I have now thoroughly circumscribed OOP as an exercise in first philosophy—an attempt to provide an ontological and categorial foundation for non-metaphysical explanation—and vindicated my contention that it is both explanatorily impotent—since it ignores explanatory anti-representationalism in favour of brute representationalism—and explanatorily regressive—since it precludes any positive analysis of properties in favour of gastronomic mysticism. Moreover, I have thoroughly articulated the sense in which OOP is methodologically uncritical in its use of basic metaphysical concepts—quality, relation, and even object itself—and the manner in which this is both engendered and obfuscated by its prioritisation of aesthetics as first philosophy—the mixture of metaphors, diagrams, and rhetoric that allusively position the ontological circumscription of thought as its own constitutive exception. However, although I have tentatively outlined the role that ontology has played within the history of metaphysics—as the study of the types of things which exist (regional/applied ontology) and as the study of what it is for them to exist (fundamental/meta-ontology)—and tentatively explained how it intersects with two distinct forms of metaphysical scepticism—Husserlian phenomenology (and its Heideggerian legacy) and Carnapian semantics (and its Quinean legacy)—I have yet to say anything about what metaphysics is, if it is not first philosophy.
The aim of this section, departing from the profoundly negative path we have walked so far, is to try to articulate some positive insights about what metaphysics is and how we should go about it. In doing so, I do not aim to say anything about what results metaphysics should achieve (what an adequate alternative to OOP would look like) but simply to try to see what Harman’s cultivated obliviousness to methodology can teach us about the process of constructing such an alternative.
So, what is metaphysics? As I have already noted, this important question does not get asked often enough, and gets answered even less. We can imagine Socrates prowling through contemporary philosophy departments demanding to know just what it is metaphysicians are up to, only to be confronted with examples of metaphysical problems (e.g., personal identity, the nature of change, the existence of the Divine, etc.) rather than definitions.1 This is not a new predicament though. Unlike logic, epistemology, semantics, or even ethics, metaphysics has always been a loosely defined grouping of problems in search of a definition. To repeat the story told earlier, even the term ‘metaphysics’ is an accident of the way in which Aristotle’s works were catalogued. It originally referred to those books which came after the books on physics, and only later came to denote a subject matter loosely understood to transcend the physical. There is of course a more detailed history of this development to be told, but I shall reach a little further back into the origins of the Western philosophical tradition before I come to it.
Wilfrid Sellars famously defined philosophy as that enterprise that aims to ‘understand how things in the broadest possible sense of the term hang together in the broadest possible sense of the term.’2 This quote is sometimes reached for when the question of defining metaphysics is raised, but it is a poor stopgap. This is not to say that it is a bad definition of philosophy, just that we should not be so quick to identify metaphysics with the discipline of philosophy as a whole, which such appeals inevitably do. To do so is to undermine the whole point of defining metaphysics as a specific philosophical discipline. However, the legitimate intuition underlying this move is that there is something eminently general about metaphysics, even in its specificity. The results of metaphysical inquiry are supposed to have an expansive, if not infinite, range of impact across other disciplines (both philosophical and extra-philosophical). Let us call this issue the question of generality.
If we return to the origin of Western philosophy in the presocratics, we see that Sellars’s definition describes what they are doing pretty well. They are engaged in the first great attempts at synoptic thinking, trying to bring together the various elements of their cultural understanding by means of unitary principles: water, air, fire, etc. However, their thinking is not yet systematic, insofar as they have yet to differentiate the numerous theoretical tasks into which philosophy will eventually divide itself (e.g., logic, ethics, aesthetics, etc.), let alone articulate the various relations between these tasks. The presocratic who deserves special mention is Parmenides, who inaugurates the first and foremost foundational philosophical distinction: that between thought—which includes what we are doing in philosophising—and Being—which names the bare unity of whatever it is that his compatriots were trying to identify with their respective primal elements. This is the conception of metaphysics, if not yet its birth, insofar as its core subject matter (Being) is separated off from the study of the means through which it is to be grasped (thought). It is thus equally the conception of logic, in its most general sense.3 This distinction inaugurates the questions regarding the methodological relation between logic and metaphysics (itself a logical matter, broadly construed) and the substantive relation between thought and Being (itself a metaphysical matter, broadly construed). These questions haunt the history of metaphysics, most clearly in the never-ending and progressively more complicated debates regarding realism and idealism (along with anti-realism, correlationism, and the other positions spawned by this debate).4 We will lump all these issues together under the heading of the question of thought.
Given the conception of metaphysics in Parmenides, we might say that Plato’s work is where it gestates. Moreover, if Socrates was the first thinker within the tradition to begin the task of thinking about the structure of thought (with his pragmatic, dialectical approach to logic), then Plato is the first to begin real systematic thinking about both the structure of thought and the structure of Being, and their relation to one another. This does not yet amount to an explicit differentiation of the systematic task of metaphysics, which is only truly born in the work of Aristotle. However, as we have already noted, the term ‘metaphysics’ appears nowhere within Aristotle’s work. Instead, Aristotle talks of first philosophy (prote philosophia), which is what ultimately becomes the template for metaphysics as it is practised by subsequent philosophers. As such, we can legitimately claim that Aristotle presents the first definition of metaphysics, insofar as he does go to some lengths to define its scope and subject matter. We have already explained this definition (the study of beings as such and as a whole) and its subsequent development (scholasticism and onto-theology),5 but it is useful to point out a further issue that Aristotle’s definition of first philosophy raises, and which I shall call the question of priority. We have already indicated that the tradition is concerned with the question of how metaphysics is related to other disciplines in terms of its generality, and how it is related to the discipline of logic more specifically, but what we have in our sights here is the issue of priority involved in all such relations. Are the results of metaphysics to be foundational for all other forms of inquiry, for some such forms, or for none at all? Those who aim to inherit the title of first philosophy lay claim to some variant of Aristotle’s answer to this question: metaphysics is the foundation of everything else.
If we take only one insight about the metaphysical tradition from Heidegger, it is that there is something important about its own lack of self-consciousness regarding its defining question (the question of Being) and that this ‘forgetfulness’ structures the way metaphysics develops throughout its history. In particular, this forgetting of Being has a complicated relationship to the more or less implicit answers to the questions of generality, thought and priority that the tradition provides for itself as it develops. It is this that we must bear in mind as we return to examining the development of metaphysics after Aristotle’s inaugural definition.
It is perhaps rather hasty to treat scholasticism as a unitary body of doctrine emerging out of the Christian appropriation of Aristotle’s work, but this is precisely what I intend to do. During the scholastic period there is obviously a massive amount of work done in fleshing out Aristotle’s metaphysics, producing numerous variants and several distinct alternatives.6 More importantly, a vast collection of specifically metaphysical problems emerge, creating and tying together the various debates that both constitute the scholastic tradition and provide the backdrop against which the rationalist and empiricist programmes that kick off the modern philosophical era are formulated. Many of these problems emerge out of the specifically theological character of the scholastic appropriation of Aristotle (e.g., the problem of evil), and they are carried forth into the rationalist metaphysics of Descartes, Spinoza, and Leibniz even as they kick away the Aristotelian methodological scaffolding within which they were initially constructed. Here then is the question: which of these purportedly metaphysical problems are genuine problems, and which are pseudo-problems? Once there is a loosely defined metaphysical tradition, whose problems are related by various historically configured relationships of family resemblance, any really novel metaphysical inquiry is faced with a demarcation problem. It must once more undertake to define metaphysics, so as to choose which of the ‘metaphysical’ problems it inherits from its forebears can be retained, and which can be regarded as mere historical artefacts.
However, as we see in the case of the rationalists, as brilliant as these thinkers are (e.g., Spinoza and Leibniz still set the terms of today’s metaphysical debates regarding the nature of modality), it is entirely possible for them to simply ignore the demarcation problem, at least in the form of the demand for a principle of demarcation. They just choose the problems they find most compelling and run with them. This strategy is the essence of what Heidegger calls the forgetting of Being. We will call it the mainstream strategy, because it is most certainly the dominant approach, even in the present day. However, there are two other strategies that also emerge at this point, and which are importantly interlinked: the sceptical strategy and the revisionary strategy. We will introduce these by continuing our dramatisation of the history of metaphysics, beginning with the first (or at least most pronounced) proponent of metaphysical scepticism: David Hume.
Hume is the first philosopher to define metaphysics simply in order to demonstrate its impossibility.7 His principled answer to the demarcation problem is that there are no genuine metaphysical problems. It is interesting to consider this response in terms of the questions of thought and priority we discussed above. Firstly, Hume uses resources from his account of thought to provide an epistemological definition of metaphysical inquiry as that which transcends scientific inquiry grounded in experience. This picks up on the scholastic reinterpretation of metaphysics as that which is concerned with what transcends the physical. He then aims to show that metaphysical questions so understood (most famously those concerning the nature of causality and normativity, or alethic and deontic modality) are strictly impossible to answer outside of the framework of empirical inquiry, and thus should be abandoned as false problems.8 Secondly, Hume gives the first (or most prominent) instance of a particular answer to the question of priority, which comes to define the subsequent tradition of empiricism and its naturalist offshoots: the idea that the results of metaphysics, whatever it is, must be subordinated to those of empirical (or natural) science.9 This position is one that will be articulated in more or less consistent ways following Hume, but it is no exaggeration to say that it has come to dominate the philosophical field, at least in the analytic tradition.10
If Hume is the paradigmatic example of a metaphysical sceptic, then Kant is his metaphysical revisionist counterpart. Kant takes Hume’s various epistemological challenges seriously, and he recognises that, against mainstream metaphysics, the only way to preserve the legitimate parts of traditional metaphysical inquiry that Hume overzealously amputates is to confront the problem of demarcation head on. This is the goal of his project in the Critique of Pure Reason: to carry out a critical delimitation of the problems of metaphysics which thereby articulates the constraints under which any future metaphysical inquiry should be carried out.11 Moreover, not only does Kant attempt to respond to the specific problems Hume raises for metaphysical inquiry (most famously defending the legitimacy of causal and normative theorising);12 he adopts the choices that frame Hume’s own attempt to define metaphysics: he provides an epistemological definition of metaphysics that gives priority to the natural sciences.13
What is fascinating here is the historical interplay between the three strategies. Hume’s epistemological challenge to mainstream metaphysics leads to a revisionary epistemological response from Kant. This then leads to a gradual mainstreaming of Kant’s approach, in which problems from the tradition prior to Hume/Kant are slowly reintegrated into his framework (e.g., Schelling’s metaphysics of Freedom and Hegel’s theology of Absolute Spirit), ultimately resulting in the excesses of German Idealism against which the next round of metaphysical sceptics will react. What is also interesting is that these excesses are largely caused by a shift in focus from the methodological to the substantive form of the question of thought: the relationship between thought and Being ceases to be epistemologically prescribed (transcendental idealism) and instead becomes metaphysically prescribed (absolute idealism). The real value of looking at the history of metaphysics in this way extends beyond giving us a clearer insight into the specific metaphysical (or anti-metaphysical) positions involved in this period: it gives us a loose dialectical schema in terms of which to understand the developments in the twentieth century that are the main topic of the present section.
Returning to the analytic tradition first, it is important to see that the anti-metaphysical correlationism of Wittgenstein and Carnap that we discussed earlier is a sceptical response to both the excesses of German idealism and the metaphysical rejoinders of British realism (e.g., Russell and Moore).14 Given our dialectical schema, we can now see that the relationship between Carnap and Quine is analogous to the debate between Hume and Kant, even though it is specifically concerned with existence rather than modality. The real difference between the two debates is that Carnap mounts a semantic challenge to the possibility of metaphysics, rather than an epistemological one. This is to say that not only does he think it impossible to determine the truth of metaphysical claims (as Hume does), but a fortiori that they make no sense whatsoever. We have already explained that Quine concurs with the greater part of Carnap’s critique of metaphysics: he agrees that metaphysics must be semantically delimited, he agrees that doing so renders most of traditional metaphysics meaningless, and he shares Carnap’s commitment to the priority of science over metaphysics; he demurs merely in providing a revisionary semantic definition of ontology that secures a tiny sliver of traditional metaphysics against Carnap’s sceptical assault. However, it is worth explaining the alternative to anti-metaphysical correlationism inaugurated by this slight divergence—what I call deflationary realism—insofar as it is this Quinean innovation that heralds the return of metaphysics within the analytic tradition.15
Deflationary realism is defined by a certain dialectical strategy for intervening in metaphysical debates between realists and anti-realists. Quine implements this strategy in relation to the traditional debate between platonism (realism) and nominalism (anti-realism) over the existence of numbers and other mathematical objects. Whereas the platonist holds that numbers ‘really’ exist in the same sense as the familiar middle-sized physical objects littering the everyday world, the nominalist denies this, holding that only the latter ‘really’ exist. Quine’s intervention is to claim that neither side knows what it means by ‘really’, and a fortiori that this meaning cannot be extracted from the confused intuitions that drive metaphysical speculation. This is the dialectical significance of Quine’s subtractive suturing of existence to the syntactic regimentation of quantificational variables deployed in natural-scientific theorising. A similar move is made by McDowell in relation to the debate between realism (e.g., Mackie) and expressivism (e.g., Blackburn) over the reality of value properties.16 Whereas the realist holds that the sunset ‘really’ is beautiful in the same sense that it is caused by the diffraction of light through the atmosphere, the expressivist denies this, holding that only the latter is a ‘real’ property. McDowell’s intervention is to claim that there is no way of distinguishing between the truth-aptness of the claims ‘the sunset is beautiful’ and ‘the sunset is caused by the diffraction of light through the atmosphere’ that would not beg the metaphysical question, and thus that their truth or falsity can’t be used to distinguish ‘real’ properties from ‘unreal’ ones. What unites these deflationary moves is the attempt to articulate a realism (or anti-anti-realism) about things in themselves without appealing to a substantive ‘reality’ that must be secured by means of a special supplement (meta-physics) to our ordinary description of them (physics). Deflationary realism thereby abandons the radical scepticism of anti-metaphysical correlationism for a more circumspect suspicion of metaphysics as a positive project.
Given this lingering suspicion, why does the lush garden of analytic metaphysics blossom in the methodological desert of Quinean ontology? This unanticipated reinflation of metaphysical debate is the dialectical parallel of the resurgence of speculative idealism in the wake of Kant’s critical idealism—the mainstreaming of the revisionary moment, in which its methodological purpose is gradually forgotten and traditional problems are slowly reimported into the new framework. To understand why this happens, it is necessary to draw out the answer to the question of thought implicit within deflationary realism, and to see how this slides from a methodological to a substantive conception of the relation between Being and thought, much as transcendental idealism slides into absolute idealism. We must first explain how deflationary realism collapses metaphysics into semantics: its answer to the question of thought is that there is no more to metaphysics—the study of the structure of Being—than semantics—the study of the structure of thought.17 This is made explicit to some extent in Davidson’s account of the relationship between metaphysics and semantics: he moves us from the mundane Quinean ontology of the entities implied by natural science to a relatively innocuous categorial ontology derived from assumptions about universal grammatical categories beyond simple quantifier phrases.18 This is the basis of his deflationary metaphysics of events.19 Quine disagrees with this,20 but he does not find it as horrific as the metaphysical menagerie that Lewis unleashes in the second half of the twentieth century. What happens in the latter is that Quine’s commitment to the priority of the natural sciences is loosened (despite the ubiquity of ‘naturalism’) to allow for greater semantic speculation—admitting naturalistically intractable entities so long as they can be used to make sense of the meaning of the claims made by the natural sciences themselves.21 In essence, once Quine allows us to deploy indispensability arguments derived from the claims of the natural sciences, he opens up the possibility of deploying indispensability arguments derived from the analyses of these claims provided by philosophical and formal semantics.
However, the reason why the real horrors of modern analytic metaphysics emerge with Lewis rather than Davidson is that Lewis’s semantic methodology abandons the residual pragmatism Davidson inherits from Quine and the constraints upon permissible assumptions this supplies. Quine and Davidson not only espouse the priority of natural science over metaphysics, but also the priority of natural science over semantics. They are very scrupulous about the entities they allow explanatory roles within their (consequently very austere) semantic frameworks. The abandonment of these scruples in the transition to Lewis’s possible world semantics is the catalyst for the metaphysical explosion that follows. On the one hand, Lewis is ontologically committed to a plurality of possible worlds containing a menagerie of possible objects existing alongside the actual world and its actual objects. On the other, he is categorially committed to understanding the actuality of the actual world in indexical terms (i.e., ‘actually…’ is equivalent to ‘here in this world…’), and the trans-world identity of individuals in terms of counterpart relations between individuals in different worlds (e.g., ‘I could have been a boxer’ is equivalent to ‘I have counterparts that are boxers’). He justifies these commitments on the basis that they enable him to interpret the semantics of the modal language deployed by natural science in terms of implicit quantifiers over worlds and objects (e.g., ‘1N of force will always accelerate 1kg of mass at a rate of 1m/s2’ can be parsed as ‘there is no 1kg mass in any possible world accessible from our own that when subjected to a force of 1N did not accelerate at a rate of 1m/s2’).22 This subtly inverts the deflationary collapse of metaphysics into semantics, such that instead of providing novel semantic answers to traditionally metaphysical questions (e.g., ‘to be is to be the value of a bound variable’), one provides novel metaphysical answers to traditionally semantic questions (e.g., ‘to have a propositional attitude is to stand in a relation to a set of possible worlds’). In essence, despite his claims to naturalism, Lewis indirectly returns metaphysics to the position of first philosophy, by using it as a dumping ground for supposedly indispensable ontological and categorial assumptions needed for his semantic framework to function.23 Lewis remains a faithful deflationary realist insofar as he denies the claim that possibilia ‘really’ exist, but simply insists that there isn’t any sense of ‘existence’ that we could use to contrast the actual with the possible. Nevertheless, his inversion of deflationism’s priorities constitutes a shift from a methodological to a substantive conception of the relation between thought and Being, wherein this relation is to be articulated by metaphysics itself, rather than being delimited by an antecedent logic. This goes some way towards explaining the more drastic methodological laxity of those he inspires, whose mainstreaming of his ideas gradually unmoors them from the minimal constraints of semantic explanation under which he operates. If metaphysics is first philosophy then it is licensed to define itself, inviting a range of vicious circles (e.g., model-theoretically defining models, appealing to a primitive metaphysical notion of ‘reality’,24 alluringly defining allure, etc.) whose conceptual flexibility allows their proponents to contort their way out of whatever methodological constraints the last round of revisionists imposed upon them.25
Returning now to the Continental tradition, it is important to see how the Heideggerian responses to Husserl’s anti-metaphysical correlationism diverge from the Quinean responses to Carnap’s anti-metaphysical correlationism. It is all too tempting to suggest that the former provide a deflationary idealism that mirrors the deflationary realism of the latter. This symmetry is not unappealing: Heidegger and his heirs are most certainly more indebted to German idealism (e.g., Hegel and Schelling) than to British realism (e.g., Russell and Moore), and there are many deflationary appropriations of its ideas on display in their work (e.g., Derrida’s tantric dialectics of deconstruction without synthesis and Heidegger’s poetic ungrounding of freedom in the ur-event of Ereignis). However, the essence of this philosophical lineage lies in its attempt to revive the Kantian thing-in-itself in response to its dismissal by German idealism and its suspension by Husserlian phenomenology. Seen in this light, the various appropriations of the noumenal that it spawns (e.g., earth/Ereignis, différance, inconsistent multiplicity, Hyperchaos, and real objects) are in many ways more inflationary than deflationary. Although it begins as a critique of metaphysics qua onto-theology, this lineage is more accurately described as metaphysical correlationism than as deflationary idealism, insofar as it aims to metaphysically circumscribe the correlation between thought and Being. This is true even of Heidegger and Derrida, who, in attempting to practically re-orient metaphysics and historically/temporally sublimate the distinction between reality and appearance, inevitably found this reorientation upon a conception of appearance (clearing/presencing) as belonging to a deeper historical/temporal reality (Ereignis/différance). This metaphysical remainder is the seed that germinates in the more obviously metaphysical projects of Badiou and Meillassoux, before blossoming into Harman’s avowedly metaphysical correlationism.
To situate this development within our guiding dialectical schema (metaphysical scepticism → metaphysical revisionism → mainstream metaphysics) it is necessary to reemphasise the distinction between Heidegger’s early revisionism and his later scepticism, in order to explain how Heidegger’s early attempt to redefine metaphysics ultimately develops into a paradoxically metaphysical anti-metaphysical correlationism.26 To this end, it is necessary to specify the type of sceptical challenge to metaphysics posed by Husserl, and to see how it shapes both Heidegger’s initial revisionist response and his ultimate return to scepticism. Although it may initially seem as if Husserl echoes Hume’s epistemological challenge by disavowing knowledge that transcends experience, we have shown that he does not disavow this knowledge so much as practically suspend it: he presents a pragmatic challenge to metaphysics that proceeds not so much by positing a definition of metaphysics as by foreclosing anything that could be used to define it (reality/existence). However, although Husserl’s project is eminently logical in the broad sense of the term, it does not present an account of the pragmatic dimension of thought, and this leaves his sceptical challenge implicit in his methodology. Heidegger’s revisionary response to Husserl’s scepticism consists not in articulating a better definition of metaphysics, but in providing the pragmatic supplement to phenomenology required to go about defining it at all. It is the failure of this attempt to ground metaphysics in a pragmatic phenomenology of the unitary temporal horizon of experience (Temporalität)27 that convinces him of the impossibility of revisionism, and thereby leads him to reconceive this temporal horizon as the ur-historical event (Ereignis) through which we are given unto different metaphysical epochs.
However, although this shift in Heidegger’s work constitutes a sceptical rejection of metaphysics, it is no longer based upon a pragmatics of thought from which this ur-historical structure could be derived, so much as an imperative to pragmatically reorient thought towards it. Heidegger’s abandonment of his earlier metaphysical project is at the same time an abandonment of Husserl’s logical project—he rejects the possibility of a transhistorical account of Being by rejecting the possibility of a transhistorical account of the structure of thought. However, this leaves him with nothing to appeal to but the ur-historical reality of the historical appearance of Being (Ereignis), and no way to appeal to it but through a reorientation of the attitude of philosophy (Gelassenheit) and a reconfiguration of its practice (poetics). The paradox of Heidegger’s correlationism is that it defines metaphysics (the epochal sending of Being) in a manner that outright refuses to define itself (practical reorientation), and thereby alludes to a deeper (metaphysical) reality that the tradition failed to grasp. This paradoxical gesture plays the same role as the methodological circularity encoded in the idea of metaphysics as first philosophy, but it performs the escapological trick more directly, by supposedly situating itself outside of metaphysics while usurping and renaming its role (das seynsgeschichtliche Denken). This usurpation is repeated at each point in the post-Heideggerian lineage that follows: Derrida’s performative evisceration of metaphysics in invoking deconstruction, Badiou’s axiomatic displacement of metaphysics in favour of meta-ontology, and Meillassoux’s tactical subversion of correlationism in the name of speculation.28 The consequence of all this is that the return of metaphysics in the Continental tradition is very much a return of the repressed—the expression of something that was always implicit in the series of partial definitions through which it was disavowed. What makes Harman’s metaphysics so unusual is that it is a return of the repressed without an end to repression—the implicit is announced but forced to remain implicit, as allusive circularity is substituted for paradoxical usurpation.
Taking these dialectical developments into account, then, the question of what metaphysics is becomes an invitation to revise the discipline by providing a response to the demarcation problem. I cannot provide a comprehensive solution to the problem here, but I will attempt to provide an outline of a response that synthesises the historical story just told with the various insights provided by my engagement with the conceptual roots of Harman’s metaphysics.
Returning to the parallel between traditions, we can now see that each lineage traces an arc from anti-metaphysical correlationism back to mainstream metaphysics, and that, although the intermediary positions this arc passes through are different (deflationary realism/metaphysical correlationism), it always begins with a methodological concern with the role of reality in traditional metaphysics (the deflation of ‘real’/the ‘reality’ of appearance), before performing a methodological sleight of hand in which the status of reality is apparently secured without its ever being defined (circularity/usurpation), and concluding with the methodological unravelling of the constraints implicit in the initial concern (semantic speculation/pragmatic allusion-subtraction). Given this, we can see that Harman’s metaphysics does not so much stand outside this trajectory as present its inevitable conclusion: it simultaneously announces the implicit essence of metaphysics by invoking the ‘real’, and subsumes it within a metaphysical opposition (with the ‘sensual’) while allusively containing the recursive bifurcation of sense unleashed by this deceptive gesture (e.g., really-[sensually-[really-[sensually-[...]]]]). I have already suggested a putatively non-metaphysical alternative to Harman’s use of ‘real’, but I have yet to situate it within my broader historical narrative. In doing so, I aim to show how it is possible to provide a non-metaphysical account of ‘reality’, and how this enables us to define metaphysics without resorting to either circularity or usurpation. I will do this by returning to the questions of generality, thought, and priority from within this perspective.
To begin with, I think that the generality of metaphysics is encapsulated by Aristotle’s original invocation of the connection between beings as such and beings as a whole, and Heidegger’s attempt to formulate the question of Being as the inquiry into the unitary structure of beings as such and as a whole. We have seen that there are roughly two strategies for addressing this connection/structure, corresponding to two different ways of objecting to the onto-theological conception of the whole: the reification objection, which encourages us to think beings as a whole in terms of beings as such (Badiou and ontological liberalism) and the de-abolutisation objection, which encourages us to think beings as such in terms of beings as a whole (Heidegger and ontological conservatism). Moreover, we have seen that each of these strategies can be pursued either subtractively (Badiou and Quine) or allusively (Heidegger and Harman), and that the common element they thereby share is an appeal to something that cannot be made explicit (i.e., Badiou: deference to axioms; Quine: deference to science; Heidegger: use of poetics; Harman: use of diagrammatics).
The task of defining metaphysics demands that we overcome these appeals to the implicit, but the question is whether it is possible for either strategy to do so without reverting to onto-theology. I doubt whether it is possible to think beings as such directly without either implicit definition, metaphorical allusion, or a reversion to thinking in terms of a highest genus of beings, although I hesitate to claim that it is strictly impossible. However, we have seen that not only is it possible to think beings as a whole directly without conceiving this whole as a being, but that extant attempts to do so (Deleuze and DeLanda) actively aim to explicate, integrate, and revise the systems of reference implicit in natural science to which Quine subtractively deferred. This sort of approach aims to think the Whole not as the totality of what can be thought (objects), but as the totality of what really exists (beings), or to think the Whole as Reality. I propose that metaphysics is eminently general precisely insofar as it is concerned with the fundamental structure of Reality, and that it therefore need not say anything about those objects that aren’t real.
Of course, this distinction between objects and beings assumes that we can provide a sufficient definition of ‘real’ in non-metaphysical terms, but it equally suggests that this problem is to be understood in terms of the relation between thought and Being. The claim that not all objects of thought are beings implies that the study of objects qua objects (logic) is independent of the study of beings qua beings (ontology). It is this independence of logic from metaphysics that promises to free us from circularity, insofar as it opens up the possibility of using logic to define metaphysics by logically distinguishing thought about the real (beings) from thought about the unreal (nonbeings) rather than metaphysically distinguishing the real from the sensual (as types of beings). This sort of separation and articulation of the relation between logic and metaphysics occurs in each of the revisionary moments we have so far considered (Kant, Quine, and Heidegger), but varies as to the manner in which the logical task is approached: in terms of knowledge about objects (epistemology); in terms of the content of this knowledge (semantics); or in terms of the practices through which these should be understood (pragmatics). These revisionary moments are mainstreamed when their initial separation/articulation comes to be conceived metaphysically rather that logically, or when metaphysical assumptions are incorporated into their account of knowledge, content, or practice.
It is worth noting that Heidegger is unique not only in mainstreaming his own revisionary moment, but in instigating the turn to usurpation as an alternative to circularity in the process. His Nietzschean insistence that theory is a form of practice enables him to stipulate that his own practice of theorising metaphysics is distinct from the practice of metaphysical theorising; this practical difference enables the former to usurp the latter without circularity. It is the same Nietzschean pragmatism that enables him to assert the impossibility of providing a transhistorical account of Being (metaphysics) by asserting the impossibility of providing a transhistorical account of thought (logic); the latter assertion is responsible for the degeneration of his earlier pragmatic phenomenology into his later poetic thinking, and therefore also responsible for the repression of metaphysics within his poetic allusions to Ereignis (as the reality of appearance). In dissolving logic into historicised pragmatics, Heidegger eschews the only way to define metaphysics that has enough reflexive purchase upon itself to avoid collapsing back into metaphysics either by embracing it (circularity) or displacing it (usurpation). If we are to avoid both of these pitfalls, our only option is to secure logic’s independence from metaphysics—and this means considering the relationship between epistemology, semantics, and pragmatics with which the revisionary moments confront us.
The truth in Heidegger’s turn to pragmatics is that it provides the only way to avoid circularity and the forgetting of Being that results: the only way to differentiate between forms of thought (i.e., logic and metaphysics) that cannot be reinterpreted as a (metaphysical) difference between their objects is to differentiate them as forms of practice. This methodology parallels Fichte’s foundational insight that the difference between practical reason and theoretical reason is itself a practical one.29 The problem with Heidegger’s turn to poetics is that it severs the link between practice and practical reason through which logic can reflexively secure itself: Heidegger abandons his attempt to ground epistemology and semantics in pragmatics by describing how knowledge and its content is situated in practice, and in so doing transforms ‘thought’ into a mysterious practice that no concrete pragmatics can describe. The alternative is to reconnect epistemology, semantics, and pragmatics in a unified logical pragmatism, and to use the account of thought this provides to define metaphysical knowledge by means of its specific content, and the specific practices that constitute this content. Given our answer to the question of generality, this means providing a pragmatic response to the semantic challenge posed by deflationary realism: we can escape the choice between circularity and usurpation by developing a pragmatically grounded semantics for the term ‘real’ that is sufficient to explain what we mean when we ask metaphysical questions (e.g., ‘What is the fundamental structure of Reality?’). This is what I have elsewhere described as moving beyond deflationary realism to transcendental realism.30
Providing this semantics is beyond the scope of the present work, but I will try to say something about how it relates to the question of priority. The important thing to understand is that the metaphysical project of organising a unitary system of reference through providing a unified account of Reality (Deleuze and DeLanda) essentially aims to make explicit, to integrate, and to revise what Quine either leaves implicit in scientific practice, or defers to the syntactic regimentation of its quantificational variables. This is to say that metaphysics and natural science both describe the same unitary structure: Reality is Nature. However, this does not mean that Reality is to be explained in terms of some antecedent conception of Nature (e.g., as material/energetic, mechanistic/vital, corpuscular/flux etc.), which would simply be more implicit metaphysics, but that we are to understand what ‘Reality’ means in terms of the pragmatic structure of natural science, and what ‘real object’ or ‘being’ means in terms of what it is to be referenced within natural-scientific explanation.31
This enables us to draw a threefold distinction between the implicit metaphysics that is more or less passively effected by natural science, the explicit metaphysics that aims to actively intervene in natural science, and the critique of metaphysics that enables us to move from the former to the latter by articulating the questions through which we explicate, integrate, and revise the fundamental structural assumptions of natural science (e.g., ‘What are beings?’, ‘What is essence?’, ‘What is causality?’, etc.). It is in terms of this relation between metaphysics and its critique that the methodological relation between Being and thought is to be articulated. Although there is certainly more to metaphysics than logic, there is an important sense in which logic constrains metaphysics. Metaphysics seeks to understand what Nature is (the question of Being), whereas logic seeks to understand what ‘Nature’ means (the question of the meaning of ‘Being’). This extends to the discussion of particular metaphysical categories and the problems that correspond to them: identity, difference, individuality, universality, quantity, quality, relation, essence, space, time, part, whole, causation, etc. Each of these corresponds to a metaphysical problem (e.g., ‘What are relations?’), the scope of which is determined by the logical analysis of the relevant category in each case (e.g., ‘What does “relation” mean?’). These metaphysical categories are derived from logical categories by means of the concept of reality, such that the metaphysics of relations is distinguished from the logic of relations insofar as it is concerned with ‘real relations’ much as ontology is concerned with real objects (beings).32 The broader range of categorial questions this opens up expands the scope of metaphysics beyond ontology by transforming its account of Reality/Nature from a system for organising scientific reference to a system for organising scientific explanation as such. This gives us some theoretical purchase upon the balance between continuity and autonomy in the relation between metaphysics and science.
The continuity between metaphysics and science consists in the fact that metaphysical ideas are already implicit in the scientific enterprise. Natural science always proceeds with some implicit understanding of what beings are, what essence is, what causality is, and so on; and this implicit understanding is itself subject to revision in the ongoing process of scientific inquiry, in more or less explicit ways. Einsteinian relativity fundamentally challenged our implicit metaphysical understanding of space and time, and the subsequent developments in physics have raised serious questions regarding how we should understand causality. The Darwinian revolution in biology has forced us to rethink the very way in which we understand the idea of types, and thus the notion of essence. Dynamic systems theory has provided us with alternative ways to conceive of the modal features of entities, and its development and extension in the field of complexity theory is forcing us to rethink our understanding of mereological relations. And this is all before we even begin to consider the conceptual puzzles generated by the counterintuitive logic of quantum mechanics.33
Thus, metaphysics is already present in natural science, it just hasn’t been made explicit as metaphysics. It is the possibility of doing metaphysics explicitly which preserves its relative autonomy within natural science. There are two features which distinguish the proper practice of metaphysics from its implicit form: criticality and systematicity. Metaphysics proper is critical insofar as it properly delimits the various questions with which it is concerned and the ways they are related, from an a priori standpoint (logic). Metaphysics proper is systematic insofar as it attempts to provide a unified answer to all of these questions which takes into account the whole variety of a posteriori considerations provided by the various natural sciences. This is just to say that it attempts to unify the various metaphysical debates implicit within the natural sciences as a whole. The autonomy of metaphysics stems from these two features. On the one hand, metaphysics has a distinctive relation to a priori considerations that are independent of the natural sciences (the critique of metaphysics). On the other hand, it is the most abstract form of a posteriori discourse, situated within the natural sciences only insofar as it plays a unifying and organisational function in relation to them. Conceived this way, metaphysics stands in a reciprocal relationship with science: it is in a position to provide the abstract conceptual foundations which organise them, while at the same time it must be sensitive to their subject matter, insofar as it is through this sensitivity that its concepts remain open to revision. This complex reciprocity is my alternative to the blunt foundationalism proposed by those who treat metaphysics as first philosophy.
1. This is invariably the way that metaphysics is taught to undergraduate students, a fact which is attested to by the lack of a solid definition in many basic textbooks on metaphysics. In its place, we find a loose taxonomy of problems, arranged by family resemblances.
2. W. Sellars, ‘Philosophy and the Scientific Image of Man’, in Science, Perception, and Reality (Reseda, CA: Ridgeview Humanities Press, 1991), §1.
3. It is this broad sense of ‘logic’ that Hegel is referring to in the title of his Science of Logic. It should not be confused with formal or mathematical logic, which is an important part of the broader study of thought. I shall have more to say about the latter in chapters 3.6 and 4.1.
4. The framing of these issues in terms of the debate between realism and anti-realism (principally in logical terms) obviously emerges out of Dummett’s work (cf. The Logical Basis of Metaphysics [Cambridge, MA: Harvard University Press, 1991]), whereas the framing of these issues in terms of the debate between correlationism, realism and idealism (principally in metaphysical terms), emerges out of Quentin Meillassoux’s work (cf. After Finitude).
5. Chapter 3.4, subsection I.
6. Neoplatonism and nominalism deserve special mention here.
7. Cf. David Hume, An Enquiry Concerning Human Understanding, in Enquiries Concerning Human Understanding and the Principles of Morals (Oxford: Oxford University Press, 1975), §1.
8. See chapter 3.2, subsection III.
9. Precisely how one defines philosophical naturalism is a difficult question. However, at minimum, I would suggest that it is precisely the disentangling of this response to the question of priority from the particular epistemological prejudices of empiricism.
10. Cf. Brandom, Between Saying and Doing, chapter 1.
11. Sadly, few people read Kant’s Metaphysical Foundations of Natural Science, which is meant to be his positive contribution to metaphysics (along with the Opus Posthumum), rather than his negative delimitation of which positions are impermissible.
12. Again, see chapter 3.2, subsection III.
13. It is because of this that eminent Kantians such as Sellars can legitimately claim a prominent naturalist pedigree.
14. This relationship to German idealism is mediated by British idealism, to which Russell and Moore were responding, and German neo-Kantianism, which was the principle reference point for the Vienna Circle in conceiving their break with the tradition.
15. For an extended discussion, see my ‘Essay on Transcendental Realism’.
16. Cf. J.L. Mackie, Ethics: Inventing Right and Wrong (London: Penguin, 1990); S. Blackburn, Spreading the Word (Oxford: Oxford University Press, 1984); and J. McDowell, ‘Values and Secondary Qualities’ and ‘Projection and Truth in Ethics’, in Mind, Value, and Reality (Cambridge, MA: Harvard University Press, 1998), 131–50 and 151–66.
17. This position is actually best articulated by Brandom as what he calls objective idealism (‘Holism and Idealism in Hegel’s Phenomenology’, in Tales of the Mighty Dead, 178–209).
18. Cf. D. Davidson, ‘The Method of Truth in Metaphysics’, Midwest Studies in Philosophy 2:1 (1977), 244–54; and ‘On the Very Idea of a Conceptual Scheme’, Proceedings and Addresses of the American Philosophical Association, Vol. 47 (1973–74), 5–20.
19. D. Davidson, ‘The Individuation of Events’, in N. Rescher (ed.), Essays in Honor of Carl G. Hempel (Dordrecht: Reidel, 1969), reprinted in Essays on Actions and Events (Oxford: Clarendon Press, 2001).
20. Cf. W. V. O. Quine, ‘On the Very Idea of a Third Dogma’, in Theories and Things (Cambridge, MA: Harvard University Press, 1991).
21. This is enabled by the paradigm of model-theoretic semantics that grew out of Tarski’s work, whose applicability to natural language semantics was initially championed by Davidson but taken in very different directions by Kripke, Lewis, and Montague, amongst others.
22. This example is obviously an oversimplification of both the relevant physical law (it takes an instance of the law [1=1/1], rather than the law itself [F=ma]) and Lewis’s semantics (it ignores both the issue of tense and temporality, and the issue of potential defeasors). However, it gives a rough idea of how the quantificational machinery underlying Lewis’s modal semantics works, and importantly, just what work is done by the accessibility relations that restrict these quantifiers. For more details, see On the Plurality of Worlds (Oxford: Blackwell, 1986) and Counterfactuals (Oxford: Blackwell, 1973).
23. Unconstrained model theory just is metaphysics as first philosophy.
24. Cf. K. Fine, ‘The Question of Ontology’, in Chalmers, Manley and Wasserman (eds), Metametaphysics, 157–77.
25. It is worth pointing out that there is a more recent metaphysical dialectic in the analytic philosophy of science. This begins with Bas van Fraassen’s constructive empiricism, which presents a distinct epistemological challenge to the possibility of metaphysics (The Scientific Image [Oxford: Oxford University Press, 1980]). Ladyman and Ross’s rejection of the mainstream analytic metaphysics represented by Lewis can be seen as a revisionary response to van Fraassen, one that attempts to engage with his epistemological framework in order to demarcate a scientifically informed metaphysics (Every Thing Must Go, chapter 1). There is more that could be said about the novelty of Ladyman and Ross’s definition of metaphysics, but I will have to leave that for another time.
26. See chapters 3, 4, and 5 of my The Question of Being.
27. Cf. Basic Problems of Phenomenology.
28. We might also include Laruelle’s axiomatic suspension of philosophy’s sufficiency in opening up the domain of non-philosophy. See above
29. Cf. J.G. Fichte, The Science of Knowledge (Cambridge: Cambridge University Press, 1982), first and second introductions; and J. Dunham, I. H. Grant, and S. Watson, Idealism: The History of a Philosophy (Durham: Acumen, 2011), chapter 6.
30. See my ‘Essay on Transcendental Realism’ for a more complete (if early) version of this story.
31. One might wonder why ‘natural’ science is privileged (including physics, chemistry, biology, and other sciences founded upon experiment), and not ‘mathematical’ science (including the various areas of pure mathematics, formal logic, and computer science that are independent of experiment). There is a more complicated answer to this question, but the simple version is that Badiou is right to say that mathematics is not strictly concerned with mathematical objects, even if his choice of set theory cuts against this insight to some extent. Rather, mathematics is concerned with mathematical structure, which, as Fernando Zalamea has pointed out, is far better viewed from the perspective of category theory than set theory (Synthetic Philosophy of Contemporary Mathematics [Falmouth and New York: Urbanomic and Sequence Press, 2012]). Mathematics is neither excluded from metaphysics nor the whole of metaphysics (or ontology), but is rather a crucial component of the project of describing the fundamental structure of Reality qua structure.
32. The other way of parsing this is to use ‘really’ as a copula modifier that makes the metaphysical character of categorial questions explicit: ‘What are properties?’ could be read as ‘What really are properties?’ as easily as ‘What are real properties?’ It is also worth pointing out that my account of the relation between logical and metaphysical categories parallels Kant’s distinction between general and transcendental logic (Critique of Pure Reason, A55–57).
33. See chapter 3.4, n. 103.