2. The Paradox of Substance
From Roy Bhaskar's early work we have learned that, if experimental practice is to be intelligible, the world must be a particular way. First, objects must be capable of behaving differently in open and closed systems. For this reason, the being or substance of generative mechanisms cannot be identified with its actualized qualities, but must be located elsewhere. It is only in closed systems, Bhaskar contends, that constant conjunctions of events in cause and effect relations obtain. In open systems, by contrast, objects can remain dormant, producing no events at all, or the intervention of countervailing causes can either a) hide events produced by objects, or b) produce events different from those that the generative mechanism or object would produce in a closed system by virtue of how entanglements of objects are woven together. Consequently, second, objects or generative mechanisms must be distinguished from events or actualities. Objects or generative mechanisms are defined not by their qualities or events, but rather by their powers or capacities. An object cannot be without its powers or capacities, but it can be without its qualities or events. Finally, third, if it is possible to form closed systems where constant conjunctions of events can obtain, then it also follows that objects or generative mechanisms must be independent of their relations.
While I readily concede that objects can enter into relations—how else would open systems be possible?—it does not follow from this that objects are their relations. In short, if it is to be possible to form closed systems in which constant conjunctions of events occasionally obtain as they sometimes do in experimental settings, then it follows that relations cannot ontologically be internal to their terms or the objects that they relate. In other words, objects are not constituted by their relations to the rest of the world. While relations to other objects often play a key role in the precipitation of events or qualities in objects, we must here recall that objects are not identical to their qualities but are rather the ground of qualities. Accordingly we must distinguish between objects and their relations, or rather the structure of objects and the relations into which objects enter. I call the former “endo-relations” (or, following Graham Harman, “domestic relations”), and the latter “exo-relations” (or, as Harman calls them, “foreign relations”  ). Endo-relations constitute the internal structure of objects independent of all other objects, while exo-relations are relations that objects enter into with other objects. Were objects constituted by their exo-relations or relations to other objects, the being would be frozen and nothing would be capable of movement or change. It is only where relations are external to objects that such change can be thought.
Insofar as what Bhaskar calls generative mechanisms are the ground of events or qualities, they deserve the archaic, Aristotelian name of substance. Because substances have the power to produce events, I shall refer to them as difference engines, for the production of an event is the production of differences in the world. Because difference engines or substances are not identical to the events or qualities they produce, while nonetheless substances, however briefly, endure, the substantial dimension of objects deserves the title of virtual proper being. And because events or qualities occur only under particular conditions and in a variety of ways, I will refer to events produced by difference engines as local manifestations. Local manifestations are manifestations because they are actualizations that occur in the world.
To this list of the properties of substances we can add a fourth: local manifestations are not to be confused with manifestations to or for a subject, but are rather events that take place in the world regardless of whether or not any subjects or sentient beings exist to witness them. Consequently, local manifestation is not equivalent to the empirical or what is experienced by a subject. Experience is a subset of local manifestation, but the set comprised of local manifestations is infinitely larger than the set consisting of experience. In this respect, the category of local manifestation shares some affinity to Badiou's conception of appearance as appearing without a subject to which appearance appears or is given. 
If, by contrast, local manifestations or events are local, then this is because the qualities or events of objects are variable depending on internal dynamisms in the object or difference engine and the exo-relations into which the object enters. Consequently, we must not say that an object has its qualities or that qualities inhere in an object, nor above all that objects are their qualities, but rather in a locution that cannot but appear grotesque and bizarre, we must say that qualities are something an object does. The concept of local manifestation is here designed to capture the context dependency—whether that context be internal or external—of the events an object produces in its manifestations.
Finally, insofar as substances are not identical to events or their qualities—nor, moreover, their exo-relations to other objects—I refer to difference engines as split-objects. The characterization of difference engines as split-objects refers not to a physical split such as the idea that objects can always be broken in half or divided, but rather to the split between the virtual proper being of objects or their powers and their local manifestations or qualities. Here the point to be borne in mind is that objects are always in excess of any of their local manifestations, harboring hidden volcanic powers irreducible to any of their manifestations in the world. In this respect, the concept of split-object captures my version of what Graham Harman has referred to as the “withdrawal” of objects. As Harman puts it, “[t]here are objects [...] withdrawn absolutely from all relation, but there is also a ubiquitous ether of qualities through which these objects interact”. 
Harman defends the withdrawal of objects in a much more radical sense than I do here; however, there are strong points of overlap between our positions. Within the framework of onticology, the claim that objects are withdrawn from other objects is the claim that 1) substances are independent of or are not constituted by their relations to other objects, and 2) that objects are not identical to any qualities they happen to locally manifest. The substantiality of objects is never to be equated with the qualities they produce. Thus, as Harman goes on to remark,
Onticology finds much to admire in this passage. Like Harman's object-oriented philosophy, onticology argues that objects or substances are withdrawn from or independent of their relations to other substances. Like Harman's object-oriented philosophy, onticology rejects the thesis of holistic interrelations where objects or substances are understood to be constituted by their relations to other substances. Finally, like Harman's object-oriented philosophy, onticology holds that no relation ever deploys all of the forces contained within an object. The point where onticology and Harman's object-oriented ontology diverge is on the issue of whether the independence of objects or substances entails that objects never touch or encounter one another, or that objects, by virtue of their withdrawal, must be vacuums. Were this the case, it seems that it would be impossible for any object to ever unleash the forces of another object. Given that objects often do unleash forces in other objects, it thus appears that objects must somehow be capable of perturbing one another, while the virtual proper being of an object forever remains in excess of this encounter and is nonetheless closed.
In this chapter, my aim is to articulate the structure of substance and the relationship between virtual proper being and local manifestation in the production of qualities. However, before proceeding to this task it is first necessary to articulate some features of the concept of substance and respond to what Kenneth Burke has called “the paradox of substance”. If Burke’s discussion of the paradox of substance in The Grammar of Motives is here relevant, then this is because what Burke treats as a paradox, and therefore critique of substance, unwittingly provides us with a fundamental clue as to the ontological structure of substance and why it is necessarily characterized by withdrawal.
2.2. Aristotle, Substance, and Qualities
It is often said that Aristotle has an analogical conception of being, holding that being is said in many senses. However, as is so often the case in the history of philosophy, the issue is more complicated than this; for while Aristotle does indeed argue that, for example, we use the term “being” differently when referring to secondary substances (qualities) and primary substances (individual things or objects), Aristotle also argues that the primary meaning of being is that of individual things. As Aristotle puts it in book Ζ of the Metaphysics,
To be, for Aristotle, is to be a substance or a thing. All other senses of being, Aristotle argues, ultimately refer back to substance for ultimately all these other forms of being reside in substances or are made possible by substances. It is this Aristotelian orientation to the being of being as substance or individual thing that onticology, and object-oriented ontology more broadly construed, defends. The question, then, is what precisely is a substance? It is this question that this book seeks to answer.
Elsewhere, in the Categories, Aristotle gives us an important clue as to the nature of substance. There Aristotle writes that, “[a] substance—that which is called a substance most strictly, primarily, and most of all—is that which is neither said of a subject nor in a subject, e.g., the individual man or the individual horse”.  In short, a substance is that which is not predicated of anything else, and which therefore enjoys independent or autonomous existence. Color, for example, is always predicated of a substance. Put differently, color must always reside in something else. The color red is never a substance in its own right, but is always in a ball or a strawberry or lipstick. Qualities reside in substances, they are predicated of substances, whereas substances are not predicated of anything.
Thus Aristotle will remark that, “of the primary substances one is no more a substance than another: the individual man is no more a substance than the individual ox”.  In short, there is an equality of objects, a democracy of objects, in the precise sense that all substances are equally substances. This does not entail that substances are equal to one another, that there are no differences among substances, and that there are not substances more or less powerful than other substances, but rather that all substances are equally substances. When I discuss the concept of flat ontology we will see that this thesis of “equal being” has profound consequences for critical theory and how we practice critical theory. In particular, it entails that we cannot treat one kind of being as the ground of all other beings.
Likewise, when I discuss mereology later, we will see that the thesis that 1) a substance is not itself predicated of anything else, and the thesis 2) that no substance is more or less a substance than any other gives rise to a host of delicate and fascinating problems pertaining to relations between parts and wholes. If objects or substances are not predicated of anything else, then it follows that substances cannot be treated as identical with their parts. Were objects identical to their parts, then this would entail that objects are predicates of their parts. This, in turn, would undermine the autonomy or independence of objects. Consequently, while substances certainly cannot exist without their parts, substantiality must be something other than the parts of which an object is composed. Here we encounter one of the ways in which the realism advocated by onticology is anathema to every form of classical materialism. The sorts of classical materialism defended by thinkers such as Democritus, Epicurus, and Lucretius hold that objects ultimately are their parts in the form of atoms and that these atoms, in their turn, are the only true substances. Onticology, by contrast, argues that scale and whether or not something is an aggregate is irrelevant to whether or not something is a substance. As Harman nicely articulates it, “[n]o privilege is granted to objects over and against mere aggregates, as though atoms were real and baseball leagues only derivative, or individual soldiers real and armies only derivative”.  “Instead”, Harman goes on to remark, “we have a universe made up of objects wrapped in objects wrapped in objects wrapped in objects” such that, “[e]very object is both a substance and a complex of relations”. 
Mereologically this entails that we must develop an ontology capable of maintaining the autonomy or independence of substances from one another such that parts are understood as themselves being substances independent of the whole to which they belong—i.e., they are not merely predicates of the wholes to which they belong—and wholes are treated as independent of their parts. A key feature of each and every object—in fact, a defining feature—is its autonomy. Regardless of whether an object is simple or compound—and onticology strongly suspects that all objects are compound—each object is nonetheless autonomous. As we will see, these seemingly arid ontological issues of the relation between parts and wholes are of surprising importance for a host of issues in social and political theory. Here it is also important to note that “size doesn't matter”. Insofar as no substance is neither more nor less a substance than another substance, it follows, as Harman points out, that atoms are no more nor less substances than molecules, aardvarks or baseball teams.
Insofar as substances are not predicated of anything else, it follows that substances are not in anything else in the sense that qualities are in substances. As Aristotle puts it, “[i]t is a characteristic common to every substance not to be in a subject. For a primary substance is neither said of a subject nor in a subject”.  Substances are not something in an individual thing, but are rather what individual things are. Consequently, all substances have the characteristic of sets whereby sets do not include themselves as a member of themselves. Thus, while all substances are “multiplicities” insofar as they contain parts that are themselves objects—though in a very different way, we shall see, than Badiou proposes in Being and Event—the substantiality of a substance is not itself a part of the substance. Substantiality, rather, is the substance.
Insofar as substances cannot be identified with their parts or the objects which compose them, it follows that substances are always numerically one. As Aristotle puts it, “[s]ubstance, it seems, does not admit of a more or a less”.  A substance is always a substance. As a consequence, a substance is neither more nor less than itself, nor is a substance ever any more than one. In the first instance, a substance is neither more nor less than itself in the sense that when a person gains weight or loses an arm they are still this substance. In the second instance, if a substance is always one then this is because, while a substance might be compounded of many parts or other objects, qua substance the substance is still one substance. Once again, it is clear that this determination of substance raises a number of delicate mereological issues revolving around problems of the one and the many.
Finally, it is a peculiar characteristic of substances that they are non-dialectical. As Aristotle remarks, “[a]nother characteristic of substances is that there is nothing contrary to them”.  Beginning with Hegel, dialectic takes on two meanings that are distinct but often conflated with one another. First, and especially in a Marxist context, dialectic can be taken to refer to thinking that is specifically relational in character. Marx, for example, shows how commodities can only exist in certain social formations characterized by wage labor and capitalism. Later, in our discussion of regimes of attraction and exo-relations we will see how some notion of dialectic in this relational sense can be retained with respect to local manifestations. Second, dialectic can be taken to mean a thinking of relation in terms of contraries and contradictions that are sublated in ever greater wholes or totalities. While onticology readily recognizes the existence of antagonisms, it sees no reason to see antagonisms as the equivalent to contraries or contradictions.
Substances are not defined by contraries or opposites, but simply are what they are. This, of course, is not to suggest that substances do not come into being or that they cannot pass out of being, only that they do not admit of opposed or contrary terms. An individual ncane toad does not have an opposite. Rather, if there is contrariety, it exists only in the domain of qualities. Later, when discussing local manifestation and virtual proper being we will see that there is reason to doubt that contrariety is a genuine ontological category. Insofar as substances are not constituted by their relations, insofar as relations are not internal to their terms, it follows that substances cannot be dialectical in either the relational sense or the sense of contrariety. Contrariety, if it exists, exists at the level of qualities, not substances. It is only through an erasure of substances, through a reduction of substances to their qualities, through the gesture of actualism as discussed in the last chapter, that it can be supposed that substance is dialectical.
This leads Aristotle to formulate another definition of substance that has wide-ranging ontological consequences. We have already seen that substance is that which is not predicated of anything else. In addition to this, Aristotle remarks that “[i]t seems most distinctive of substance that what is numerically one and the same is able to receive contraries”.  Aristotle goes on to illustrate this point with an example: “[A]n individual man—one and the same—becomes pale at one time and dark at another, and hot and cold, and bad and good”.  In short, a substance is that which is capable of actualizing a variety of different qualities while remaining one and the same substance. Later in the Categories Aristotle will remark that qualities are “that in virtue of which things are said to be qualified somehow”.  Here we find confirmation of the onticological thesis that substances are not identical to their qualities for, insofar as substances are able to take on different qualities while remaining the same substance, it follows that objects must be distinct from their qualities.
However, here we must take care. For, in claiming that substances are distinct from their qualities, we do not mean to imply that they are numerically distinct from one another, as if the qualities were one entity and the substance another entity. Speaking of the difference between real distinctions, numerical distinctions, and formal distinctions, Deleuze writes,
While I do not wish to follow Deleuze in his thesis that the difference between numerical distinction and formal distinction is merely a semiological distinction that refers to nothing ontological, Deleuze nonetheless draws attention to a difference between two very important forms that real distinctions take. On the one hand, two things are numerically distinct when they exist independently of one another. As we have seen, all substances are numerically distinct insofar as they are independent of one another. On the other hand, two things are formally distinct if they really are distinct from one another, but they cannot exist independently of one another.
In the case of the relation between substances and their qualities, there is a real distinction insofar as substances are never identical to their qualities. However, the distinction between substances and their qualities is not a numerical distinction but a formal distinction. Here, however, I hasten to add that the formal distinction between substances and their qualities is not symmetrical but rather asymmetrical. As we saw in the last chapter, substances can exist unactualized or without producing any events. As a consequence, substances are not dependent on their qualities, but can exist without any qualities at all (in a form yet to be specified). The contrary, however, is not true. Where substances can exist without their qualities or without producing any events, qualities can never exist without substances in which to exist. Finally, to this we must add that the distinction between substance and quality is not a distinction between what is real and what is not real. Both substances and qualities are entirely real. The point is merely that substances can never be reduced to any of their local manifestations or actualized qualities.
2.3. The Paradox of Substance
As we saw in the last chapter, the ontological category of substance is indispensable to rendering our account of experimental practice intelligible. The practice of experiment is premised on the existence of generative mechanisms, difference engines, or substances that act in open systems, that they can be out of phase with the events they are capable of producing, and that they are separable from their relations to other substances. Nonetheless, it is clear that the concept of substance has fallen into disrepute within philosophy, often being equated with a metaphysical ghost or fiction with no warrant whatsoever.
One of the roots of the disdain with which the concept of substance is today received can be found in Locke's An Essay Concerning Human Understanding. There Locke writes that,
Locke's criticism of the concept of substance spins on the manner in which substance and qualities are split. Within the Aristotelian framework, substance is the ground of qualities, yet we never encounter substance as such, but rather only ever encounter the qualities of substance. From this observation, two problems emerge for Locke: first, what warrant is there for supposing the existence of substance at all? If substance is never encountered at all, if all we ever encounter are qualities, how is substance any different from a reference to Zeus to explain lightning? Second, if substance differs fundamentally from its qualities, what could it possibly be? If substance is stripped of all its qualities aren't we left with a bare substratum, leading to the bizarre and absurd conclusion that all substances are ultimately identical?
Elsewhere, in The Grammar of Motives, Kenneth Burke, in a discussion of Locke, will call this “the paradox of substance”. There Burke writes that, “the word ‘substance,’ used to designate what a thing is, derives from a word designating something that a thing is not. That is, though used to designate something within the thing, intrinsic to it, the word etymologically refers to something outside the thing, extrinsic to it”.  Burke's point is that substance is supposed to be that which is intrinsic to an object, that which makes an object what it is, but that oddly substance ends up being external to the object. If substance, according to Burke, turns out to be external to the object, then this is because we only ever encounter the qualities of the object, and never the substance of the object. If, then, the object is equated with its qualities, then substance turns out to be strangely other than the object.
Locke's critique of substance precipitates something of a crisis that reverberates all the way down to contemporary philosophy today. As Meillassoux remarks, in prior philosophy “one of the questions that divided rival philosophers most decisively was, 'Who grasps the true nature of substance? He who thinks the Idea, the individual, the atom, God? Which God?'”  However, with Locke's critique of substance, this entire debate is thrown into crisis as there no longer seems to be any epistemic warrant for the ontological concept of substance. However, while the ontological concept of substance seems to be banished to the world of occult and unwarranted suppositions with no place in philosophy, individual things nonetheless persist in the world of our experience. Having banished the ontological concept of substance—viz., substances as they exist in their own right, independent of any cognition—philosophy thus finds itself confronted with the question of how to account for individual things without recourse to mind-independent substances inaccessible to experience. Hume, for example, will argue that substance is not a feature of the world—or, at least, any world that we can know—but rather arises from the operations of mind. Having experienced the combination of many similar sensations occurring together in the past, the mind comes to associate these impressions or sensations with one another. In this respect, the object itself, for Hume, is not a substance, but rather the sense that one encounters a substance when encountering an object is instead an effect of how the mind associates impressions and ideas together in a unity. In this way, Hume responds to Locke's challenge by making no reference to “occult entities” independent of what is given in sensation.
We encounter a similar move from world to mind in Kant's Critique of Pure Reason. As Kant observes,
For Kant, the realm of empirical intuition (sensation) is a sort of confused chaos and therefore cannot, contra Hume, provide us with any ordered or structured experience. “[O]ur entire sensibility is nothing but the confused representation of things, which contains solely that which pertains to them in themselves but only under a heap of marks and partial representations that we can never consciously separate from one another”.  Or, as Kant will write when discussing the first analogy, “[o]ur apprehension of the manifold of appearance is always successive, and is therefore always changing. We can therefore never determine from this alone whether this manifold, as objects of experience, is simultaneous or successive”. 
It will be noted that Kant fully carries over the premise of Locke's critique of substance, leaving this critique itself unquestioned. Beginning with the premise that we have no access to substances but only qualities as they are experienced, and with the thesis that the “manifold of intuition” or empirical sensation is unformatted, Kant has no other recourse than to claim that the substantiality of substances is not an ontological feature of objects themselves, but rather issues from our mind. To be sure, Kant endorses the thesis that things-in-themselves exist, but maintains that we have no access to these objects and therefore no means of determining whether, like the objects of our experience, things-in-themselves are autonomous, individual unities, or whether the things-in-themselves are, in reality, really a thing-in-itself, a primordial unity or One, that is then subsequently formatted or “cut up” by our minds. Since the substantiality of substance must issue from somewhere, and since we cannot appeal to being itself to ground substance, Kant contends that substance is instead an a priori category of mind that is imposed on the chaotic manifold of intuition giving it structure or formatting it.
What we have here is what Harman has referred to as the “overmining” of substances. Where undermining dissolves objects in a something that is purported to be more fundamental such as atoms, water (Thales), the One, the pre-individual, and so on, overmining dissolves objects in something that is treated as being more immediate. Of overmining, Harman writes, “it is said that [objects] are too deep. On this view the object is a useless hypothesis, a ‘je ne sais quoi’ in the bad sense”.  In the case of Hume, substances are overmined in favor of impressions or sensations that are then bundled together by associations of the mind, while in the case of Kant, substances are overmined in favor of the manifold of intuition (sensations), along with the pure a priori forms of space and time and the a priori categories of the mind. In both instances, objects or substances are treated as effects of something more immediate or accessible (empirical experience and mind).
It would be difficult to overestimate the impact of Locke's critique of substance and Hume’s and Kant's proposed solution to the paradox of substance on the subsequent history of philosophy and theory. For while direct reference to Locke, Hume, and Kant in subsequent philosophy and theory will often be absent, we nonetheless encounter Locke's critique of substance as an implicit presupposition, and Hume’s and Kant's style of solving this problem throughout contemporary philosophy and theory. Wherever, for example, we are told that it is language that structures reality, we are encountering a variant of Kant's response to Locke. While, to be sure, the content of the critique and the proposed solution differs, the form of the critique remains the same. Here the premises that 1) the ontological category of substance should be banished because we have no direct access to substance, and 2) that the manifold of intuition is a chaotic rhapsody of sensation have been fully embraced and Kant's mind and a priori categories have been replaced by society and language.
Ironically, however, Kant's reasoning is based on an amphiboly, though of an ontological rather than a transcendental sort. In the Critique of Pure Reason Kant tells us that a transcendental amphiboly is, “a confusion of the pure object of the understanding with [an empirical] appearance”.  For Kant, there is both a rationalist and an empiricist way of falling into amphibolous reasoning. Drawing on Leibniz and Locke as examples, Kant argues that “Leibniz intellectualized the appearances, just as Locke totally sensitivized the concepts of the understanding, i.e., interpreted them as nothing but empirical or abstracted concepts of reflection”.  Kant's charge is that Leibniz is guilty of amphibolous reasoning by virtue of finding, directly in sensation, what only issues from the a priori concepts of the understanding. Leibniz treats sensations as if they were identical to what is found only in concepts. For example, sensations are always particular and require the presence of a thing, whereas concepts allow me to think a plurality of things with a shared characteristic in their absence. By contrast, Locke, according to Kant, falls prey to amphibolous reasoning by virtue of arguing that categories that can only be found a priori in the understanding, concepts that can only be generated by mind, can be abstracted from sensation or the domain of the empirical. In both cases, Kant contends, Locke and Leibniz conflate the transcendental structure of mind and the empirical dimension of sensation.
If Kant (and Locke) are guilty of amphibolous reasoning, this arises not from conflating the transcendental (as understood by Kant as a structure of the mind) and the empirical, but rather from conflating the ontological and the empirical. For on the one hand, Locke infers that because substances are not given in experience but rather only empirical qualities are given, we are warranted in banishing substance from our ontology. Likewise, Kant infers that because substance is not given in the manifold of sensation, we must reject the claim that substance pertains to things-in-themselves, but must instead see substance as a category issuing from mind. An ontological amphiboly thus consists in confusing two distinct domains of inquiry: the ontological and the epistemological. The ontological is here subordinated to the epistemological, and the epistemological is then used to determine what is and is not. The problem is that what we can and cannot know cannot be used to legitimately legislate what is and is not. The being of a thing is independent of our ability to know a thing.
However, the problem with the Humean and Kantian solution is much more serious than a mere conflation of two distinct sets of questions or domains of inquiry. Let us take the example of Kant to illustrate this point. Kant's thesis is that the manifold of intuition, being a sort of rhapsodic chaos, cannot deliver the determinations necessary for experience. Rather, this delirious manifold must be structured by a priori categories of mind. And for this reason, these a priori categories of mind cannot be drawn or abstracted from experience, but must instead spring from the mind alone. Whether these categories represent reality as it is independent of our mind is, according to Kant, forever beyond our knowledge because we cannot sneak up on ourselves from behind to see how we see reality and determine whether our experience corresponds to reality. Consequently, whenever we experience an individual thing or speak of an individual thing, this thing is the result of how our mind has formatted the chaotic manifold of intuition through the application of the a priori categories of unity, reality, substance, and existence. And here we must note that these four categories issue from mind not world.
Initially it would seem that Kant provides a clever solution to the question of why our experience is formatted in the way that it is, thereby evading Locke's critique of substance as a sort of occult concept by showing how these concepts issue a priori from the mind (to which we do have access). However, a moment’s reflection reveals that Kant's solution is far more problematic than it first appears. Speaking in the context of Heidegger's early 1919 discussions of being where being is distinguished between “being as a whole” and “something at all”, Harman observes that “no explanation is offered of why certain specific qualities should be assigned to one 'something at all' rather than another”.  This same criticism applies equally to Kant's proposed solution to Locke's critique of substance. Kant has no way of explaining how or why a priori categories such as unity, substance, and existence get applied to one manifold of sensations rather than another. Why, for example, are the categories of substance and unity not applied to an aggregate consisting of my daughter, my parents' dog Rula, and the United Nations? Insofar as the categories are purely a priori, they themselves have no content. What is it then that leads an a priori category to be applied to one thing rather than another? The same problem emerges with those variations of the Kantian solution that would have language rather than pure a priori concepts of the understanding do this work. In both cases we are left without the means of explaining how the “something at all” is ever specified as a concrete entity. As Deleuze puts it in the context of his discussion of Bergson's critique of dialectic and the category of possibility, these categories are “like baggy clothes, [that] are much too big”. 
The point here is not that we have incorrigible knowledge of substances and access to them in our experience, nor that the way we parse the world is the way the world is actually formatted. Rather, the point is that 1) questions of substance are ontological questions absolutely distinct from how we know substances, and 2) that questions of substance cannot be dissolved in questions of access or knowledge. As we saw in the last chapter, ontology cannot be erased by epistemology, nor can ontological questions be transformed into epistemological questions revolving around our access to beings. Wherever one attempts to erase ontological questions in this way, we end up with a variant of Harman's “something at all” problem.
Locke, Kant, Hume and much of the subsequent philosophical tradition ends up where they do precisely because they fall into what Bhaskar calls the “epistemic fallacy” and actualism, confusing questions of our access to beings with questions of what beings are. Beginning with the actualist thesis borne out of a desire for secure foundations (i.e., a desire secondary to the demands of ontology), they restrict discourse to what is given in experience. They then find that they are unable to account for the furniture of the universe precisely because substance is that which withdraws from any givenness, experience, or, indeed, actuality. As such, substance is not something that can anywhere be found in experience—no one has ever seen or experienced, I contend, a single substance—but is rather an irreducible ontological premise necessary if our commerce with the world and experimental activity is to be intelligible. The existence of substance is not something that can be arrived at through an experience or a direct observation, but can only be arrived at as a premise through transcendental argumentation. When we adopt the actualist gesture of restricting knowledge to what is directly given in experience, this way of reaching substance is irrevocably foreclosed.
Returning then to what Burke called “the paradox of substance”, we should not so much argue that Burke is mistaken in his characterization of substance, as that Burke articulates the very essence of substance. In short, we should embrace Burke's characterization of substance as split between qualities and substantiality. It is only when we begin from the standpoint of epistemology, from the standpoint of what is given in experience, that substance appears paradoxical. And if this is the case, then it is because beginning with epistemology leads us to simultaneously claim that the object we experience is its qualities and that it is something radically other than its qualities. However, if we begin from the other end with ontology and note that substance is such that 1) it can actualize different qualities at different times (Aristotle), and that 2) it can fail to actualize qualities (Bhaskar), we can now argue that the very essence or structure of substance lies in self-othering and withdrawal. Insofar as objects or substances alienate themselves, as it were, in qualities, they are self-othering. They generate differences in the world. However, insofar as objects are never identical to their qualities, insofar as they always harbor a volcanic reserve in excess of their qualities, they perpetually withdraw from their qualities such that they never directly manifest themselves in the world. As Harman remarks, it's as if all objects are vacuums populating the universe. It is precisely for this reason that the being of substance is essentially split.
And here it should be noted that onticology and object-oriented philosophy are both metaphysics or ontologies that thoroughly escape what Derrida refers to as ontotheology and the metaphysics of presence. Far from being a signifier that denotes presence or the fullness of being, the very essence of substance is to withdraw from presence and to be in excess of all actuality. However, this overturning of the metaphysics of presence occurs not through a demonstration of the manner in which being always harbors deferral and difference for us such that presence is forever unobtainable, but rather by showing that being as such, being in itself, withdraws in this way. Let us look more closely at this split between virtual proper being and local manifestation through a concrete example.
- Jane Bennett, Vibrant Matter: A Political Ecology of Things (Durham: Duke University Press, 2010) p. 5.
- Graham Harman, The Quadruple Object (Winchester, UK: Zer0 Books, forthcoming).
- Alain Badiou, Logics of Worlds, trans. Alberto Toscano (New York: Continuum, 2009) p. 119.
- Graham Harman, Guerrilla Metaphysics: Phenomenology and the Carpentry of Things (Chicago: Open Court, 2005) p. 76.
- Ibid., p. 81.
- Aristotle, Metaphysics, in The Complete Works of Aristotle (Two Volumes) ed. Jonathan Barnes (Princeton: Princeton University Press, 1984) p. 1028a10–30.
- Aristotle, Categories, Complete Works, p. 2a10–15.
- Ibid., p. 2b25.
- Harman, Guerrilla Metaphysics, p. 76.
- Ibid., p. 85.
- Aristotle, Categories, p. 3a10.
- Ibid., p. 3b35.
- Ibid., p. 3b25.
- Ibid., p. 4a10.
- Ibid., 8b25.
- Gilles Deleuze, Difference and Repetition, trans. Paul Patton (New York: Columbia University Press, 1994) p. 35.
- John Locke, An Essay Concerning Human Understanding (Oxford: Clarendon Press, 1979) pp. 295–296.
- Kenneth Burke, A Grammar of Motives (Berkeley: University of California Press, 1969) p. 23.
- Meillassoux, After Finitude, p. 6.
- Kant, Critique of Pure Reason, p. A176/B219.
- Ibid., p. A43/B60.
- Ibid., p. A182/B225.
- Harman, The Quadruple Object, chapter 1.
- Kant, Critique of Pure Reason, p. A270/B326.
- Ibid., A271/B327.
- Harman, The Quadruple Object.
- Gilles Deleuze, Bergsonism, trans. Hugh Tomlinson and Barbara Habberjam (New York: Zone Books, 1991) p. 44.