Daniel Pucciarelli[1]


Abstract: The paper explores the meaning and fate of the category of critique within the so-called “ontological” or “speculative turn”. First, the article addresses the question of knowing what could give consistency to the very concept of a contemporary ontological turn, in order to present, in a second moment, the tension that, since Kant, constitutes the relationship between critique and ontology. Finally, the paper explores – particularly in the work of Quentin Meillassoux – the possibility of a new ontology that claims to be both non-dogmatic and immune to the neutralization of ontology carried out by critique.


Keywords: Critique. Criticism. Ontology. Correlationism. Ontological turn.



Despite its explicit anti-metaphysical orientation, contemporary philosophy has not abandoned ontological research, traditionally defined as a primary subdiscipline of metaphysics. On the contrary, what we see in diverse traditions of 20th century philosophy is, rather, a manifest return to ontology, even if coupled with an effort to untie it from the scope of traditional metaphysics. In this regard, it is enough to think of two famous ontologists as philosophically disparate as Heidegger and Quine: in both cases, there seems to be an explicit attempt of refoundation of ontology as discipline without subordinating it, or even in outright opposition to classical (pre-Kantian) metaphysics. It would be difficult to overestimate the relevance of these philosophers – and, consequently, of the very place occupied by the concept of “ontology” – for contemporary philosophy.

Presumably, a new movement of return to ontological speculation is taking place in some original currents of thought of the last two decades.[2] Obviously, we are dealing here with a different theoretical environment and also with other motivations in comparison with the ones of the 20th century. But what is taking place now is arguably a new type of ontological turn or a speculative turn that loosely brings together in a common umbrella philosophies as disparate as the speculative philosophy of Quentin Meillassoux, the new realism of Markus Gabriel, the transcendental naturalism of Iain H. Grant, the object-oriented ontologies of Graham Harman and Tristan Garcia, to name just a few. Here too we observe an effort of rethinking ontology in a strictly contemporary framework, and not to rehash it as a subdiscipline of classical metaphysics.

Evidently, there is no clear consensus on what this “strictly contemporary” framework would be. Moreover, there is even no consensus if an ontology in its own sense, e.g., a doctrine of “being qua being”, would be possible at all inside the basic anti - metaphysical horizon which seems to dominate contemporary philosophy. Indeed, the mere concept of a science of “being qua being” sounds to the contemporary ear as something anachronistic and irrecoverable. Therefore, we can understand the concept of ontological or speculative turn not so much in a descriptive sense, but rather in a desiderative or programmatic sense: first of all, there seems to be an open field of problems and debates around what would be the concept and the current status of ontology.

This text is part of a larger research on this open field of problems and debates. In particular, I am interested in the meaning and fate of the category of critique itself – and of those traditions of thought that, since Kant, are umbilically associated with it – in contemporary ontological thought. In this paper, I will take especially Meillassoux's thinking as my object of analysis, as it has been consistently indicated as the most influential among these new ontologies. I will start with an exposition of his concept of correlationism, which arguably could give consistency to the very notion of an ontological or speculative turn taking place in the last decades (I). This exposition enables me to explore the intratheoretical effects of the correlationist thinking matrix – taking Kant and Adorno as paradigms – to the very idea of ontology in general (II). Finally, I will discuss in what sense we can speak of a post-correlational or post-critical ontology[3] that at the same time would abandon and preserve the critical device (III).


1 Correlationism

Given the plurality and heterogeneity of the various currents of thought explicitly or implicitly affiliated with the so called ontological or speculative turn, these currents of thought could hardly be subsumed to a cohesive school or an univocal orientation of thinking. Indeed, the authors associated with the movement have repeatedly rejected the very idea of belonging to a school. If there is a general convergence – albeit precarious – between them, it seems to be an unity of diagnosis about the state of the art of 21st century philosophy. In general, this diagnosis indicates the general exhaustion of what Meillassoux called correlationism. Thus, it is perhaps not excessive to say, with Graham Harman, that correlationism has become something like a common enemy unifying the various ontological trends in development today: “Speculative realism is best understood as a loose umbrella term for a series of vastly different philosophical enterprises. What all have in common is their rejection of what Quentin Meillassoux first termed ‘correlationism’.” (Harman, 2011, p. vii).

The term correlationism designates a set of philosophies based on the same theoretical assumption, namely: that being and thinking are unsurpassably entangled. Historically, Berkeley would have been the first philosopher to formulate a model of correlational thinking; Kant’s Copernican Revolution, however, would have been the first major systematic monument of a true Era of Correlation that – the diagnosis follows – has become the hegemonic horizon of thinking from the 19th century until today[4]. In Meillassoux's own definition:

By “correlation” we mean the idea according to which we only ever have access to the correlation between thinking and being, and never to either term considered apart from the other. We will henceforth call correlationism any current of thought which maintains the unsurpassable character of the correlation so defined. Consequently, it becomes possible to say that every philosophy which disavows naïve realism has become a variant of correlationism. (Meillassoux, 2008, p. 5).


Of course, this general thesis about the necessarily relational status of thinking and being doesn’t need to be explicitly discussed – and often it is not – for a philosophy to be structurally correlational. Likewise, the specific manner by which such a philosophy individually defines the correlation terms is also irrelevant for its correlationist setting, since that philosophy endorses the supposedly insurmountable entanglement or co-belonging between the terms – which can be and have effectively been defined as a correlation being-thought, subject-object, sense-reference, noesis-noema, mind-world, etc.

Thus understood, we can see how correlationism can effectively be considered, as Graham Harman (2011, p. 87) and others have argued, as a philosophy of access: if the correlation between being and thinking is postulated as unavoidable, then the very access to an instance of reality or knowability radically apart of thought is in principle undermined. In fact, correlationism makes an astute and seemingly irresistible correlational two-step which consists in affirming that, by thinking any given object, by converting it into a mere object of thought, that object will necessarily be posed in this constitutive entanglement with thought. Every object of thought, in other words, for the simple fact of being an object of thought, must always be found within the correlation, regardless of how we conceive this interiority: the immanence of consciousness, intersubjectivity, language, the horizon of a culture... The essential consequence of this procedure is quite evident, and it has profound effects for all later philosophy: the tendency to convert all in-itself independent of thought into a for-us, the dissolution of all supposedly nonrelational absolutes in correlational entities or realities.

It is evident that, as a kind of umbrella concept, correlationism cannot encompass the specificities of all models of thought that, in one way or another, followed Kant's Copernican Revolution. At the outset, we must distinguish between correlations that preclude any concept of the Absolute, on the one hand, and correlations that operate with critical (or intracorrelational) notions of the Absolute, on the other, as it is the case in German Idealism. Likewise, we should be able to conceptually articulate the materialistic – that is, non-correlational – desiderate of certain models of post-Kantian thought that, spite materialists, did not abandoned the critical paradigm. Indeed, Meillassoux himself has sought to develop a broad typology to systematically classify the main subtypes of correlationism: weak and strong correlationisms, hybrid correlationisms, subjectalism, among others. It is not my goal here to exhaustively discuss this typology and not even to rule out a possible initial schematic outline of the concept.[5] Its virtue, I believe, relies on the fact that, on the one hand, the concept of correlationism is wide enough to encompass historical and philosophical trends that, being so hegemonic, have been gradually naturalized, and, on the other, that concept is sufficiently determined and consistent to be operational – and to be able to base a diagnosis on the state of the art of current philosophy.

For if there is something that unifies the ontological turn or the contemporary speculative turn, we argued, it is precisely the general diagnosis that correlationism has reached a state of profound exhaustion. This exhaustion results, on the one hand, from correlationism’s internal dynamics, which, since Kant's first critics, effectively involved a permanent movement of self-criticism that fatally saturated its explanatory possibilities. On the other hand, this exhaustion results also from external constraints such as the emergence of scientifically founded discourses on extracorrelational realities – as, for instance, an ancestral past prior to the emergence of all consciousness and even all life in Earth, which Meillassoux calls ancestrality –, which seems to put in question the very epistemic foundations of correlationism.[6] Hence, the task of current philosophy would be that of creating conceptual resources to abandon the horizon of correlationism and finally regain the Great Outdoors, that is, a radically nonrelational reality, severed from the scope of thought by critique.

Obviously, however, it is not a question here – the diagnosis follows – of going back to a pre-critical matrix of thought, such as pre-Kantian metaphysics or any kind of naive ontological realism. Rather, the challenge of the ontological turn consists, as already mentioned, of overcoming correlational thinking without falling into a reheated variant of dogmatism. In a sense, it is a matter of formulating a new model of post-critical or post-correlational ontological reflection that, at the same time, abandons and preserves criticism – if that is possible at all. Indeed, for those who seek to think with their ears, the very concept of post-critical ontology already poses a hard tension. Then it seems to belong to one of the most elementary meanings of the concept of critique, as it has been formulated at least since Kant, its emphatic opposition to the very idea of ontology. Before turning to a first effort of explaining what such an ontology can be, let us quickly remember the fundamental meaning of critique in its relation to the ontological discipline.


2 Two correlational criticisms of ontology

2.1. Kant

Kant also develops a diagnosis about the exhaustion of the thought of his time – the famous alternative between dogmatism and skepticism – in order to open a third way for philosophy. Kant called this third way the critical path, and sentenced: “[…] the critical path is the only one which is still open.” (KrV, B884). As is well known, the critical procedure consists in using the self-reflective dimension of rationality to investigate the formal limits of its properly cognitive function – limits, however, which are systematically exceeded by the ordinary use of reason. In other words, to the extent that rationality itself is endowed with a self-reflective function allowing it to turn on itself and investigate its own formal procedures regardless of its application to objects, it becomes possible to establish a second-order discourse that will draw a dividing line between the thinkable and the knowable. This second-order discourse is the critical investigation itself in its Kantian sense. And it is precisely this dividing line between the thinkable and the knowable, in turn, that will allow us to avoid the consequences of both the dogmatic path and its inability to demonstrate the foundations of its own assertions about transcendent objects, on the one hand, and the skeptical path and its virtual dissolution of the cognitive potential of rationality, on the other.

This mere summary description of the Kantian critical project already allows us to visualize its profound consequences for virtually all philosophical disciplines. Let us limit ourselves to ontology. In determining the limits of the knowable in relation to the functions of our epistemic faculties – now understood as constituents in the knowledge process –, the critical procedure submits the classic philosophical question about what it is to the critical question par excellence about how it is possible to know. As a result, a reconversion or, more precisely, a virtual replacement of ontology by epistemology takes place. As Kant stated in a famous passage: “[…] the proud name of ontology, which presumes to offer synthetic a priori cognitions of things in general in a systematic doctrine (e.g., the principle of causality), must give way to the modest one of a mere analytic of the pure understanding.” (KrV, A247 / B304, our emphasis). In the place once occupied by ontology, Kant introduces this new discipline that cannot research “being qua being”, but only the concepts of understanding through which any predicative relation to entities is possible at all. In contrast to traditional ontology, moreover, analytics of understanding cannot relate its concepts to things in general, but only to objects of possible experience. Accordingly, its function consists in presenting the principles of the exposition of phenomena, and not the fundamental laws of being as such. Also in contrast to rationalist metaphysics, an analytics of pure understanding presupposes a transcendental aesthetics that can only provide understanding with the material of phenomena; both are required for knowledge production.

In other words: one indeed couldn’t say goodbye more clearly to traditional ontology. Kant seems to insist here that an ontology as theory of the modes of being as such is simply not possible according to critique. Indeed, an analytics of understanding is a completely redesigned discipline that now substitutes traditional ontology, taking its systematic role and especially making completely different knowledge claims. Of course, this does not mean that “being qua being” (or, in Kant's terminology, the thing itself) is not thinkable and, even more, not thinkable without contradiction – as it is declared non-knowable as such, however, critique carries out a dissolution of the ontological discipline that had it as its object, at least in its traditional configuration.[7]

            If the Kantian critique undermines the foundations of traditional ontology comprehended as a discipline imbued with extracorrelational traction, it cannot be said, however, that it avoids the formulation of any ontology at all. In fact, the very immediate history of the effects of Kantian critique saw the emergence of modified concepts of ontology that were still understood under the auspices of criticism – think, in this regard, of the Hegelian Absolute Idealism, which gives ontological status to logic, or of the great ontologists of the 20th century already mentioned, such as Heidegger and Quine.

Is it the same ontology here that “must give way to the modest one of a mere analytic of the pure understanding”? In other words, are these ontologies intended to demonstrate a truly extracorrelational dimension or should they rather be called "ontologized epistemologies"? In this latter case, why should we call them ontologies at all? As Herbert Schnädelbach states:

Is all ontology that respects Kant's saying that “the critical path is the only one which is still open” – and the new ontology intends to do so – not condemned to failure? Any answer to the question of whether there was an ontology in general, after Kant, which is more than the dogmatic conservation of outdated positions, will depend on what is understood under the term “ontology”. (Schnädelbach, 1983, p. 235).


2.2 Adorno

In general, it is this kind of questioning from which departs Adorno in his dialectic modality of critical theory; dialectics, moreover, which didn’t understand ontology merely as an object of criticism among others, but which understood itself in a structural mediation with ontology: “[…] the transition to dialectics must, in fact, reside in the critical self-reflection of ontology.” (Adorno, 2008, p. 12s).

Let us summarize the main steps of Adorno's criticism of ontology. Despite all polemics against Heidegger, his criticism is not intended to be directed against an ontology among others, but to the very conditions of possibility of ontology in a post-Kantian theoretical environment. It is directed, therefore, both against a fundamental ontology of Heideggerian matrix and against an eventual materialist ontology of the “social being”, for instance, such as that of the last Lukács.[8] It is supposed to challenge, in other words, the very conditions of ontology understood as second (or third) order discourse, logically superior in relation to the different discourses of ontic sciences.

In terms of Heidegger's technical terminology, Adorno believes he can demonstrate a fallacy underlying the very idea of an ontic-ontological difference. The core of his argument is eminently correlationalist: any sphere of meaning supposedly detached from the ontic can only be understood, Adorno argues, as an abstraction of the latter – the ontological, in other words, cannot be rationally explained except in terms of a hypostasis of ontic elements, which are extracted from its specific ontic constellation and raised to ontological status. This holds true for the concept of “being” operative in any ontology, whose necessarily mediated character Adorno seeks to unveil through a categorical analysis of the logical functions – copula and position – of the concept:

“Is” establishes a context of existential judgment between the grammatical subject and the predicate, thus suggesting something ontical. Taken purely by itself, however, as a copula, it means at the same time the general, categorical fact of a synthesis, without representing anything ontical. Hence it can be entered straightways on the ontological side of the ledger. From the logicity of the copula, Heidegger gets the ontological purity that suits his allergy to all things factual, and from existential judgment he gets the memory of things ontical – which will permit the categorial achievement of the synthesis to be hypostatized, then, as given. (Adorno, 1973, p. 100 - 101).


The critique consists in unveiling the operations of abstraction and hypostasis which turn ontology possible, bringing its supposed ontological categories back to the ontic sphere from which they came. In other words: unveiling its undeniably correlational dimension, which entails its ontological neutralization. As Adorno states unequivocally: “[…] the seeming jailbreak [of ontology] terminates in what the flight is from; the Being it flows into is θέσει.” (Adorno, 1973, p. 85).[9]

Finally, we should mention that the consequences of this criticism of any ontology lies at the heart of Adorno’s own critical theory. It is this critical self-reflection of ontology, in short, that leads to the very idea of dialectics Adorno wants to sustain: against all ontology, negative dialectics is an essentially autarchic and anti-systematic structure of mediating subject and object, which is not to be confused with any type of foundationalism or first philosophy: “In criticizing ontology we do not aim at another ontology, not even at one of being nonontological. If that were our purpose we would be merely positing another downright ‘first’ – not absolute identity, this time, not the concept, not Being, but nonidentity, facticity, entity.” (ADORNO, 1973, p. 136).


3 Post-correlational ontologies?

Let's recapitulate our argumentation so far. Despite the heterogeneity of positions which characterize the various currents of thought associated with the so called ontological or speculative turn, it is possible to unify them based on the diagnosis of the exhaustion of correlationism, understood as the hegemonic horizon of post-Kantian philosophy. So correlationism is highly multifaceted and has been developed concretely in a series of subtypes – what is common to them all, however, is the thesis about the essential relatedness of thinking and being, and therefore the procedure of a consequent dissolution of the absolute understood as an extracorrelational ontological stance. With regard to the ontological discipline in particular, we can see the theoretical effects of this correlationist device, on the one hand, in Kant’s dissolution of ontology as metaphysica generalis and its consequent substitution by an analytical of understanding; and, on the other hand, in the criticism of any ontology in its very claim to base a logically superior discourse in relation to the ontic, as seen, for instance, in Adorno's critical theory. As we have tried to show, the very idea of critique involved in these two thinkers nurtures an essential relation of opposition to ontology. Taken together, both intend to refute not only the possibility of a traditional discourse about the in-itself independent of thought, but also the ontologization of any intracorrelational concepts or categories. Hence, the fate of correlationism is an eminently post-foundational thought (“anti-systematic”, in Adorno's terms; or “factual”, in Meillassoux's).

What divides the representatives of the ontological turn is precisely the question concerning how to deal with this exhaustion of correlationism or even to overcome its thinking matrix altogether. As we already mentioned, in addition to the speculative ontology of Meillassoux, there are also naturalistic (Iain Grant), object-oriented (Harman, Garcia), perspectivist (Viveiros de Castro), neorealist (Gabriel), nihilist (Ray Brassier) attempts, among others. Mutatis mutandis, all of them seem to endorse the same imperative to leave correlationist modes of thinking without regressing to pre-critical dogmatism; in other words, any eventual post-correlationist ontology has to retain, in some way, the critical component – and it is this apparently paradoxical operation that is our object of analysis.


3.1. Meillassoux: the necessity of contingency

For Meillassoux, there is only one way to follow the imperative of abandoning the correlacionist thinking matrix without falling back in dogmatism, namely: to radicalize the theoretical decisions of correlationism and, from within, extract a principle of absolute grounding that is simultaneously distinct from the absolute principles of traditional metaphysics and immune to the critical device of correlationism. In other words, it will be necessary to conceive a theoretical procedure that accesses an absolute which is simultaneously extracorrelational and non-metaphysical from within correlationism. As the book's subtitle already indicates, this is the necessity of contingency.

Let us summarize Meillassoux’s demonstrative argumentation. First of all, it starts with the very procedure of de-absolutization of the most radical (strong) correlationisms, which converts any supposed in-itself into a for-us without reconverting this for-us into a new type of absolute. What is at stake in this procedure? Well, the fact that one might demonstrate the relativity of a supposed absolute to the one who knows it. In other words, the fact that one might demonstrate that this supposed absolute could be otherwise, if also the related instance were also different – which means that this supposed absolute is not absolute at all. Think of Kant’s and Adorno's criticisms of ontology outlined above: in both cases, what is at stake is, in general, a demonstration of the relativity of ontology to epistemology, the relativity of any doctrine of “being qua being” to the knowing subject – and, therefore, that ontology is not really ontology in its classical sense, since it could be otherwise if the related instance were also different. Now, Meillassoux concludes, this could-be-otherwise is nothing more than contingency, which is already internally operative in all critical-correlationalist thinking.

Indeed, one of the most elementary meanings of critique, as we have seen, is the dissolution of pseudo-objectivities and the demonstration of their mediated character. Or, in other words, the demonstration that supposed facticities are not absolute, but related to (or mediated by) a knowing subject, a certain collectivity or a sociocultural formation – and that, as such, they could be otherwise, or even (especially for critical theory) are essentially open to transformation. In other words: they are contingent. According to this reasoning, it can be said that contingency is, therefore, the (ontological) condition of possibility of critique itself. But, unlike all objects of critique – and this is Meillassoux’s decisive non-correlationist step –, contingency can only be necessary. In other words, Meillassoux believes he can demonstrate via an indirect – anhypothetical – argument that the very refutation of the necessary character of contingency necessarily presupposes it. In fact, in order to think contingency as relative to us, as contingent only for us, we need to assume that what we conceive as a contingent could be otherwise; well, here we see contingency again under this very assumption. It is possible to show, therefore, that it is necessary to assume that contingency is a non-correlational absolute in order to try to demonstrate that it is relative to us.[10]

If we follow Meillassoux, contingency is, therefore, the Great Outdoors of correlationisms. Basically, it is the necessity of contingency (and only of contingency) that they implicitly access when they postulate the relativity of any in-itself to a certain for-us. It is the experience of the maximum potency of thought – that is, that of touching a non-correlational absolute – which is indirectly done when correlationism believes only to draw its supposedly insurmountable limits. And what is contingency? It is precisely the necessary could-be-otherwise of another being, an ontological regime of the pure possible that is an unavoidably condition of the being of every entity. Every entity, every intramundane being, the very laws that govern this world as well as the world itself – in short, everything that is must be thought of as necessarily contingent. In order to illustrate this concept of contingency, Meillassoux draws on the concepts of an omnipotent time capable of creation and annihilation of virtually everything ex nihilo, as well as on the concept of hyper-Chaos not subjected to (almost) any law or regularity. It is this contingency thus conceived, in short, that will occupy the position of primary (that is, foundational) absolute in Meillassoux's ontology.

I will not elaborate further on the exposition and eventual criticism of this demonstration; a lot of material has been published about it in the last decade.[11] It can already be foreseen that a large part of Meillassoux’s effort consists in demonstrating the structural compatibility of a chaotic universe with the phenomenal stability of the world and, therefore, with experience and rationally articulated discourse. Let us emphasize once again that Meillassoux’s ultimate goal is to ground the ontologically realistic vector of science, and thereby to radicalize the project of modern rationalism. The demonstration of this compatibility will be the task of a factial or derivative ontology, which should deduce secondary and tertiary absolutes from that primary absolute which is contingency and that will also operate as a normalization of the power of chaos. I will develop the principles of this derivate ontology very briefly – only to the necessary extent to explore the idea of critique within it.


3.2 Elements of derivative ontology

What characterizes a derivative ontology is, first of all, its positivity. Unlike certain correlationist ontologies that believe to infer ex negative some kind of knowledge about the absolute, Meillassoux’s derivative ontology claims to be nothing less than a positive knowledge about the in-itself. Although the principle of the necessity of contingency is conquered by an anhypothetic method, a derivative or factial ontology should positively elaborate it from there.[12] Secondly, what differentiates derivative ontology from the traditional one is precisely the procedure whereby this positive development of the principle of absolute contingency is render possible: the principles taken from the absolute contingency are extracted from it precisely by derivation, that is, they clarify a posteriori the conditions of the primary absolute itself and, as such, they normalize it as well. However, they are not "less absolute" or less potent than the primary absolute. Although we are dealing here with a type of robust neofoundational ontology, it does not follow from there a neoplatonic ontology which explains the Hypostases derived from the One in terms of its progressive ontological weakening.

Meillassoux believes he can derive three absolute principles from the necessity of contingency, namely: the principle of non-contradiction (1), the principle of the existence of “something in general” (2) and the Cantorian transfinite (3). The derivation of the latter, however, was only indicated in his published work and remains to be done. Taken together, Meillassoux argues, they render possible to reconcile an ontology of hyper-Chaos not only with our ordinary experience and the possibility of rational discourse, but also with the conditions of possibility of modern science.

1. The derivation of the absolute character of the principle of non-contradiction is counterintuitive, but very sophisticated. It goes like this: if only contingency is necessary, then the principle of non-contradiction must also be absolutely true, because a contradictory entity would be necessary. In other words: a contradictory entity, that is, an entity being A and non-A at the same time and under the same aspect, would contain in itself its own otherness, including its own non-being, which means that it could not become otherwise nor could cease to be. In other words: it would be necessary. Therefore, if only contingency is necessary, then the principle of non-contradiction is absolutely true (MEILLASSOUX, 2008, p. 110ff).

2. Here is the derivation of the second principle: if only contingency is necessary, then it is also necessary that there is at least one contingent entity. In other words: the necessity of contingency imposes the existence of a contingent entity, because otherwise contingency would itself be contingent, which is self-refuting according to the anhypothetic demonstration above. This is to say: the necessity of contingency imposes the existence of contingent entities. Therefore, if only contingency is necessary, then it is absolutely necessary that there is something contingent and not nothing (MEILLASSOUX, 2008, p. 116ff).

3. With these two derivations Meillassoux claims to have an argument for the principle of non-contradiction and for the existence of something in general (therefore, for the degree zero of rational discourse). Meillassoux still needs, however, to explain how it is possible for the outside world to be phenomenally stable, since physical laws are necessarily contingent. To do so, it would be necessary to demonstrate that it is absolutely false to infer the necessity of the laws governing the physical world from their stability.

Following Jean-René Vernes (1981), Meillassoux believes that this inference is of probabilistic nature and lies at the heart of the problem of causality by both Hume and Kant. It is the following premise, implicitly present in the two philosophers: if physical laws could change without reason, then they would change frequently without reason, so that its change would make experience impossible (Kant) or would be attestable within experience (Hume). For Meillassoux (and Vernes), this premise is nothing more than a probabilistic calculus that takes the universe as the totality of the possible and our stable world as a random case within that totality. Given the reiteration of this same random case, that is, given the phenomenal stability of our world, it follows that there must be an extra-mathematical reason that explains this highly improbable reiteration of the same case in a potentially unlimited range of possibilities. Hence the necessity of laws is inferred without a rational explanation of this extra-mathematical reason.

Now, if we demonstrate that the possible is untotalizable, that is, transfinite, then we will have also demonstrated that this probabilistic calculation cannot be applied to the universe as a whole, precisely because the range of possibilities is not totalizable. It is true that this can be demonstrated by means of the standard axiomatic (ZFC) of the Cantorian set theory – which prohibits the existence of a set of all sets –, but nothing prohibits the use of other axiomatics that allow a totalization of the possible.[13] In other words, Meillassoux believes to demonstrate that the procedure that authorizes the application of the probabilistic calculus to the universe as a whole and then to the necessitarian inference is not the only possibly one (and not even the most rational one, since it generates an insurmountable explanatory deficit that is the extra-mathematical ratio indicated above). But it would be necessary to derive the Cantorian transfinite itself from the contingency in order to demonstrate that this probabilistic calculus is absolutely false. This derivation has not yet been made by the philosopher. His partial solution of the problem via ZFC axiomatic, however, already suggests a consistent compatibility of absolute contingency with the phenomenal structure of our world according to laws (contingent, but stable).


3.3 The fate of critique

If we follow Meillassoux, this ontology and its derivative principles correspond to the only possibility of rehabilitating a full-blown ontology in a post-correlational environment. According to his claim, exclusively through this whole procedure we would have incorporated what is essential to critique from a speculative point of view and, at the same time, demonstrated the fallacy inherent to the speculative (critical) decision all correlationisms share.

Now it is easy to see why this ontology claims not to be dogmatic. First, its primary absolute – contingency – is of essentially distinct nature of both the absolute of metaphysics and the intracorrelational absolutes. In effect, unlike the absolutes of traditional metaphysics, contingency is not an entity whose necessary existence one may attempt to deduce from a variant of ontological argument, as it occurs with virtually all metaphysics (if we agree, like Meillassoux, with Heidegger's well known assertion that metaphysics is essentially reducible to ontotheology[14]). In other words, it is not intended to affirm the necessity of contingency based on an argument of the type “X being an entity of type A, therefore X necessarily exists”. The principle of the necessity of contingence is thus immune to the Kantian demolition of the metaphysical absolutes, according to which existence is not a real predicate one could impute to a certain entity based on the mere concept of that entity.[15] It seems that we are here, effectively, at a higher level of reflexivity.

Furthermore, unlike the intracorrelational absolutes (like the idealist, for example, or any ontologized intracorrelational principles), one could demonstrate that contingency can only be thought of as extracorrelational, because any attempt to the contrary is self-refuting. It is equally immune, therefore, to Adorno's critique on the abstract and hypostasizing character of all ontology outlined above. One could demonstrate, moreover, that the same goes in relation to the secondary principles (non-contradiction and the existence of a contingent entity) derived from the primary absolute: strictly speaking, they are little more than a posteriori explanations of the ontological conditions of that absolute, which are also immune to the counterargument that they are projections or hypostases of epistemology onto ontology.

If it is true that the main lessons of critique, at least from the speculative point of view, are incorporated in the very constitution of Meillassoux’s ontology, it is also worth asking what is lost here of critique. Some interpreters[16] claim that the speculative ontology ends up depolitizing critique, to the extent that it leads us back to a principle of intelligibility that is entirely oblivious to human powers: contingency as such, with its emergencies and annihilations ex nihilo. Indeed, a consequence of correlationism, especially in its widespread postmodernist caricature, seems to be a kind of hyperpolitization of knowledge – since, if radicalized, correlationism itself not only replaces ontology by epistemology, but also tends to equate epistemology and politics (or truth and power). From this point of view, there is no doubt that speculative ontology is in fact depoliticizing. This hyperpolitization of knowledge is even generally seen by speculative philosophers as a symptom of the exhaustion of correlationism. It will be necessary to consider, not only in epistemic but also in political terms, whether this is effectively a symptom of exhaustion or, rather, of the very potency of correlationalist forms of thought.

Finally, it is also worth mentioning an apparently inevitable consequence of speculative ontologies with regard to the sciences. For one of the great goals of Meillassoux's endeavor is, as already said, to give an ontological foundation to the realistic vector – that is: in its truly extracorrelational dimension – of modern science. In order to do this, the third derivation of contingency – namely: the transfinite – would have a decisive function, namely: to ground the ontological potential of mathematics. There is no doubt that Meillassoux sees this as an antidote against the perverse fusion of correlationism – which is always skeptical-fideistic – with different forms of contemporary obscurity and New Age religiosity, which seem to dominate even the political scene of several countries today. One could wonder, however, if we would not hereby end up by rehabilitating a king of neopositivist conception of science, which, ultimately, would be placed beyond the scope of rational criticism. These are the questions, I believe, that must be addressed to speculative ontology from the point of view of critique.




Resumo: O objetivo do artigo é escrutinar o sentido geral e o destino da categoria de crítica, no interior do pensamento ontológico contemporâneo, particularmente no quadro da assim chamada “virada ontológica” ou “giro especulativo”. Considerando que essas vertentes de pensamento pretendem reabilitar a ontologia em um ambiente filosófico não dogmático, pergunta-se: em que se converte a crítica, em seu interior? Para tal, o artigo aborda primeiramente a questão de saber o que poderia dar consistência ao conceito de uma tal virada ontológica ora em curso, para então apresentar a relação tensa que vigora desde Kant entre crítica e ontologia. Por fim, investiga-se, a partir sobretudo da obra de Quentin Meillassoux, a possibilidade de se fundar uma ontologia que se queira simultaneamente não dogmática e imune à neutralização da ontologia efetivada pela crítica.


Palavras-chave: Crítica. Ontologia. Correlacionismo. Virada ontológica. Giro especulativo



ADORNO, T. W. Negative Dialektik. Frankfurt am Main: Suhrkamp, 1970.

ADORNO, T. W. Negative Dialectics. Translated by E. B. Ashton. London and New York: Routledge, 1973.

ADORNO, T. W. Ontologie und Dialektik. Frankfurt am Main: Suhrkamp, 2008.

BADIOU, A. O ser e o evento. Rio de Janeiro: Zahar/Editora UFRJ, 1996.

BRYANT, L. et al. (ed.). The speculative turn: contemporary realisms and materialisms. Melbourne: Free Press, 2011.

COOLE, F. (ed.). New materialisms: Ontology, agency, and politics. Durham: Duke University Press, 2011.

DOLPHIN, R.; VAN DER TUIN, I. New materialism: Interviews and cartographies. Michigan: University of Michigan Library, 2012.

GABRIEL, M. O sentido da existência: Para um novo realismo ontológico. Tradução de Bernardo Romagnoli Bethonico. Rio de Janeiro: Civilização Brasileira, 2016.

HAAG, K. H. Kritik der neueren Ontologie. Stuttgart: W. Kohlhammer Verlag, 1960.

HÄGGLUNG, M. Radical Atheist Materialism: A critique of Meillassoux. In: BRYANT, L. et al. (ed.). The speculative turn: contemporary realisms and materialisms. Melbourne: Free Press, 2011.

HARMAN, G. Quentin Meillassoux: Philosophy in the Making. Edinburgh: Edinburgh University Press, 2011.

HEIDEGGER, M. Identität und Differenz. Stuttgart: Klett-Cotta, 2002.

HINSKE, N. Kants langer Abschied von der Ontologie. In: Esposito, C. (ed.). Origini e sviluppi dell'ontologie (secoli XVI-XXI). Turnhout: Brepolls, 2009.

JOHNSTON, A. Hume’s Revenge: À Dieu, Meillassoux? In: BRYANT, L. et al. (ed.). The speculative turn: contemporary realisms and maerialisms. Melbourne: Free Press, 2011.

KANT. I. Kritik der reinen Vernunft. Hamburg: Meiner, 1998.

MEILLASSOUX, Q. Après la finitude: essai sur la nécessité de la contingence. Paris: Seuil, 2006.

MEILLASSOUX, Q. After Finitude: An Essay on the Necessity of Contingence. Translated by Ray Brassier. London: Continuum, 2008.

MEILLASSOUX, Q. Iteration, Reiteration, Repetition: A Speculative Analysis of the Meaningless Sign. In: AVANESSIAN, A.; MALIK. S. (ed.). Genealogies of Speculation. Materialism and Subjectivity since Structuralism. London: Blumsbury, 2016.

NUNES, R. O que são ontologias pós-críticas?. Revista EcoPós, v. 21, n. 2, p. 111 - 142, 2018.

OUATTARA, B. Adorno et Heidegger: Une controverse philosophique. Paris: L‘Harmattan, 2000.

PUCCIARELLI, D. Ontologie und Ontologiekritik bei Kant. In: KISNER, M. et al. (org). Das Selbst und die Welt. Beiträge zu Kant und der nachkantischen Philosophie. Würzburg: Königshausen und Neumann, 2019.

RIVERO, G. Zur Bedeutung des Begriffs Ontologie bei Kant. Kantstudien-Ergänzungshefte Band 180. Berlin: De Gruyter, 2014.

SCHNÄDELBACH, H. Philosophie in Deutschland: 1831 - 1933. Frankfurt am Main: Suhrkamp, 1983.

TERTULIAN, N. L'ontologie chez Heidegger et Lukács: Phénoménologie et dialectique. Kriterion, Belo Horizonte, n. 119, 2009.

VERNES, J.-R. Critique de la raison aléatoire. Paris: Aubier-Montagne, 1981.


Recebido: 23/8/2020

Aceito: 17/9/2021



[1] Doctor of Philosophy at the University of Munich. Professor at the State University of Minas Gerais (UEMG), Belo Horizonte, MG – Brazil. ORCID: E-mail:

[2] Among the textbooks and manuals already produced about this movement (and its denomination attempts), see especially the following ones: Bryant et al. (2011), Coole (2012) and Dolphin (2011).

[3] As Rodrigo Nunes formulated in a recent paper (2018), to which I am highly indebted.

[4] Berkeley is not mentioned in Après la finitude, but only in a later – and complexifying – review by Meillassoux of the terminology and historical milestones of correlationism. Despite this, Kant never lost his centrality as the first great systematizer of correlationism, especially if we consider the history of the effects of Critique. See Meillassoux in this respect (2016).

[5] Perhaps correlationism is the most celebrated and also the most discussed concept in Meillassoux's thought in particular and in the new ontologies in general. Meillassoux himself revised and reformulated some developments of the concept, first, in the English translation of Après la finitude (MEILLASSOUX, 2008) and, later, in a conference already cited (MEILLASSOUX, 2016). Criticisms of the concept tend to challenge its heuristic potential by listing singular models of thought that do not exactly fit its conceptualization or that reveal its schematism. See Harman (2011, p. 7-14) for a summary of these criticisms and Meillassoux's responses, and Nunes (2018) for a recent criticism.

[6] The fact that natural sciences can produce empirically founded statements about realities prior to the emergence of consciousness and even of life in Earth puts in question the epistemic bases of correlationism, for Meillassoux, insofar as these very statements seem to instantiate the existence of an object of knowledge prior to its manifestation for a thinking subject. In order to deal with this object of knowledge, correlationism must, according to Meillassoux, either weaken the literal meaning of ancestral statements or operate through retrojection of the past on the basis of the present. In both cases, what results is a falsification of the true – non-correlationist – meaning of ancestral statements. – This is one of the most criticized elements of Meillassoux's thinking by Kant scholars. According to them, Meillassoux incurs here in a confusion between the empirical and transcendental. For a response to these criticisms, see Meillassoux, 2008, p. 34ff.

[7] The concept of ontology has undergone several architectural changes from the pre-critical Kant until the formulation of his critical project. During the 1770s, for example, the very concept of "transcendental philosophy" was alternatively called "criticism" and "ontology". This terminological oscillation, which only stabilizes with the Critique itself, is already an indication of a structural transformation in the very concept and scope of what classical ontology was. It is also at that moment, in fact, that Kant uses for the first time the concept of Kategorie, originally from Aristotle’s ontology, in order to designate the concepts of understanding. In this regard, see Hinske (2009), Rivero (2014) and Pucciarelli (2019).

[8] See in this respect the work of Nicolas Tertullian (2009) and his disciple Bourahuma Ouattara (2000), that develop the differences and similarities between Heidegger and Lukács, also in relation to Adorno's critique.

[9] Here, as in virtually all critique of ontology of his Negative Dialectics, Adorno echoes the work Kritik der neueren Ontologie by his student Karl Heinz Haag, published in 1960 and quoted a few times in Negative Dialectics: “The ontological discussion of our days silences that question that no one can avoid unless he openly professes dogmatism. It is the question of whether ontological structures are something in themselves, φύσει, or whether they are merely products of thought, θέσει. As long as it deviates from this issue, ontology will renounce the discussion of its own validity.” (HAAG, 1960, p. 7).

[10] “The sceptic is only able to conceive of the difference between the 'in-itself and the 'for-us' by submitting the 'for-us' to an absence of reason which presupposes the absoluteness of the latter. It is because we can conceive of the absolute possibility that the 'in-itself could be other than the 'for-us' that the correlationist argument can have any efficacy. Accordingly, the anhypotheticity of the principle of unreason pertains to the 'in-itself as well as to the 'for-us', and thus to contest this principle is already to have presupposed it. Similarly, to query its absoluteness is already to have presupposed the latter.” (MEILLASSOUX, 2008, p. 101).

[11] We could schematically summarize the criticisms on the concept of absolute contingency according to two orders of questions. On the one hand, we have criticisms that challenge its own absolute character based on some variation of the argument of the finitude of human knowledge (JOHNSTON, 2011); on the other hand, we have arguments that challenge the terms in which Meillassoux make the concept itself explicit, such as the reference to the idea of ​​a virtually omnipotent temporality or to hyper-Chaos (HÄGGLUNG, 2011). The second type of criticisms seems to me to be pertinent and relevant; the first, in most cases, and in particular in Johnston himself, seems to me to disregard the core of Meillassoux's argument.

[12] A factial or derivative ontology “[…] is not intended to be a 'negative ontology' – we do not wish to confine ourselves to maintaining that contingency as we understand it is not accessible to this or that type of reasoning; rather, our aim is to elaborate an ever more determinate, ever richer concept of contingency. Accordingly, every difficulty encountered by factial speculation should be converted into a means for identifying a determinate condition of chaos capable of allowing us to overcome the obstacle.” (Meillassoux, 2008, p. 163).

[13] Although Alain Badiou's Being and Event is Meillassoux's explicit reference of the philosophical treatment of the Cantorian thought, we must remember that Meillassoux himself does not subscribe his former professor's thesis which equates mathematics and ontology. In fact, Badiou argues that only Cantor's invention would make it possible to fulfill the conditions of existence of a discourse about being qua being which were not subjected to the primacy of the One. For Badiou (1996, p. 16f), this implies that ontology itself is an axiomatic and non-philosophical discipline. Differently, Meillassoux uses Cantor almost strategically to conceive of the transfinite. For Meillassoux, mathematics, in its formal and above all “meaningless” dimension, has a specifically speculative scope, but they are not ontology. – Johnston (2011, 105s) formulates four major criticisms to the use made by both Meillassoux and Badiou of mathematics. According to him, both incur in a fetishization of mathematics with its supposed speculative potential (1); they collapse the basic categorical distinction between pure and applied mathematics (2); moreover, they circumscribe in an unjustified and even dogmatic way the terrain of "mathematics" (in the plural) to the Cantorian thought (3) and, within the latter, to the axiomatic of ZFC, as a privileged subdiscipline of its ontologies (4). It is not my aim to discuss these criticisms individually. It must be said, however, that if all of them can be effectively attributed to Badiou, only (2) could be consistently attributed to Meillassoux. As for the others, Meillassoux is well aware that his use of ZFC is strategic for his argumentative purposes, and this does not seem to pose a problem for his argument.

[14] See in this regard, for instance, the second part of the work published in 1957 entitled Identität und Differenz (HEIDEGGER, 2002, pp 51-79), in which Heidegger discusses the "onto-theo-logical constitution of metaphysics” (Die onto-theo-logische Verfassung der Metaphysik).

[15] In this regard, see Pucciarelli, 2018, p. 89f.

[16] Like Johnston (2011) and Hägglung (2011).