, 16b10 in nΒΔΣΛαd: καὶ ἀεὶ καθ’ἑτέρου λεγομένων σημεῖόν ἐστιν, and 16b11 in nΒΤiαd reads: τῶν καθ’ὑποκειμένου ἢ ἐν ὑποκειμένω. 39 in the world. So σύμβολα appear to be things which signify, such as words and expressions, while πράγματα are things in the world. In Aristotle’s distinction from the beginning of DI 1, then, a grammatical subject would be a σύμβολον (symbol) or σημεῖον rather than a πρᾶγμα, while in 16b10-11 a ὑποκείμενον is a πρᾶγμα. This is evidenced in part by the way Aristotle treats universals and particulars in DI 7. Actual things, πράγματα, are universals and particulars (17a40), for “it must sometimes be of a universal that one states something holds or does not, sometimes of a particular.” (17b2-4) What something holds of is not a word, but a thing in the world, what is siginified by an ὄνομα. This comes across a little earlier, when Aristotle distinguishes statement-making sentences from other sentences. Statement-making sentences are either affirmations, in which one thing is said to hold of something else, or negations, in which one thing is said not to hold of something else. (DI 5,6) Further, he says that “Every statement-making sentence must contain a verb [ῥῆμα]” (17a10), for, as the definition of a verb given above says, a verb signifies something’s holding of something else. Thus when Aristotle says at 16b10 that verbs signify something which holds of a subject, it is clear that that subject is itself a πρᾶγμα, and not a linguistic or referring item. Similarly, then, the universals and particulars at 17a40 are πράγματα. It is clear, then, that Aristotle is thinking of ὑποκείμενα as being πράγματα, or things in the world. One might wonder: are not words, sentences, and expressions also things in the world? Do they not count as πράγματα? This suggests a more careful way of characterizing Aristotle’s distinction between σύμβολα and πράγματα. In the way Aristotle is thinking of πράγματα, it is not that linguistic items such as words or sentences never count as πράγματα, it is that when they do count as πράγματα they are being treated as non-referring. The difference between “significant spoken sounds” or “symbols” [σύμβολα] and πράγματα is not the distinction between linguistic and non-linguistic items, but between items which are taken as referring, insofar as they refer, and items which are not treated as referring (or whose ability to signify or refer to something else is not relevant to the respect under which they are being considered). Nouns and verbs, for example, as significant spoken sounds, refer or signify; they are signs of something else. Πράγματα, on the other hand, at least insofar as they count as πράγματα, are not signs or 40 signifiers. Something is treated as a πράγμα when it is not treated as referring, especially in the context in which it is something which is itself signified or referred to. That Aristotle chose to use the term πράγμα rather than something even more neutral such as τὰ ὄντα, “things there are,” as he does at Categories 1a20, helps to support this. Words and expressions are, for Aristotle, signs or symbols14, and as signs they are among the things that are, just as are those things which they signify. In distinguishing between σύμβολα and πράγματα Aristotle is distinguishing between signifiers and things signified, rather than between linguistic entities and non-linguistic entities. In its first appearance in DI, then, the term ὑποκείμενον seems to coincide with my term “sentence-subject,” and has something in common with Plato’s use of the related verb in Protagoras. Since it is not itself a linguistic item, it might seem misleading that I am calling this usage of the term ὑποκείμενον the sentence-subject. I do not mean “sentence-” here to indicate that the ὑποκείμενον is a word or expression, or anything which is a part of a sentence. I mean only to indicate that the ὑποκείμενον is what the grammatical subject in the sentence picks out. It is the subject of the sentence. But further, it is what the grammatical subject picks out on a de dicto reading; it is the thing in the world, inasmuch as it fits the description of it provided in the grammatical subject. If the sentence is “Jane runs”, Jane herself is the subject. If the sentence is “This year’s women’s triathlon champion is running in the charity race,” the subject is this year’s women’s triathlon champion, even if this year’s champion happens to be Jane. After all, it might not be, or might not have been. If the sentence-subject were just whatever item in the world the grammatical subject picks out, de re, the subject of both of these sentences would be the same. But it is important to be able to distinguish the sentence-subjects of these two sentences, for reasons which will become clear in what follows If the second sentence is true and Jane is this year’s women’s triathlon champion, the sentence is limning the co-occurrence of, or the relation between, her being the

14 Walz, in his 2006 “The Opening of On Interpretation: Toward a More Literal Reading,” has argued that more should be made of the difference between σύμβολον and σημεῖον in Aristotle. If he is correct about this, and about the implications for how Aristotle thinks about linguistic entities, I think this helps my case. 41 champion and her running in the charity race, rather than between her being Jane, herself, and her running in the race. In other words, there is a sense in which different items in the world are being put together, or signified as put together, in the different sentences. Being able to distinguish subjects in this way will also be important for understanding the difference between predications κυρίως and predications κατὰ συμβεβηκός in An. Po. A.22. For Aristotle presents that distinction as relying on whether or not there is a genuine ὑποκείμενον in the predication or not, and his examples of genuine and non- genuine ὑποκείμενα will only make sense if we can distinguish between sentence- subjects de re and de dicto. Relatedly, sometimes Aristotle’s use of ὑποκείμενον seems captured rather by what I will call the ontological subject rather than the sentence-subject, and one must be able to distinguish between the subjects of the two sentences to make the distinction between ontological subjects. In characterizing sentence-subjects as the items in the world which grammatical subjects pick out de dicto, I am not here attempting to ascribe to Aristotle a theory of reference. Perhaps the “prima facie evidence” Phil Corkum puts forward is persuasive (“Aristotle on Predication,” pg. 3), and Aristotle does claim, though not in these modern terms, that “the semantic role of a term is reference, the mapping of linguistic items onto extra-linguistic entities.” (ibid) Even if Aristotle does hold this, his view about reference is irrelevant to my current interests, and I think the text shows that Aristotle is also interested in the metaphysical question I am interested in here. Perhaps he has a theory of reference; what is important for the topic of my dissertation here is not his view (if he has one) about how words refer, but rather his view about the metaphysics of their referents. I think that, even if he has additional interest in semantic theory, in his discussion of affirmations and negations in DI Aristotle is thinking about the way in which claims mark out relations between different entities. This is a part of his project of investigating which relation between which entities is limned by statements when they are true, part of his investigation into what Code called “the connection between the correct use of language and an objective, language-independent reality.” (“Aristotle: Essence and Accident,” pg. 411) He is preparing the way for his logic, which is aimed at characterizing a finished science, one in which the propositions we know limn the structure of things in the actual world. Aristotle is building toward the likeness claim he 42 adumbrates in the opening of DI, that when we know something, in some way the “affections in the soul” are likenesses of the things in the world which are the objects of our knowledge.

§.4 Ontological subjects in Categories

Aristotle’s Categories inquires into beings15 by considering the many ways in which things are said to be or are called beings. It does so by examining the way simple statements in ordinary language carve up the world, and attending to the ontological relations in virtue of which we make true predications.16 The idea is that linguistic predications, when they are true, point to, and are true because of, real metaphysical relations between the items referred to in the predications. In order for any linguistic predication to be true, the item(s) referred to in the predicate must stand in some appropriate metaphysical relation to the item(s) referred to in the grammatical subject. In Categories Aristotle mentions both “things that are said” (1a6) and “things there are” (1a20). However, even when ostensibly discussing the “things there are,” he discusses these in terms of predication. In what follows, for the sake of clarity I will adopt the terminology of linguistic versus ontological predication; for something to be ontologically predicable of something else is just for it to stand to its ontological subject, or ὑποκείμενον, in a metaphysical relation which is such as to make it true to linguistically predicate it of its subject. In brief, a ὑποκείμενον is what something is ontologically predicated of. Thus in Categories as in DI, a ὑποκείμενον is not a kind of word or symbol, but rather a πρᾶγμα, one which receives the title of ὑποκείμενον because of the position it stands in as what is limned by or picked out by the grammatical subject of a predication. Further, a ὑποκείμενον is always the ὑποκείμενον of some ontological predicate. Being a ὑποκείμενον means standing in a certain kind of relation to something

15 That in Categories Aristotle inquires into beings, and not only into matters of speech or language, is no longer a matter of much controversy in commentary on the work. Cf. J. Ackrill (1963), R. Dancy (1975), A. Code (1986), F. Lewis (1991), M. Loux (1991), M. Wedin (2000), M. Kohl (2008). 16 Here I am following Alan Code’s proposal from his (1986) “Aristotle: Essence and Accident” in Richard E. Grandy & Richard Warner (eds.) Philosophical Grounds of Rationality, pg. 411-439. 43 else. To become clearer about it and what it means to stand as a ὑποκείμενον to something, we must look into the ontological predication relations Aristotle describes. Aristotle presents two ways in which something is ontologically predicable of something else: said-of a subject and in-a-subject. In this way he outlines two sorts of relationships in which ὑποκείμενα might stand to what they are ὑποκείμενα of. In Categories 2 Aristotle introduces said-of a subject with examples which suggest that genera are said-of whatever is included under them. To use Aristotle’s examples, man is said-of an individual man, and knowledge is said-of a particular kind of knowledge, such as knowledge of grammar. In Categories 3 Aristotle adds to this, saying that said-of predication is transitive.( 1b10-13) In other words, if X is said-of Y as its ὑποκείμενον, and Y is said-of Z as its ὑποκείμενον, X is also said-of Z. In Categories 5 Aristotle further clarifies that things which are said-of a subject are ontologically predicable of their subject, the ὑποκείμενον, such that their definitions as well as their names may be truly linguistically predicated of that subject. (2a20) For example, man is ontologically predicable of Socrates in such a way that it is true both to linguistically predicate the name “man” and the definition of man of Socrates. Aristotle calls the second ontological predication relation “in a subject,” and introduces it with a definition in Categories 2 as follows: “By ‘in a subject’ I mean what is in something, not as a part, and cannot exist separately from what it is in.” (1b25) What this definition amounts to is a subject of much debate, but it implies at least that what is in-a-subject ontologically depends on its subject in some way.17 Aristotle illustrates the difference between something being said of a subject and in a subject by using examples; paleness and musicality are predicable of Socrates as in a subject, while color is predicable of paleness and man of Socrates as said-of a subject. These examples further show that, for Aristotle, something which is in a subject cannot be said-of the very same subject it is in, for when something is in a subject, “it is impossible for the

17There is a voluminous literature on this definition, and on the debate about the status of non-substantial individuals which has tended to be rooted in it. The literature begins with Ackrill’s discussion in his 1963 translation and commentary of Categories, and Owen’s challenge to Ackrill’s view in his 1965 “Inherence.” James Duerlinger (1970), Michael Frede (1987), Daniel Devereux (1992), and Michael Wedin (2000) have all made contributions to the debate. 44 definition to be predicated” of the same subject.( 2a31) Genera are never in what falls under them as a subject. In other words, for Aristotle what something has can never be what it is. So, for example, it is true to say that Socrates is musical or Socrates is pale because musicality and paleness are “in” Socrates; Socrates has musicality and paleness. But it is not the case that Socrates is musicality or paleness; it is false to say “Socrates is musicality,” or “Socrates is paleness.” Aristotle illustrates the difference between the two metaphysical relations of being and having, or being said-of-a-subject or in-a-subject, not only by providing examples, but also by pointing to the different ways they are (usually) exhibited in speech. He thinks that there are linguistic markers for when one relation or the other is being picked out, and he uses these as support for the metaphysical schema he is proposing. Though he does not argue from the markers for this schema, he does present the fact that his schema maps onto certain patterns of speech as an advantage for his view. For example, Aristotle’s discussion of paronyms suggests that when things which are ontologically predicable as in-a-subject are linguistically predicated of their subjects, the name in the grammatical predicate of the sentence has its ending changed so that it is paronymous. Now, as Aristotle admits, paronymous naming may not always occur. The linguistic markers may not always be present. Inasmuch as Aristotle is not presenting a semantics, this is untroubling for him. Nevertheless, the use of paronymous names marks or is distinctive of cross-categorial predication. The ontological relation of said-of, on the other hand, is marked by the fact that when something is said of something else, its definition can also be truly linguistically predicated. This distinction between the two kinds of ontological predication relations results in a four-fold metaphysical schema, underlying the categorial schema of the whole work. Items which are said-of ὑποκείμενα stand to their ὑποκείμενα as genera to what falls under them. This sets up the vertical structure of each of the categories; each category is comprised of genera and sub-genera, each said-of what falls under them, down to the lowest level of particulars, which are not said-of anything else. Items which are in-a- subject thus have as ὑποκείμενα things which do not fall under them in the same category. Further, Aristotle presents primary substances, particulars in the category of substance, as being those items which are not in-a-subject or said-of-a-subject (2a11-13), 45 while he says that no substances are ever in-a-subject (3a8-10). Thus the relation of said- of-a-subject distinguishes genera from species and particulars, while the relation of in-a- subject distinguishes the members of the non-substance categories from the members of the substance categories. This produces the four-fold ontology: universal and particular substances, and universal and particular non-substances. It is clear that in Categories Aristotle uses the notion of ὑποκείμενον to characterize the role of substance in his metaphysics. Aristotle introduces substance in Categories 5 as follows: “A substance—that which is called a substance strictly, primarily, and most of all—is that which is neither said of a subject [ὑποκείμενον] nor in a subject.” (2a11-13) In other words, primary substances are never themselves predicates. Aristotle later clarifies that primary substances are not only never predicates, but all ontological predicates ultimately take a primary substance as a ὑποκείμενον (2a34-35, 2b3-5). It is because of this that they are called substances most strictly (2b15-17, 2b38- 3a1), since on these grounds “if the primary substances did not exist it would be impossible for any of the other things to exist” (2b5). 2b5 shows that for Aristotle both the relations of being said-of-a-subject and in-a-subject are dependence relations, where what is predicated depends on its ὑποκείμενον.18 Elsewhere in his corpus Aristotle states in different ways that substance is what grounds his ontology; here he expresses it by asserting that primary substances are primary in the sense that they are what every ontological predicate finally depends on as ὑποκείμενα. It is because of this that commentators often take subjecthood to be the criterion of substancehood in Categories.19

18 It is a much-discussed question whether being in-a-subject and being said-of-a-subject indicate two different kinds of dependence. 19 See, for a recent summary of views on this topic, Markus Khol (2008), and the bibliography in his footnote 14. The textual locus for discussion of whether Aristotle eventually rejects the subjecthood criterion is Metaphysics Z, especially Z.3. Mary Louise Gill, in her 1992 Aristotle on Substance: A Paradox of Unity argues that Aristotle does not reject subjecthood as a criterion for substancehood in Metaphysics Z, but that he corrects a misunderstanding of his notion of subjecthood. Michael Wedin, in his 2000 Aristotle’s Theory of Substance, argues that Aristotle does not reject subjecthood as a criterion, but he does modify his view of subjecthood. Alan Code in his 1986 “Aristotle: Essence and Accident” argues that Aristotle rejects subjecthood as a criterion. Frank Lewis in his 1991 Substance and Predication in Aristotle 46 Nevertheless, we should be careful to distinguish what it is to be a substance from what it is to be a ὑποκείμενον in Aristotle’s Categories. For, firstly, it is an open question whether all substances are subjects. Aristotle’s characterization of primary substances at 2a11-13 is a negative characterization; primary substances are neither said of a subject nor in a subject. He does not here assert that all of them, or indeed any of them, are subjects. 2b5 also does not go so far as explicitly to claim that all primary substances are subjects. By 2b5 all Aristotle has said is that every predicate ultimately takes a primary substance as a subject. This leaves it open that there are some things which fulfill the original terms of being a primary substance by not being predicates of any kind, while also not being subjects. At 2b15-17 Aristotle says that “it is because the primary substances are subjects for all the other things … that they are called substances most of all.” Now, this is ambiguous; it may still leave it open that there are primary substances which are neither predicates nor ontological subjects. If we take 2b5 as claiming only that the ultimate subjects are all primary substances, that primary substances are the only sorts of things to stand as ultimate subjects may be a sufficient indicator of their primacy as substances. This is all that 2b15-17 need assert. This leaves open the question of what makes them to be such that only they can serve as ultimate subjects. This puts into sharper relief the question of what makes them to be such that only they can serve as ultimate subjects, a question to which as yet there are no clues in Categories. This line of thought is strengthened if we take it that in Categories Aristotle is limiting the scope of his claims to those items which stand in ontological predication-relations to each other, that is, those items whose relations are limned by the basic sorts of patterns in speech to which he is attending. And there is reason to think this. Aristotle’s project in Categories, as stated above, is to lay out how things in the world must be related to each other if we can say true things about them, true statements whose truth lies in how they limn, or express, the metaphysical relations they are about. Aristotle’s examples show that in Categories he is interested in the metaphysical relations which hold between things of ordinary

argues that Aristotle maintains the subjecthood criterion. Michael Loux in his 1991 Primary Ousia argues that in Metaphysics Z Aristotle abandons the subjecthood criterion. 47 experience, and which might be expressed in ordinary ways of speaking. But there might be other things, beyond ordinary experience, whose structure our natural ways of thinking and talking about do not limn so well. It is possible to find elsewhere in Aristotle’s corpus at least one substance which does not also seem to be a subject: the perfectly simple divine being20 found in Physics VIII and Metaphysics XII (e.g. Phys. VIII.10.267b26-28, and M XII.7.1073a3-7). Having no properties, it is not truly a subject of ontological predication, even though it is the sort of thing upon which other things depend.21 It is just not the sort of thing which

20 According to Aristotle, this divine being is unchanging, and thus is not receptive of contraries. Since being receptive of contraries is proposed in Categories 5 as what is distinctive of substances, this might seem to suggest that if the divine being is a substance, it is a substance on a conception of substancehood inconsistent with the one proposed or developed in Categories. But I think not. Being subject of contraries was proposed as distinctive of substances in Categories in the sense that only substances are subject to contraries in this way, not in the sense that all substances are so subject. Similarly, it might be distinctive of dogs that they bark, but this can be true even if some dogs do not bark, just as long as only dogs bark. The divine being could still be a substance on the Categories view, even if it does not itself bear this mark. In other words, being subject to contraries was not a necessary condition for substancehood in Categories. This is similar to the way Aristotle presents there being certain linguistic markers for the two different ontological predication relations. It is not that these are always present, but rather that when they are present they clearly indicate or mark just one ontological predication relation. 21 Here I am, admittedly, taking a certain line of interpretation on some very perplexing passages. It seems important for Aristotle’s argument in both Physics VIII and Metaphysics XII that this divine being be perfectly simple. Yet he also seems to ascribe certain properties to it, e.g., being a thinking being (1072b18-21), being alive (1072b26-27), being unmoved (1072a25), being eternal (1072b28-30). I am inclined to take seriously the need for simplicity in order for Aristotle’s god to explain what it is called upon to explain. One reason for this is that it is difficult to see how something whose very essence is actuality (1071b20) could be a ὑποκείμενον in the sense I have laid out in this dissertation, which seems to imply some potentiality. Consequently, I think that the positive ascription of properties to the divine being, as though they are predicates, is a result of Aristotle’s recognition that knowers such as us cannot grasp a thoroughly simple being, since we understand or cognize through predication, and an attempt to nevertheless try to point us toward this. On the other hand, recognizing that most of the “properties” Aristotle seems willing to ascribe to his god are rather peculiar, it could be that they are not properties the way my pallor is property of me; they are not predicates of the divine prime mover as a subject, and so do not compromise the divine being’s simplicity. 48 our ways of speaking and thinking grasp as well as things closer to our experience. Of course, this raises interesting questions about the dating of these sections from Metaphysics and Physics, and whether these texts are evidence of Aristotle having changed his mind about what counts as a substance. That Aristotle presents something as being a substance but not a subject in Metaphysics will come as no surprise to those who argue that Aristotle revised his view of substancehood by the time those treatises were written. But I take it that such a substance being presented in Physics is more surprising. Most of those scholars who argue that Aristotle changed his mind about ὑποκείμενον as a criterion for substance argue that he did so as a result of his discoveries in Physics, and not before. In Physics Aristotle posited matter as ὑποκείμενον. In so doing, the story goes, Aristotle posits complexity in the substances, or subjects, which he had originally treated as simple, in such a way that now it looks as though one of the parts of substances is itself a substance on the grounds that it is ὑποκείμενον. This prompts him to take up again the question of what it is to be substance, and which among the items in his ontology wears that title. These are not questions Aristotle addresses within Physics, but on this story they are raised by what he proposes there. If in Physics Aristotle posits as substances items which are not also subjects, before having engaged in the discussions on the basis of which he supposedly revises his view of substance, this challenges the narrative that Aristotle at that time accepted ὑποκείμενον as a necessary and sufficient criterion for subjecthood. This raises the somewhat vexed question of the history of the text which we now have as Physics, specifically regarding the dating of Physics VIII. If Physics VIII is a work of sufficiently late composition that in writing it Aristotle might have been influenced by the same train of thought that led him to renounce ὑποκείμενον as a criterion in Metaphysics VII, if indeed he does, then it presents no very helpful evidence for the case I am making now. Aristotle could have changed his mind. Jaeger, in arguing that the passages regarding the prime mover in Metaphysics XII.8 are later additions to that text, seems to argue the same holds of the passages in Physics VIII.6. (Aristotle, 351-354) Since the whole of Metaphysics is generally thought to be later in composition than Physics, then Physics mention of the prime mover would not, if Jaeger is correct, help to show that in that and earlier works 49 Aristotle does not propose ὑποκείμενον as being coextensive with substance. Sir David Ross, however, surveys the same evidence for the dating of those passages, and also evidence concerning the dating of Physics VIII as a whole, and concludes that not all of the passages mentioning a prime mover in Physics VIII are later additions. He dates the form of the book without later additions as being no “earlier than Aristotle’s final residence in Athens, from 334 to 323 B.C.” (Ross, Physics, pg. 10). Yet he goes on to say that references in De Caelo to Physics VIII “in view of the obviously early character of that work, lend colour to the view that Phys. VIII may itself be early [Ross’s emphasis].” (Ross, Physics, pg. 11). On this account, Aristotle’s offering up of the prime mover as a substance, but one which is not a ὑποκείμενον, in Physics VIII, helps to confirm my reading that in earlier works Aristotle did not assume that every substance was a ὑποκείμενον. There remains room for doubt about the precise dating of Physics VIII, and what stage of Aristotle’s thought it might represent. It is perhaps not very certain support for the idea that in the view of substance proposed in Categories something can be a substance without being a ὑποκείμενον. Yet Aristotle’s divinity being a substance does not clearly support the opposite view, either. Without assuming the developmental story, the prime mover as a non-ὑποκείμενον substance in Physics VIII does not indicate that Aristotle changed his mind about whether every substance was a ὑποκείμενον. In the light of this, I think that the only text one can turn to for solving this problem is Categories itself. Given this, that Aristotle does not claim in Categories that all substances are ὑποκείμενα becomes more significant. In this case, it seems best not to assume that Aristotle means anything stronger than what he actually says. But this will find further support, I hope, in what follows. Further, even if 2b15-17 does not leave it open that some primary substances are not subjects, so that all substances are subjects, there remains a problem for the view that being a subject is the criterion for being a substance. For primary substance’s substancehood is explained both by their not being predicates and by their being ultimate subjects: it is not clear how secondary substances gain the status of being substances on this ground. Secondary substances are still themselves predicates, and thus still dependent beings; that all predicates ultimately take as subjects some primary substances does not 50 help their own claim to substancehood, especially if what 2b5 and 2b15-17 are really aimed at presenting is the way in which substances ground everything else. Loux (1992), among others, argues that secondary substances get to be called substances because they reveal what the primary substances are, rather than because they fit the same criterion as the primary substances by being basic subjects. (pg. 26,28) Loux’s view seems to undermine the claim that being a ὑποκείμενον is the criterion for substancehood, for it does not explain why secondary substances count as substances, even if it does explain why Aristotle still thought it useful to call them substances. Kohl (2008), on the other hand, argues that if we understand the kind of dependence Aristotle is saying everything else has on primary substances, we will discover an analogous dependence also holds between every other predicate and the infimae species, and in a more qualified way, the other secondary substances. “The analogy,” Kohl says, “is that non-reciprocal subjecthood is assigned to substance species for substance genera, and thus the former, within the class of secondary substances, can be regarded as being more of a substance.” (pg. 160-161) According to Kohl, this is not just a matter of how many predicates primary substances vs. infimae species vs. higher-level secondary substances underlie. Kohl’s view, I think, is undermined by the fact that the analogy which he proposes seems very weak, weak enough, in fact, that once it is seen that items which Aristotle does not count as substances do still count as ὑποκείμενα for him, it appears as though they ought also to count as substances by the same lights. Thus for this reading, too, there is a problem taking ὑποκείμενον as a criterion for substancehood. It is also not the case that only substances are ὑποκείμενα in Categories. This is evident from the examples of ὑποκείμενα which Aristotle employs. For instance, in Categories 2 Aristotle says that knowledge is said-of knowledge-of-grammar (γραμματική) as a subject (1b1-2). Devereux (1992), points out another example: in 4a10-17 “Aristotle says that a particular action which is one in number cannot ‘receive’ both bravery and its opposite; one might argue that since bravery can be predicated of particular actions, it is implied that non-substance particulars may be subjects of the ‘in a subject’ relation(pg. 128, footnote 24) For further examples, consider 5a9-37, where numbers and language as well as lines, surfaces, bodies, time, and place are subjects which “quantity” [ποσά] is said of. Of these, only bodies could easily be counted as 51 substances. In the next lines (5a38-39) Aristotle also treats actions and colors as subjects, pointing out that quantity can only be predicated of them accidentally [συμβεβηκός] rather than strictly [κυρίως]. The brief mention of accidental vs. strict predication in 5a38-39 points to the An. Po. I.22 discussion of ὑποκείμενον within the discussion of accidental [κατὰ συμβεβηκός] and strict [κυρίως or ἁπλῶς] predication. For just as Aristotle has not explained what it is to be a ὑποκείμενον in Categories, he also has not explained the difference between accidental [κατὰ συμβεβηκός] and strict [κυρίως] or simpliciter [ἁπλῶς] predications. In 5a38-39, however, he qualifies the status which actions and colors can have as ὑποκείμενα for quantity by pointing out in what sort of predications they can serve as subjects. He seems to be gesturing at details of what it is to be a ὑποκείμενον and what sort of framework this notion is a part of which are not evident in Categories. It is in An. Po. that Aristotle explains the distinction between the two types of predication. 5a38-39 suggests that the notion of ὑποκείμενον which arises from that An. Po. discussion of accidental and strict predication is the one to which Aristotle is helping himself throughout Categories. It is to this passage that I suggest we turn next. I have argued that we do not get a complete or thorough picture of what it is to be a ὑποκείμενον in Categories, and that Aristotle’s way of characterizing substance in terms of the notion of ὑποκείμενον at best leaves us in doubt about what it could mean for being a ὑποκείμενον to be the criterion for being a substance in Categories. If it is a criterion, it is not one by being a sufficient condition, for there are ὑποκείμενα which are not substances in Categories. Further, it does not appear to be a criterion by being a necessary condition for being a substance, for it is an at least an open question whether all substances are subjects. Aristotle never claims this in Categories, and there is some mixed evidence that he does not actually hold this view. Yet it remains that Aristotle takes himself to be communicating something important about substances by pointing out that primary substances are ultimate subjects, and not themselves predicates. To better understand what Aristotle is trying to reveal about substance by relating it to subjecthood,

52 I think it is important to get clearer about what it is to be a ὑποκείμενον.22 5a38-39 points us toward where turn next: the Posterior Analytics discussion of strict and accidental predication.

§.5 Ontological Subjects in Posterior Analytics

In Posterior Analytics A.4 Aristotle is investigating what sort of propositions we can have knowledge [ἐπιστήμη] of by thinking about what sort of relations must hold between ontological subjects and predicates in order for propositions about them to be knowable. He argues that the relation must be necessary and universal, and he spells this out by distinguishing between essential and accidental predication. In this text, then, Aristotle is once again interested in what sort of relations things in the world stand in when the linguistic predications about them are true, but he is adding one more feature to the consideration. He is interested not only in what is the case when predications are true, but also in how things in the world must be related if they, and propositions about them, are to be knowable. Compare this to Parmenides’ claim in Plato’s Parmenides that

If someone, having an eye on all the difficulties we have just brought up and others of the same sort, won’t allow that there are forms for things and won’t mark off a form for each one, he won’t have anywhere to turn his thought, since he doesn’t allow that for each thing there is a character that is always the same. In this way he will destroy the power of dialectic entirely [τὴν τοῦ διαλέγεσθαι δύναμιν]. (135bc)

22 Michael Wedin in his (2000) Aristotle’s Theory of Substance, argues that the Categories theory of substances proposes the explanandum for the Metaphysics Z theory of substance, rather than serving as a rival to it (e.g., pg. 4). On his view, Categories raises questions which become the focus of the discussion of Metaphysics Z. Whether or not one agrees with the details of his view, I think that one can agree with the spirit of Wedin’s proposal: it seems that what comes out of Categories foremost are questions, and things to be explained, rather than explanations. 53 “Dialectic” here may mean either the science of dialectic or dialogue and speech more generally. In Republic 511b, Plato used the articular infinitive τοῦ διαλέγεσθαι to name the science by which the Forms are grasped, but especially the Good itself, at the top of the line in his Line analogy. The expression there names the highest kind of knowledge. Thus it appears that in Parmenides Plato has the character Parmenides suggest that there is no knowledge without Forms. But διαλέγεσθαι is also a word used in ordinary Greek for conversation or discourse, as Cooper reminds readers in a footnote to his edition of Parmenides (pg. 369, footnote 12). Because of this, Plato might be making the even stronger claim that not only is knowledge not possible without Forms, but neither is meaningful speech or ordinary conversation. I think that in An. Po., Cat., and DI Aristotle is following in Plato’s footsteps. A part of this is Aristotle’s exposition in 73a26-73b16 of the different ways one thing can belong to another καθ’αὑτό, i.e., the different kinds of essential predication. In 73a34-73b5 Aristotle lays out two kinds of essential predication, and distinguishes these from accidental predication. According to Aristotle, something is predicated essentially in one of two ways. In the first way, something is predicated essentially when it appears in the definition of the subject of which it is predicated: “something belongs to another in itself [καθ’αὑτό] … if it belongs to it in what it is [ἐν τῳ τί ἐστιν]—e.g. line to triangle and point to line (for their substance depends upon these and they belong in the account which says what they are).” (73a34-36) In the second way, something is predicated essentially when the subject of which it is predicated is in the definition of the predicate, as line belongs in the definition of straight and curved. Jonathan Barnes, in his (1993) commentary on Aristotle’s Posterior Analytics summarizes these as follows:

(1) A holds of B in itself =df A holds of B and A inheres in the definition of B. (2) A holds of B in itself =df A holds of B and B inheres in the definition of A. (Aristotle: Posterior Analytics, pg. 112)

In this passage Aristotle defines accidental predication (predication κατὰ συμβεβηκός) entirely negatively: “what belongs in neither way I call accidental, e.g., musical or white to animal.” (73b5) Accidental predication occurs when something is 54 ontologically predicated of something else, but neither what is predicated nor what it is predicated of belongs in the definition of the other.23 This way of distinguishing accidental from essential predication is only one of two different ways of using the term “accidental predication” Aristotle offers in his Posterior Analytics, however. In 83a1-18 in Posterior Analytics I.22, Aristotle distinguishes between predications which are συμβεβηκός and predications which are haplōs. It is important to see that this is a different distinction from the one between predications συμβεβηκός and καθ’αὑτό which he had proposed in I.4. In particular the senses of “accidental” are different. In Ι.4 accidental predication is any ontological predication within which the predicate and subject are not definitionally related (if they are, it is essential predication). In I.22 the opposition is between what is a strict ontological predication vs. what is a predication only qualifiedly, where what makes the difference is whether or not a ὑποκείμενον is picked out in the grammatical subject of the sentence; in strict predications one is, while in accidental predications one is not. Some predications which counted as accidental predications in the sense from I.4 will count as predications ἁπλῶς in this later passage, and thus not as accidental predications in the I.22 sense, since according to the latter distinction there can be genuine predications which are not essential predications. This can be clarified by thinking about how Aristotle distinguishes accidental from essential beings, in the light of his first distinction between essential and accidental predication. Given the two kinds of predication, he suggests that we are presented with two kinds of beings. DI had claimed that linguistic predication put together symbols (words) of likenesses (thoughts) of things in the world (πρᾶγμα) together in thought, in such a way that true sentences limn or mirror what actually is together in the world. Essential and accidental predication are two ways of putting things together in speech and

23 Metaphysics V.30 offers a similarly negative, but nevertheless more informative, description of the same: And a musical man might be white; but since this does not happen of necessity nor usually, we call it an accident. Therefore since there are attributes and they attach to a subject, and some of them attach in a particular place and at a particular time, whatever attaches to a subject, but not because it is this subject, at this time or in this place, will be an accident. (1025a17-25) 55 thought which, when true, thus limn two different kinds of composite facts or πράγματα. These two different kinds of composite Aristotle names from the two different kinds of predication which, when true, point to them.

Again, what is not said of some underlying subject—as what is walking is something different walking (and white) while a substance, and whatever signifies some this,’ is just what it is without being something else. Thus things which are not said of an underlying subject I call things in themselves, and those which are said of an underlying subject I call accidentals. (73b6-9)

“Things in themselves” here translates καθ’αὑτό, which I have been rendering as “essential” in this context. Aristotle is distinguishing between essential and accidental things. Accidental beings are such as what is walking, or the walking thing. The walking thing is what it is, walking, because it is also something else. Let us say that in our example the walking thing is a person. Thus there is a walking thing on the sidewalk because there is person there; were there no person there would be no walking thing. In other words, the walking thing is ontologically predicated of the person. A somewhat awkward English idiom is apropos here; there being a walking thing is predicated on there being a person. Similarly for the pale thing. Or, again, the walking thing can be pale because the person who is walking is pale. Accidental beings such as the walking thing and the pale thing are what they are in virtue something else, in the sense that they are themselves ontologically predicated of something else. Essential beings, by contrast, are not themselves ontological predicates. Each of them “is what it is without being something else”; Callias is a man without being something else, for example, a pale thing, however much indoors he is, first. Nevertheless, Aristotle is not suggesting that only essential beings can be ὑποκείμενα, as can be seen in 83a9-17. It is not being what it is essentially that makes something suited to be a ὑποκείμενον at 83aff. In other words, it is not because the predicate belongs to it ἐν τῳ τί ἐστιν. Rather, what makes something suited to be a ὑποκείμενον for something is standing in the appropriate relation to its predicate by being 56 ὁπερ τί what it is. In 83aff Aristotle suggests that to be a genuine ὑποκείμενον for something else, in such a way that a predication ἁπλῶς is achieved, it must be the case that by being what it is, or ὁπερ τί, the candidate ὑποκείμενον is predicationally related to its predicate. When this holds, this means that there is a predication ἁπλῶς; the linguistic predication limns the ontological structure of the thing in the world. Aristotle proposes the following examples to explain this: “the white thing is walking,” “the large thing is a log,” and “the musical thing is white” on the one hand, and “the log is large” and “the man is walking” on the other. Aristotle says the following about these two kinds of prediction:

Well, if we must legislate, let speaking in the latter way be predicating and in the former way either no predicating at all, or else not predicating simpliciter [ἁπλῶς] but predicating accidentally [συμβεβηκός]. (83a15)

Thus he offers “the log is large,” and “the man is walking” as examples of predications ἁπλῶς, while the other examples are either not truly predications, or predications only accidentally. Consider this example: “the white thing is a log.” There is nothing wrong with the sentence in English or in Greek. The problem with the predication “the white thing is a log” is not linguistic, it is that the sentence’s structure does not mirror or limn the metaphysical relations, the ontological predication relations, which underlie its truth: “For when I say that the white thing is a log, then I say that that which is accidentally white is a log; and not that the white thing is the underlying subject for the log; for it is not the case that, being white or just what is some white [ὅπερ λευκόν τι], it came to be a log, so that it is not a log except accidentally.” (83a7-9) The difficulty, in other words, is that while in this sentence the white thing is the linguistic subject, the thing being talked about, it is not the sort of thing which can properly serve as an ontological subject or ὑποκείμενον for log as a predicate. The white thing is an accidental, a συμβεβηκός. It is

57 something which is predicated of, or on, something else, the log.24 The reason the white thing is not a ὑποκείμενον in this case is that it is not suited to be the grounds for the predicate log by being just what it is (ὅπερ x τι). In fact, it is not suited to be the grounds for being a log at all. In the case of the sentence “the white thing is walking” the situation is not so bad. The white thing is not suited as such to ground the predicate, walking. But insofar as the white thing is also a person, and the person can ground the property of walking, the white thing is accidentally the subject of the predicate, i.e., walking. For the white thing is accidentally the same as the person.25 In his commentary on Physics A William Charlton explains ὅπερ τι thus: “ὅπερ τι in Aristotle normally means, I think (for a fair selection of examples v. Bonitz 533b39- 534a23), ‘precisely what is something’ in the sense in which a certain bodily condition might be said to be precisely what is healthy.”26 A comparable English expression for the sense Charlton gives to ὅπερ τι would be “model”, as the Gilbert & Sullivan’s character used the term when he proclaimed “I am the very model of a modern major general.” In the play, Major General Stanley is claiming that he exhibits perfectly, lacking nothing, the characteristics of the (ideal) modern major general. I think Aristotle has in mind something different from this in the An. Po. passage, something which also fits with the other texts Bonitz points out. Something being ὅπερ τι what it is in An. Po. I.22 is not so much a matter of it perfectly presenting the characteristics of a type, but rather of its bearing the property predicated of it in virtue of being of that kind. In the example in 83a9-10, the white thing is not a log in virtue of being white, and because of this it is not the subject for the log. Nevertheless, there might be a white thing because there is a log.

24 The ontological predication of the musical thing of the white thing is a case of what some ancient commentators called an unnatural predication. The thing about unnatural predications, says Barnes, is that “the subject of an unnatural predication is ontologically parasitic on other entities.” (pg. 116) It is ontologically parasitic because it is itself ontologically predicable of something else, for it is an accidental rather than an essential being. It is what it is, e.g., walking, pale, musical, because it is already something else. While the pale thing may in fact be musical, musicality only belongs to the pale thing because it belongs to the person who also happens to be pale. Thus the genuine ontological subject is the person herself. 25 Cf. Metaphysics V.9 1017b27-30 26 Charlton, William. Aristotle’s Physics I, II. Oxford, Oxford University Press (1970), pg. 60 58 Similarly, “the white thing is walking” is not a predication ἁπλῶς because the person, not the white thing, is the real ὑποκείμενον. It is tempting to think that in these passages in An. Po. the expression is synonymous with καθ’αὑτό. If this were the case, then since substances are the essential beings only substances could truly be ὑποκείμενα. If so, there would be a conflict with Categories, where Aristotle clearly allows non-substances to be ὑποκείμενα. But we need not worry. The point Aristotle is making here is not that the white thing is not essentially what it is, but rather that what it is, as white, does not ground its being a log. Similarly with the other examples. The problem with the linguistic predication of musical of the pale thing is not that the pale thing is not an essential being, but rather that there being a musical thing is not ontologically predicated on there being a pale thing. Thus An. Po. provides us with a necessary condition for something being a ὑποκείμενον in a genuine or unqualified predication; its being just what it is, or ὅπερ x τι, must ground its having or being what is predicated of it. Another example of Aristotle’s helps to clarify this. In Physics A.7 Aristotle famously uses the example of a person becoming musical or cultured to explain his account of the principles of change. On my reading of An. Po., predicating cultured of a person counts as a genuine predication. According to Aristotle, humans are distinctively rational creatures, and in virtue of this they are teachable. Thus in virtue of being human, a person can become cultured. A person is a genuine ὑποκείμενον for being cultured (and thus, as we shall see, an important element in the change which is “becoming cultured”). The ontological predication of cultured of a person is genuine in that it reveals the structure of the properties that person has. Similarly, to use a present-day stereotype as a further example, if some academic philosopher were called “analytical” or “argumentative”, these would also signify cases of genuine ontological predication, picking out a genuine subject. Having a certain manner of conversation which might be called argumentative, or a certain bent in reflection which might be called analytical, are generally marks or features of a person much of whose life is spent in the academy, and whose work consists in philosophical analysis and argumentation. Support for this reading can be drawn from seeing how fruitful this criterion for subjecthood is. Making the requirement for being a ὑποκείμενον being ὅπερ x τι in this 59 sense, rather than καθ’αὑτό, reveals the complex metaphysical structure of a substance’s properties. The metaphysical structure of things in the world is not merely two-level, with the substance or essential being on the first level and all of its properties arrayed above it on the next. Because the properties themselves can have a structure, ὑποκείμενα can appear at multiple levels. For example, while being human is a genuine ὑποκείμενον for being cultured, being cultured might be a genuine ὑποκείμενον for certain habits of address or conversation. What matters is not that being cultured is itself an accidental being, predicated on an essential being, but that by being what it is, cultured, a cultured thing bears some other properties. Consider the example of the log which is white. Perhaps being a log (a piece of driftwood, one wonders, if its color is so pale), is a genuine ὑποκείμενον for being white. But being white might be a genuine ὑποκείμενον for having a certain albedo. In some scientific investigations, such as climate science and some branches of engineering, the object’s being white may be more relevant information than the object’s being a log. Because of this, Aristotle’s notion of ὑποκείμενον plays an important role in scientific explanation and the structuring of the various sciences. To take one example, consider the importance Aristotle gives to the question of the difference between physics, or natural science, and mathematics. On the surface, these are obviously distinct sciences, since the one concerns natural things and the other concerns numbers and geometrical things. Numbers, by Aristotle’s lights, are not natural, at least not in the sense in which the items under investigation in his Physics are natural; numbers do not have internal principles of change and rest, and do not in fact change at all. Yet Aristotle also holds that numbers and geometrical items are not substances. According to Aristotle, physics and geometry, for example, both seem to have physical bodies as their subjects, since geometry is just the study of extended things, and the things which are extended are, primarily, bodies. If this is so, then it is not immediately clear how the subject of natural science can be distinguished from the subject of mathematics. To distinguish the subjects of sciences which seem in this way to be the same, in his Metaphysics Aristotle introduces a qualifier: qua or ἧ. Natural science investigates bodies insofar as they are natural, that is, qua natural. What it means for a body to be natural Aristotle explains in book II of his Physics. But mathematics, Aristotle says, 60 studies bodies qua extended (in geometry) or numerous (in arithmetic). That is to say, natural science investigates what belongs to them on the grounds that they are extended or numerous. The ὑποκείμενον in natural science will thus be different from the ὑποκείμενον in geometry. For being inscribable in a circle, for example, and following certain laws of motion, belong to something or are predicated of it, in virtue of different more basic properties. Jonathan Lear in his “Aristotle’s Philosophy of Mathematics” characterizes the use of qua here as a “predicate filter” (pg. 168): “Thus to use the qua- operator is to place ourselves behind a veil of ignorance: we allow ourselves to know only that b is F and then determine on the basis of that knowledge alone what other properties must hold of it.” (ibid) Aristotle treats knowledge-of-grammar and numbers in the same way. Thus, in my view Lear’s qua operator allows us to treat as a ὑποκείμενον something such as a line or a kind of knowledge which is itself also a predicate, and to ask of it what belongs to it just by its being what it is, or ὅπερ x τι. Both mathematicals and knowledge of grammar can be ὑποκείμενα, if only in a qualified way in the context of inquiry in which they are the subject of some science. Though the same bodies can be the subjects of physics, in Aristotle’s sense, and of geometry, the two sciences outline the structures of different sets of properties, which depend on different more basic properties. These basic properties pick out different ὑποκείμενα. For mathematics, the ὑποκείμενον is extended body, or extended thing; for it is just as (ὅπερ τι) extended thing that each triangle has three sides, and interior angles that add up to 180 degrees. For natural science, it is just as natural bodies that each body follows the laws of motion. In other words, the qua-operator or predicate filter which Lear points to does the work of meeting the ὅπερ x τι criterion on subjecthood which Posterior Analytics points out. This is consistent with the Categories treatment, which allowed items such as grammatical knowledge to be ὑποκείμενα. Knowledge of grammar is ὅπερ τι a suitable subject for some predicates, specifically in the context of a discussion of or inquiry into the different kinds of sciences. In this context, one can ask what its definition is, or what its characteristics are, such as whether it should be taught to children in elementary school; it is itself an object of inquiry. It is treated as something which is something (e.g. a branch of knowledge) and which has various characteristics (e.g., it is important for children to learn it). 61 It may be that knowledge of grammar is, from the perspective of the science of metaphysics, ultimately a dependent sort of being. For Aristotle the context of metaphysics may even be a privileged context; other sciences mark off some kind of being, while metaphysics deals with being ἁπλῶς. (1015a8) But distinguishing substancehood from subjecthood illustrates that the properties a substance possesses can themselves be structured, in such a way that shows the possibility of the variety of extant sciences. Still, metaphysics alone has as ὑποκείμενα items which are not also predicates, for metaphysics is the science of substance. This, arguably, is why being a ὑποκείμενον is presented as informative about substancehood in Categories. This, I argue, is why being a ὑποκείμενον is presented as so distinctively related to substance in Categories.

62 Chapter 3: Physics A.1-6

§.1 Introduction: Organon in Physics A

Having drawn from Categories and Posterior Analytics what I take to be Aristotle’s logical-metaphysical use of the term ὑποκείμενον, in this chapter I turn to the first book of Physics. I will discuss how Aristotle’s concept of ὑποκείμενον enters his dialectical search for the principles of change in A.1-6. In the final chapters of my dissertation I will argue for my reading of how ὑποκείμενον is presented as a principle of change in Physics A.7. In this chapter I start in Physics A.1-3. My intent in doing so is two-fold. First, since I wish to read ὑποκείμενον as it appears in Physics A.4-9 as synonymous with the sense I have just explicated from Categories and Posterior Analytics, in this chapter I aim to provide support for this by showing how Aristotle helps himself to key elements of the view of scientific knowledge from Posterior Analytics in the opening of Physics A. In this way I hope to emphasize the continuity between Aristotle’s thought in his logical works and in his Physics. Second, I will argue that Aristotle’s response to Parmenides and Melissus in A.2-3 provides more positive evidence that the notion of a ὑποκείμενον is key to the project Aristotle is undertaking in Physics A, for in his response to the Eleatics Aristotle suggests that the reason they were unable to provide a satisfactory (to him) account of the principles of natural things was precisely because they did not distinguish between ontological subjects and predicates, or between a ὑποκείμενον and what is said of it. After discussing the logical themes in A.1-3, I will turn to A.4-6. I will argue that in those chapters Aristotle draws from his predecessors insights into the number and nature of the principles, but also develops from this a paradox or puzzle which both motivates his proposal in A.7, and which that proposal solves. In A.7, as I will argue, Aristotle relies on his concept of ὑποκείμενον in order to solve the puzzle. To begin, it is worth asking what Aristotle’s project is in Physics A. Physics is addressed to the student of nature. Its audience is composed of those seeking to become knowledgeable in the science [ἐπιστήμη] of natural things [περὶ φύσεως]. Physics A & B 63 concern the discovery of the principles of natural science, rather than the investigations and inferences which a scientist would undertake as being properly part of the science of natural things. The opening lines of Physics A set the stage for this:

When the objects of an inquiry, in any department, have principles, causes, or elements, it is through acquaintance with these that knowledge and understanding is attained. For we do not think that we know a thing until we are acquainted with its primary causes or first principles, and have carried our analysis as far as its elements. Plainly, therefore, in the science of nature too our first task will be to try to determine what relates to its principles. (184a10-16)27

Aristotle’s plan for investigating the principles is given in A.2 184b15-22. He sketches how he will go about seeking the principles by listing the options for how many, and what kind, one might posit. Ross diagrams the division as follows (pg. 458):

μία ἀρχή πλείονες ἀρχαί

ἀκίνητος κινουμένη πεπερασμένοι ἄπειροι

δύο τρεῖς τετταρες κτλ. homogeneous heterogeneous

Fig. 1: The number and nature of the principles

Aristotle considers each of these possibilities by reviewing the views of his predecessors, and classifying each of their views under the division in 184b15-22. Aristotle discusses the view that the principles are one and unmoveable in A.2 & 3 by responding to

27 Ross translation. Charlton construes the first three lines somewhat differently: “In all disciplines in which there is systematic knowledge of things with principles, causes, or elements, it arises from a grasp of those.” 64 Parmenides and Melissus. He discusses the view that there is one moveable principle, as well as the views which hold that there are infinite homogeneous or heterogeneous principles, in A.4. Each of these views he dismisses, though neither without argumentation nor without drawing from his discussion of them some lesson he will carry forward. In A.5 he begins to examine how many principles there might be if there is a limited number, and what these principles are. He concludes, somewhat provisionally, that the principles must be contraries. In A.6 Aristotle surveys the views of his predecessors to consider what might be missing if it is proposed that the principles are only a pair of contraries. In A.7 Aristotle offers a view which builds upon what he has concluded in A.2-6, and posits two equivalent ways of characterizing the principles. In A.8 he shows that this view additionally solves a problem about change which had flummoxed some of his predecessors; in A.9 he concludes his discussion.

§.2 Physics A.1

As scholars agree, Physics A “begins with material which is quite familiar and intelligible to us from other passages in Aristotle, particularly from the early chapters of the Posterior Analytics.”28 Yet there are also some passages in Physics A.1 which seem to be in tension with Posterior Analytics. If the tensions are great enough, they might suggest an independence between Aristotle’s discussions in his logical works and in his Physics which would undermine my claim that ὑποκείμενον should be read consistently across these works. In this section I argue that the overwhelming impression from A.1 is that in this opening chapter of Physics Aristotle is building on his discussion of scientific knowledge from Posterior Analytics, and I offer a solution to the apparent tensions between them. Physics A.1 begins with Aristotle rehearsing his claim that scientific knowledge arises from knowledge of the principles, causes, and elements of the subject of the science. This echoes, if in a compressed way, the conclusion of the discussion of principles with which Aristotle opens Posterior Analytics A. In this way, the theoretical

28 Bolton, “Aristotle’s Method”, pg. 3 65 underpinning for the investigation in which Aristotle is engaged in Physics A is one of the key theses from the account of knowledge developed in in An. Po. This suggests that Aristotle is taking for granted that understanding of scientific knowledge in Physics A; it provides the strongest reason for the reader to take for granted, in their turn, that when Aristotle introduces into Physics A terms which signified key concepts belonging to Aristotle’s account of scientific knowledge from An. Po., such as ὑποκείμενον, that he means to bring those concepts into his discussion of natural principles. In A.1 Aristotle sets the stage for his search for the principles of natural things. He points out that scientific knowledge is achieved only through the principles pertaining to its subject matter, and advises that these principles will be found by proceeding from what is more known to us to what is more known by nature. In other words, the method for discovering the principles is to begin by attending to the universal, which, though not as yet grasped clearly, announces the subject of the science. From this ones can advance toward knowledge of the principles or elements belonging to the subject. For the reader of Physics, this raises two questions: what is the subject of the science whose principles Aristotle is seeking? And what is meant by principles? In Metaphysics Γ Aristotle says that while metaphysics is the science [ἐπιστήμη] which investigates beings qua beings, the other sciences “cut off a part of being and investigate the attributes of this part.” (1003a20-25) Similarly, in An. Po. A.10 he claims that each science has a proper subject, the genus whose existence it assumes. (76b10-13) He says, further, that in every science one assumes not only the existence of the subject matter, but also its chief attributes, such as odd or even for arithmetic (76b5-10). What, then, is the subject genus whose principles are being sought in Physics A, and what are its chief attributes? It is not until 184b25-26 more than ten lines into Physics A.2, that the reader discovers what the subject is, and Aristotle announces it only in passing when he says that “to investigate whether what exists is one and motionless is not a contribution to the science of nature.” Physics concerns natural things; Physics A seeks the principles of natural things in order to establish a foundation for natural science. When in 184b25-26 Aristotle so vehemently avers that monism makes no contribution to natural science, the grounds he offers for this suggest that the chief attributes of natural things are that they are many, and that they are moveable or in motion. Noticing this, however, is still to see 66 the universal or genus which is the subject of the science indistinctly and unclearly. The reader must wait for Physics B.1 for a more precise characterization of the subject matter. That in Physics A the subject of the science is grasped only unclearly fits with the An. Po. characterization of the search for principles. Physics A constitutes a search for the principles of natural things, and thus of natural science. An. Po. suggests that this kind of project begins from a confused grasp of the subject matter, which becomes more clear as a result of the search for principles. So in Physics A the subject of natural science is understood to be changeable and consisting of a plurality of items; but this is not to say very much. It is only once we have grasped the principles as a result of the invstigation in Physics A that in Physics B the subject of the science can be stated more clearly. In Physics B.1, Aristotle says that natural things are those which have an inner principle of change and rest:

By nature the animals and their parts exist, and the plants and simple bodies… All the things mentioned plainly differ from things which are not constituted by nature. For each of them has within itself a principle of motion and of stationariness (in respect of place, or of growth and decrease, or by way of alteration). On the other hand, a bed and a coat and anything else of that sort, qua receiving these designations—i.e. in so far as they are products of art—have no innate impulse to change. But in so far as they happen to be composed of stone or of earth or of a mixture of the two, they do have such an impulse, and just to that extent –which seems to indicate that nature is a principle or cause of being moved and of being at rest in that to which it belongs primarily, in virtue of itself and not accidentally. (192b9-23)

Thus natural things are things which change (or rest) because they have an inner principle to do so. From A.2 184b25-26 we can glean that natural things change, but 192b9-23 implies that they change in a distinctive way, that their changes are natural or by nature. Sarah Waterlow offers a helpful way of thinking about what kind of change is 67 characteristic of natural things in her Nature, Change, and Agency. She makes the following comparison:

In attempting an account of Aristotle’s concept of change from an inner principle, we shall find it useful to consider two limiting notions: (a) that of change entirely beyond the power of external conditions to affect; (b) that of change entirely dependent on such conditions. I call these “limiting” since each represents a conceptual extreme not coherently applicable to any change having place in a system of natural phenomena. (pg. 5)

Natural change, she says, falls between these two extremes. External conditions can affect natural change, but natural changes arise from and must be explained by something else, that is, by internal principles. External conditions can act as impediments or as aids to natural changes, but they are not themselves causes of natural changes. Thus if there is not enough sun or the soil is too shallow an acorn may never grow very tall or very sturdy, and in the worst case may never grow into a tree or a sapling. But sun and soil alone will not produce a tree out of a lump of stuff unless that stuff is already of the sort to develop into a tree, given the right conditions. Natural things, then, are the sorts of things which will develop and change in predictable, orderly, ways unless some external impediment to their development is in place. From An. Po. A.2-8 it is clear that the principles for each science must be universal, explanatory, prior to those things of which they are principles, and more known by nature, and indemonstrable at least within the context of the science of which they are principles. This leaves several questions open, however. Are the principles entities, or kinds of entities? Is Aristotle searching for which particular things are responsible for the changes in the natural world, or which kind or kinds are responsible? The monists, insofar as they had a view about the principles of natural change or of the natural, observable world, seemed to think there is a single principle, a single being which is the principle. The atomists, on the other hand, posited a number of things as principles, and disagreed about whether they were of the same kind. 68 I propose that what Aristotle seeks when he is searching for principles in Physics A is different from all of these options. Aristotle aims to provide a schema of explanation, and roles within that schema which items in the world must play to feature in an explanation of the relevant phenomena. The principles are those roles. Bostock puts it this way: Aristotle is “trying to lay down in advance the general form which any physical enquiry must have.” (pg. 4) Once Aristotle has figured out what the principles are and how they are related, he will have a guide for constructing proper explanations in natural science. Physics A.1 continues by appealing to another familiar Aristotelian distinction, that between what is “more knowable and clear to us” vs. “those which are clearer and more knowable by nature.” This is a distinction which is familiar from, e.g., An. Po. A.2 71b29-72a5. He appeals to it here to advise how the one investigating the natural world should proceed. Both of these references, to the role of principles in scientific knowledge and to the distinction between what is more knowable to us and more knowable by nature, are appeals to notions developed by Aristotle, and not to concepts or distinctions which would have been part of the assumed background among those philosophers he takes as his interlocutors (e.g., Parmenides, Anaxagoras). Since Aristotle appeals to these distinctions without arguing for them or much elaborating upon them, this adds an air of respectability to the presumption that Physics assumes the understanding of scientific knowledge given in An. Po. Nevertheless, the specific advice Aristotle gives at 184a18-25 for how to go about searching for the principles seems to be in tension with the referenced passages of An. Po. There are two apparent tensions. The first concerns whether one should move from what is more known to us to what is more known by nature, or the reverse. In Physics A.1 Aristotle says that the would-be scientist should “advance from what is more obscure by nature, but clearer to us, towards what is more clear and more knowable by nature.” (184a18-21) In An. Po. he had said the opposite. The second source of tension concerns which things are more known to us, and which more known by nature. In Physics Aristotle says that what is καθόλου [usually taken to signify what is universal], is more known to us, and what is καθ΄ ἕκαστα [usually taken to signify what is particular] is less known (184a23-24: ἐκ τῶν καθόλου ἐπὶ τὰ καθ’ ἕκαστα); in Posterior Analytics he had 69 said that what is καθ΄ ἕκαστα is more known to us, while what is καθόλου is less (An. Po. A.2 72a1-5). In fact, in Physics A.1 Aristotle says that what is καθόλου is closer to sensation, and thus more known to us, while in An. Po. it is what is καθ΄ ἕκαστα which he claims is closer to sensation. It appears as though Aristotle is simply contradicting himself. The first of these tensions is quickly dealt with, and shows the way to the resolution of the second. Ross has this to say about the different directions advised in An. Po. and Physics:

It is true that in An. Post. i.2 it is implied that we should proceed from τὰ γνωριμώτερα τη φύσει, not from τὰ γνωριμώτερα ἡμῖν. But that is because Aristotle is there stating the nature of scientific proof. Here, on the other hand, he is describing the method of attaining knowledge of the ἀρχαι (πειρατέον διορίσασθαι πρῶτον τὰ περὶ τὰς ἀρχάς a15; cf. a22 f.). This is the very reverse of scientific proof. It is the method of reasoning back from what is confusedly given in experience to what that presupposes; cf. E.N. 1095b3… (Ross pg. 457)

Jonathan Barnes agrees with Ross’s assessment of the focus of Posterior Analytics in his later translation and commentary. In his introduction, Barnes argues that Posterior Analytics is not really a work concerning scientific methodology, but rather a “formal account and presentation of the finished system” (pg. xii), that is, what a fully developed science looks like when all the discoveries have been made and can be presented in orderly proofs. In other words, Posterior Analytics concerns the order of knowledge, not the order of discovery; the way down the Line, as it were, rather than the way up. The first book of Posterior Analytics “does not contain a theory of scientific methodology. Aristotle does not pretend to be offering guidance to the scientist… Rather, it [sci., Posterior Analytics] is concerned with the organization and presentation of the results of research.”29 In Physics A, by contrast, Aristotle is trying to discover the first principles

29 Barnes, Jonathan. Aristotle’s Posterior Analytics, pg. xii. 70 of a science of natural things, for “in the science of nature too our first task will be to try to determine what relates to its principles.” (184a15-16) He is not trying to organize or present the results of research into the natural world; he is trying to present the background and foundation for that research. In Physics A Aristotle is beginning the program of research into the natural world which ideally would result in or produce a finished science of the kind he describes in An. Po. Physics A is concerned with discovery. He begins with the principles. Thus there is no tension. Instead, the different approach recommended in Physics complements the one proposed in Posterior Analytics. In looking at the two side by side, the different projects of engaging in research versus presenting its results are clarified by the comparison. This solution to the first problem shows the way to the solution to the second. To better understand the second problem, I quote both the Physics and An. Po. passages here:

Now what is to us plain and clear at first is rather confused masses [συγκεχυμένα], the elements and principles of which become known to us later by analysis. Thus we must advance from universals [ἐξ καθόλου] to particulars [καθ΄ ἔκαστα]; for it is a whole that is more knowable to sense-perception, and a universal is a kind of whole, comprehending many things within it, like parts. (Phys. A.1 184a16-26)

I call prior and more familiar in relation to us what is nearer to perception, prior and more familiar simpliciter what is further away. What is most universal is furthest away, and the particulars are nearest; and these are opposite to each other. (An. Po. A.2 72a1-5)

In An. Po. A.2 71b33-72a5 Aristotle also discusses the difference between what is more known to us and what is more known by nature. As Barnes points out in his commentary, this is one of many places in his corpus in which Aristotle draws the

71 distinction.30 In this passage, he claims that what is more familiar, that is to say, more known, to us is nearer to sensation, while what is more known by nature is further away from sensation. What is universal, or καθόλου, is furthest from sensation and more familiar by nature. In Physics A.1, on the other hand, after re-introducing the distinction between ways of being more knowable or more familiar, Aristotle goes on to characterize those things which are more clear to us both as closer to sense perception and as καθόλου (184a24-25). One possible solution to this tension is that in An. Po. Aristotle allows that in some cases what is more familiar to us are universals. Jonathan Barnes’ commentary on Posterior Analytics supports this suggestion. In it Barnes says that “the question ‘What things are more familiar to us?’ is empirical, and as Top. Z 4 141b36-142a4, recognizes, it has no unitary answer.” (pg. 97) The strongest thing we can say about generality or about universals vis a vis what is more knowable to us, says Barnes, is that “if P is more familiar by nature than Q, P cannot be less general than Q.” (ibid) In other words, what is more familiar to us, compared to something else, might itself be a universal; it is just that what it is more familiar than must be at the same or a greater level of generality or universality. If this is the right reading, there is no tension between the two works on this topic as long as in Physics A.1 Aristotle is claiming only that certain universals among the natural kinds are more familiar to us than others. However, in Physics Aristotle seems to be making a much broader claim than this. He does not limit his claim to just some of the things which are καθόλου. Even if Barnes is right that the most we are licensed to say, according to Aristotle in An. Po., is that if P is more known by nature than Q, Q is a universal at the same or higher level of generality, this will not help resolve the tension between the two texts. I suggest that, instead, in the Physics passage the contrast between καθόλου and καθ΄ ἔκαστα should be understood in terms of the contrast between wholes which are “confused masses” or mixed together [συγκεχυμένα, 184a22] and the characteristics which are mixed together in them, rather than between universals and particulars. In this I take some support from Robert Bolton, who suggests that καθόλου, for Aristotle, is not

30 Barnes, pg. 96 72 always best translated “universal,” and especially not in Physics A. In keeping with this, I propose that we should not understand καθ’ἕκαστα at 184a24 as referring to particulars, as opposed to universals, but rather to the principles or elements which are appropriate to each whole or universal. The idea that in Physics A.1 Aristotle is using καθόλου in such a way that it is not referring to universals (vs. particulars), is not new. Ross, for example, proposes a version of this in his commentary on Physics:

It is clear that καθόλου is not used in its usual Aristotelian meaning. The reference must be not to a universal conceived quite clearly in its true nature, but to that stage in knowledge in which an object is known by perception to possess some general characteristics (e.g. to be an animal) before it is known what its specific characteristic is (e.g. whether it is a horse or a cow). (pg. 457)

Here Ross suggests that Aristotle’s use of the term is non-technical just in the sense that Aristotle is not attempting to suggest that the items which are closer to sense and to our experience are universals grasped as universals. Yet he is still claiming that universals are in a way closer to sensation than particulars, which conflicts with An. Po. This may mitigate, but it does not entirely resolve, the tension. It may even be a source of further tension, for it makes the problem worse for how Aristotle could consistently say that particulars are farther from sensation and more knowable in themselves. If they are farther from sensation in a way comparable to that by which univerals can be, according to Ross, closer to it, then it is particulars grasped as particulars— but not clearly— which are more knowable in themselves. This makes little sense, and I think should not be attributed to Aristotle. Yet there is something important in the suggestion Ross sketches, for surely there is a stage of coming to know or understand something at which what we aim to understand is not grasped quite clearly; a particular is what one is paying attention to, but what it is, what features it has, how they are related, is not quite clear. Bolton argues for a similar approach as Ross, though his interpretation differs in that he does not take Aristotle to be distinguishing between the stages at which universals 73 and particulars are grasped.31 He argues that in both An. Po. A.2 and Physics A.1 καθόλου refers to complex wholes whose parts are as yet undifferentiated to our perception, but not to universals. He thinks that in both passages it is better translated as “whole” than “universal.” Bolton argues that this special sense of “whole” is the same one Aristotle has in mind in An. Po. A.31 and B.19. Thus he offers a new reading of the An. Po. texts as well. Bolton’s approach is similar to Ross’s in that he reads Aristotle as pointing to a stage in coming to know in which what one is attending to is grasped but not clearly. It differs in that he does not take Aristotle to be reflecting on what role universals or particulars might play in this. I think Ross and Bolton are right that καθὀλου in Physics A.1 is not used in the technical sense of “universal”. To accept this and not to accept that ὑποκείμενον in Physics A is used in a non-technical way, or at least in a way different from that in which it was used in Posterior Analytics, might seem ad hoc. But Bolton’s readings of An. Po. A.2, A.31, and B.19 help here; these references suggest that in key parts of the discussion of scientific knowledge in his Analytics Aristotle also uses the term τὸ καθόλου to mean something other than “universal”. Reading it as meaning “according to the whole” or “as a whole” in Physics A.1 is not inconsistent with An. Po., on this evidence. In fact, given the topic of the discussions in An. Po. which Bolton references, this implies a great continuity between the texts than can be found on Ross’s or Barnes’s proposals. The An. Po. references are to texts in which Aristotle is discussing the search for principles, as he is in Physics A.1. This is complemented by reading καθ’ἔκαστα as “according to each” or “what is proper to each”, rather than as “particular” (or even “part,” as Bolton seems tempted to do).

§.3 Physics A.2: Why begin with the monists?

One wonders why Aristotle begins the search for the principles of natural things by discussing the monism of the Eleatic philosophers, especially given his claim at

31 Bolton, Robert. “Aristotle’s Method in Natural Science: Physics I.” In Aristotle’s Physics: A Collection of Essays, Ed. Lindsay Judson. Oxford, Clarendon Press (1991): 1-30 74 184b25-26 that they made no contribution to natural science. Natural science presumes the existence of change and plurality, but according to Parmenides and Melissus change is impossible since it is inconsistent with the monism which they espoused. Perhaps Aristotle is compelled to begin with them by the order of investigation he lays out in the opening lines of A.2; if the search for principles is organized by how many principles one might propose, the first possibility is that there is only one principle. Aristotle justifies this briefly by asserting that one who investigates the number of existents is engaged in a similar enquiry as the one investigating the number of principles. (184b23-25) Yet it is immediately after this that Aristotle denies that the monist proposal is a proposal about the principles of natural things: “for if what exists is just one, and one in the way mentioned, there is a principle no longer, since a principle must be the principle of some thing or things.” (185a3-5) Aristotle says that, because of this, “to inquire therefore whether what exists is one in this sense would be like arguing against any other position maintained for the sake of argument … or like refuting a merely contentious argument— a description which applies to the arguments both of Melissus and Parmenides.” (185a5- 9) Why, then, does he spend the next two chapters responding to the monists? Aristotle even offers a second reason not to engage the monists in Physics. The second reason is that what distinguishes the subject of the science of nature from the subjects of other sciences is that natural things are changeable. It is a necessary presupposition of the science that there are things in motion (Aristotle also thinks this is evident by induction [185a12-14]). Thus, according to Aristotle, someone engaged in the kind of study in which he himself is engaged does not “inquire… whether what exists is one in this sense.” (185a5) Perhaps, then, it is someone else’s job to respond to the Eleatics, and not the natural philosopher’s. Bolton, for example, has argued that it would be the job of the dialectician (“Aristotle’s Method,” pg. 14-15), while Ross suggests instead that the science in which that question truly belongs is metaphysics. (pg. 461) Yet he says the following to justify Aristotle’s including the discussion in Physics A:

Since such a view denies the essential presupposition of physics, it is not the business of physics to discuss it. Nevertheless, to clear the ground for 75 physics, Aristotle proceeds to refute the view which would make physics impossible by denying its very presupposition.” (pg. 20)

Charlton, similarly:

Aristotle says that it is improper to inquire what a thing is, until you have established that it is, i.e. established that there is such a thing (e.g. An. Po. II.93a19-20, but cf. Met. E 1025b17), and his practice in Physics reflects this view: thus with change, the infinite, place, void, time, and cf. on nature at 193aff. Now chapter 7, which is the kernel of Physics I, is in fact an analysis of becoming; according to his principles, then, Aristotle ought to show that there is such a thing as becoming, that things do come to be. Chapters 2-3 fulfill this need.” (pg. xv)

In other words, in the end Aristotle finds it useful to respond to the monists in order to head off an objection to his whole project in Physics, an objection whose source’s fame and influence make it rhetorically pressing that he should respond. This seems right, but I think something more, and more directly pertinent to the enquiry into natural things, is also accomplished in Aristotle’s response to the monists. Aristotle’s refutation of the Eleatics is instructive for the search for principles, for in his refutation he emphasizes what the Eleatics had missed, in his view. What they had missed was something so basic that, without it, they rightly conclude that a science of nature is impossible. What they missed thus presages what Aristotle will find in his search for principles: that on which the intelligibility of the natural order rests, or the principles of natural science. Aristotle is keen to show why the monists cannot contribute to natural science, and to expose which of their tenets is really the impediment. From his response to them, one can reconstruct that natural things, if they are to exist at all, are complex. Each is complex firstly in the sense that although it is one in number it is polyeidetic or multi-aspectual, and not statically so. It is the mark of natural substances that they are changing things. This, I think, is what Aristotle has in mind at the beginning of Physics A.1 when, having said that we should begin our inquiry from what is clear to us, 76 he says that “what is to us plain and clear at first is rather confused masses.” 32 (184a21- 22) The word he uses, which Hardie and Gaye translate as “confused masses”, is συγκεχυμένα. It means “mixed together”, but especially, “mixed together confusedly”. Natural substances are multi-aspectual, and their aspects or properties are changing. Until we sort through and analyze their principles and elements (184a23-24) what we can say about them will remain as confused and mixed up as their various aspects at first appear. We will not understand their behavior, or be able to predict their changes, or see the connections between their various aspects or characteristics. Understanding the principles enables us to organize and see the connections among the true things we can say about natural things; it allows us to organize a science of nature. Natural things appear to be confusedly complex. Discovering their principles enables us to see that though they are complex, they are not confusedly so. Because the Eleatic monists seemed, or were interpreted by Aristotle as being, committed to a certain sort of simplicity, Aristotle’s response to them is instructive of the kind of complexity which he wishes to ascribe to natural things. In A.2 Aristotle responds to the Eleatic monists by cataloguing and addressing the various ways their claim that all things are one could be interpreted. The ostensible aim is to show that no matter what kind of one-ness the Eleatics are committed to, no matter how Aristotle interprets their ambiguous central claim that “all things are one”, every version is untenable. (185a26-27) For example, Aristotle first tries to show how little sense a monist can make if their monism requires rejecting the pluralism espoused in the 10 genera given in Categories. In Categories Aristotle claims that “of things said without any combination, each signifies either substance or quantity or qualification or a relative or where or when or being-in-a-position or having or doing or being-affected.”33 (1b25-26) That is to say, there are roughly ten ways being can be said. If all beings are called beings on only one ground, Aristotle says that the monist must determine which ground, among the ones

32 Unless otherwise noted, the quoted English translations of passages from Physics are from R.P. Hardie and R.K. Gaye’s translation, edited by Jonathan Barnes in The Complete Works of Aristotle (1984), volume 1. 33 Trans. J.L. Ackrill, in The Complete Works of Aristotle vol. 1. 77 Aristotle himself has pointed out, is the privileged one. Suppose, then, that all things are qualities, or all things are quantities, or all things fall into any one of the non-substance categories. Aristotle argues that this is impossible, for to be a quantity or quality or any kind of property is to be the sort of thing which depends upon a substance for its existence. He says in Physics A.2, reminding his readers of his claim in Categories, that “none of the others can exist independently (χωριστόν) except substance; for everything is predicated of substance as subject” (185a31-32). Some beings are the kind which are “in a subject”, according to Categories, in the sense that they cannot exist apart from that in which they are (1a25), and which are such that, ultimately, they depend on some particular substance (2a34-35, 2b3-5). Quantities and qualities, indeed the items in every category except substance, are like this; it is distinctive of substances not to be like this. Aristotle does not think that it makes any more sense to claim that there is only substance and not also quantities or qualities, etc. In the words of Categories, while substances and their properties might mutually imply each other’s existence, the latter depend upon the former for their existence. (14b10-14) Still, substances and properties do mutually imply each other. There are neither substances without properties nor properties without substances. In considering the proposal that all things are substance, or that there is only one substance, Aristotle points out that if this were the case the one substance could have no magnitude of any kind, or be infinite (or finite). Here he is directly addressing Melissus, who seems to have argued that the One was infinite. Thus by Melissus’s lights, in Aristotle’s categorial scheme the One cannot be substance. But a similar argument might apply to any view; if what-is is only substance, there would be nothing more to say about it. Intuitively, this is a problem. Thus Aristotle concludes contra the monists that there is no single suitable ground upon which τὰ πάντα might be called beings. To try to suggest otherwise is to make it so that nothing we say makes any sense, including our attempt to state the monist position. Regardless of the merits of this particular argument, for my purpose in this dissertation two things should be emphasized. First, Aristotle is helping himself to views from his logical works, including paraphrasing some of its key claims. He continues to do so throughout that chapter. Second, in so doing he points to the different ways things are called beings, and to the metaphysical schema he proposed in his Categories, which 78 relies on his understanding of ὑποκείμενον. More particularly, it relies on the kind of logical complexity according to which one thing may be predicated of another such that it has one account as ὑποκείμενον, and another as what is said of that. This, as I will argue in my next chapter, is key for Aristotle’s account of the principles of change. The same themes can be found throughout the chapter as Aristotle proceeds in a more or less orderly way through different versions of monism that might be ascribed to his interlocutors. For example, after the argument just summarized, Aristotle considers the versions of monism that say that the universe (τὸ πάν) or what is (τὸ ὄν) might be one by way of being continuous (συνεχές), or because it is indivisible (ἀδιαίρετον), or because all beings have the same definition (λόγος) (185b7-9). To respond to the first, he relies on his own understanding of continuity, according to which continua are one actually, but two (or more) potentially, inasmuch as they are divisible (186a1-5). This is a view bruited in Cat. 6, though addressed elsewhere in his corpus more thoroughly (Metaphysics Δ.5, Physics Z). That he relies on the view of mathematical continuity which also underwrites his discussion of quantity in Categories 6 helps to support the interpretational strategy I wish to follow, inasmuch as it shows a unity of thought between the two works. The appeal to his notions of actuality and potentiality, absent from the mention of continuity in Categories though inextricable from more thorough discussion of continuity elsewhere in his works, helps in a different way. For after giving his account of the principles of change in Physics A.7 in terms of ὑποκείμενον + form and lack, Aristotle will say that the same thing can be said in terms of potentiality and actuality. In fact, as I will argue, Aristotle uses his concept of ὑποκείμενον to illuminate how he thinks about what it is to be something potentially. This has to do especially with complexity in λόγος. This is brought home in the final version of monism Aristotle considers in I.2. The final monist proposal Aristotle considers is that all things are one in the sense that the same definition belongs to each of them, even though they are called by different names. Recall the Categories presentation of beings which are synonyms, homonyms, or paronyms in Cat. 1. Synonyms are those things which have the same name, and for whom the account telling what it is to be for something by that name is the same; 79 homonyms are those things which have the same name, but for whom the account/λόγος of what it is to be according to that name is different. For example, bat and bat are homonyms since for the one the account of what it is to be a bat is “flying rodent”, while for the other the account is “baseball implement.” This understanding of a λόγος of what it is to be according to a name clarifies the final option for monism Aristotle considers in Physics A.2. In Categories 1 Aristotle lays out the options for the λόγος of things having the same name. In Physics A.2 he is is concerned with the possibilities for the λόγος when things have different names; what it is to be according to the name could be the same, or it could be different. Thus, in Aristotle’s example, λώπιον and ἱμάτιον (185b20) are different words for clothing, and this means that though their names are different, the account of what it is to be for each name turns out to be the same. What it is to be the one is the same as what it is to be the other. One way of understanding monism is to read it as proposing that there is only one λόγος, no matter what name something goes by. If this were the case, then nothing could be complex by being a ὑποκείμενον for some predicates; nothing is such that it has some characteristic or characteristics such that it has others which are founded upon this one. There is a reason Aristotle takes up this proposal last in A.2. It is, for him, the underlying problem for monism. It even underlies the more particular problems he addresses earlier in the chapter. The puzzle about how a monist might account for mathematical continuity, for example, turns on whether a monist can accept the potentiality vs. actuality distinction, for Aristotle, which requires that things have the kind of logical complexity which is captured by Aristotle’s notion of ὑποκείμενον. The success of Aristotle’s response to this version of monism in Physics A.2 is up for debate, and frankly more than can be taken on in this dissertation. But for now it serves to show how, again, Aristotle relies on notions from his logical works..

§.4 Physics A.3 and what is ultimately at stake

That Aristotle’s notion of ὑποκείμενον, and the kind of logical complexity belonging to things which goes along with it, is for him the key thing Parmenides and his students are missing is borne out in Physics A.3. In A.3 Aristotle makes two claims 80 about Parmenides's view: (1) it is based on a false assumption, and (2) even granted that false assumption his further conclusions do not follow (186a23-24).34 Parmenides’ false assumption, says Aristotle, is that he takes being [τὸ ὄν] to be said simply [ἁπλῶς].(186a23-24) The conclusion which Aristotle argues does not follow from this is that what is is one. Even if being is said simply, and not in many ways, Aristotle argues that it does not follow from this that what is will be continuous or one in account (τῶ λόγῳ). Aristotle’s strategy in showing that Parmenides infers badly from his first premise has two parts. In the first part, 186a25-28, he argues that even if being [τὸ ὄν] has one meaning or signals one thing [σημαίνοντος ἕν, a26-7] it does not follow that being is one either by being continuous or by being one in account. He argues this by substituting τὸ λευκόν for τὸ ὄν, and suggesting that what is clear in the case of τὰ λευκά follows also for the case of beings. He supposes that τὸ λευκόν has one meaning [σημαίνοντος ἕν], where this is standing in for being said simply; this leaves open, he says, that “it is different to be for the white and for what receives [it].(my translation)” [186a28-29: ἄλλο γὰρ ἔσται τὸ εἶναι λευκῷ καὶ δεδεγμένῳ]. The opposite of Parmenides’ assumption that being is said simply is that being is said in many ways. This is a claim Aristotle makes in many places throughout his corpus. Charlton suggests that when Aristotle uses the expression “said in many ways” in the Physics passage, he has in mind things being called or said to be X not “on the same grounds” (Aristotle’s Physics I, II, pg. 55). Being called something on the same grounds is like the case in which “all the things which are called spherical are called spherical for the same reason, that all points on their surface are equidistant from a single point: we have a single idea of what it is to be spherical, and they are called spherical in so far as they accord with this.” (ibid). Thus Charlton suggests that we read “signifying one thing” [σημαίνοντος ἕν, a26-7] as meaning the same thing as “said simply”. In the case of τὸ λεύκον, if the expression signifies one thing this is because if one calls a dog white and a cat white, or even just a color (an attribute) white, “white” there has the same meaning; it

34 Charlton’s translation: “The answer to him [Parmenides] is that he assumes what is not true and infers what does not follow.” Ross’s translation: “the answer to him being that this is not true and that does not follow” (emphasis not mine). 81 is a color at the opposite end of the spectrum from black. It does not matter for this that the attribute and the thing which possesses it are not identical, for both get to be called white for the same reason or on the same grounds. The sense in which they are different is not about the meaning or signification of the term; it is not about what is signified or pointed to or drawn attention to by use of the term “white”, for the use of the term picks out something according to the same color. Yet even if it is the very same meaning or definition of white which is the grounds for calling both the attribute “white” and the thing which bears it, the definition belongs to both of these in different ways. One of these is a white thing as something whose λόγος is to be white, while the other is a white thing as something which is accidentally one in number with the thing for whom what it is to be is to be white. The attribute or predicate and the ὑποκείμενον have different λὀγοι since “to be for white and that to which it belongs is different” [τῷ εῖναι ἕτερον τὸ λευκῶ καὶ τῷ ὑπάρχει ](186a31). This is in accord with the Met. Z.6 passage (1031b22-28), where Aristotle announces that the essence of each of these is different. Hardie and Gaye seem to agree, for they translate 186a31 “For whiteness and that which is white differ in definition.” Their use of the term “definition” is somewhat misleading , but aims to capture the sense of Aristotle’s expression “to be” [τῷ εῖναι] in 186a31. Charlton translates the same line differently, but to the same effect: whiteness and that which bears or receives it “differ in their being” (pg. 5). This is supported by 186a32-34, where Aristotle supplies to Parmenides the premise which he thinks is lacking in Parmenides’ argument. Aristotle says, as Hardie and Gaye translate it, that “it is necessary for him, then, to assume not only that ‘is’ has the same meaning, of whatever it is predicated, but further that it means what just is and what is just one [ὅπερ ὄν καὶ ὅπερ ἕν].” Hardie and Gay’s translation suggests that Aristotle is talking about the word or verb “is.” What is in the text,however, is τὸ ὄν, which means not “is” but “being” in the sense of a thing which exists. Aristotle is claiming that Parmenides assumes that “being” is said synonymously. The function of ὅπερ in this context is to clarify that whatever is picked out as x (white, or being), is being picked out just inasmuch as it answers to the definition of being x, or ὅπερ x. It is being used in just the way it was in An. Po.. E.g., to refer to 82 something as ὅπερ λεὺκον τι is to pick out something which is white, inasmuch as it is white, regardless of whatever else it might be. In this argument it helps clarify that to get the conclusion he wants Parmenides must not only assume that, to use Aristotle’s example, there is only one definition for what it is to be white, but also that there is no difference between being ὅπερ white and being what is a ὑποκείμενον for white (itself being ὅπερ something else). The same goes for beings as for white things. Even if one allows that “being” has the same signification when ontological subjects are called beings and when the things which are ontologically predicated of subjects are called beings, nevertheless when the subject is called a being and when what belongs to it is called a being, they are called so differently. Thus even if “being” signifies one thing, this does not rule out its being said in many ways such that substances as ὑποκείμενα and what are predicated of these can be distinguished. Yet Parmenides needs to assume this in order to draw his monist conclusion.

§.5 Physics A.4 and the physicists

After A.3 Aristotle’s readers should expect the distinction between ontological subjects and predicates to appear again in the context his search for principles in natural philosophy. We should expect Aristotle to bring to bear this distinction deliberately, and with the same awareness of and sensitivity to methodological concerns as he showed in A.1-3, in order to clarify his investigation into the natural world. Aristotle’s readers should have something of the experience of the reader of a detective novel. At some point near the beginning of the novel the astute reader suspects that a key character has been introduced. The perpetrator of the deed has been foreshadowed, but the reader does not yet know how the details of the crime, or its discovery, are going to work out. I think that in Physics A.3 the perpetrator has been foreshadowed. We suspect that, somehow, ontological subjects and predicates are the key, responsible for all the action, even though Aristotle has not quite said so yet. We still have to allow the plot to unfold. We have to wait for Aristotle to work out the two ideas, the one from A.2 about complexity and difference, and the one from A.3 about the importance of the notion of an ὑποκείμενον. We are not yet certain how the notion of an ὑποκείμενον is going to help explain change, 83 but we begin to suspect that it will. My goal in what remains of this chapter is to show how Aristotle develops these two thoughts in his investigation into the principles of natural things in the latter part of Physics A. In A.4 Aristotle moves on to consider the views of those he calls the “physicists” (οἱ φυσικοί). One of the key premises for the argument of A.5 arises from Aristotle’s reflections here; in A.5 Aristotle takes as a premise that “in nature nothing acts on, or is acted on by, any other thing at random, nor may anything come from anything else” (188a31-34). Aristotle argues for this premise in A.4 by dismantling Anaxagoras’s view. By Aristotle’s lights, Anaxagoras offers a view according to which anything comes from anything, because everything is in everything (187b1ff). Aristotle suggests that Anaxagoras thinks we can discover that anything comes from anything at all (“any chance thing”); even blood can be gotten from a stone, for there is some blood mixed in everything, just as there is some stone mixed in everything. Whenever we observe some change, we note that we can describe the terminus a quo in a variety of ways, since natural things are polyeidetic, and the same for the terminus ad quem. As Aristotle presents it, Anaxagoras does not see any reason to think any one description of either terminus is more apt than any other.35 As a consequence, it seems that we can observe that the just comes to be from the musical, or the pale from the studious, or the tall from the round, or whatever. Thus anything can come from anything, and there is no observable pattern. But since Anaxagoras accepts the dictum from Parmenides that nothing can come from nothing, he holds that in order for anything to come from anything it must be the case that what it comes to be was already present, but hidden. Thus blood can come from a stone because there was already blood in the object we were calling a stone. There was a preponderance of stone, to be sure; for this reason it was called a stone. Yet because there was a bit of blood in it, just as there is a bit of everything in everything, it was possible to get blood from a stone. Aristotle has several responses to Anaxagoras, but I am going to highlight only a couple of them here. After offering some epistemological reasons to prefer a finite

35 Note that, as Code (1976) and Waterlow (1982) have argued, Aristotle himself argues that some descriptions of the termini are more apt than others in the first part of Physics A.7. 84 number of parts to an infinite number, Aristotle argues at 187b14-21 that the parts of things cannot be infinitely small or infinitely large. Aristotle thinks that the kinds of stuff, the elements or principles which are in everything, include homoiomerous things such as flesh and bone. And because they must be mixed in everything, according to Anaxagoras, Aristotle takes it that Anaxagoras’s principles or elements must, by Anaxagoras’s lights, be indefinitely small. Further, there must be an infinite number of such elements mixed into everything, otherwise it would not be the case that you could separate out from anything some portion of any kind of stuff that you choose. If Aristotle can show that some parts, at least, cannot be infinitely small, then he can use this to show that there also cannot be an infinite multitude of elements mixed into everything. And if things of finite size in the natural world can neither have an infinite number of parts, nor can their parts be infinitely or indefinitely small, then Aristotle can show that it is not the case that everything is mixed in everything. At 187b14 Aristotle argues as follows. First, it is necessary that if the parts can be of any size then so can the whole, and vice versa. Second, we can discover from observation that it is not the case that natural things such as animals or plants can be of just any size. That is, natural wholes such as humans, dogs, insects, and flowers are not such that they can be of any size at all. A dandelion which is 20 feet tall would not be able to support itself. Pace Mary Norton and her “borrowers”, humans cannot be only a few inches high. But if natural things such as animals and plants are such that they cannot be of just any size, then it follows that neither can their parts. And the same argument can be run for things such as hands and feet and their parts. Hands cannot be of just any size, therefore neither can their parts. Their parts are such as flesh and bone. But flesh and bone were among those principles which, according to Anaxagoras, were such as to be mixed in everything, and consequently must be able to be as small you please. Now, because Aristotle has shown that parts such as flesh and bone have some smallest size, he can now argue from straightforward Euclidean principles it is not the case that everything is in everything. To argue this Aristotle imagines performing a process of extraction, extracting flesh, for example, from some finite body. Then, since each piece of flesh is of some finite size, then either the process of extraction will end or 85 it will not. If it ends then it is not the case that everything is in everything, for there will be some of the original body left in which is there is no flesh. If the process does not end, then it must be the case that there is an infinite number of finite equal parts in a finite body, but this is impossible. So it must be that the process ends and there is a finite number of equal parts in any finite body. Because of this, it cannot be the case that everything is in everything. Since Anaxagoras’s claim that anything can come from anything depended on his claim that everything is in everything, Anaxagoras no longer has grounds to claim that anything comes from anything. This highlights an important point for Aristotle. For Aristotle, to make sense of change we must be able to see the seeds of change in what has come before; the natural scientist must be able to see why this has come about, and not that, or how what came before has in it the necessary precondition for what comes to be. Anaxagoras allowed that anything could come from anything, but only because what comes to be was already present. There was some reason that could be given for why it came to be. Aristotle shows that Anaxagoras’s way of meeting this constraint is unsatisfying; taking the time to do so helps the reader to see why it was important for Anaxagoras to claim that everything is in everything in the first place. It also provides an important premise for Aristotle’s argument in A.5.

§.6 Physics A.5 and the contraries

Aristotle opens chapter 5 with the observation that “All make contraries principles”36 (188a19), even Parmenides in his Way of Opinion. He goes on to say that this is reasonable (εὐλόγως, 188a27), for contraries have the sort of primacy which is required of principles. Aristotle suggests that contraries are suitable candidates for principles because “principles are neither from each other nor from anything else, and all are from them” (188a27-28), and contraries meet these conditions. He goes on to clarify; contraries are not “from” each other, since they are contraries, and if they are primary contraries are not from anything else.

36 My translation. 86 As I understand these lines Aristotle is not offering what he takes as a definitive argument that the principles he is searching for in Physics A are contraries. Rather, he is saying that positing contraries as principles is εὐλόγως (reasonable). The reasonability which Aristotle is pointing to consists in the way that Parmenides and the rest seem to have grasped the kind of primacy which the first principles of natural philosophy must have. Here at 188a27 Aristotle suggests that their primacy has three conditions: (1) they are not “from” [ἐκ] each other, (2) they are not “from” anything else, (3) everything else in the science is “from” them. Contraries, at least the way Aristotle’s predecessors have treated them, fit this. Hardie and Gaye, in their translation, add in “derived,” so that they translate the simple preposition “ek” as “derived from”: “first principles must not be derived from one another nor from anything else.” Sean Kelsey suggests that an appropriate translation of ἐκ in these lines is “consists of”, to help mark the contrast between ἐκ +ἐστι [ ἐκ + the verb for “to be”] and γίγνεται + ἐκ [ek + the verb for coming to be] (footnote of his 2008). Ross, similarly, in his commentary translates it “composed of”. I think that these attempts to clarify the conditions for things being principles make Aristotle to be too precise, too quick. At 188a27 Aristotle offers a set of conditions which he sees all the views he has just mentioned as meeting. Aristotle is saying of Parmenides, Democritus, Empedocles, and the rest of those he calls the “physicists” that they all posit contraries which are such that they meet the criteria above. If we read the “from” as “consists of”, then Aristotle would be suggesting that all of these thinkers claim that the natural world and everything in it is composed of their chosen pair of primary contraries. But a look at the way Aristotle presents their views in the surrounding passages does not uphold this. For example, the Empedoclean contraries Aristotle points to are Love and Strife. Aristotle presents the Empedoclean view as being that Love and Strife organize and disorganize, respectively, the elements from which the cosmos and all the things in it are composed. That is, they seem to function more like efficient causes than as what the cosmos is composed of or consists of. In the way that Aristotle presents Democritus’s view, plenum and void are the Democritean contraries, as well as “differences in position, shape and order, and these are the genera of which the species are contraries, namely, of position, above and below, before and behind; of shape, angular and angle-less, straight 87 and round.” (188a22-26) Surely Aristotle is not suggesting that for Democritus the cosmos is “composed of” the void, or even of above & below. Rather, the cosmos is composed of atoms which are above and below each other in the void. Thus reading ἐκ at 188a27 as “consists of” does not fit with the range of views Aristotle is canvassing. On what sense of “from,” then, is it reasonable that contraries should be principles, such that their not being from each other follows from their being contraries? If the “from” in condition (1) is “comes to be from”, in the sense in which an oak comes to be from an acorn, then there is another way in which Aristotle contradicts himself. Just a few lines later, Aristotle argues that in the natural world contraries always come to be from contraries. For example, pale things always come to be from dark things (something comes to be pale from having been dark), and musical things come to be from unmusical ones, “for how could white come from musical, unless musical happened to be an attribute of the not-white or of the black?” (188a35-36). Thus he concludes that “everything that comes to be or passes away comes from, or passes into, its contrary.” (188b21-23) This is a key part of his own argument that contraries are principles. He cannot mean by “from” [ἐκ] at 188a27 the same thing as “comes to be from” (γιγνεται ἐκ). If he does, then in the course of the same chapter he would both argue that the fact that contraries do not come to be from each other or anything else is part of their qualification for being principles, and that the fact that contraries always come from each other qualifies them to be principles. This makes Hardie & Gaye’s “derived from” a happier choice than Kelsey’s “consists of”, in that “derived from” does not obviously exclude either the Empedoclean or the Democritean principles, and neither does it tend to impute to Aristotle contradictory claims. I think that ἐκ in 188a27-31 means “derived from” in an epistemic sense. Principles, if they are indeed to explain those things of which they are principles, must be such that they are not explained by or logically derived from anything else in the science, while it must be the case that everything else in the science is made known from, or by, these. It is because contraries are contraries that they are not explained by each other, one being prior in explanation to the other; they come as a pair. It is because in the theories in which they appear the Democritean and Empedoclean contraries are given as primary that they are not explained by anything other than each other; and their being 88 posited as principles just is to propose them as explaining everything else. The reasonableness of positing contraries, then, comes in great part because they are contraries. Aristotle has already argued against the possibility that there is only one principle and the view that the principles are infinite in number. In considering the remaining option that there are more than one principle, but not an infinite (or even a very great) number of principles, he notes that any view positing multiple principles faces the objection that ultimately some of the principles can be derived from the others. Yet if the principles are contraries, this objection can be avoided. Attractive as this is, this reasoning is not sufficient for Aristotle to posit that the principles of natural change, and of natural science, indeed are contraries. Aristotle presents his argument for this in the next part of A5 (“we must see how this can be arrived at as a reasoned result”, 188a30-31). He does this by appealing to the claim he argued for in A4, that “in nature nothing acts on, or is acted on by, any other thing at random, nor may anything come from anything else, unless we mean that it does so accidentally.” (188a31-34)37 That is to say, he takes as a premise that natural things change and develop in regular, orderly ways, such that it is possible to understand and even predict their changes. Oak trees always come from acorns, never (at least immediately) from stones, for example. Aristotle thinks that our observation of the natural world supports a more precise claim, too. This is what he claims in A.5. It is not just that natural things change in orderly ways, but that the order involved is that something always comes to be from its opposite. If we attend to any case of change, Aristotle thinks that we will discover that what comes to be does so from its contrary. This begins to become clear when we take note of the seemingly obvious point that what something comes to be it was not before; what is x comes from what is not-x, e.g., the pale thing comes to be from the not-pale. Aristotle thinks there is more to be said about the orderly pattern of all changes, however, than simply that what is x comes to be from what is not-x. The contrary from which something comes to be is not merely something characterized by the contradictory property, what is not-x, but something having a contrary property where this is a definite,

37 Barnes edition Hardie & Gaye translation. 89 positive characteristic bearing an appropriate relation to the characteristic which comes to be. What Aristotle means by contrary here takes some spelling out, and he attempts to do so by using examples. Aristotle uses several examples: the pale coming to be from the not-pale, the musical from the not-musical, what is tuned from what is not-tuned. His analysis of the tuned coming to be from the not-tuned is especially helpful, especially for English-speaking philosophers, since there is a way in English to distinguish between the contradictory of being in tune and the contrary which Aristotle is trying to point to. Aristotle says the following:

The same holds of other things also: even things which are not simple but complex follow the same principle, but the opposite state has not received a name, so we fail to notice the fact. For what is in tune must come from what is not in tune, and vice versa; the tuned passes into untunedness—-and not into any untunedness, but into the corresponding opposite. It does not matter whether we take attunement, order, or composition for our illustration; the principle is obviously the same in all, and in fact applies equally to the production of a house, a statue, or anything else. A house comes from certain things in a certain state of separation instead of conjunction, a statue (or any other thing that has been shaped) from shapelessness—each of these objects being partly order and partly composition. (188b9-21)

In the last remark, that each of these is “partly order and partly composition” Aristotle attempts to capture what is common among all the cases of the relevant contrary: what is not-tuned, or not-shaped, or not built, must already have a certain character or composition in order that it might be put together (put into the right order) to become its contrary. In the house example, it is things which are in “a certain state of separation instead of conjunction” which form the relevant contrary. 188b17-19 might be better translated “The house comes to be out of things not put together but dispersed in this very way” [ἥ τε γὰρ οἰκία γίγνεται ἐκ τοῦ μὴ συγκεῖσθαι ἀλλὰ διῃρῆσθαι ταδὶ ὡδί]. Not just any not-built or not-constructed or not-house thing comes to be a house; not just any not-

90 constructed thing is the relevant contrary which is the terminus a quo of the change from which a house results. This can be seen by thinking about the way in which English implicitly distinguishes between being out-of-tune and being not-tuned. What is out-of-tune is a subset of what is not-tuned, with the distinguishing characteristic that what is out-of-tune has some features such that it actually can be tuned. Not every not-tuned object is this way. The children’s joke works well here: you can tune a piano, you can tune a harp, but you can’t tune a fish. A piano can be out of tune, a harp can be out of tune. A fish can be not-tuned, but it is never out-of-tune, for it cannot be tuned. Musicologists explain the features of things which make them such that they are tunable (that is, out-of-tune rather than merely not-tuned); it has something to do with their physical construction such that when they are sounded the sound waves occur in certain ratios, or can be made to do so. Regardless of what the relevant feature or features actually are, Aristotle’s point with this example is that there are some. The contrary of what is tuned is something having a certain characteristic, not merely lacking the characteristic of being tuned. This leads Aristotle to a very general conclusion. Rather than concluding, as many of his predecessors had, that some particular pair of contraries such as hot and cold is involved in every change, Aristotle infers the more general claim that some pair of contraries is involved in every change. Thus he agrees with his predecessors that contraries are among the principles, but, unlike them, he does not settle for any particular pair. At least, he does not settle for any pair in the way that they do. Aristotle points out that, inasmuch as all pairs of contraries are contraries, they are analogous to each other. Their analogy consists in their bearing a certain relationship to each other, which he characterizes in terms of form and lack. Form and lack, then, are the most general pair of contraries; it is the determinable of which every other instance of contraries is a determinate instance. Every other instance of contraries is not only a determinate instance of form and lack, it is also analogous to form and lack. It is this pair of contraries which Aristotle calls the primary contraries. Form and lack are primary in that this pair of contraries will be present in every change, in some determinate form. Further, finding out what determinate instance of these contraries is involved in any particular change will be an important part of 91 understanding that change, and what its outcome is. To return to the tunedness example, if one wants to have a tuned instrument it is important to start with something which has the right features. That the sounds which are musical are those which stand to each other in certain ratios, and are produced by instruments having certain features, was an important discovery among the ancient Greeks. Realizing that an octave, a key interval in music, is produced by strings whose lengths stand in the ration of 2:1 means, among other things, that something which is constructed differently cannot be tuned appropriately. This is borne out by looking at the development of musical instrument- making . Historians of music can point to a development whereby musicians have chosen to make musical instruments from things with features which are better and better suited to becoming, and staying, tuned. This development came about partly through better understanding of the features required for something to become tuned. In other words, this development came about partly through scientists better grasping the relevant contrary to tunedness. Thus finding the right pair of contraries is important both for musicologists, scientists who study music, and for musicians; it makes a difference both to understand the phenomenon of being tuned, and to producing things which are tuned. This helps to show why it is reasonable for Aristotle to posit form and lack as principles. They are principles in the sense that, since there is something analogous to them in every change, they are the most general formulation of two of the elements which a natural scientist must look to in giving an account of some natural change. Thus when Aristotle says at the end of A5 that contraries are principles, this is a way of reformulating his earlier claim that there are contraries involved in every change, while pointing to the fact that if you want to understand some particular kind of natural change you need to understand which contraries the change is between. He’s turning that general observation about natural changes into a principle for natural science. Bostock, in his “Aristotle on the Principles of Change in Physics I”38, says of form and lack as principles that “It is as if one were to say that the fundamental opposition in physical enquiry is that between ‘a thing and its opposite’, for there is no more content to the pair ‘form and privation’ than this – and in fact there is less.” (pg. 3-

38 In Language and Logos, Ed. Malcolm Schofield and Martha Nussbaum (1982): 179-196 92 4) That is to say, Bostock thinks that form and privation, or form and lack, are best understood as contradictories: “form and privation are practically the negations of one another, for anything which is of the right sort to have a certain form but does not have it will be said to have the corresponding privation, and vice versa.” (pg. 9) However, I think Bostock is misunderstanding what Aristotle means by form and lack, and the examples to which Aristotle appeals in A.5 show this. It is this understanding of the contraries which Aristotle is trying to avoid when he says that “white comes from not- white—and not from any not-white, but from black or some intermediate” (188a36- 188b1), and that “musical comes to be from non-musical, but not from any thing other than musical, but from unmusical or any intermediate state there may be” (188b1-3) or again “white does not pass into musical (except, it may be, accidentally), but into not- white—and not into any chance thing which is not white, but into black or an intermediate” (188b3-5) In each case he emphasizes that it is not just something lacking the property which counts as the relevant contrary; it is something which has some other relevant features (e.g., black or some intermediary, untuned or some intermediary, the right stuff in a certain state of disarray and needing to be put together).

§.7 Physics A. 6 and ὑποκείμενον

In A.6 Aristotle presents considerations to motivate positing a third principle in addition to the contraries, and specifically for ὑποκείμενον as a principle. Yet he does not present a proof, or even arguments which he takes to be dispositive. He says that the view that there are three principles (and that these are a pair of contraries and a ὐποκείμενον) is reasonable: he says that it would seem to have a certain reasonableness [δόχειεν ἂν ἔχειν τινὰ λόγον (189b17-18), which Hardie & Gaye translate as “would seem…a plausible view”], and he concludes by saying that “whether two or three is, as I said, a question of considerable difficulty” (189b27-29), thus leaving the question somewhat open. The conclusions of the chapter have something of a tentative tone, and this is for good reason. In A.6 Aristotle presents considerations in favor of positing a ὑποκείμενον in addition to the contraries as a principle. These take two forms, each based on one of the functions which his notion of ὑποκείμενον performs: (1) as that 93 which is acted upon so as to undergo change, as in Cat. 5 (4a10), (2) as that which is presupposed by and so underlies other items in his ontology, especially as substance. The first he proposes in 189a21-26, and brings up again at 189b3-16, the second in 189a27- 33. The second motivation Aristotle points to for positing a ὑποκείμενον is concerned with his claim in Cat. 5 that nothing else can be unless it is predicated of a substance as ὑποκείμενον, mediated or immediately. This same consideration, however, seems to undermine the contraries’ standing as principles. Thus Aristotle presents some considerations in favor of positing a ὑποκείμενον as an additional principle, but in doing so discovers himself in ἀπορία regarding his original posit of contraries as principles as well as the suggestion that there should be a third principle. A.6 ends in ἀπορία which A.7 must solve. The first consideration Aristotle raises in favor of positing a ὑποκείμενον as a third principle takes it that there must be something else which can come to have, or to be characterized by the contraries. This is a rather rough intuition, and one which Aristotle justifies by surveying the views of those predecessors of his who posited contraries, and noting that they all seemed sensitive to this intuition, too. In 189a21-26, for example, Aristotle takes Empedocles’ view to illustrate the kind of consideration which was missing from his proposal in A.5. Empedocles, he points out, seems compelled to posit not only Love and Strife, his primary contraries, but also something else for these to affect and “from which they construct the world of nature.” (189a27). This is a curious example for Aristotle to take—Aristotle’s contraries are not such that they need some third thing to act upon, since they are the termini of natural changes rather than the agents of it. Why should the musical and the unmusical, or the pale and the not-pale, need to “act upon” some third thing? To see why Aristotle thinks this is a helpful example, we should read Aristotle as employing the metaphorical language Empedocles did in characterizing Love and Strife as agents of change, while pointing to the idea which the metaphor aims to capture. If Aristotle is taking Empedoclean Love as order, and Empedoclean Strife as disorder, then his appeal to Empedocles is helpful because he sees Empedocles as realizing that there must be some third principle which can be organized or disorganized, or which can take on the

94 characteristic of being organized or disorganized. This is what it would be to be acted on by Love or by Strife, respectively, as in the metaphor. That Aristotle is thinking of the third thing needed as what is characterized by the contraries is borne out by his return to the topic in 189b3-16. In those lines Aristotle points to those who posited some one kind of stuff, from which everything else is made by being differentiated by opposites. For example, Anaximenes proposed that everything was made of air, which is differentiated by density and rarity. Because he is thinking of the third thing as being differentiated by the contraries, Aristotle praises those who propose it to be something which is intermediate, or whose own characteristics will not interfere with its ability to be differentiated by the contraries. He says, for example, that it is best if the third thing is something other than earth, air, fire, or water, “for fire, earth, air, and water are already involved with pairs of contraries.” (189b4-5) Similarly, he says that “the next best choice is air, as presenting sensible differences in a degree less than the others.” (189b7-8) In pointing to these views, however, Aristotle seems to have done little more than show that, on their grounds, it makes sense for them to posit a third thing. At best, he has pointed to a general intuition which seems to have been captured in the different views of principles with which he is in dialogue, a hint that maybe the wise have all grasped something to which he should also attend. Since his own view of what the contraries are and why they are needed is so different from that of his predecessors, it is not yet clear how whatever insight motivating them applies also to Aristotle and to his own account as he has set it out so far. It is for this reason that Aristotle turns to the second consideration he raises in A.6. Here he tries to precisify the insight he pointed to with his appeal to the views of the wise by considering why by his own lights, he might need to posit a ὑποκείμενον as a third principle in addition to the contraries he has proposed. The passage in which he does so is as follows:

We do not find that the contraries constitute the substance of any thing. But what is a first principle ought not to be predicated of any subject. If it were, there would be a principle of the supposed principle; for the subject is a principle, and prior presumably to what is predicated of it. Again, we 95 hold that a substance is not contrary to another substance. How then can substance be derived from what are not substances? Or how can non- substances be prior to substance? (189a27-34)

The appeal to the features of being a substance which were pointed out in Categories, as well as to the concept of ὑποκείμενον, is clear here. In Categories Aristotle had said that no substance was contrary to any substance 3b24-25. Further he had claimed that no substance was in any subject (3a7-8), and no primary substance was said of any subject, either (2a12-13) In this passage Aristotle is thinking about how the proposal that there must be a third principle in addition to the contraries fits with features of his own view, not only with the contraries as he understands them, but with the background of logical concepts to which he has been appealing throughout. As I understand it, his reasoning is as follows. Something is a primary substance according to the definition of being which is presumed or presupposed by every other definition or λόγος which belongs to it. As I argued in my chapter 2, this is what is unique to primary substances as ὑποκείμενα, and what Aristotle is getting at when he claims in Cat. 5 that everything else is ultimately predicated of a primary substance as a ὑποκείμενον (2b6-8). Thus the contraries presuppose and are dependent on a substance as ὑποκείμενον, if they are not themselves substances. However, if the principles of natural change are contraries, then it seems to follow that they cannot be substances, since no substance is contrary to any other substance. A fortiori it follows that no primary substance can be a contrary, so the principles will not be primary substances. This seems to suggest that there must be a third principle, a ὑποκείμενον. Yet there remains a problem. If the motivation for positing a third principle is the need to posit a substance, which is presupposed by the other principles, then the other erstwhile principles, the contraries, seem to lose their claim to the primacy in virtue of which they counted as principles in the first place. For a third thing is needed, a substance as ὑποκείμενον, because the other putative principles depend on this. This opens up a tangle of issues for Aristotle to try to resolve in A.7. He has some reason to think that the contraries need a ὑποκείμενον, for the rest of the philosophers whom he has consulted have all posited one in a way that seems reasonable 96 within their views. Further, Aristotle has some reason by his own lights, for the contraries he has posited seem to need a substance as a ὑποκείμενον. Yet he has gotten himself into a puzzle. Solving this puzzle in A.7 involves getting clearer about what it is for the contraries Aristotle has posited as principles to be contraries. Aristotle clarifies their nature in such a way that he can say, without contradicting his claim in Categories that no substance is a contrary, that a substance could be one of these contraries. It also involves seeing why Empedocles and co. were right to see that there must be something which can be characterized by either of the contraries. Aristotle’s concept of ὑποκείμενον as what is presupposed by something, and the kind of logical complexity this imparts to natural things, is the key to this solution.

97 Chapter 4: Matter as Ὑποκείμενον in Physics A.7

§.1 Terminological consistency between Physics A and Organon

In the course of this dissertation I have been arguing that Aristotle uses the term ὑποκείμενον in the first book of his Physics synonymously with his technical use of it in his Categories and Posterior Analytics. I have argued this by pointing out how Physics A.1 relies on ideas about scientific understanding and knowledge which Aristotle developed at length in his Posterior Analytics, then by showing how he appeals to the notion of ὑποκείμενον from those prior works in his response to the Eleatic monists, especially to Parmenides. Finally, I offered a reading of Physics A.4-6 which shows how Aristotle introduces his notion of ὑποκείμενον when he is developing his account of the principles of change and of changeable things. I argued that he agreed with many of his predecessors and contemporaries that the principles are contraries, but he does so by emending his predecessors’ views about what sort of contraries are principles. Rather than proposing a pair such as hot and cold or wet and dry, or any set of such pairs, Aristotle proposes form and lack, the most general version of the kind of opposition which he thinks is pertinent to understanding change. Aristotle goes on to suggest reasons for proposing a ὑποκείμενον in change, but runs into difficulty; the reasons he has for positing a ὑποκείμενον also undermine his claim that the contraries are principles. This apparent conflict arises because Aristotle is employing his notion of ὑποκείμενον, the one he developed in his An. Po. and also employed in his Categories. I think we can conclude from these considerations that when Aristotle names the ὑποκείμενον as one of the principles of change in A.7, we must understand him to be thinking of something which is a ὑποκείμενον consistently with the sense developed in his logical works. Nevertheless, it may be helpful to respond to a final line of defense at which an opponent might stand. Even if Aristotle has been appealing to logical terms and notions developed in his Organon in the previous chapters of Physics A, one might argue that A.7 marks a re-start in Aristotle’s investigation into the principles of change, such that there is no reason to continue to read ὑποκείμενον as the same technical term going forward. On such a reading, in the previous chapters Aristotle presented a doxography of 98 his predecessors’ (and contemporaries’) views about the principles of the natural world. While he includes some reflections of his own on the δόχα therein, the purpose of A.2-6 is historical, to show that he has done his due diligence before entering his own view into the arena. At best, these chapters provide only preliminary considerations which he sets aside in A.7 to argue afresh for his view about the nature and number of the principles. The opening lines of Physics A.7 seem to provide the best evidence for this. I argue that these lines finally provide no such evidence. If anything they suggest the opposite view, that in A.7 Aristotle takes himself to be continuing and building on the discussion in the preceding chapters. In the Barnes edition of The Complete Works of Aristotle, Hardie and Gaye translate the opening line of A.7, 189b30, as follows: “We will now give our own account…”. Charlton offers something similar in his translation and commentary on Physics A & B, if in more colloquial English: “This is how I tackle it myself.” Both of these translations give the impression that in the previous chapters Aristotle has been rehearsing or assessing the views given by others, while now, at last, he can set them aside to investigate the principles of change and of natural things in his own way. Before he showed other thinkers tackled the issue; now he will show how he “tackles it himself.” Yet these translations are misleading, for the Greek text has only the following: Ὡδ’ οὖν ἡμεῖς λέγωμεν πρῶτον περὶ πάσης γενέσεως ἐπελθόντες. The two expressions quoted above (“We will now give our own account” and “This is how I tackle it myself”) seem to be translating the first two words in the Greek text, ὡδ’ οὖν, while picking up on the first person in the subject and verb of the sentence: ἡμεῖς λέγωμεν. Glen Coughlin offers a different rendering of ὡδ’ οὖν in his 2005 translation of Physics: “So, then, having come to this point,” and continuing the line he translates: “let us first speak about every coming to be.” Hardie & Gaye and Charlton seem to take ὡδ’ οὖν as helping to emphasize the first person plural of λέγωμεν and to mark a strong division between the discussion in the previous chapters and the discussion to come, while Coughlin instead takes the ὧδε as a metaphorical instance of its signification regarding place (LSJ, “hither” 39), supplementing “having come”. Coughlin’s translation softens the implied contrast

39 99 from the Hardie & Gaye and Charlton translations between the preceding and forthcoming discussions, for on his rendering Aristotle is reminding his readers of what has been accomplished so far, before pointing out the next stage of his investigation. One way to determine which of these translations better reflects the transition signified by ὡδ’ οὖν is to figure out whether that phrase serves to emphasize the use of the first person plural in ἡμεῖς λέγωμεν. It will be helpful to look at the kinds of verbal constructions Aristotle used in the more doxographical discussion in the preceding chapters, to see whether the use of the first person in the opening sentence of A.7 marks a contrast. When Aristotle is discussing the views of others earlier in Physics A he tends to use more impersonal constructions (e.g., σκεπτέον, 185b6; δεῖ ,188A.31). These constructions have a normative tone: “one must consider”, “it is necessary to see.” Thus, while in these passages Aristotle avoids the use of the first person, his use of these normative, impersonal constructions still imply a judgment made in his own voice. Hardie & Gaye’s translation as a whole reflects this, for they render each of these impersonal constructions as first person plural verbs in English, in doing so suggesting that even these constructions do not imply that Aristotle is not speaking in his own voice or giving his own opinion about how to understand or assess his predecessors’ and contemporaries’ views. This undermines the strong contrast between Aristotle’s approaches within A.2-6 and A.7 which their translation of 189b30 might seem to suggest. It should be noted, too, that Aristotle does not eschew the use of the first person before A.7. At 187b11-13, for example, he says “for it is when we know [ὑπολαμβάνωμεν] the nature and quantity of its components that we suppose we know [εἰδῶμεν] a complex.” In these lines, that “we” take ourselves to know a complex when we know its parts does not imply that “they”, Aristotle’s philosophical interlocutors, do not. Rather, this is the inclusive “we” which Aristotle uses when he is presenting an insight which he thinks anyone might have upon reflection, or which he thinks all or most take for granted. Nor does this use of the first person mark a strong contrast with other elements of his analysis. The commentary he offers on Parmenides, Melissus, Anaximander, etc., is offered as Aristotle’s own assessment of their views. Thus the use

100 of the first person at the beginning of A.7 does not mark a strong break with the content or approach of the previous chapters. Ross and Charlton both seem convinced of this, despite Charlton’s own translation, for in their commentaries both emphasize the continuity of the argument throughout Physics A. Ross, for example, says that “The whole structure of book i [sci., Physics A], if we eliminate incidental digressions, is the establishment of matter, form, and privation as the factors involved in all change.” (pg. 24) While Charlton presents Aristotle as “reviewing” others’ positions in A.4 (pg. 63), and “criticizing” the views of the Eleatics in A.2&3 (pg. 53), he characterizes Aristotle as “arguing” (pg. 67) or “presenting arguments for” (pg. 65) some claims in A. 5&6, which together result in a “mild antinomy” (pg. 67) that A.7 must solve. Charlton views A.5-7, at least, as a continuous argument for Aristotle’s own view. Agreeing that Aristotle’s argument for these three principles begins early in Physics A, and not at the beginning of A.7, I think a better way of translating ὡδ’ οὖν in the opening line is as follows: “Things being so, then …” This avoids awkwardly trying to bring in the locational significance of ὧδε as Coughlin does, but captures that Aristotle is taking stock of the discussion in the preceding chapters in order to make use of it in what follows. In referencing the preceding discussion Aristotle indicates his willingness both to take on its conclusions and to engage with the puzzles elicited by it. This indication of continuity already helps to show that if Aristotle has introduced his notion of ὑποκείμενον into the discussion before A.7 (in fact in the immediately preceding chapter), it would be surprising if he began, without warning, using the term equivocally in A.7.

§.2 Argumentative Structure of A.7

Seeing A.7 to be continuous with the preceding chapters in this way helps to make sense of how Aristotle pursues the goal he announces in the opening sentence of A.7, to “first address all comings-to-be, for we shall be following the natural order of inquiry if we speak first of common characteristics, and then investigate the

101 characteristics of special cases.”40 The ὡδ’ οὖν reminds the readers to cast their minds backwards over the discussion that brought them to this point, while the programmatic remarks which follow point to how Aristotle intends to continue the discussion. Just as one would expect, Aristotle is not constrained by the terms of the discussion set out by his predecessors, though he does set himself up to benefit by them. Now he will re-cast the investigation in such a way as to make use of what he has discovered from his predecessors and to advance towards his stated goal: discovering the principles of all change. His strategy in A.7 is janus-faced in the way this opening sentence is; Aristotle draws from the preceding discussion, but also brings to bear his own, new, way of considering their content to draw the investigation to a close. The means by which Aristotle pursues this plan is also two-fold. In A.7 he proposes that (a) we think about all changes in terms of a framework of simples and complexes present at the termini of the change, and (b) analyze this framework using logical tools which he developed in An. Po.. The framework gives him a new vantage from which to examine his conclusions from the preceding chapters, and the logical tools provide the means for doing so—the chance to move beyond the impasse at which his reflections on others’ views have left him. In particular, it allows him to clarify what kind of opposition (or what way of being contrary) is important to understanding change, and also how this is consistent with a clarification of why a ὑποκείμενον is indeed needed. The proposal of the framework is part of the backwards-looking theme of A.7, since it arises as a way of putting together the conclusions of the previous chapters (provisional as they might be, in the case of A.6). The forward-looking aspect comes from Aristotle’s analysis of the complex items and simple items of the framework using the kind of predicative analysis he had employed in An. Po. Particularly important for this project is Aristotle’s notion of ὑποκείμενον, as providing structure to the logical complexity which Aristotle uses the framework to point to. Aristotle proposes the framework of simples and complexes as follows:

40 My modification of Hardie & Gaye’s translation. 102 We can say that one thing comes to be from another thing, and something from something different, in the case both of simple [τὰ ἁπλᾶ] and of complex things [τὰ συγκείμενα]. I mean the following. We can say that the man becomes musical, or what is not-musical becomes musical [ἔστι γὰρ γίγνεσθαι ἄνθρωπον μουσικόν], or the not-musical man becomes a musical man. Now what becomes in the first two cases—man and not- musical—I call simple, and what each becomes—musical—simple also. But when we say the not-musical man becomes a musical man, both what becomes and what it becomes are complex. (189b32-190A.5)41

Aristotle presents this framework as a result of his investigations in A.5 & 6; he aims to capture the insights from those discussions, and present them in a way which will enable him to move forward in the enquiry. With this framework, Aristotle points to the beginning of change and the outcome of change, or the terminus a quo and terminus ad quem, and proposes that we see them as having a certain structure. At both termini he points to a pair of simples which together form a complex:

Terminus a quo → Terminus ad quem

ὁ ἄνθρωπος/the man (simple) ἄνθρωπος/the man (simple) + + τὸ μὴ μουσικόν/the unmusical (simple) τὸ μουσικόν/the musical (simple) = = the unmusical man (complex) → the musical man (complex)

Fig. 2: The simples and complexes in change

The simples are ὁ ἄνθρωπος (the man/person), τὸ μουσικόν (the musical/cultured), and τὸ μὴ μουσικόν (the not musical/cultured), while the complexes are τὸ μὴ μουσικὸν

41 Hardie & Gaye translation in the Barnes edition of The Complete Works of Aristotle. 103 ἄνθρωπον (the unmusical man/person) and τὸ μουσικὸν ἄνθρωπον (the musical man/person). The term ἄνθρωπος has the non-gendered sense of “person.” In many translations, especially older translations, this sense is intended to be captured by the generic term “man”, used to refer to an instance of humankind, not necessarily male. While the use of “man” for “person” has fallen out of favor in formal English writing and is now, at best, something of an anachronism, I preserve it here for its continuity with most translations. Similarly, the expression μουσικόν is closer in meaning to the English “cultured” or “educated” than the English “musical.” The sense of the word in Greek is of someone trained in the arts of the Muses, which include poetry and dance, and not only of someone who has musical training in the present-day sense. However, most translations use “musical”, and in continuity with them so will I. A more philosophically significant issue, at least for the present purposes, is that the expression Aristotle uses for the two simples which are opposites is famously ambiguous. Τὸ μουσικόν, for example, could be translated either as “musicality” or as “the musical thing”, while its opposite could be translated as “unmusicality” or “the unmusical thing.” Hardie & Gaye seem to vacillate between these two translations, rendering τὸ μὴ μουσικόν at 189b35 as “what is not musical”, while at 190A.2 as “not musical” (i.e., the characteristic of not being musical). Charlton takes a different approach by trying to preserve the ambiguity of the Greek expressions in his translation. I will follow Charlton’s lead, and translate τὸ μουσικόν, for example, as “the musical” and τὸ μὴ μουσικόν as “the not musical.” The musical and the unmusical Aristotle sometimes calls contraries [ἀντικείμενα] (e.g. 190a18), thus capturing the conclusion from A.5. On the other hand, he also sometimes calls the two complex termini contraries (e.g. 190b20, together with 190b29- 30), seeing the framework as fitting with A.5 in a different way. Similarly, Aristotle calls the complex at the terminus a quo the ὑποκείμενον (as at 190a15, 190a34, and 190b24); since it is what is affected so as to undergo change, in this way it answers two of the considerations from A.6. He also sometimes calls only one of the simples at the terminus a quo the ὑποκείμενον, sci., the one which is not a contrary, as at 190b20, and 190b3- 34together with 191a1-3. The simple which is not a contrary reflects the insight from A.6 that there must be something present to be “acted upon” by each of the contraries so as to 104 be changed. Thus there are two different ways in which the framework seems to capture the conclusions and insights of A.5 & 6, which will be reflected finally in the two different ways in which Aristotle presents his own account of the principles (e.g. 190b35- 36). Yet the positing of the framework does not itself provide the conclusion. In the first instance it only summarizes the results of the previous two chapters. In the second instance it serves to clarify the analysis which Aristotle gives subsequently in A.7, in which Aristotle applies logical tools from his Organon to the framework. Many commentators, such as Bostock, Waterlow, and Charlton, have claimed that Aristotle infers the framework he presents from a survey of the ways that we speak about change. On this reading, in his examination of everyday locutions about change Aristotle discovers a pattern in our ways of speaking, one which reveals which things in the world are present in change, and from which, consequently, an account of principles can be gleaned. Yet Aristotle never claims to have given an exhaustive survey of locutions about change42. Instead, he presents the framework as a synthesis of the conclusions of A.5 & 6, and then proposes that this rough framework is consistent with some common ways of talking about change. That it is consistent with these provides him with some additional support for the account of the principles which he will propose in terms of that framework; at least our usual ways of speaking do not flatly contract the account he will give. More straightforwardly, however, it presents him with a toy example he can use in his analysis: the change of the unmusical man into the musical man. This is supported by looking at the way in which Aristotle introduces his initial conclusion. He says this: “These distinctions drawn, from all cases of becoming this is clear, if someone should look at things just in the way I am saying.”43 (190a13-14) The conclusion is only clear if one looks at things in the way that he has suggested. Aristotle proposes that if we think about change in the way he suggests, and analyze it according to

42 Cf. Code, “The Persistence of Aristotelian Matter”, pg. 360. 43 My translation. 105 the distinctions he has proposed, then we will be able to make sense of it. This is the reason he proposes in 190a13-14 for his reader to accept his analysis.44 Aristotle’s task in A.7 is to use it to spell out the insights of the previous chapters in a more precise and compelling way so as to finally arrive at a consistent and compelling view concerning the principles of change. He does so by extending the kind of logical analysis he had developed and employed in his Organon. The conceptual tools he has from the discussions recorded in those works allow him to re-characterize the complexity of the terminus a quo as its being two in account though one in number, and to be more precise about the way in which there are ἀντικείμενα and a ὑποκείμενον represented within that framework. It leads to this initial conclusion:

These distinctions drawn, one can gather from surveying the various cases of becoming in the way we are describing that there must always be an underlying something [δεῖ τι ἀεὶ ὑποκεῖσθαι], namely that which becomes [τὸ γιγνόμενον], and that this, though always one numerically, in form at least is not one. (By ‘in form’ [εἴδει] I mean the same as ‘in account’ [λόγῳ].) For to be a man is not the same as to be unmusical. (190A.13- 17)45

In 189b32-190A.5, Aristotle concludes less provisionally than he did in A.6 that there must be a ὑποκείμενον in every change46, and posits that this ὑποκείμενον is one in number while being two in form or account.

44 This was first suggested to me in a draft of a paper by David Charles. However, I have since come to realize that Waterlow, for example, in her Nature, Change, and Agency, reads this passage similarly when she presents Aristotle as supporting his account of the principles with the thinking that “a hindrance exists only for those who desire some goal, and they, once a path is cleared, do not require to be further persuaded of the rationality of following it.” (pg. 19) 45 Hardie & Gaye translation in the Barnes edition of The Complete Works of Aristotle. 46 Sheldon Cohen, in his (1984), argues that it is an error in translation (or “overly aggressive translations”, such as Hardie & Gaye’s (pg.182)) which causes scholars to attribute to Aristotle the claim that there must be a single (persistent) substratum (ὑποκείμενον) in every change. According to Cohen, since “the ei kai in line 15 can be taken to be concessionary rather than assertive,” the line should really be read to say that 106 Building on this initial conclusion, Aristotle draws the further and final conclusion that there are in a way two, and in a way three, principles of change (190b35- 36). He first lists the principles of change as subject and form, or ὑποκείμενον and μορφή, at 190b20. He immediately reminds his readers that the ὑποκείμενον is complex, and goes on to say the following:

There is a sense, therefore, in which we must declare the principles to be two, and a sense in which they are three; a sense in which the contraries are the principles—say for example the musical and the unmusical, the hot and the cold, the tuned and the untuned—and a sense in which they are not, since it is impossible for the contraries to be acted on by each other. But this difficulty is also solved by the fact that what underlies is different from the contraries; for it is itself not a contrary. The principles therefore are, in a way, not more in number than the contraries, but as it were two; nor yet precisely two, since there is a difference of being, but three. (A.7. 190b29-191a1)

Aristotle concludes the chapter by saying “So much then for the question of the number and the nature of the principles” (191A.21-22), in this way laying to rest the investigation which he had begun in A.1. He follows this up in A.8 and A.9 by pointing out an advantage of his view—that it solves a paradox which had puzzled other philosophers—and making final remarks on what matter is, in preparation for his discussion of the causes in Physics B. In particular, in A.9 Aristotle calls matter “the primary substratum [ὑποκείμενον] of each thing” (192A.31), for the first time identifying matter and ὑποκείμενον.

there is always some substratum in every change, “and the substratum, ‘even if it is one in number, is not one in form.’” (pg. 182) This allows that there might be more than one substratum, so that there might be “one substratum at the terminus a quo and another at the terminus ad quem.” (ibid) However, even if the ei kai can be taken to be concessionary rather than assertive, this in itself is no reason to do so, and on balance the text of A.7 supports the reading of it as assertive. 190b17-29, for example, seems to summarize the findings of the previous section, and 190b23-24 is not concessionary, but rather assertive. 107

§.3 Logical analysis within the framework of simples and complexes.

On my reading, in A.7 Aristotle is directing his attention to speech in much the same way as he did in Posterior Analytics and Categories. In Categories he asked how the world must be if the basic subject-predicate sentences he was examining are true; in An. Po., if they are knowable. He assumes that when or if we grasp how the world is, our thought (and therefore our speech) about it will conform to or mirror the reality. Yet he does not look simply at what is referred to in each sentence to infer from this what categories or kinds of things are involved. Instead, he constructs an elegant logical framework which aims to capture precisely how the structure of our thoughts limns the structure of the reality it grasps. On my reading, in Physics A.7 Aristotle is now thinking about what must be the case in order for simple locutions about change to be true and knowable. In doing so, he can rely on the background of logical tools and distinctions which he had developed in those earlier discussions, e.g. the distinction between πράγματα and σύμβολα from DI, and his analysis of predication, especially his notion of ὑποκείμενον, from Posterior Analytics. I call this the logical analysis of change. My reading takes as a central text 190a13-17, already quoted in part, which Hardie & Gaye translate as follows:

These distinctions drawn, one can gather from surveying the various cases of becoming in the way we are describing that there must always be an underlying something [τι ὑποκεῖσθαι], namely that which becomes, and that this, though always one numerically, in form [εἲδει] at least is not one. (By ‘in form’ I mean the same as ‘in account’ [λόγῳ] For to be man is not the same as to be unmusical. (190a13-17)

The complexity of what underlies, τι ὑποκεῖσθαι at 190a14-15 and τὸ ὑποκείμενον elsewhere, is that it answers to more than one λόγος. What underlies is the complex at the terminus a quo, such as the unmusical man, or the unformed bronze. That it is the whole complex which Aristotle first calls the ὑποκείμενον is apparent from the fact that 108 he choose the unmusical and man as examples of the two λόγοι belonging to it. In announcing that the ὑποκείμενον is one in number but two in account, Aristotle thus makes more precise the kind of complexity he wishes to ascribe to the terminus a quo. What is it for this initial terminus to be two in account? Or, put in another way, what kind of complexity is this, and in what way are the simples of which it is composed simple? A first possibility is that items like the man or the musical are simple because the expressions which name them are simple, or unitary, while the musical man is a complex because the expression which names it is complex: “musical + man.” This has some appeal if one is already attracted to the idea that what Aristotle is doing is focusing on what we say, where what we say are linguistic items, such as words or sentences. If Aristotle is trying to lay out certain patterns of speech, a relevant detail of those patterns might be the complexity or simplicity of the terms employed. On this reading, when Aristotle says that we say something comes from something else ἤ τὰ ἁπλᾶ λέγοντες ἢ τὰ συγκείμενα at 189b33-34, the simples and complexes which we are saying [λέγοντες] should be read as simple and complex expressions. A translation of 189b32-34 along these lines might be as follows: “For we say something comes to be from something else and something different from something different saying either complex terms or simple terms.” However, as we have seen before, Aristotle often uses λέγω in such a way that what is said is a thing, not a term or a word. Categories 2 is a representative example of this. At 1a16 Aristotle mentions “things that are said” [τῶν λεγομένων], but it is clear from the surrounding passages that these are πράγματα, in the sense spelled out in my chapter 2. What are “said” in this sense are non-signifying items, rather than words or σύμβολα. In Categories 2 Aristotle is talking about beings. Similarly, in Physics A.7 the context suggests that we take them as things, or πράγματα. The immediately following lines help to show this:

We can say the man becomes musical, or what is not-musical becomes musical, or the not-musical man becomes a musical man. Now what becomes in the first two cases—man and not-musical—I call simple [emphasis not mine], and what each becomes—musical—simple also. But 109 when we say the not-musical man becomes a musical man, both what becomes and what it becomes are complex. (189b34-190a5)

At 190a1-3 Aristotle gives as examples of “what becomes” (that is, the thing which changes so as to become what is present at the end of the change) the man and the not musical, and as his example of what these become, the musical. At 190a3-5 the unmusical man is given as what becomes, and the musical man as what it becomes. But of course Aristotle is not suggesting that the words “man” or “unmusical” are undergoing any change, or that the expression “musical man” is the result of any change; clearly the simples and complexes discussed here are not words or expressions but things. If the simplicity and complexity is not the simplicity or complexity of the linguistic expressions, then it might be the metaphysical simplicity or complexity of the items signified by those expressions. Some difficulty with this suggestion appears from the outset. If Aristotle is thinking of man and the unmusical as both being wholly metaphysically simple this conflicts with what is ostensibly the ultimate conclusion of his reflections in Physics A, that ultimately man is metaphysically complex as a composite of matter and form. Further, if one takes the expressions “the musical” and “the unmusical” as signifying the properties of musicality and unmusicality, it is difficult to see why these, or any properties, should be thought of as being metaphysically simple. The property of being rectangular, which belongs to my desk surface, does not seem to be simple, for example. It seems to consist in being four-sided, with two of the sides shorter than the other two, and all the interior angles being right angles. It is not even clear that the property of being musical, or being cultured, as Waterlow translates it, is simple. Does it not have several elements itself, e.g., knowing certain facts, as well as possessing a certain facility in reasoning or inference? On the other hand, if “the musical” and “the unmusical” are understood to signify a musical and an unmusical thing, it is difficult to see what kind of simplicity they could share with the man. One’s attention is immediately drawn to the thought that, ultimately, the musical = man + musical, for there is no musical thing unless there is a man who is musical. This suggests that a greater specificity is needed to grasp what kind of simplicity and complexity Aristotle thinks we should see in the members of the framework. 110 Aristotle characterizes the needed specificity in 190a13-17, where he tells us in what way the ὑποκείμενον is simple, and in what way it is complex. I will call it logical complexity, and logical simplicity. The complexes are complex in the sense in which Aristotle says that the ὑποκείμενον is complex; they have more than one account or form, and thus as a whole a complex account or form. The simples are simple in the correlative way; they have only one account or form. The sense in which the complexes have more than one account, or the simples a single account, is related to the way Aristotle introduced λόγοι in his Categories. In Cat. 1, the first distinction Aristotle draws among beings is concerned with what in Ackrill’s translation is given as “the definition of being which corresponds to the name [ὁ κατὰ τοὔνομα λόγος τῆς οὐσίας].” (1a1-2) This notion of a λόγος corresponding to a name enables him to distinguish between synonyms and homonyms in Cat. 1. Synonyms are beings which share a name, and for whom the definition of being corresponding to the name is the same. For example, a man and an ox are synonyms in this way when they are both called animals, because what it is to be an animal is the same for each of them (e.g., to be a living thing). Homonyms are those beings for whom the definition of being corresponding to the name are not the same, as for example a bat, the baseball implement, and a bat, the winged rodent. This is the sense of λόγος Aristotle has in mind in Physics A.7, too. The ὑποκείμενον is complex in that it has two different definitions of being corresponding to its two different names. Together these form a complex λόγος corresponding to its being the ὑποκείμενον. This is why Aristotle explains the ὑποκείμενον’s not being one in form or account by saying “to be man is not the same as to be musical.” Note the way this statement appears in the Greek: οὐ γὰρ ταὐτὸν τὸ ἀνθρώπῳ καὶ τὸ ἀμούσῳ. This uses a shortened form of the same kind of locution Aristotle uses in An. Po. to express essence: εἶναι + dative. In the Physics passage, Aristotle is claiming that τὸ εἶναι ἀνθρώπῳ is not the same as τὸ εἶναι ἀμούσῳ, even while the man and the unmusical are numerically the same individual. If we read 190a13-17 as I have done above, this fits well with the suggestion that τὸ μουσικόν and τὸ ἄμουσον be taken in these passages not as referring to the properties musicality and unmusicality but instead as referring to the thing which is musical, qua musical, and unmusical qua unmusical. Charlton agrees, and points to scholarly 111 consensus about this at the time he wrote his commentary. (Aristotle’s Physics I,II, pg. 70) The musical counts as simple just in that it answers to one definition or λόγος belonging to the name. Roughly, it is one thing to be musical (and another to be unmusical). In this way the simplicity of the simples in A.7 is correlative to the complexity of the complexes. Further, it is a kind of simplicity equally possessed by the man, qua man, and the musical, qua musical. It does not impart any sense of metaphysical simplicity which would compromise Aristotle’s hylomorphism.47 Here is one way one might think of a “logical analysis of change”, differently from how I propose to do so, but instructive as a contrasting alternative. By dividing the termini of the change into pairs of simples which form complexes, we can see the predicative structure which is present at both termini of the change. At one terminus, for example, the unmusical is predicated of man, while at the other the musical is predicated of man. This allows us to see that, at least in the example Aristotle has chosen, the same thing is the ontological subject at either terminus of the change. Noting that there is ultimately a single ὑποκείμενον involved in that change, since there is a ὑποκείμενον present at the beginning of the change which persists to be the ὑποκείμενον of the state of affairs present at the end of the change, one might generalize to conclude that in every change there is some one thing which persists through the change as the ὑποκείμενον of both termini. What is revealed about change on such a reading is that there is one ὑποκείμενον, that is, one substance, which possesses first one predicate and then another.

47 Kelsey (2008) disagrees. He suggests that hylomorphism is a certain kind of complexity in the definition. For my purposes this might not be a problem. The complexity belonging to the ὑποκείμενον belongs also to the terminus ad quem, and thus reveals the complexity of every generated thing. I see no problem with suggesting that every generated or generable thing shows hylomorphic complexity, and that this is just what makes it to be generable or changeable. The man in Aristotle’s example of the musical man can still count as logically simple in the way Aristotle requires, even if, when it is understood as itself being a generable thing, it must be seen to be complex. That is, there is one thing it is to be a man, and one definition that goes with that name, at the level of precision which is relevant when the man is what persists through the change. But when the man is the result of the change, the level of precision or analysis of the λόγος belonging to the name might be such that it is more appropriate to call the man a complex, and to investigate what goes into his definition. 112 Yet in Physics A.7 Aristotle is no longer trying to show how simple sentences of the form “X is Y” limn reality when they are true. This was the sentence structure which he had relied on primarily in his Organon to develop and communicate the logical framework he presented there. The kinds of claims Aristotle investigates in Physics are such as “X becomes Y,” and “Y came from X”, rather than “X is Y.” One way in which the logical framework developed in An. Po. does not straightforwardly apply to sentences of the sort “X becomes Y” is that these kinds of expressions do not easily reveal ontological subjects and predicates. This, I take it, is part of what Aristotle is pointing to in 190a5-12 and 190a21-33 in his discussion of the framework he proposes. By canvassing the different ways the simples and complexes are slotted into the basic structure of “X becomes Y”, he makes it clear that in none of these locutions does either X or Y obviously stand as the ὑποκείμενον of the other. If, for example, the pair that are put into the X and Y slots are the unmusical and the musical, then Y cannot be the ontological predicate of X when the change is complete, for in that case the X does not persist through the change to be the ontological subject. Yet simultaneity of ὑποκείμενον and predicate is necessary on the kind of subject-predicate analysis in An. Po. The predicate presupposes its ὑποκείμενον in such a way that it is founded on it; the thing in question cannot have the characteristic given by the ontological predicate unless it has the characteristic given by the ὑποκείμενον. In the case of the musical and the unmusical in the change Aristotle is discussing using the framework in A.7, even if it were the case that both were present after the change such that this hindrance to the predication relation were overcome, that ontological predication relation would be an impossibility—a contradiction. Nor can the X be the ontological predicate of Y in this case, for the same reasons. If the complexes Aristotle refers to in that passage are brought into play in the “X becomes Y” framework an added layer of difficulty is brought in. Consider the sentence “the unmusical man becomes a musical man.” The whole complex, “the unmusical man,” cannot be the ontological predicate for “the musical man,” since it does not persist through the change such that they could be simultaneous. Even if the musical man and the unmusical man were simultaneous it is impossible that one should be the predicate or

113 subject of the other. Further, even if it did not result in a contradiction, taking either as the predicate of the other results in man being predicated of man. Of the statements describing change which Aristotle considers, it is only for statements of the same form as “the man becomes musical” that it seems possible to read off the ὑποκείμενον from the locution, and this only if the ὑποκείμενον is recognized from the static depiction of the result of the change: “the man is musical.” An. Po. provides the tools for understanding that the man is the ὑποκείμενον for the musical in the case in which the man in fact is musical, and the fact that this is one of the termini of the change being described may help. However, on its own it is not clear why it should be the case that the statement “the man becomes musical” points to either the man or the musical as a ὑποκείμενον, or that it limns the relation of ὑποκείμενον and predicate at all, since the sentence “X becomes Y” does not point to a static condition. In the logical works, the way in which statements of the form “X is Y” limn reality is by pointing to a static condition and a static relation between the ontological subject and predicate. Locutions of the sort “X becomes Y” depict a dynamic event, something which takes some period of time. Why should any of the items in the dynamic event limned by such a locution stand to each other in the same relation as the static one represented by locutions of the sort “X is Y”? It is not clear that one should expect there to be a ὑποκείμενον of the sort which can be uncovered in “X is Y” statements in “X becomes Y” statements, too. It does not appear that analysis of the sort presented in An. Po. helps Aristotle at all here. Nevertheless, the division into simples and complexes in 189b32-190A.8 paves the way for the logical analysis of change. Since the analysis which fit “X is Y” does not apply straightforwardly to statements about becoming, Aristotle must do some work to show how it does, or how the principles of that analysis also reveal truths about the nature of becoming. He lays the groundwork for his logical analysis by proposing the framework of “simples” [τὰ ἁπλᾶ, 189b33] and “complexes” [τὰ συγκείμενα, 189b33-34] which he lays out in 189b32-190A.8. This framework enables him to clarify the logical relationship between what is present at the beginning of the change and at the end of the change, and in so doing to reveal the structure of every change. With this strategy he hopes to finally unveil the principles of change, and thus of changing things. 114

§.4 Philosophical significance of the complex ὑποκείμενον

In this section I argue that in introducing ὑποκείμενον as a principle of change in the way he does Aristotle is able to clarify and harmonize the insights from A.5&6 to develop his notion of matter. To do this, I will start by explaining the importance of reading Aristotle as counting the complex at the terminus a quo as the ὑποκείμενον. First, some texts. At 190a15 and 190a34 it is the whole γιγνόμενον, the complex at the terminus a quo, which is said to underlie. Again at 190b24 Aristotle seems to be thinking of the whole complex as the ὑποκείμενον, for he points out again that the ὑποκείμενον is one in number but two in form. In the previous lines (190b20-23) he even explains further what this means, when he points out that “the musical man is composed in a way [ἐξ ... τρόπον τινά] of man and musical: you can analyse its definition into the definitions of these.”48 On the other hand, the use of ὑποκειμένου in 190b20 is clearly a case in which the persistent simple is called ὑποκείμενον. I think this use of the term helps to highlight both the connection with ὑποκείμενον from An. Po. and what has been discovered through Aristotle’s reflections on change in Physics A. Here are those lines in their context:

Plainly then, if there are causes and principles which constitute natural objects and from which they primarily are or have come to be—have come to be, I mean, what is said to be in its substance, not what each is accidentally—plainly, I say, everything comes to be from both subject and form [μορφὴς]. For the musical man is composed in a way of man and musical: you can analyse its definition into the definitions of these [διαλύσεις γὰρ [τοὺς λόγους] εἰς τοὺς λόγους τοὺς ἐκείνων]. (190b17-23)

Here the form [μορφή] is one of the simples from the terminus ad quem. The subject [ὑποκείμενον], however, is present at both termini. Clearly Aristotle is counting

-48 My modification of Hardie & Gaye translation. 115 as ὑποκείμενον the simple which persists through the change. The kind of analysis given in An. Po. in which the notion of ὑποκείμενον is developed seems to be concerned primarily with static states of affairs rather than with changes or dynamic states; all the examples have to do with locutions of the sort “X is Y.” If one thinks about the musical man in the way Aristotle suggests in An. Po., then one can see that the man is the ὑποκείμενον, the ontological subject, of the musical. After all, it is only because he is rational that the man can possess musicality as a property, or that the musical can be one in number with the man; his musicality is predicated on his rationality (his being a human). As was argued in chapter 2, the subject or ὑποκείμενον in An. Po. is the thing which, by being just what it is (having the definition belonging to it according to the name) is suited to take on the predicate. To think of the ὑποκείμενον in this way, however, is to see it as something which is susceptible to change, in fact something which is such that it can change in the very way required to take on the ontological predicate. In other words, the ὑποκείμενον picked out in An. Po. will also be something which persists through change. This helps to show why Aristotle calls what persists through the change in the framework in Physics A.7 the ὑποκείμενον; it persists through the change not because it is a substance but because it is the ὑποκείμενον of the outcome of the change. As such, it must pre-exist that final terminus. Thus the kind of complexity pointed out by Aristotle’s analysis of change in An. Po. just is the kind of complexity belonging to changeable things. The ὑποκείμενον of the state of affairs present at the end of the change is also the ὑποκείμενον which is a principle of the change. In An. Po., however, Aristotle had not presented the ὑποκείμενον as being complex, or as having a complex or two-part λόγος. That he does in Physics A.7 does not reveal a point of philosophical inconsistency, but rather an attempt to further clarify the nature of the ὑποκείμενον as what is presupposed, and to show how it is needed in the explanation of change. Sometimes, as at 190a17-20, Aristotle characterizes one of the simples at the terminus a quo as a contrary (ἀντικείμενον, a18), and the other simple as what is not a contrary; under one account the ὑποκείμενον is a contrary, and under another account it is not. Yet both accounts are needed to clearly see what thing is the ὑποκείμενον for the outcome, and why it counts as this.

116 On the one hand, that the ὑποκείμενον is the man (what is not a contrary) points to the characteristic which is presupposed by the outcome. There must be a rational thing for there to be a musical thing. Yet to see how the man is suited to take on the predicate, the musical, it is important not only to think about what it is to be a man, that is, about the rational, in himself, but to realize that, as rational the man is such as to be or to become musical. To be a ὑποκείμενον is to stand in a certain relation—to be what is presupposed by some predicate. The terms in which the ὑποκείμενον is given, such as man, or log, are not themselves relational terms, and do not of themselves imply a relation. Explaining the change requires seeing that they are related to, that is, presupposed by, the outcome. Attending to the second λόγος of the complex ὑποκείμενον accomplishes this by directing the mind to the thing which the ὑποκείμενον has not yet become, by negating it. Simply pointing out that there is a human does not clearly indicate that there is present the necessary precondition for the change which is becoming musical; it must also be pointed out that the human is unmusical, otherwise there is no change. Put another way, while, as I argued in the previous chapter, the kind of opposition that Aristotle thinks characterizes the contraries is not that of a thing and its opposite, but to emphasize that it is a kind of contrariety for most of his examples Aristotle relies on that way of naming them: musical & unmusical, pale and not pale. Yet Aristotle points out that for some cases “the opposite state has not received a name, so we fail to notice the fact [that there are contraries involved].” (188b10-) Naming the contraries using x and not-x kinds of locutions helps us to see that they are contraries. Thus positing the second λόγος as Aristotle does points to the kind of opposition which Aristotle thinks is key to understanding any change, namely, the opposition which pertains to form and lack. It is important to see that the whole complex at the terminus a quo is the ὑποκείμενον for the change, for the two λόγοι under which it is grasped each provide an important piece of information about the change. One might worry that positing the second λόγος, the one which clarifies that the ὑποκείμενον is a contrary, is in tension with the definition of the ὑποκείμενον from An. Po. There the ὑποκείμενον was given as being what is presupposed by or suited to take on some property, just by being what it is. If “what it is” is the λόγος according to which the ὑποκείμενον persists through the change, then the λόγος by which it becomes clear 117 that it is a contrary seems something additional. If it is required in order to see how the thing which is the ὑποκείμενον counts as a ὑποκείμενον in A.7, this might seem to suggest that the thing is a ὑποκείμενον not just by being what it is, but by being what it is plus something else. But this is to misunderstand the way in which two λόγοι of the complex at the terminus a quo are related, and the sense in which there are and are not two principles present at that terminus. Even in A.5 Aristotle had suggested that the contrary at the terminus a quo being named according to the not-x pattern, or by a name which clarifies that it is a contrary, is not important. What is important is seeing the thing as being related to the outcome of the change by the appropriate kind of opposition. There are two ways in which the second λόγος can be seen to enable this. On the one hand, we can think of the two λόγοι as belonging to different items which are accidentally one in the complex at the terminus a quo, e.g., the unmusical and the man. Then, since the item which is the contrary does not persist (“the unmusical”), then the ὑποκείμενον will be just that item which does persist, and whose λόγος is such that it is presupposed by the outcome. Taken in this way, there are three principles: ὑποκείμενον, form, and lack. The ὑποκείμενον will count as ὑποκείμενον just according to the λόγος by which it also persists. Still, seeing that it was one in number with the contrary helps to explain why it could become what it does in the terminus ad quem, so the second λόγος is an important part of explaining the change. On the other hand, we might see the two λόγοι of complex at the terminus a quo as really spelling out parts of the same λόγος; the one giving the λόγος which is presupposed by the λόγος of the outcome, the other clarifying that this λόγος is to be grasped inasmuch as it is related to the outcome in the right way. The relation is that of being the ὑποκείμενον; it is also the way of being opposed according to the kind of opposition described in A.5.49 Considered this way, the whole complex at the terminus a

49 Arguably Aristotle distinguishes when he is thinking of the principles in a way conducive to enumerating them as three, and when as two, throughout A.7 by the way he names the contrary which is the lack. For example, when he first proposes the framework he uses the expression τὸ μὴ μουσικόν, “the not musical,” as at 189b35. He uses this expression again at 190a1, a2, a11, and a19. However, in 190a19-20 he says τὸ μὴ μουσικόν δὲ καὶ τὸ ἄμουσον. He uses τὸ ἄμουσον, “the unumusical,” at 19021, a23, a28-30, b14, and b31. At 190b31 Aristotle is calling one of the two principles (when he is enumerating the principles as 118 quo is the proper ὑποκείμενον, and in treating its λόγος as complex in the way that Aristotle, does Aristotle is trying to show how by being a man (or being some x) something can also be a ὑποκείμενον, that is, bear the right kind of logical relationship to something else. For to be a man is not the same as to be a ὑποκείμενον, even if it is by being a man that something can be a ὑποκείμενον. This is why Aristotle concludes from this that “if there are causes and principles which constitute natural objects and from which they primarily are or have come to be [ἐξ ὧν πρώτοων εἰσι καὶ γεγόνασι] … plainly, I say, everything comes to be from both subject and form [ἐκ τε τοῦ ὑποκειμένου καὶ τῆς μορφῆς].” (190b17-20) On this way of thinking about the case, there are two principles: a pair of contraries, such as Aristotle had argued for in A.5. Here ὑποκείμενον, properly understood, is standing in as the lack. Or, rather, the στέρησις, properly understood,is the ὑποκείμενον. Of course, the thing itself which is the ὑποκείμενον might have even more λόγοι, corresponding to more names which may be truly ascribed to it. In addition to being a man and not-musical, the thing which is the ὑποκείμενον might be humorous, or swarthy, or rather tall. Pointing to these does not help to explain the change by which it becomes musical however. This was part of Aristotle’s contention in A.4. To say that the tall becomes musical does not provide any insight into the change, for there is no particular reason why a tall thing should, or should not, become so. The λόγος belonging to something as tall is not that which is related to its being musical in the kind of opposition depicted in A.5; it is not that by which it stands as ὑποκείμενον to the musical. There are only two λόγοι which are relevant to the change, and which together pick out the ὑποκείμενον: the λόγοι belonging to it insofar as it is a man and insofar as it is unmusical. The trick for explaining any particular change is to be able to determine which two names and corresponding λόγοι the ὑποκείμενον comprises.

two), τὸ ἄμουσον, just after he has called τὸ ἄμουσονan opposite or ἀντικείμενον. One gets the sense that when Aristotle is thinking of the simple, the item whose λόγος, such as it is, is its not being musical, he uses τὸ μὴ μουσικόν, the not musical. This is what we must realize is accidentally one with the man in order to understand the change, when we count the principles as three. The unmusical, τὸ ἄμουσον, by contrast, imports the full notion of a contrary which was given in A.5; it compares to “the not musical” the way that “untuned” compares to “not tuned”. 119

§.5 Substantial change

Though Aristotle has presented his plan as being to discuss what is common to all changes before turning to the features of particular kinds of changes, the fact that the example which he relies on to illustrate his account is a case of change of accidents might seem to belie this. Aristotle himself is aware that it may not be obvious how his analysis applies to cases of change of substance. One might reasonably wonder, for example, how to fill out the framework that he proposed if it is not the substance which persists through the change. For this reason, in 190b1-10 Aristotle takes a little time to show how his analysis fits with the less obvious case of substantial change. Here cases of substantial generation and destruction are treated as possible counterexamples to his analysis, counterexamples which he wishes to refute: “But that substances too, and anything that can be said to be without qualification, come to be from some underlying thing, will appear upon examination.” (190b1-3) To see how Aristotle deals with this counterexample, it will be helpful to pause for a moment to see how Aristotle organizes or classifies the variety of kinds of changes. There are several ways one might distinguish among kinds of changes. One division separates substantial change (change of substance: the coming to be and destruction of substances) and accidental change (change of accident: something coming to have some property or characteristic which is not constitutive of a substance). If this is what Aristotle has in mind, then in the opening lines of A.7 he is proposing to uncover what is common to both substantial change and change of accidents (περὶ πάσης γενέσεως) before addressing in another place what is proper to and distinguishes each kind (περὶ ἕκαστον ἴδια). Another way of dividing up the kinds of change is to divide them into changes with respect to each of the categories, for example, change of substance vs. change of quality, vs. change of quantity, etc. On this view he has taken an example from change of quality, but implies that change in each of the other categories besides substance easily fits into the analysis he has proposed. Again, another way to divide up changes is into kinds such as locomotion, change of shape, growth, diminution, addition, and putting together. 120 Clearly these different groupings of kinds of changes overlap in certain ways. For example, growth, diminution, and addition all seem to be instances of change with respect to the category of quantity. Further, one is tempted to group locomotion, change of shape, etc., as well as change of quality, quantity, place, and the rest under this way of dividing up change, as being different kinds of change of accident, vs. substantial change. If this is right then the primary distinction between kinds of change is between change of substance and change of accidents, and the other divisions mentioned are really subdivisions under change of accident. But a better reading of the text shows that for Aristotle the different ways of distinguishing the kind of change are really all orthogonal. Each of the kinds of change of the latter two divisions might be cases of accident change or of change of substance. Evidence that Aristotle holds this comes from 190A.31-190b10, where he briefly discusses both ways of dividing up the kinds of change, and points out that substantial coming-to-be can be accomplished in many ways.

Things which come to be without qualification, come to be in different ways: by change of shape, as a statue; by addition, as things which grow; by taking away, as the Hermes from the stone; by putting together, as a house; by alteration, as things which turn in respect of their matter. It is plain that these are all cases of coming to be from some underlying thing [ἐξ ὑποκειμένων γίγνεται]. (190b5-10) 50

The suggestion is that each case of substantial generation (or destruction) is also an instance of one of these other kinds of change, such as change of shape, or growth, or even alteration.51 If there is no problem fitting each of these into the framework of

50 Translation from the Loeb edition. Note that at 319b26 Aristotle uses the term ἀλλοιώσις to name a kind of change which is not substantial coming to be. When Aristotle uses the same term in Physics A.7 at 190b8, he is pointing to a kind of change which could be an instance of substantial change. 51 This provides some support for Code’s reading in his “The Persistence of Aristotelian Matter” that Aristotle thinks that every change which is a coming-to-be of something is also the alteration (that is, the modification, not necessarily the change of quality) of something else. 121 simples and complexes from A.7, then there is no problem with substantial change, either. The case of air becoming water, or alteration, he puts into this framework himself in DGC A.4. In Generation of Animals (henceforth, GA) A.20-23 he fits the example of the sperm and menses producing the embryo into the framework of simples and complexes, too. In Physics A.7 Aristotle does not go into detail; he seems to think it will be obvious as soon as he points to these various kinds or instances of substantial coming- to-be that they will all fit the analysis he has provided. Thus analysis of each of these shows that there must be a ὑποκείμενον in every change, a ὑποκείμενον which is one in number and two in form. For example, according to Aristotle, an embryo comes to be from the female’s menses and from the male’s sperm in that sperm imparts a certain movement and heat to the menses. (GA A.22 73b9-31).In the framework from Physics A.7, this means that the complex at the terminus a quo for the coming to be of the embryonic human is the menses + a certain lack of heat and movement. The sperm is a cause as efficient cause. It brings about the contrary by imparting the appropriate heat and movement to the menses. Thus if one looks just at the case of the terminus a quo, statically, as it were, its ὑποκείμενον is the menses, for one can only get this thing, the precondition for the coming to be of the embryo, from menses + the relevant lack. Once the semen imparts the appropriate motion and heat to this ὑποκείμενον (the menses lacking heat and movement), the embryo comes to be. The complex at the terminus a quo is the unheated and unmoving menses, and the complex at the terminus ad quem is the heated and moving menses, that is, the embryo. .52

52. Aristotle usually characterizes the matter of a human, or any living thing, as its body. How does this fit with what I have just described? In GA Aristotle suggests that the first result of the motion and heat imparted by the semen is the formation of the heart, which then sustains the motion of the heat so that the rest of the menses is transformed, gradually, into the other organs of the body. This suggests that, for Aristotle, there is a body as soon as there is an embryo in that the heart begins to develop. Yet if the body of an animal is something with organs, it looks like this will not pre-exist the animal. Jennifer Whiting provides a possible solution to this in her “Living Bodies”, where she suggests Aristotle distinguishes between generative and constitutive matter. The scholarly discussion of Aquinas’s purported “delayed homonization” might also provide some guidance here, by transforming the discussion into a question of 122 In Physics A.9 Aristotle makes explicit that the ὑποκείμενον of A.7 is the material principle in change and natural things. He does so when he says “for by matter I mean that primary underlying thing in each case, out of which as a constituent and not by virtue of concurrence something comes to be [τι ἐνυπάρχοντος μὴ κατὰ συμβεβεκός]”53 (192a31-32) By “primary” in “primary substratum” here Aristotle does not mean “primary” in the sense of “ultimate” as in the expression “prime matter.” Instead, it implies the proximate or immediate substratum, the one immediately underlying the change or the μορφή at the outcome of the change. One reason for taking “primary” here to point to proximate matter is that Aristotle has as yet done little more than pave the way for that multi-level subject-form or matter- form analysis which he will employ so casually elsewhere in his corpus, as in the opening chapter of DGC. His focus has been on the ὑποκείμενον which immediately underlies the change or the terminus ad quem; in fact, this has been the only item to which he has applied the term. Thus he has not yet offered any other ὑποκείμενον with which that one might contrast as being primary. For this reason, I think both that what counts as primary subject in the sense Aristotle has in mind here is not some subject arrived at through further analysis beyond the immediate subject. It also indicates that primary here is not meant to contrast with those subjects which might be arrived at by a further level of analysis. On the other hand, given that what counts as ὑποκείμενον in A.7 is either the complex or either of the simples which comprise it, and it is the complex which Aristotle first calls the ὑποκείμενον and which seems to justify the simples also being so called, the sense of “primary substratum” is perhaps best construed as being the thing which primarily or first deserves the name of ὑποκείμενον, or which best fits what it is to be a ὑποκείμενον. The contrast between primary and secondary that Aristotle is implying might reasonably be read as being between the complex and the two simples, rather than between the immediate substratum of a change and that substratum’s substratum.

how to understand the developmentod Aristotelian matter. Haldane and Lee provide a bibliography in footnote 3 of their “Aquinas on Human Ensoulment, Abortion and the Value of Life”. 53 Charlton’s translation. 123 If this is what is meant by “primary substratum”, then matter is that principle involved in change which Aristotle presented through his analysis of the framework of simples and complexes in Physics A.7. It is that principle which in A.7 he has explained using his notion of ὑποκείμενον from An. Po. It is the complex at the terminus a quo, inasmuch as this both gives the necessary precondition for the outcome of the change and is itself the sort of thing which has a tendency toward that outcome. It is also that which is present in the outcome, at least under one of its two λόγοι. Because of this, it imparts a complexity in λόγος to the outcome of the change. It pre-exists and is present in [τι ἐνυπάρχοντος] the outcome of the change. Matter is the ὑποκείμενον.54

54 Despite Aristotle’s identification of matter and ὑποκείμενον in A.9, Sheldon Cohen suggests that we should “drive a wedge between the notions of a persisting matter and a persisting substratum, to allow for cases in which form, not matter, is the more rightful substratum.” (pg. 173) His main textual support for the position that form is sometimes the persisting substratum is his claim that in Metaphysics Z Aristotle says that “form...has a greater right than matter to be taken to be the substratum” (pg. 188), and that in Physics A.7 190b24-30 “Aristotle says that the substratum comprises both matter and form” (pg. 189). The interpretation of the text of Z he suggests here is not obvious, as indeed is no interpretation of that text. For this reason, I will set aside Z to avoid arguing from he less clear. But the text Cohen cites from Physics does not seem to help his case very much, either. If I am right about how to understand Aristotle’s characterization of the persisting matter in A.9, there is a wedge between the persisting subject and matter; the persisting subject is not the primary subject, and thus counts as matter derivatively. It counts as matter the same way as it counts as ὑποκείμενον; derivatively, by leaving out part of the complex λόγος which belongs to matter. Still, Cohen might point out that since on my reading both of the simples which comprise the ὑποκείμενον can be called by that name, what he is calling the “form” of the ὑποκείμενον will also count as matter derivatively. Even if this is the case, the natural reading of Aristotle’s examples suggests that it is never the “form” element of the ὑποκείμενον which persists, but the matter. To see this, note that the first time Aristotle uses the term ὕλη in Physics A is at 190b25, where it appears in the phrase ὕλη ἀριθμητή, or “countable matter”. Aristotle says that

Now the subject is one numerically, though it is two in form[ἔιδει]. (For there is the man, the gold—in general, the countable matter; for it is more of the nature of a ‘this’, and what comes to be does not come from it accidentally; the privation, on the other hand, and the contrariety are accidental.) And the form [ειδος] is one—the order, the art of music, or any similar predicate. (ref)

124

§.6 Some alternative readings, and some responses

It is generally proposed in scholarship on A.7 that both the framework of complexes and simples Aristotle proposes and his subsequent proposal for the principles of change are inferred from a survey of ordinary patterns of speech about change. Though the readings which follow this line of thought vary among themselves in how this inference works and what additional reasoning, if any, it depends on, there is a common thread. Each of these readings takes Aristotle’s approach to be the analysis of speech about change for the sake of uncovering a universal pattern which would reveal what kinds of things are always referred to, and thus always present, in a change. Seeing how my approach differs from this will help to clarify how I think Aristotle argues in A.7, so I will begin by presenting in outline three prominent interpretations of Aristotle’s reasoning in A.7, each of which takes his argument to derive from reflections on ordinary patterns of speech about change. Bostock, for example, says of the passage 189b32-190A.8 that “it is perfectly clear throughout these passages that Aristotle is concerned to comment simply on the way we speak, and to show that it conforms to a general scheme.” (“The Principles of Change,” pg. 5) Some textual support can be found for this. Hardie and Gaye

In this passage, it is clear that the opening claim re-iterates what Aristotle had said at 190b13, where he clarified that by two in form he meant two in account. In the section of the translation which Hardie & Gaye have in parentheses, Aristotle clarifies and comments on the two accounts. Ostensibly, the two accounts belonging to the ὑποκείμενον which he gives in this passage are the countable matter and privation. These are the two forms (190b24). The man and the gold are “countable matter” inasmuch as they are more τόδε τι ((xx), but even as countable matter they are still one of two forms. In the examples he uses to illustrate his point, it is the man or the gold which persists through the change. Thus in the examples Aristotle uses in Physics A, it is only the “countable matter” that persists through the change. The man persists through accidental changes; the gold persists through the substantial change which is the coming to be of the statue. In none of the examples does the privation persist, nor could it logically do so. Thus the textual support for the claim that Aristotle allows form to persist as substratum or ὑποκείμενον even in Physics does not hold up.

125 interpolate “we can say” into their translation 189b34, but the first half of A.7 is dotted with conjugated versions of λέγω and φήμι (e.g., φάμεν in 189b32, λέγοντες in 189b33, φῶμεν in 190A.4, and λέγεται 190A.6). Thus it is tempting to read the whole passage as being concerned with how we would ordinarily talk about change, as many, including Bosock, do. On Bostock’s view, this linguistic analysis is complemented by a sort of conceptual analysis. According to Bostock, Aristotle deduces from his review of the ways that we generally speak about change the concepts which we always employ in describing or conceiving of change, and from this infers that what they signify are always present in change. In this way Aristotle concludes that matter (or ὑποκείμενον), form, and lack are always present in change. In my previous chapter, chapter 3, I argued against Bostock’s claim that the form and lack which Aristotle posits as being present in change are nothing more than a thing and its opposite, or x and not-x. There is also a difficulty with how Bostock understands ὑποκείμενον in Aristotle’s proposal. Bostock reads ὑποκείμενον as doing “double duty” (pg. 9) in A.7. Bostock says that for most of A.7 it signifies the subject of predication, synonymous with its usage in Aristotle’s logical works and implying a substance. It also has the separate signification of whatever persists through change; eventually, this is the only signification Aristotle preserves for the term in his account of principles. That is, on Bostock’s reading Aristotle begins the chapter by using a familiar technical term from his own logical works, but in his conclusion shifts the signification of that term so that it no longer means the subject of predication, but has instead only a new technical meaning: whatever persists through change. The ὑποκείμενον which is a principle is not the ontological subject, but whatever persists through change, just insofar as it does. Bostock’s suggestion that ὑποκείμενον as a principle of change is just whatever persists through the change, and not the ontological subject, is motivated by a difficulty he finds in interpreting Aristotle’s argument. When Aristotle concludes that in a way there are three principles of change (190b30), one way of taking this final trio is as ὑποκείμενον, form, and privation.55 For Bostock, this would mean that the three

55 Here is the relevant passage: “There is a sense, therefore, in which we must declare the principles to be two, and a sense in which they are three; a sense in which the contraries are the principles—say for 126 principles are substance, form, and privation, as he points out (“The Principles of Change,” pg. 14).56 But, as Bostock points out, this is in conflict with 191A.19-20, where Aristotle says that it is not yet clear whether, of the principles, form or the ὑποκείμενον is substance. Bostock points out that, on the one hand, we might avoid attributing to Aristotle this contradiction by allowing that he is revising “the doctrine of the Categories on predication, to say that there are subjects of (accidental) predication which are not substances.” ((“The Principles of Change,” pg. 14). That is, Aristotle does not contradict himself within A.7 if he changes his mind about the doctrine that all ὑποκείμενα are in the category of substance. Then it would not follow from Aristotle’s posited three principles that he may as well enumerate them as substance, form and privation, thus leaving open the question of 191A.19-20. Bostock dismisses this suggestion as proposing too radical a break with Aristotle’s prior work. He proposes a different solution. In his preferred solution, Aristotle uses ὑποκείμενον to signify both whatever persists through change and the subject of predication (it has “two senses” (pg. 14), or does “double duty” (pg. 9)). It signifies both these things in A.7 until the end, when ὑποκείμενον is named as a principle. Bostock argues that there, in his conclusion, Aristotle narrows down the term so that only ὑποκείμενον as what persists serves as a principle of change. If Aristotle uses ὑποκείμενον deliberately ambiguously in A.7 in the way Bostock says, his abandoning the logical signification of the term at the end is insufficiently motivated. More precisely, it is motivated from an interpreter’s point of view, but not from the author’s. The conflict Bostock points out is an interpreter’s conflict. Working with one understanding of how the term is used throughout the argumentation of A.7, the

example the musical and the unmusical, the hot and the cold, the tuned and the untuned—and a sense in which they are not, since it is impossible for the contraries to be acted on by each other. But this difficulty is also solved by the fact that what underlies is different from the contraries; for it is itself not a contrary. The principles therefore are , in a way, not more in number than the contraries, but as it were two; nor yet precisely two, since there is a difference of being, but three. For to be man is different from to be unmusical, and to be unformed from to be bronze.” (190b29-191A.3 56 Note that here Bostock assumes that ὑποκείμενον and substance are coextensive, so that to posit a ὑποκείμενον is to posit a substance; I have already argued against this view in my chapter 2. 127 interpreter comes to see that this is quite clearly in conflict with a claim Aristotle makes at the end of the passage. Unless the interpreter concludes that Aristotle made an obvious error, this would seem to argue that the understanding of the term which the interpreter has been working with has been mistaken, for it does not fit all the textual evidence. But instead of this, Bostock argues that Aristotle changes his mind about how to use the term. No philosophical motivation within the text is given for this change. Bostock suggests no principle according to which we could see Aristotle divorcing his use of ὑποκείμενον as subject of predication from his use of it as what persists, to show when the reasoning he employs applies only to one and not to the other, and it is difficult to see what such a principle would be. In the light of my chapter 2, I think the move for the interpreter to make is to take Aristotle as allowing that something can be the ὑποκείμενον without itself being a substance; but I think this is consistent with Aristotle’s treatment of ὑποκείμενον in Categories. As I argued in chapter 2, in Categories Aristotle presents several examples of ὑποκείμενα which are not substances; he does not hold that all or only substances are ὑποκείμενα. A further, textual difficulty with Bostock’s reading is that it glosses over the fact that what Aristotle first calls the ὑποκείμενον is the whole complex at the terminus a quo, rather than the persistent simple. When Aristotle introduces ὑποκείμενον in A.7, he does so saying that “it is necessary that there is always something underlying [τι ἀεὶ ὑποκεῖσθαι] what comes to be, and this is one in number, but not one in form (for by ‘in form’ [ἔδει] I mean the same as ‘in account’ [λόγῳ]), for to be for the man and for the unmusical are not the same.57” (190a13-17) What underlies in this way in Aristotle’s pet example is the unmusical man. This means that the item in the framework which Aristotle is calling the ὑποκείμενον is the complex at the terminus a quo. The complex is also what Aristotle calls the subject (ὑποκείμενον) at 190b23-25, when he repeats that the ὑποκείμενον is one in number but not one in form. Seeing that Aristotle calls the complex the ὑποκείμενον is necessary for seeing why Aristotle should allow that, on a

57 My translation. 128 final count, there are in a way two and in a way three principles, as I will argue in the next section. That Aristotle is willing to count the complex at the initial terminus as the ὑποκείμενον undermines Bostock’s claim that to be the ὑποκείμενον as a principle in change is just to be whatever persists through change. Just a few lines before the above- quoted text, at 190a10, Aristotle had pointed out that the man persists through the change when a man becomes musical; man is ὑπομένον. But ὑπομένω should not be taken as communicating the same as ὑποκεῖσθαι in these lines. Aristotle reiterates that the man persists through the change at 190a17-21, yet this is clearly in the context of explaining the way in which what underlies, the ὑποκείμενον, is not one: “For to be a man is not the same as to be unmusical. One part survives [ὑπομένει], the other does not: what is not an opposite survives (for the man survives), but not-musical or unmusical does not survive, nor does the compound of the two, namely the unmusical man.” In these lines Aristotle has in mind the sense in which what underlies is not one, and is distinguishing between its accounts or forms. As man, the ὑποκείμενον persists, while as the unmusical, it does not. That is to say, in a way the ὑποκείμενον persists, and in a way it does not; to be a ὑποκείμενον in this context cannot simply be whatever persists through change. This suggests that if Bostock is right about how Aristotle uses the term when he states his conclusion about the principles of change, Aristotle must have introduced the new technical meaning for it somewhere else in the discussion. There is no clear place in the text where he does so. Ross proposes that Aristotle begins A.7 with the intention “to establish more securely and more definitely the triad of principles already reached” (Aristotle’s Physics, pg. 22), and that he does so by appealing to the three ways of “describing” (ibid) a change. On Ross’s reading, similarly as in Bostock’s, Aristotle infers from his analysis of the ways we speak which kinds of things or which elements are always involved in change. Unlike Bostock, Ross does not emphasize a conceptual level of analysis. Rather, he suggests that Aristotle arrives at his proposal for the principles by thinking about which of the different ways we speak about change describes the “complete fact.” (pg. 22)

129 Waterlow’s reading is very close to Ross’s, but goes further into detail than Ross does in his commentary to offer an explanation on Aristotle’s behalf for why Aristotle might (reasonably) prefer one of the ways of describing change which he points out over the others, or why only one presents, in Ross’s locution, the “complete fact.” On her view, in A.7 Aristotle develops his account of the three principles of change in part as a way of overcoming or solving a paradox which had stymied his predecessors. She presents the paradox as follows:

On the one hand, let us then say, if X comes into being, X could not have existed before, since X would then be nothing new in the world, and would not have come to be. But on the other hand, it is impossible to accept that the X which comes into being was not somehow rooted in what went before: but in what? If we say: in things other than X, this leaves us no better off than if we had said: in nothing at all. … So if X comes into being but is at the same time rooted in what went before, it cannot have been rooted in nothing, nor in what was other than itself; so the only remaining possibility is in itself: but not in its own prior non-existence…; therefore in itself as previously existing. Thus it must have been before coming to be, and therefore did not come to be. (Nature, Change, and Agency, pg. 9)

On Waterlow’s reading, 189b32-190A.8 presents three ways of describing change: (1) the man becomes musical, (2) the not-musical becomes musical, (3) the not- musical man becomes a musical man. According to Waterlow, the three ways of describing change “show that the paradox is by no means a recherche one which reveals itself only to those who abandon ordinary speech for some esoteric terminology.” (pg. 15) Option (2), if relied on as a proper or best way of describing the change, suggests that the coming to be of the outcome of the change, the musical, is rooted in its own prior non-existence, as though its non-existence were what primarily explained its coming to be. From (3) it appears that one can derive the description that the man becomes a man, which, if it were allowed as an apt description of the so-called change, would suggest 130 there finally is no coming to be. Because of this, Waterlow argues that the paradox of change is only solved if one way of describing the change is preferred, namely, (1) “the man becomes musical”. Expressions of type (1) show the solution to the paradox because they highlight what “does the becoming” (pg. 18) in the proper way, placing it as the subject of the sentence describing the change:

So Aristotle is saying that what does the becoming should properly be described as what it was, is, and will be before, through, and after the change, i.e. as ‘man’, although it would not be false to describe it as not being what it will become, i.e. as ‘not cultured.’ And he is saying that what does the becoming should properly be described as becoming what (as yet) it is not (cultured), although it would not be false to say that it comes to be what it is (a man; since what it comes to be is a certain sort of man—a cultured one). (pg. 18)

Once one sees that (1)-(3) all describe the same change, one can see why each gives a true description of the change, but only (1) describes the change in such a way as to dissolve the paradox. For on that way of describing change the change is rooted in or determined or delimited by the subject which undergoes it and which persists through the change, as is exhibited by the subject of the sentence, and the outcome of the change is highlighted by only pointing out what is different between the termini, as the predicate of the sentence shows. In other words, when this description of change is properly understood and compared to the others, it makes sense of change. It does so because it appeals to the substance, specifically, to the substance-constitutive characteristics, in whom the change is rooted. On Waterlow’s view, that preferring this way of speaking about change solves the paradox is grounds enough to accept it, for “a hindrance exists only for those who desire some goal, and they, once a path is cleared, do not require to be further persuaded of the rationality of following it.” (pg. 19) The goal was to reveal the principles by which we can grasp change—to defend our claim to being (potential) knowers. If one accepts

131 description (1) as the best description in the terms Waterlow offers, this preserves the possibility that we can understand change. Waterlow reads Aristotle’s discussion of change in Physics A as being centrally concerned with the contribution which a substantial form makes in directing the change, both in the case of change of accident and change of substance. It sets limits for the kinds of accidental change a substance can undergo, for example; a bit of wood and a pale human will both change colors in the sun eventually, but only one of these is likely to turn red in these conditions.58 Thus on Waterlow’s reading Aristotle argues that “every becoming involves a subject which remains, and this subject is a substance.” (pg. 46) It is important that there be as substance which remains, for it is this substance’s substantial form which (partially) determines or delimits the outcome of the change.59 One difficulty this reading encounters, by my lights, is that it relies on taking ὑποκείμενον as always being or signifying a substance; I argued against this understanding of ὑποκείμενον in my chapter 2. On her reading, that Aristotle appeals to substance is the key point, since it is this which shows how a natural change can be rooted in what came before. On her reading ὑποκείμενον as substance fulfills this

58 E.g.: “Thus if something equally truly describable as ‘a human being’, ‘a pale object’, ‘something in Athens’, ‘something weighing ten stone’, is said to become (say) hot, Aristotle’s preference for describing this subject primarily as ‘a human being’ can be justified as follows: that description tacitly connects the subject, thus described, to the predicated change in a way in which none of the others do. … The point is, rather, that for a language user of normal experiences, the first description on the list provides some indication, however sketchy and incomplete, concerning various possible circumstances under which, and processes by which, the change is likely to have taken place, as well as concerning possible conditions for reversing it; whereas the other descriptions, again for someone of normal experience, in themselves provide no such information.” (Nature, Change, and Agency, pg. 25) 59 A difficulty arises, then, for how to explain substantial change, since intuitively exactly what does not remain through the change is the substance. Waterlow argues that “the subject-substance in generation is the same individual as the eventual mature creature” (pg. 47), and it is this which persists through the change, though it goes by a different name at the beginning of the change. The immature form is present at the beginning of the change; the same individual substance thus persists through the change, though at the end of the change it is present in its mature form. What is common between the immature version at the terminus a quo and the mature form at the terminus ad quem is that “in both its phases the creature embodies the same nature in the sense of principle of life and change typical of members of its kind.” (ibid) 132 desideratum of an account of change. I argued in the previous section that ὑποκείμενον need not be a substance for it to fulfill the same desideratum; ὑποκείμενον on my reading will also do so, in virtue of the logical complexity which Aristotle ascribes to it in Physics A.7. A second difficulty is the initial textual difficulty I pointed out for Bostock’s view: this reading takes ὑποκείμενον in Physics A as referring primarily to the persistent simple which is not an opposite, for it takes the ὑποκείμενον to be the item which persists through the change, to the exclusion of the simple from the terminus a quo which does not persist. As I argued above, Aristotle introduces the whole complex at the terminus a quo as the ὑποκείμενον, not only the persistent simple. Given the persuasiveness of Waterlow’s account, more must be said, and will be in what follows. Perhaps more significantly, Waterlow’s reading seems to be a characterization of the development of a substance rather than of the substance’s coming to be. It helps to explain how a child becomes an adult, for example, but not, to borrow one of Aristotle’s own examples of substantial generation from GA A.4, how menses and sperm become an embryo. Yet the latter is a case of the coming to be of a new human; the former is, instead, a case of the development of a human.

133 Chapter 5: Matter as Analogous

§.1 Introduction

In this chapter I argue that the exegesis in the previous chapters of the dissertation of Aristotle’s introduction of matter as ὑποκείμενον, show that matter, for Aristotle, is something grasped only by analogy; it is an analogous term or concept. This is not merely an epistemological point, but tells something about what it is to be matter, according to Aristotle. Something can only be matter analogically. This is consistent with Aristotle’s claim in Metaphysics Λ 4 that “the causes and the principles of different things are in a sense different, but in a sense, if one speaks universally and analogically [analogian], they are the same for all.” (1070a31-21) In Metaphysics Λ 3 Aristotle had just been discussing matter and form, and the way in which these each are principles of sensible substances. In this chapter I will show that the introduction of matter reveals that in his Physics, too, Aristotle thinks of matter as analogous. Aristotle adumbrates his understanding of matter as analogy in his treatment of the contraries, lack and form, in A 5. In 191a17-21 he states it outright. The way in which he introduces matter helps to show what this means for him, and why it matters. In this chapter I will also present some of the ramifications of matter being ὑποκείμενον and so being analogous—both immediate consequences and some ramifications for future work concerning Aristotle’s theory of substance. In explaining the way in which matter is analogous according to Aristotle’s Physics, I will rely on the way Mary Hesse presents Aristotle’s understanding of analogy and analogical terms in her “Aristotle’s Logic of Analogy” (1965). In that essay, Hesse provides an exegesis of Aristotle’s understanding of analogical terms or concepts. She additionally criticizes it, and points to a way forward for our own understanding of scientific analogy. The latter concern is beyond the scope of my dissertation; here I attend only to how Aristotle treats analogical concepts, according to Hesse. I argue that the details of Aristotle’s concept of matter as it is presented in Physics A, as well as features of the argument by which Aristotle introduces it, fit with how Aristotelian analogy works according to Hesse. Matter being ὑποκείμενον is a key part of this. 134 One ramification of this understanding of matter is that Aristotle cannot consistently posit a basic or fundamental matter of the kind Scholastic philosophers called materia prima. By materia prima I understand a basic stuff which has no characteristic of its own, except its being potentially whatever form or account it should take on. As I understand it, this is inconsistent with the concept of matter as it is introduced in Physics. Of course, this is not enough to say that nowhere in Aristotle’s corpus does he posit a fundamental matter like this, only that should he do so this would be inconsistent with the understanding of matter he introduced in Physics, though it does provide some reason. I think that others, such as Charlton in the appendix to his 1971 commentary on Physics A & B provides a multitude of textual evidence suggesting that Aristotle nowhere admits of something like materia prima, but for my present purpose this is neither here nor there. This understanding of matter also has ramifications for the debate about how to understand Aristotle’s deliberations about substance in the central books of his Metaphysics. In a key part of these deliberations, Metaphysics Z 3, Aristotle worries whether the claim that substance is “that which is not predicated of a subject, but of which all else is predicated” requires that matter is primary substance in virtue of its being ὑποκείμενον. Interpreters disagree widely about how to read this passage. If I am right about how to understand matter as ὑποκείμενον, this affects how M Z.3 should be read. Further, it has the possibility of shedding some light on other contentious debates about Aristotle’s metaphysics of substance, such as how to understand the relation between matter and form as substance, as Aristotle characterizes it in Metaphysics H 6. Though there is no space to fully enter into either of these debates in this dissertation, at the end of this chapter I will suggest some ways in which the reading of matter as ὑποκείμενον provided here help to point the way forward.

§.2 Hesse on Aristotle on Analogy

In “Aristotle’s Logic of Analogy” Hesse argues that in Aristotle’s mature thought analogy has a “dual nature” (pg. 330) in that things which are analogous have a property or form in common, and also exhibit a similarity of relation. Hesse points to two cases 135 which best exemplify this. The first comes from Topics 108a, where Aristotle says that windlessness in the air is analogous to calmness in the sea. The common form is rest, while the similarity in relation is between the wind and the air in the one case, and the waves and the sea in the other. Another example comes from De Anima 419b25-30, where Aristotle describes echoes and the reflection of light as analogous (Hesse, pg. 332). Echoes are the rebounding of air from a wall, says Aristotle; reflections are the rebounding of light from surfaces. Hesse says that “here we have both the senses (i) and (ii) of analogy—both the common features of situations: ball bouncing, echo, reflection, giving the form or generic cause ‘repercussion’, and also the likeness of relation in the causal chain of events connected with sound, light and bouncing balls, namely throwing, impact and rebounding.” (pg. 332) Terms or concepts which are analogous in this way are such that there is a common form which is realized in different ways in different things, and which is itself discovered through seeing a certain sameness of relation. Hesse points to several cases in which Aristotle explains the analogy not simply by pointing to or naming the form which is possessed by each of the items which are analogous, but by doing a sort of induction whereby by reflecting on several cases side by side, a similarity of relation arises. The relation enables one to grasp the property. For example, in explaining actuality and potentiality, Aristotle proposes a series of examples from which his reader is to grasp what it is to be actual and what it is to be potential—building : capable of building = awake : asleep = seeing : eyes shut = differentiated matter : matter = finished article : raw material = actuality : potentiality. (Hesse, pg. 336-337) Hesse has this to say about what this sort of induction is supposed to show:

It is, however, a relation, not a universal, that is grasped. And yet not only a relation, for actuality and potentiality are not indissolubly hyphenated in Aristotle’s account. On the contrary, he claims that actuality and potentiality are defined by the sets of relata taken separately. Something more than a relation appears to be involved (…) Building, waking, seeing, etc., are not “actual” merely in virtue of being relata of a relation, say R, of which “actuality” is also a relation, they are “actual” in virtue of 136 themselves, just as particulars are particular cases of the universal. And yet ‘actuality’ is not a universal, since it is not a member of the categorial classification, and, more important, it is not predicated of the particulars in the same sense but only analogically, whereas genus and species are always predicated literally (Cat. 3 a-b). (pg. 337)

Reflecting on the examples helps one to understand the similarity of the relation, in virtue of which one can grasp the property or form that is held in common by all the instances of the analogous item. As Hesse presents it, for Aristotle the similarity of the relation and the common form are inseparable, and both required, for grasping the analogous concept or term, whether it is motion, rest, actuality, or matter. Returning to the examples of the windless day and the calm seas, Hesse contends that to grasp the analogy it is important both to “see the similarity of relation of calm to sea and windlessness to air, and … the form common to calm and windless.” (pg. 331) That Aristotle is thinking of matter as being analogous in this way in Physics A can be seen as early as A 5. In A5, as I argued in my chapter 3, Aristotle reflects on the fact that many of his predecessors proposed some set or sets of contraries as principles of change and of naturally changeable things, and tries to uncover what sort of insight they show in this. His predecessors propose such a variety of contraries (hot and cold, above and below, fire and water, love and strife), that one might feel at a loss trying to choose from among them or from similar contraries. Instead, in examining all the cases Aristotle comes to see a similarity of structure between all the pairs of contraries which have been proposed; ultimately, it is to this that he points when positing his contraries. His pair of contraries are the basic form of this. Yet, as he indicates in his examples, he is not simply pointing to the relation of opposition, or to the pair which is simply a thing and its opposite. Instead he suggests that each of the contraries has a sort of distinctive character, one which he sees in the relevant contrary of the pairs posited by his predecessors. From this, the structural similarity between the contraries which arises from a comparison of his predecessors’ views and the characteristic which he sees each of them as exhibiting or pointing toward, he proposes his own pair: the comparatively abstract lack and form. In other words, in A 5 Aristotle presents himself as going through 137 the process of induction from which he grasps the relevant analogy to discover lack, and form, as principles. As we see in A 7, the lack re-appears as the ὑποκείμενον, or as one of the λόγοι or εἴδη of it. In A7 Aristotle says that the ὑποκείμενον has two forms, or two accounts, of which the lack is one. As I argued in my last chapter, he presents the ὑποκείμενον as having a complex account (λόγος) or form (εἴδος), such that to understand what it is to be a ὑποκείμενον and how it is related to the outcome of a change, one has to grasp both what the thing is, e.g., a human, and the respect in which it is a lack, e.g., being unmusical. This leads to the enumeration of the principles as two at the end of A 7: a pair of contraries which he names the ὑποκείμενον and the form. Here the ὑποκείμενον is standing in as the lack; it is so by having the complex acccount just characterized. When Aristotle enumerates the principles as three (the ὑποκείμενον, the lack, and the form) he emphasizes the way in which the ὑποκείμενον and the lack are two, rather than the way in which they are one. In either enumeration Aristotle emphasizes the dual nature of the analogy by which the ὑποκείμενον, as matter, is grasped. As the ὑποκείμενον, the human exhibits a character which it has in its own right; as the lack, the unmusical (which is really the same thing as—one in number with—the ὑποκείμενον) stands in a certain relation to the form. The whole which these comprise, the complex at the terminus a quo, thus exhibits both features of Aristotelian analogy. This is borne out when Aristotle says the following:

The underlying nature [ὑποκειμένη φύσις] can be known by analogy. For as the bronze is to the statue, the wood to the bed, or the matter and the formless before receiving form to any thing which has form, so is the underlying nature to substance, i.e., the ‘this’ or existent. (191a7-12)

Here Aristotle presents another set of examples to try to more clearly indicate what it is to be the underlying nature, or the matter. He is not using these examples to try to point to some specific thing which is the underlying nature; he is presenting the various examples of an underlying nature side by side so that his reader can grasp what is in common among them, or what makes them all to be instances of an underlying nature. One might 138 read the review of possible counterexamples in 190b1-10 as additionally serving the same purpose. Aristotle asks his reader to “survey the various cases of becoming in the way we are describing” (190a14), and is confident that if they do so they will se that it is always the case that what comes to be does so from a ὑποκείμενον. The survey of cases helps him make his case, incomplete though it is, because it helps to clarify the nature of the thing he is pointing to, matter as ὑποκείμενον, by setting out putative instances of it side by side to make its character stand out. That matter is analogous in just this way also helps to explain why Aristotle says that “matter comes to be and ceases to be in one sense, while in another it does not.” (192a25-26). Something which will count as matter for something else might itself come to be. For example, a log might come to be when a tree is felled; later that log might come to be pale in color by drifting about in the seas for a while before washing ashore again (here I’m thinking about Aristotle’s “the log is pale” example from An. Po. A.22). But the log coming to be is not, properly, matter coming to be. The log itself will not count as matter until the change by which it comes to be pale. When the change takes place by which the log becomes pale, the log will then come to play the role of matter for that change. Yet this, too, is not a case of matter coming to be, strictly speaking; it is a case of a pale thing—a piece of driftwood—coming to be. Matter only comes to be or ceases to be accidentally, when something else comes to be or ceases to so that there is something which counts as the matter for that change. Or, alternatively, it comes to be when something else comes to be, which is such that it will count as the matter for some other change. The understanding of matter as ὑποκείμενον which I have presented here is also better suited to fill out Aristotle’s characterization of material necessity in Physics B 9 than are rival views according to which matter is merely what persists through change. As ὑποκείμενον, the matter is what the thing at the outcome of the change is built upon, requires, or presupposes. In other words, Aristotle is spelling out what is already implied by matter being ὑποκείμενον when he says that

If then there is to be a house, such-and-such things must be made or be there already or exist, or generally the matter relative to the end, bricks 139 and stones if it is a house. But the end is not due to these except as the matter, nor will it come to exist because of them. Yet if they do not exist at all, neither will the house, or the saw—the former in the absence of stones, the latter in the absence of iron—just as in the other case the principles will not be true. (…) The necessary in nature, then, is plainly what we call by the name of matter. (200a24-31)

Further, it follows from this understanding that matter is not an ontological category. There is nothing which counts as matter the way something counts as a plant, for example, by falling under the genus of plant. For one thing, matter is said analogously rather than synonymously of whatever it is predicated of; universals which are genuine categories are predicated synonymously of everything of which they are predicated. Further, for something to be matter it must already fall under some other category of being, not matter, in virtue of which it stands in the right sort of relation to the outcome of a certain kind of change. In other words, to call something matter or to treat it as matter is not to reveal what it is in itself, but to point to that in virtue of which it stands in a certain way towards some change or the outcome of some change. In some sense the outcome of the change is what determines what counts as matter, for what counts as matter will be what it presupposes or requires (195a18-19, cf. An. Po. B.11 94a21-22). Thus it is part of matter’s being ὑποκείμενον that the end of the change, the τέλος or the that for the sake of which, is a principle of the reasoning (200a22-24). In this way my analysis of matter as ὑποκείμενον, and as analogous, differs from Gill’s exegesis of Aristotle’s concept of matter in her Aristotle on Substance: A Paradox of Unity. In that work, Gill argues that Aristotle uses the term “matter” ontologically, as well as what she calls “analytically.” On her view, when some X is called matter ontologically, this implies something about what kind of a being X is; there are some things which are ontologically matter and others which are not. So, for example, bronze is matter ontologically because it is the sort of thing from which many brazen things can be made. But Socrates is not called matter ontologically because he is not the sort of thing which enters into composition with any substantial form, or from which something else can be made that might still be called human the way a statue is called brazen. My 140 reading of Aristotle’s introduction of matter as ὑποκείμενον shows that for Aristotle matter is never an ontological category, and seeing how matter is an analogous concept for Aristotle (a la Hesse), supports this. To call something matter does not reveal anything about the kind thing it is in itself. Rather, it reveals that plays a certain explanatory role. On the other hand, my reading differs from Code’s in that Code separates the roles of matter and subject in substantial change. On his reading, in substantial change the matter persists while the ὑποκείμενον, the matter-form composite from which the new substance comes to be, does not. That is to say, according to Code to be matter is not to be the ὑποκείμενον in some change, though it may happen to be the case that what counts as matter is also accidentally the same as the ὑποκείμενον. This requires some clarification. Since there are hierarchical levels of composition in the hylomorphic analysis of substances, there are hierarchical levels of matter; what is called matter in the context of hylomorphic analysis depends upon which level of analysis one is attending to. The matter which immediately comes into composition with a substantial form is a substance’s proximate matter; according to Code, the proximate matter alone persists through a substantial change. More generally, on Code’s view whatever persists through change, either in substantial change or in accidental change, can properly be called matter, though in each case the matter which persists will be the proximate matter relative to the form picked out at that level of analysis. In other words, on Code’s view matter is just whatever persists through some change, or whatever comes into composition with some form. In reading Aristotle’s introduction of matter in this way, Code seems to be taking it that Aristotle calls only the persistent simple matter, while he calls the complex at the terminus a quo the ὑποκείμενον. In my previous chapter, chapter 4, I argued that Aristotle calls both items both names. Both the complex at the terminus a quo and the persistent simple count as the ὑποκείμενον, the latter being so called derivatively or secondarily, insofar as it is one in number with the complex whole. Which one Aristotle picks out by the term ὑποκείμενον depends upon which enumeration of the principles he is employing: two, or three. Further, when in A.9 Aristotle offers a definition of matter he presents the primary substratum as matter, which I argued referred to the complex 141 ὑποκείμενον. Nevertheless, for the same reason that the persistent simple also counts as ὑποκείμενον, it will count as matter or ὕλη, too. Thus Aristotle does not divorce matter and ὑποκείμενον in the way Code suggests.

§.3 Ramifications

A consequence of this reading is that there is no materia prima, or prime matter in the Scholastic sense, in Aristotle. This is because, on my reading, a ὑποκείμενον cannot be bare in the way that Scholastic materia prima is. This is not only a textual but a philosophical point, for nothing which is bare can play the explanatory role required of matter as ὑποκείμενον. To see this, first it is important to clarify what doctrine it is that I am denying Aristotle holds. Scholars who attribute to Aristotle a doctrine of prime matter do not all agree on what this doctrine is. One reason for this is that the expression “prime matter” is hard to find in Aristotle’s corpus. The expression is a term of art which is found chiefly in commentary on Aristotle, so that what is meant by it sometimes varies between commentators. There are two general approaches. The first is the traditional interpretation of Aristotle, owing its traditionality in great part to Thomas Aquinas’s commentaries on Aristotle. This interpretation presents a doctrine which I will call materia prima. As Aquinas has it in his De Principiis Naturae caput 2, prime matter, or materia prima, is pure potentiality which has no actuality or positive character at all in its own right, but which is such as to receive form or actuality. It is passive potentiality; it is not itself a principle of generation or motion by having any tendency toward any form. Aquinas argued that this is what immediately underlies any substantial form; materia prima for Aquinas is the proximate matter for substantial form. Other philosophers have argued instead that while Aristotle does not posit materia prima as the proximate matter of substantial form, he does posit it as the matter of the elements. Thus materia prima is, on either reading, the ultimate matter of any material thing. The second class of interpretations also treated prime matter as being the ultimate matter for any material thing, emphasizing that it is found as the proximate matter for Aristotle’s elements. But proponents of this view suggest that prime matter has some 142 characteristics in its own right, and is not pure passive potentiality.60 I argue that it follows from the interpretation of ὑποκείμενον for which I have argued, and for Aristotle’s matter as ὑποκείμενον, that there can be no materia prima as something featureless in itself, or as bare, passive potentiality, whether it is the proximate matter for all substantial form or the ultimate matter of the elements. I also argue that on Aristotle’s view there is no one kind of thing which stands as a fundamental matter, and which is itself not the product of any change. Thus neither kind of prime matter is consistent with Aristotle’s concept of matter in Physics A.7. First, it follows straightforwardly from my understanding of matter as ὑποκείμενον that matter is never pure potentiality. To be a genuine ὑποκείμενον in a case of change, something must be the precondition for the terminus ad quem of the change just by being what it is, according to a which it possesses. To meet this criterion, it must have some λόγος. It must have some character x which belongs to it so that by being ὅπερ x ti, to use Aristotle’s expression, the outcome can come to be. That is to say, it must have a definition in the same way that Aristotle allowed λόγοι to non-substances in Categories and in his examples in Physics A. Having a definition in this way is not to have a full-on definition corresponding to an essence, but merely to have a what-it-is-to- be-x corresponding to the name it is given or the way it is categorized. Cf. Physics 190b24-26, where Aristotle says that the ὑποκείμενον, at least under one of its λόγοι, is a τόδε τι. Its being able to be a ὑποκείμενον presupposes it has some such λόγος, in virtue of which it stands in a certain relation to another λόγος. A bare substrate does not fit with this criterion for being a ὑποκείμενον from An. Po. This prompts me to explain a difficult bit of text which forms a key part of Aquinas’s understanding of prime matter. According to Aquinas, “Et quia omnis definitio et omnis cognitio est per formam, ideo materia prima per se non potest cognosci vel definiri sed per comparationem ut dicatur quod illud est materia prima, quod hoc modo se habet ad omnes formas et privationes sicut aes ad idolum et infiguratum. Et haec

60 E.g. Sorabji and Sokolowski, who argue that prime matter is extension, or Chrstopher Byrne (1995), who argues that prime matter is extended, physical, and moveable, or Sheldon Cohen (1984), who argues that it is per se potentially hot, heavy, light, etc.. and thus νot bare (pg. 178).

143 dicitur simpliciter prima.”61 (De Principiis Naturae, caput 2) Here he seems to be commenting on Physics A.7 191a7-12: “The underlying nature [ὑποκειμένη φύσις] can be known by analogy. For as the bronze is to the statue, the wood to the bed, or the matter and the formless before receiving form to any thing which has form, so is the underlying nature to substance, i.e., the ‘this’ or existent.” On Aquinas’s reading prime matter is not in itself knowable, since all knowledge and understanding is through the form. Because of this, he thinks, Aristotle says that the underlying nature can only be known by comparationem. Though comparationem is the Latin word translating the Greek ἀναλογίαν, Aquinas’s use of it here suggests that he understands the relevant notion for Aristotle differently than does Hesse. For he claims that the only way that prime matter can be known is by seeing it in comparison to Forms, in contrast with the way matter at other levels of analysis can be grasped. To understand matter by comparationem with Forms is to see that it stands in a relation like that of bronze to the statue. Prime matter, as matter, always implies form and must be understood in relation to form. But because prime matter has no form or other account in its own right by which it might be grasped, it can only be understood through this relation. By contrast, Aquinas says, matter at other levels of analysis can also be understood in terms of its genus, e.g., as bronze or flesh, or even man as the matter for whiteness.62 This shows that while Hesse argues that ἀναλογίαν for Aristotle implies both a relation and a property, for Aquinas comparationem implies only a relation. It seems to me to be a mistake to read 191a7-12 as making a claim about prime matter in the sense of some ultimate matter, in distinction from what scholars now generally call proximate matter. Proximate matter is, roughly, that matter which is arrived at first in hylomorphic analysis. It is proximate in that it immediately underlies the form,

61 “And since all definition and all cognition is through form, for this reason prime matter cannot be known or defined through itself, but only by comparison, as it is said “that is prime matter which holds itself toward all forms and privations just as bronze to the statue and as the unshaped.” (My translation)

62 See also his commentary on Aristotle’s Physics, Lectio 13.

144 or the outcome of the change; it is the ὑποκείμενον of the change. It contrasts with matter which might be arrived at by further levels of analysis. To use Aristotle’s example from the opening chapter of Generation of Animals, the proximate matter of an animal is the organized body—the parts which are organs. The matter for this is the homoiomeries, etc. The ultimate or final level of analysis, if there is one, would be the prime matter. Though it is consistent with Aristotle’s presentation in A7 that matter-form analysis can be performed at various levels (something he confirms in the opening chapter of GA), at no place in A7 does Aristotle distinguish between levels of analysis or suggest that something is true of matter one level, but not at another. In fact, doing so would be contrary to his stated aim of showing what is common to all change. In 191a7-12 itself Aristotle also does not distinguish between different levels of matter. He merely makes a claim about the “underlying nature”, or ὑποκείμενον φύσις. If Aristotle has anything to say about prime matter uniquely, he is not saying it in A7. Second, even if he were trying to make a claim about prime matter, as opposed to proximate matter or matter at other levels of analysis, Aquinas’s presentation of how matter is to be understood through comparationem is inconsistent with the understanding of Aristotelian analogy which I gave in the previous section. If analogy, for Aristotle, is to be understood as Hesse characterizes it, a way which I argued is borne out in Aristotle’s introduction of matter as ὑποκείμενον in Physics A 5-7, then Aristotle should not be read as claiming that prime matter has no character or feature in virtue of which it can be grasped. He especially should not be understood to be claiming that this is so on the grounds that matter is understood by analogy. Its being understood by analogy in fact requires that it have some character or property, other than its relation to form, by which it is grasped. When Aristotle includes “the formless before receiving form to anything which has form” as one of the relations by which the analogy can be seen he is not proposing that matter is completely formless in itself, he is reminding his reader that the matter also stands as the lack to the form, in the way he described in A 5. In A 5 he chose strategic examples to clarify that by “lack” he did not mean what is not, or even merely what is not x, as I argued in my chapter 3. Having incorporated this into his view in A 7, especially in the enumeration of the principles as two, as we have seen, he should not be read as 145 changing his mind to propose that matter is what is completely formless. Reading 191a7- 12 in this way fits better with the surrounding text, in which Aristotle is trying to summarize his conclusion about the number and nature of the principles. He is not trying to suggest new features of the principles which he has not already discussed. Rather, he is reviewing his findings and showing how they fit with the considerations which brought them to light. In A7 191a7-12 Aristotle is showing that the underlying nature, or ὑποκείμενον, stands to form in the way that the primary contraries should. Note that even though in A.9 Aristotle will call this underlying nature “matter”, at 191a7-12 he is content to use the term ὑποκείμενον, and to speak of the underlying nature [ὑποκειμένη φύσις]. That is, regardless of what the underlying nature is in any particular case, i.e., regardless of what sort of thing serves as the ὑποκείμενον in that change, it can still be grasped as the underlying nature through the analogy. In other words, its being grasped as the underlying nature is different from grasping it in its own right. For example, it is different to grasp something as a human and to see the human as the ὑποκείμενον for some change. Similarly, it is different to conceive of bronze as bronze and to see bronze as the matter for some change or some product of change. This is an important point for Aristotle to make in the context of the dialectic in which he has been engaging throughout A 7. With it he reiterates that we need not make the mistake of those he calls the “physicists”; our account of the principles will be complete even if it does not arrive at one distinctive kind of stuff from which everything is made or in which everything ultimately consists. Rather, our account is complete when we arrive at principles which are present in every change and every changeable thing. His making this point is also helpful for the present case against materia prima, since it presumes that whatever counts as ὑποκείμενον already has some λόγος, in virtue of which it counts as ὑποκείμενον. Thus it presumes that the ὑποκείμενον is not bare in the way that materia prima is generally understood to be.

§.4 Ramifications for Future Work

The understanding of matter as ὑποκείμενον which I have presented here also has ramifications for future work, especially for understanding Aristotle’s investigation into 146 substance and primary substance in his Metaphysics. For example, it has ramifications for how to understand the so-called “stripping argument” of Metaphysics Z.3. Metaphysics Z.3 is one of the textual loci of the debate about Aristotle’s later theory of substance. Z.3 can be outlined as follows. Aristotle begins by presenting ὑποκείμενον, genus, essence, and the universal as items which are candidates for primary substance, or as possible criteria for substancehood. He focuses on ὑποκείμενον first, saying that it “especially [μάλιστα]” seems to be substance. (1029a1) He clarifies that he is thinking of the ultimate or final ὑποκείμενον, not just anything which is a ὑποκείμενον: “the ὑποκείμενον is that of which other things are said, but which is itself never predicated of anything else”63 (1028b36). Aristotle then presents what has been called a “stripping argument,” in which, taking as his starting point a sensible particular, Aristotle “strips away” all the predicables, in order to show that matter is the ultimate ὑποκείμενον. Having done so, he points out that the matter arrived at by this analysis cannot be substance because it is neither separable nor definable (1029a27-28). For the matter which is uncovered in this way is “that which in itself is neither a particular thing nor of a certain quantity nor assigned to any other of the categories by which being is determined.”64 (1029a20-21) The strategy of the argument in the body of Z.3 has often been read as being to offer a reductio ad absurdum purporting to show that since matter, or rather prime matter, is the ultimate ὑποκείμενον and it fails to meet the criteria for substancehood, the ultimate ὑποκείμενον is not primary substance, contrary to what Aristotle maintained in Categories. This is presented to show that Aristotle discovers a tension between his early theory of substance and the hylomorphism which he proposes in his Physics with the introduction of matter. Yet different interpretations abound, hinging especially on the question of whether or not the matter which is uncovered as a result of the stripping argument is indeed Aristotle’s own view, or is instead a different view which he presents for didactic reasons.

63 My translation. 64 Ross translation in Barnes edition of The Complete Works of Aristotle. 147 Code, for example, understands the subject criterion as it is carried over from the Categories as a logical criterion concerned only with the sorts of ontological predication outlined in the Categories. As such, on his view it is both unhelpful and misleading in a causal inquiry like the one in Metaphysics Z. (“Science of Principles,” pp. 375) Code argues that Aristotle uses the result of the stripping argument as a sort of reductio ad absurdum to show that, when applied to the parts of the hylomorphic composite, the subject criterion only leads to something impossible, and thus should be abandoned. The reading which he proposes is complex; the upshot of it is that he thinks Aristotle claims at 1029a9 that the subject criterion does not help settle what plays the role of primary substance. Gill (1989), on the other hand, reads Z.3 as first proposing the subject criterion and then clarifying it. On her view, like Code’s, the stripping argument is a sort of reductio. But she claims Aristotle concludes from it not that the subject criterion should be abandoned but that it should be refined and clarified in terms of separability and thisness. Another interpretive question which arises is whether the reader is to take the picture of matter which appears as the result of the stripping argument as Aristotle’s own, and what this means for the way in which matter is finally eliminated as a candidate for primary substance. There are three options. The matter which is described in the stripping argument and disqualified as neither χωριστόν nor τόδε τι is either (1) someone else’s view of matter, (2) a view of matter which Aristotle ascribed to at least in earlier works. My reading of Aristotle’s introduction of matter has ramifications for both of these interpretive issues. First, it suggests that the stripping argument employs a concept of ὑποκείμενον which is not Aristotle’s, close as it may be. For, as I argued in this chapter, Aristotle’s ὑποκείμενα can never be bare. Something which counts as a ὑποκείμενον for Aristotle must always have some characteristic in its own right in virtue of which it stands as a ὑποκείμενον for something else. Thus if Aristotle is consistently appealing to his own concept of ὑποκείμενον and its relation to its predicates, the stripping away of predicates in the mind will never reveal something which is “of itself neither a particular thing nor of a particular quantity nor otherwise positively

148 characterized; nor yet negatively, for negations also will belong to it only by accident.” (1029a24-26) One possibility that arises from Aristotle’s understanding of ὑποκείμενον as I have presented it in this dissertation is that in the stripping argument Aristotle is presenting a reductio ad absurdum to show how the relation between a ὑποκείμενον and its predicate does not work. He is clarifying how he understands being a ὑποκείμενον and its relation to being a substance by contrasting it with a different, relevantly similar idea. One might think of being a ὑποκείμενον as simply being a property-bearer, something which has characteristics which are said of it. If this is all there is to being a ὑποκείμενον, then one might conceivably separate the bearer from all its properties or descriptors, so that in the end there remains only something which is nothing in its own right; it is merely a property bearer. If this were Aristotle’s understanding of being a ὑποκείμενον, then this would conflict with his characterization of a substance. Primary substances, even things which counted as primary substances in Categories, would not fit with this. Thus there would be a tension between Aristotle’s discussion of substance in Physics and Categories which he brings out by uncovering the ultimate ὑποκείμενον. Since, as I have argued, this is not how Aristotle understands being a ὑποκείμενον, he can be read as using Z 3 to clarify this. In doing the stripping argument Aristotle carries the mistaken understanding through to its logical conclusion to clarify both that it is not his view and why he has reason not to accept it. Analysis of the stripping argument in the light of the understanding of being a ὑποκείμενον I have presented in this dissertation might also bear fruit by shedding light on how Aristotle wants to think about substance and primary substance. The reductio in Z.3 is part of a long investigation by Aristotle into what it is to be a substance. That the first issue he takes up is the relationship between being a ὑποκείμενον and being a substance is significant. The very next question Aristotle takes up is the question of how essences are related to substances, especially to primary substances. My reading of Aristotle’s concept of ὑποκείμενον leaves Z.3 poised to clarify the difference between being something being a ὑποκείμενον and its being a substance, in a way which can be read as following through on the train of thought which Aristotle begins with the stripping argument in Z.3. In Z.3 he clarifies that to be a ὑποκείμενον is not simply to be 149 something which has a predicate, but is something which has an account of being which is relevant to, or related in the right way to, the presence of some particular predicate. In so doing he avoids the problem of the bare ὑποκείμενον, which cannot be a substance. In the next chapter, however, he contrasts being a ὑποκείμενον with being a substance by distinguishing having an account of being from having an essence. Read in this way, Z.3 and Z.4 do not conflict with Aristotle’s metaphysical schema from Categories, but rather fill it out and clarify it. This is merely a sketch, but it suggests a non-developmental interpretation of Aristotle’s thought from Categories to Physics and through to his Metaphysics—non-developmental in the sense that Aristotle need not be read as changing his mind about some fundamental tenets of his metaphysics, though he may be read as unpacking or investigating them.65 )

65 Similarly, if my understanding of Aristotle’s introduction of matter as ὑποκείμενον and as analogous is correct, then Aristotle is not referring to his own concept of matter when he says “By matter I mean that which in itself is neither a particular thing nor of a certain quantity nor assigned to any other of the categories by which being is determined.” (1029a20-21) Thus the result of Z.3 cannot be read as being that matter is not substance, for the argument by which matter is purportedly rejected as substance attacks a different concept, and not his own. This leaves open the question of how matter is related to substance, for Aristotle. This makes sense of why Aristotle should take it up again later in Z, and again in H, where, famously (going with Ross’s edition of the text) he says that “the proximate matter and the form are one and the same thing, the one potentially, the other actually.” (1045b17-19 150 Bibliography

Ackrill, J.L. “Aristotle’s Definitions of Psuche.” Proceedings of the Aristotelian Society New Series, Vol. 73 (1972 - 1973), pp. 119-133.

Aquinas, Thomas. De Principiis Naturae. .

Barnes, Jonathan. Aristotle: Posterior Analytics. Oxford, Clarendon Press. (1975).

Barnes, Jonathan, Ed. The Complete Works of Aristotle: The Revised Oxford Translation, Vol. I & II. Princeton, Princeton University Press (1984).

Blackwell, Richard. “Matter as a Subject of Predication in Aristotle.” The Modern Schoolman, Vol. 33, Issue 5, November 1955, pp. 19-30.

Bogen, James. “Introduction” in How Things Are: Studies in Predication and the History of Philosophy and Science.” Ed. James Bogen and James E. McGuire. Dordrecht, Holland, D. Reidel Publishing Company (1985).

Bolton, Robert. “Aristotle’s Conception of Metaphysics as a Science.” In Unity, Identity, and Explanation in Aristotle’s Metaphysics. Oxford, Clarendon Press (1994), pp. 321-354.

Bolton, Robert. “Aristotle’s Method in Natural Science: Physics I.” In Aristotle’s Physics: A Collection of Essays, Ed. Lindsay Judson. Oxford, Clarendon Press (1991), pp. 1-30.

Bostock, David. “Aristotle on the principles of change in Physics I”. In Space, Time, Matter, and Form: Essays on Aristotle’s Physics. Oxford, Clarendon Press (2006), pp.1-18.

Bostock, David. “Aristotle’s Theory of Matter.” In Space, Time, Matter, and Form: Essays on Aristotle’s Physics. Oxford, Clarendon Press (2006), pp. 30-47.

151 Bostock, David. “Aristotle on the Eleatics in Physics I. 2-3.” In Space, Time, Matter, and Form: Essays on Aristotle’s Physics. Oxford, Clarendon Press (2006), pp.103- 115.

Byrne, Christopher. “Compositional & Functional Matter: Aristotle on the Material Cause of Biological Organisms.”Apeiron, Vol. 48, No. 4 (2005), pp. 387-406.

Charles, David. “Matter and Form: Unity, Persistence, and Identity.” In Unity, Identity, and Explanation in Aristotle’s Metaphysics. Oxford, Clarendon Press (1994), pp. 75-106.

Charlton. Aristotle’s Physics I,II. Oxford, Clarendon Press (1970).

Charlton, William. “Aristotle on Identity.” In Unity, Identity, and Explanation in Aristotle’s Metaphysics. Oxford, Clarendon Press (1994), pp. 41-54.

Cohen, Sheldon. “Aristotle’s Doctrine of the Material Substrate.” The Philosophical Review, Vol.93, No. 2 (April 1984), pp. 171-194.

Cooper, John, Ed. Plato: The Complete Works. Indianapolis, Hackett Publishing Company (1997).

Corkum, Phil. “Aristotle on Ontological Dependence.”Phronesis, Vol. 53 (2008), pp. 65- 92.

Corkum, Phil. “Aristotle on Predication.” European Journal of Philosophy, Vol. 23, No. 3 (2013), pp. 793-813.

Dancy, Russell. “On Some of Aristotle’s First Thoughts About Substances.”Philosophical Review, Vol. 84, No. 3 (1975), pp. 338-373.

Dancy, Russell. “On Some of Aristotle’s Second Thoughts About Substances: Matter.” Philosophical Review, Vol. 87, No. 3 (1978), pp. 372-413.

Devereux, Daniel T. “Inherence and Primary Substance in Aristotle’s Categories.” Ancient Philosophy, Vol. 12 (1992) pp. 113-131.

152 Devereux, Daniel, and Demoss, David. “Essence, Existence, and Nominal Definition in Aristotle's "Posterior Analytics" II 8-10.” Phronesis Vol. 33, No. 2 (1988). Pp. 133-154.

Code, Alan. “Aristotle: Logic and Metaphysics.” Routledge History of Philosophy, Vol. 2, E. D. Furley. Routledge (1999), pp. 40-75.

Code, Alan. “Aristotle’s Metaphysics as a Science of Principles.” Revue Internationale de Philosophie (1997), pp. 353-378.

Code, Alan. “Aristotle: Essence and Accident.” In Philosophical Grounds of Rationality, Ed. R.E. Grandy and R. Warner. Oxford, Oxford (1986), pp. 411-439.

Code, Alan. “On the Origins of Some Aristotelian Theses about Predication.” In How Things Are, Ed. J. Bogen & J.E. McGuire. Reidel (1985), pp. 101- 131, 323-326.

Code, Alan. “The Persistance of Aristotelian Matter.” Philosophical Studies, Vol. 29 (1976), pp. 357-367.

Ebrey, David. Dissertation. .

Ferejohn, Michael. “The Definition of Generated Composites in Aristotle’s Metaphysics.” In Unity, Identity, and Explanation in Aristotle’s Metaphysics. Oxford, Clarendon Press (1994), pp. 291-317.

Ferejohn, Michael T. "Aristotle on Focal Meaning and the Unity of Science.” Phronesis, Vol. 25, No. 2 (1980), pp. 117-128.

Fine, Kit. “A Puzzle Concerning Matter and Form.” In Unity, Identity, and Explanation in Aristotle’s Metaphysics. Oxford, Clarendon Press (1994), pp. 13-40.

Gill, Mary Louise. Aristotle on Substance: The Paradox of Unity. Princeton, Princeton University Press (1989).

153 Gill, Mary Louise. “Individuals and Individuation in Aristotle.” In Unity, Identity, and Explanation in Aristotle’s Metaphysics. Oxford, Clarendon Press (1994), pp. 55- 71.

Gershenson, Daniel E & Greenberg, Daniel A. “Aristotle’ Confronts the Eleatics: Two Arguments on ‘The One’.” Phronesis, Vol. 7, No. 2 (1962), pp. 137-151.

Hesse, Mary. “Aristotle’s Logic of Analogy.” The Philosophical Quarterly, Vol. 15, No. 61 (Oct., 1965), pp. 328-340.

Irwin, Terence. Aristotle’s First Principles. Oxford, Clarendon Press (1988).

Jones, Barrington. “Aristotle’s Introduction of Matter.”Philosophical Review, Vol. 83, No. 4 (1974), pp. 474-500.

Kelsey, Sean. “Hylomorphism in Aristotle’s Physics.” Ancient Philosophy, Vol. 30, No. 1 (2010), pp. 107-124.

Kelsey, Sean. “The Place of I 7 in the Argument of Physics I”. Phronesis, Vol. 53 (2008), pp. 180-208.

Kelsey, Sean. “Aristotle Physics I 8.” Phronesis, Vol. 51, No. 4 (2006), pp. 330-361.

Kerferd, G.B. The Sophistic Movement. Cambridge, Cambridge University Press (1982).

G.S. Kirk, J.E. Raven and M. Schofield. The Presocratic Philosophers. 2nd Edition. Cambridge, Cambridge University Press (1957).

Kohl, Markus. “Substancehood and Subjecthood in Aristotle’s ‘Categories’.” Phronesis, Vol. 53, No. 2 (2008), pp. 152-179.

Kung, Joan. “Aristotle on ‘Being Is Said in Many Ways’.”.History of Philosophy Quarterly, Vol.3, No. 1(1986), pp. 3-18.

Kung, Joan. “Aristotle on Thises, Suches and the Third Man Argument.” Phronesis , Vol. 26, No. 3 (1981), pp. 207-247.

154 Lear, Jonathan. “Aristotle’s Philosophy of Mathematics.” Philosophical Review, Vol. 91, No. 2 (1982), pp. 161-192.

Lewis, Frank. Substance and Predication in Aristotle. Cambridge, Cambridge University Press (1991).

Lewis, Frank. “Aristotle on the Relation between a Thing and its Matter.” In Unity, Identity, and Explanation in Aristotle’s Metaphysics. Oxford, Clarendon Press (1994), pp. 247-278.

Loux, Michael. “Aristotle on Matter, Form, and Ontological Strategy.” Ancient Philosophy Vol.25, No. 1 (2005), pp. 81-123.

Loux, Michael. Primary Ousia: An Essay on Aristotle’s Metaphysics Z & H. Ithaca, NY, Cornell University Press (1991).

Loux, Michael. “Composition and Unity: An Examination of Metaphysics H.6.” In The Crossroads of Norm and Nature, Ed. May Sim. Lanham, Rowman & Littlefield Publishers, Inc. (1995), pp. 247-280.

Loux, Michael. “Understanding Process: Reflections on Physics III.1.” In The Crossroads of Norm and Nature, Ed. May Sim. Lanham, Rowman & Littlefield Publishers, Inc. (1995), pp. 281-304.

Modrak, Deborah. “Theories of Meaning and Ontology in Aristotle’s Metaphysics.” In The Crossroads of Norm and Nature, Ed. May Sim. Lanham, Rowman & Littlefield Publishers, Inc. (1995), pp. 221-234.

Moravscik, J.M.E. “ Aristotle on Predication.” Philosophical Review, Vol. 76, No.1 (1967), pp. 80-96.

Moravscik, Julius. “What Makes Reality Intelligible? Reflections on Aristotle’s Theory of Aitia.” In Aristotle’s Physics: A Collection of Essays, Ed. Lindsay Judson. Oxford, Clarendon Press (1991), pp. 31-48.

155 O’Connor, S., ‘The Subjects of Natural Generation in Aristotle’s Physics I 7’, Apeiron, Vol. 48 (2015), pp. 45-75.

Owen, G.E.L. Logic, Science, and Dialectic: Collected Papers in Greek Philosophy. Ithaca, NY. Cornell University Press. (1986).

Owens, Joseph. “Aristotle on Categories.” In Aristotle: The Collected Papers of Joseph Owens, Ed.John R. Catan. Albany, State University of New York Press (1981), pp. 14-22.

Owens, Joseph. “Matter and Predication in Aristotle.” In Aristotle: The Collected Papers of Joseph Owens, Ed.John R. Catan. Albany, State University of New York Press (1981), pp. 35-47.

Peramatzis, Michail. “Matter in Scientific Definitions in Aristotle” in Oxford Handbooks Online. New York,Oxford University Press (2014).

Philoponus. On Aristotle’s Physics I.1-3. Trans. Catherine Osborne. Ithaca, NY. Cornell University Press (2006).

Philoponus. On Aristotle’s Physics I.4-9. Trans. Catherine Osborne. Ithaca, NY. Cornell University Press (2009).

Pernot, Laurent. Rhetoric in Antiquity. Washington, DC:, Catholic University of America Press (2005).

Polis, Dennis F. “A New Reading of Aristotle’s Hyle.” The Modern Schoolman, Vol. 58, No. 3 (1991), pp. 225-244.

Ross, W.D. Aristotle’s Physics. Oxford, Clarendon Press. (1936).

Ryle, Gilbert. “The Timaeus Locrus.” Phronesis, Vol. 10, No. 2 (1965), pp. 174-190.

Scaltsas, Theodore. “Substantial Holism.” In Unity, Identity, and Explanation in Aristotle’s Metaphysics. Oxford, Clarendon Press (1994), pp. 107-128.

156 Schaeffer, Jonathan. “On What Grounds What.” In Metametaphysics: New Essays on the Foundations of Ontology, Ed. David Manley, David J. Chalmers & Ryan Wasserman. Oxford,Oxford University Press (2009), pp. 347-383.

Solmsen, Friedrich. “Light from Aristotle’s Physics on the text of Parmenides B 8 D-K.” Phronesis, Vol. 22, No. 1 (1977), pp. 10-12.

Spangler, G.A. “Aristotle’s Criticism of Parmenides in Physics I.” Apeiron, Vol. 13, No. 2 (1979), pp.92 103.

Walz, Matthew D. “The Opening of On Interpretation: Toward a More Literal Reading.” Phronesis, Vol. 1, No. 3 (2006), pp. 230-251.

Waterlow, Sarah. Nature, Change, and Agency in Aristotle’s Physics. Oxford, Clarendon Press (1982).

Whiting, Jennifer. “Form and Individuation in Aristotle.” History of Philosophy Quarterly, Vol. 3, No.4. (1986), pp. 359-377.

Whiting, Jennifer, “Living Bodies.” In Essays on Aristotle’s De Anima, Ed. Martha C. Nussbaum and Amélie Oksenberg Rorty. Oxford, Clarendon Press (1995), pp. 75- 92.

Witt, Charlotte. Substance and Essence in Aristotle. Ithaca, Cornell University Press (1989).

Wedin, Michael. Aristotle’s Theory of Substance. Oxford, Oxford University Press (2000).

Woods, Michael. “The Essence of a Human Being and the Individual Soul in Metaphysics Z and H.” In Unity, Identity, and Explanation in Aristotle’s Metaphysics. Oxford, Clarendon Press (1994): 279-290.

157