Anything is possible

Quentin Meillassoux, After Finitude: An Essay on the Necessity of Contingency, preface by Alain Badiou, trans. Ray Brassier, Continuum, London and New York, 2008. viii + 148 pp., £16.99 hb., 978 0 826 49674 4.

Philosophical speculation can regain determinate knowledge of absolute reality. We can think the nature of things as they are in themselves, independently of the way they appear to us. We can demonstrate that the modality of this nature is radically contingent – that there is no reason for things or ‘laws’ to be or remain as they are. Nothing is necessary, apart from the necessity that nothing be necessary. Anything can happen, in any place and at any time, without reason or cause.

Such is the ringing message affirmed by the remarkable French philosopher Quentin Meillassoux in his first book, After Finitude, originally published by Seuil in 2006. Against the grain of self-critical and self-reflexive post-Kantian philosophy, Meillassoux announces that we can recover ‘the great outdoors, the absolute outside of pre-critical thinkers’, the utterly ‘foreign territory’ that subsists in itself, independently of our relation to it. And when we begin to explore this foreign land that is reality in itself, what we learn is that

there is no reason for anything to be or to remain thus and so rather than otherwise.… Everything could actually collapse: from trees to stars, from stars to laws, from physical laws to logical laws; and this not by virtue of some superior law whereby everything is destined to perish, but by virtue of the absence of any superior law capable of preserving anything, no matter what, from perishing.

Neither events or laws are governed, in the end, by any necessity other than that of a purely ‘chaotic becoming – that is to say, a becoming governed by no necessity whatsoever’. (Meillassoux, ‘Potentiality and Virtuality’, Collapse 2, March 2007, p. 59)

For Meillassoux, as for Plato or Hegel, philosophy’s chief concern is with the nature of absolute reality, but as Meillassoux conceives it the nature of this reality demands that philosophy should think not ‘about what is but only about what can be’. The proper concern of a contemporary (post-metaphysical, post-dogmatic but also post-critical) philosophy is not with being but with may-being, not with être but with peut-être. If Meillassoux can be described as a ‘realist’, then, the reality that concerns him does not involve the way things are so much as the possibility that they might always be otherwise.

It is the trenchant force of this affirmation, no doubt, that accounts for the enthusiasm with which Meillassoux’s work has been taken up by a small but growing group of young researchers exasperated with the generally uninspiring state of contemporary ‘continental’ philosophy. It’s easy to see why Meillassoux’s After Finitude has so quickly acquired something close to cult status among some readers who share his lack of reverence for ‘the way things are’. The book is exceptionally clear and concise, entirely devoted to a single chain of reasoning. It combines a confident insistence on the self-sufficiency of rational demonstration with an equally rationalist suspicion of mere experience and consensus. The argument implies, in tantalizing outline, an alternative history of the whole of modern European philosophy from Galileo and Descartes through Hume and Kant to Heidegger and Deleuze. It is also open to a number of critical objections. In what follows I reconstruct the basic sequence of the argument (also drawing, on occasion, on articles published by Meillassoux in the last few years), and then sketch three or four of the difficulties it seems to confront.

The simplest way to introduce Meillassoux’s general project is as a reformulation and radicalization of what he on several occasions describes as ‘Hume’s problem’: that pure ‘reasoning a priori’ cannot suffice to prove that a given effect must always and necessarily follow from a given cause. There is no reason why one and the same cause should not give rise to a ‘hundred different events’. Meillassoux accepts Hume’s argument as unanswerable, as ‘blindingly obvious’: ‘we cannot rationally discover any reason why laws should be so rather than otherwise.’ Hume himself, however (along with both Kant and the main thrust of the analytical tradition), retreats from the full implications of his demonstration. Rather than ditch the concept of causal necessity altogether, he affirms it as a matter of ‘blind faith’. Whether this belief is then a matter of mere habit (Hume) or an irreducible component of transcendental logic (Kant) is, as far as Meillassoux is concerned, a secondary quarrel. Ever since, analytical philosophers have tended to assume that we should abandon ontological speculation and retreat instead to reflection upon the way we draw inductive inferences from ordinary experience, or from ordinary ways of talking about our experience.

In keeping with a tactic he deploys elsewhere in his work, Meillassoux himself quickly turns Hume’s old problem into an opportunity. Our inability rationally to determine an absolute necessity or sufficient reason underlying things, properly understood, can be affirmed as a demonstration that there is no such necessity or reason. Rather than try to salvage a dubious faith in the apparent stability of our experience, we should affirm the prospect that Hume refused to accept: there is no reason why what we experience as constant laws should not break down or change at any point, for the simple reason there is no such thing as reason or cause. The truth is not just that a given cause might give rise to a hundred different effects, but that an infinite variety of ‘effects’ might emerge on the basis of no cause at all, in a pure eruption of novelty ex nihilo.

The vision of the acausal and anarchic universe that results from the affirmation of such contingency is fully worthy of Deleuze and Guattari’s appreciation for those artists and writers who tear apart the comfortable normality of ordinary experience so as to let ‘a bit of free and windy chaos’ remind us of the tumultuous intensity of things:

If we look through the aperture which we have opened up onto the absolute, what we see there is a rather menacing power – something insensible, and capable of destroying both things and worlds, of bringing forth monstrous absurdities, yet also of never doing anything, of realizing every dream, but also every nightmare, of engendering random and frenetic transformations, or conversely, of producing a universe that remains motionless down to its ultimate recesses, like a cloud bearing the fiercest storms, then the eeriest bright spells, if only for an interval of disquieting calm.… We see something akin to Time, but a Time that is inconceivable for physics, since it is capable of destroying, without cause or reason, every physical law, just as it is inconceivable for metaphysics, since it is capable of destroying every determinate entity, even a god, even God.

Without flinching from the implications, Meillassoux attributes to such ‘time without development [devenir]’ the potential to generate life ex nihilo, to draw spirit from matter or creativity from stasis – or even to resurrect an immortal mind from a lifeless body.

Rational reflection encourages us to posit the absence of sufficient reason and to speculate about the potentialities of this absolute time: it is only our experience, precisely, that holds us back. Our ordinary sensory experience discourages us from abandoning a superstitious belief in causality. Conversion of Hume’s problem into Meillassoux’s opportunity requires, then, a Neoplatonic deflation of experience and the senses. However far we might push such deflation, though, it obviously remains the case that the world we experience is not chaotic but stable. How might we explain everyday empirical consistency on the basis of radical contingency and the total absence of causal necessity? If physical laws could actually change for no reason, would it not be ‘extraordinarily improbable if they did not change frequently’?

This question frames a second stage in Meillassoux’s argument. Since the earth so regularly rotates around the sun, since gravity so consistently holds us to the ground, so then we infer that there must be some underlying cause which accounts for the consistency of such effects. Meillassoux claims to refute such reasoning by casting doubt on the ‘probabilistic’ assumption that underlies it. An ordinary calculation of probabilities – say, the anticipation of an even spread of results from a repeated dice-throw – assumes that there is a finite range of possible outcomes and a finite range of determining factors, a range that sets the criteria whereby a given outcome is more or less likely in relation to others. At this point, following Badiou’s example, Meillassoux plays his Cantorian trump card.

It is precisely this totalization of the thinkable which can no longer be guaranteed a priori. For we now know – indeed, we have known it at least since Cantor’s revolutionary set-theory – that we have no grounds for maintaining that the conceivable is necessarily totalisable.

Cantor showed that there can be no all-inclusive set of all sets, leaving probabilistic reason with no purchase on an open or ‘detotalized’ set of possibilities: ‘laws which are contingent, but stable beyond all probability, thereby become conceivable’ (‘Potentiality and Virtuality’).

On this basis, Meillassoux aims to restore the rights of a purely ‘intelligible’ insight – that is, to reinstate the validity of pre- or non-critical ‘intellectual intuition’ and thereby challenge the stifling strictures of Kant’s transcendental turn. Rather than elaborate a merely ‘negative ontology’, he seeks to elaborate ‘an ever more determinate, ever richer concept of contingency’, on the assumption that these determinations can then be ‘construed as so many absolute properties of what is’, or as so many constraints to which a given ‘entity must submit in order to exercise its capacity-not-to-be and its capacity-to-be-other’.

A first constraint required by this capacity entails rejection of contradiction. The only law that survives the elimination of causal or sufficient reason is the law of non-contradiction. A contradictory entity would be utterly indeterminate, and thus both contingent and necessary. In order to affirm the thesis that any given thing can be anything, it is necessary that this thing both be what it is here and now, and forever capable of being determined as something else. In other words, where Kant simply posited that things-in-themselves existed and existed as non-contradictory, Meillassoux claims to deduce the latter property directly from the modality of their existence.

What does it mean, however, to say that such things exist? Meillassoux’s approach to this question circumscribes a second, more far-reaching determination of contingency: absolute and contingent entities or things-in-themselves must observe the logical principle of non-contradiction, and they must also submit to rigorous mathematical measurement. Here again, Meillassoux’s strategy involves the renewal of perfectly classical concerns. In addition to an affirmation of the ontological implications of the scientific revolution, it involves the absolutization of what Descartes and then Locke established as a thing’s primary qualities – those qualities like its dimensions or weight, which can be mathematically measured independently of the way an observer experiences and perceives it – that is, independently of secondary qualities like texture, colour, taste, and so on. But whereas Descartes conceived of such qualities in geometric terms, as aspects of an extended substance, Meillassoux takes a further step, and isolates the mathematizable from extension itself, so as then

to derive from a contingency which is absolute, the conditions that would allow me to deduce the absolutization of mathematical discourse [and thus] ground the possibility of the sciences to speak about an absolute reality …, a reality independent of thought. (‘Speculative Realism’, Collapse 3, 2007, p. 440)

Meillassoux admits that he has not worked out a full version of this deduction, but the closing pages of After Finitude imply that his approach will depend on the presumption that ‘what is mathematically conceivable is absolutely possible’, coupled with an appreciation for the absolutely arbitrary, meaningless and contingent nature of mathematical signs qua signs (e.g. signs produced through pure replication or reiteration, indifferent to any sort of pattern or ‘rhythm’). Perhaps an absolutely arbitrary discourse will be adequate to the absolutely contingent nature of things (‘Time Without Becoming’, talk at CRMEP, Middlesex University, May 2008).

The main obstacle standing in the way of this anti-phenomenological return ‘to the things themselves’, naturally, is the widely held (if not tautological) assumption that we cannot, by definition, think any reality independently of thought. Meillassoux dubs the modern currents of thought that accept this assumption ‘correlationist’. A correlationist humbly accepts that ‘we only ever have access to the correlation between thinking and being, and never to either term considered apart from the other’, such that ‘anything that is totally a-subjective cannot be’. Nothing can be independently of thought, since here ‘to be is to be a correlate’. Paradigmatically, to be is to be the correlate of either consciousness (for phenomenology) or language (for analytical philosophy).

Kant is the founding figure of correlationist philosophy, of course, but the label applies equally well, according to Meillassoux, to most strands of post-Kantian philosophy, from Fichte and Hegel to Heidegger or Adorno. All these philosophies posit some sort of fundamental mediation between the subject and object of thought, such that it is the clarity and integrity of this relation (whether it be clarified through logical judgement, phenomenological reduction, historical reflection, linguistic articulation, pragmatic experimentation or intersubjective communication) that serves as the only legitimate means of accessing reality. The overall effect has been to consolidate the criteria of ‘lawful’ legitimacy as such. Correlationism figures here as a sort of counter-revolution that emerged in philosophy as it tried, with and after Kant, to come to terms with the uncomfortably disruptive implications of Galileo, Descartes and the scientific revolution. Post-Copernican science had opened the door to the ‘great outdoors’: Kant’s own so-called ‘Copernican turn’ should be best understood as a Ptolemaic attempt to slam this door shut.

How, then, to reopen the door? Since a correlationist will assume as a matter of course that the referent of any statement ‘cannot possibly exist’ or ‘take place [as] non-correlated with a consciousness’, so then Meillassoux claims to find the Achilles heel of correlationism in its inability to cope with what he calls ‘ancestral’ statements. Such statements refer to events or entities older than any consciousness, events like the emergence of life, the formation of Earth, the origin of the universe, and so on. In so far as correlation can only conceive of an object that is given to a subject, how can it cope with an object that pre-dates givenness itself?

Now Meillassoux realizes that in order to overcome the Ptolemaic–correlationist counter-revolution it is impossible simply to retreat from Kant back to the ‘dogmatic’ metaphysics of Descartes, let alone to the necessity- and cause-bound metaphysics of Spinoza or Leibniz. He also accepts that you cannot refute correlationism simply by positing, as Laruelle does, a mind-independent reality. In order to overcome the correlational obstacle to his acausal ontology, in order to know mind-independent reality as non-contradictory and non-necessary, Meillassoux thus needs to show that the correlationist critique of metaphysical necessity itself enables if not requires the speculative affirmation of non-necessity.

This demonstration occupies the central and most subtle sections of After Finitude. The basic strategy again draws on Kantian and post-Kantian precedents. Post-Kantian metaphysicians like Fichte and Hegel tried to overcome Kant’s foreclosure of absolute reality by converting correlation itself, the very ‘instrument of empirico-critical de-absolutization, into the model for a new type of absolute.’ This idealist alternative to correlationist humility, however, cannot respond in turn to the ‘most profound’ correlational decision – the decision which ensures, in order to preserve the ban on every sort of absolute knowledge, that correlation too is just another contingent fact, rather than a necessity. As with his approach to Hume’s problem, Meillassoux’s crucial move here is to turn an apparent weakness into an opportunity. The correlationist, in order to guard against idealist claims to knowledge of absolute reality, readily accepts not only the reduction of knowledge to knowledge of facts: the correlationist also accepts that this reduction too is just another fact, just another non-necessary contingency. But if such correlating reduction is not necessary then it is of course possible to envisage its suspension: the only way the correlationists can defend themselves against idealist absolutization requires them to admit ‘the impossibility of giving an ultimate ground to the existence of any being’, including the impossibility of giving a ground for this impossibility (‘Speculative Realism’).

All that Meillassoux now has to do is absolutize, in turn, this apparent failure. We simply need to understand ‘why it is not the correlation but the facticity of the correlation that constitutes the absolute. We must show why thought, far from experiencing its intrinsic limits through facticity, experiences rather its knowledge of the absolute through facticity.’ In knowing that we know only contingent facts, we also know that it is necessary that there be only contingent facts. We know that facticity itself, and only facticity itself, is not contingent but necessary. Recognition of the absolute nature or absolute necessity of facticity then allows Meillassoux to go on to complete his deduction ‘from the absoluteness of this facticity those properties of the in-itself which Kant for his part took to be self-evident’ – that is, that it exists (as radically contingent) and that it exists as non-contradictory. By affirming this necessity of contingency or ‘principle of factuality’, Meillassoux triumphantly concludes, ‘I think an X independent of any thinking, and I know it for sure, thanks to the correlationist himself and his fight against the absolute, the idealist absolute’ (‘Speculative Realism’, p. 432).

Unlike Meillassoux, I believe that the main problem with recent French philosophy has been not an excess but a deficit of genuinely relational thought. From this perspective, despite its compelling originality and undeniable ingenuity, Meillassoux’s resolutely absolutizing project raises a number of questions and objections.

First, the critique of correlation seems to depend on an equivocation regarding the relation of thinking and being, of epistemology and ontology. On balance, Meillassoux insists on the modern ‘ontological requisite’ which stipulates that ‘to be is to be a correlate’ of thought. From within the correlational circle, ‘all we ever engage with is what is given-to-thought, never an entity subsisting by itself.’ If a being only is as the correlate of the thought that thinks it, then from a correlationist perspective it must seem that a being older than thought can only be ‘unthinkable’. A consistent correlationist, Meillassoux says, must ‘insist that the physical universe could not really have preceded the existence of man, or at least of living creatures’. As far as I know, however, almost no one actually thinks or insists on this, apart perhaps from a few fossilized idealists. They don’t think this because correlationism as Meillassoux defines it is in reality an epistemological theory, one that is perfectly compatible with the insights of Darwin, Marx or Einstein. There’s nothing to prevent a correlationist from thinking ancestral objects or worlds that are older than the thought that thinks them, or indeed older than thought itself; even from an orthodox Kantian perspective there is little difference in principle between my thinking an event that took place yesterday and an event that took place six billion years ago. As Meillassoux knows perfectly well, all that the correlationist demands is an acknowledgement that when you think of an ancestral event, or any event, you are indeed thinking of it. I can think of this lump of ancient rock as ancient if and only if science currently provides me with reliable means of thinking it so.

Genuine conquest of the correlationist fortress would require a reference not to objects older than thought but to processes of thinking that proceed without thinking, or objects that are somehow presentable in the absence of any objective presence or evidence – in other words, processes and objects proscribed by Meillassoux’s own insistence on the principle of non-contradiction. This is the problem with using a correlationist strategy (the principle of factuality) to break out of the correlationist circle: until Meillassoux can show that we know things exist not only independently of our thought but independently of our thinking them so, the correlationist has little to worry about. Anyone can agree with Meillassoux that ‘to think ancestrality is to think a world without thought – a world without the givenness of the world.’ What’s less obvious is how we might think such a world without thinking it, or how we might arrive at scientific knowledge of such pre-given objects if nothing is given of them.

Along the same lines, Meillassoux’s rationalist critique of causality and necessity seems to depend on an equivocation between metaphysical and physical or natural necessity. The actual target of Meillassoux’s critique of metaphysics is the Leibnizian principle of sufficient reason. He dispatches it, as we’ve seen, with a version of Hume’s argument: we cannot rationally demonstrate an ultimate reason for the being of being; there is no primordial power or divine providence that determines being or the meaning of being to be a certain way. What Meillassoux infers from this critique of metaphysical necessity, however, is the rather more grandiose assertion that there is no cause or reason for anything to be the way it is. This inference relies on a contentious understanding of the terms ‘reason’, ‘cause’ and ‘law’. It’s been a long time since scientists confused ‘natural laws’ with logical or metaphysical necessities, and it is perfectly possible, of course, to reconstruct the locally effective reasons and causes that have shaped, for instance, the evolution of aerobic vertebrate organisms. There was nothing necessary or predictable about this evolution, but why should we doubt that it conformed to familiar ‘laws’ of cause and effect? What does it mean to say that the ongoing consequences of this long process might be transformed in an instant – that we might suddenly cease to breathe oxygen or suffer the effects of gravity? Although Meillassoux insists that contingency applies to every event and every process, it may be that the only event that might qualify as contingent and without reason in his absolute sense of the term is the emergence of the universe itself.

Meillassoux’s acausal ontology, in other words, includes no account of an actual process of transformation or development. There is no account here of any positive ontological or historical force, no substitute for what other thinkers have conceived as substance, or spirit, or power, or labour. His insistence that anything might happen can only amount to an insistence on the bare possibility of radical change. So far, at least, Meillassoux’s affirmation of ‘the effective ability of every determined entity’ to persist, change or disappear without reason figures as an empty and indeterminate postulate. Once Meillassoux has purged his speculative materialism of any sort of causality he deprives it of any worldly historical purchase as well. The abstract logical possibility of change (given the absence of any ultimately sufficient reason) has strictly nothing to do with any concrete process of actual change. Rather like his mentor Badiou, to the degree that Meillassoux insists on the absolute disjunction of an event from existing situations he deprives himself of any concretely mediated means of thinking, with and after Marx, the possible ways of changing such situations.

The notion of ‘absolute time’ that accompanies Meillassoux’s acausal ontology is a time that seems endowed with only one dimension – the instant. It may well be that ‘only the time that harbours the capacity to destroy every determinate reality, while obeying no determinate law – the time capable of destroying, without reason or law, both worlds and things – can be thought as an absolute.’ The sense in which such an absolute can be thought as distinctively temporal is less obvious. Rather than any sort of articulation of past, present and future, Meillassoux’s time is a matter of spontaneous and immediate irruption ex nihilo. Time is reduced, here, to a succession of ‘gratuitous sequences’. The paradigm for such gratuitous irruption, obviously, is the miracle. Meillassoux argues that every absolute ‘miraculous’ discontinuity testifies only to the ‘inexistence of God’ – that is, to the lack of any metaphysical necessity, progress or providence. It may be, however, that an argument regarding the existence or inexistence of God is secondary in relation to arguments for or against belief in this quintessentially ‘divine’ power – a supernatural power to interrupt the laws of nature and abruptly reorient the pattern of worldly affairs.

The argument that allows Meillassoux to posit a radically open miraculous time depends on reference to Cantor’s ‘de-totalization’ of every attempt to close or limit a denumerable set of possibilities. A still more absolute lack of mediation, however, characterizes Meillassoux’s appeal to mathematics as the royal road to the in-itself. Cantor’s transfinite set theory concerns the domain of pure number alone. The demonstration that there is an open, unending series of ever larger infinite numbers clearly has decisive implications for the foundations of mathematics, but Meillassoux needs to demonstrate more exactly how these implications apply to the time and space of our actually existing universe. In what sense is our material universe itself infinite? In what sense has the evolution of life, for instance, confronted an actually infinite (rather than immensely large) number of actual possibilities? It is striking that Meillassoux pays little or no attention to such questions, and sometimes treats the logical and material domains as if they were effectively interchangeable.

Admittedly, you can make a case for the equation of mathematics and ontology in the strict sense, as Badiou does, such that post-Cantorian theory serves to articulate what can be thought of as pure being-qua-being (once being is identified with abstract and absolute multiplicity, i.e. a multiplicity that does not depend on any preliminary notion of unit or unity). Such an equation requires, however, that ontological questions be strictly preserved from merely ‘ontic’ ones: as a matter of course, a mathematical conception of being has nothing to say about the material, historical or social attributes of specific beings. A similar ‘ontological reduction’ must apply to Meillassoux’s reliance on Cantorian mathematics. Here again he seems to equivocate, as if the abstract implications of Cantorian detotalization might concern the concrete set of possibilities at issue in a specific situation – for example, in an ecosystem or in a political conflict. He seems to think that the Cantorian transfinite – a theory that has strictly nothing to do with any physical or material reality – might underwrite speculation regarding the ‘unreason’ whereby any actually existing thing might suddenly be transformed, destroyed or preserved.

In short, Meillassoux seems to confuse the domains of pure and applied mathematics. In the spirit of Galileo’s ‘mathematization of nature’, he relies on pure mathematics in order to demonstrate the integrity of an objective reality that exists independently of us – a domain of primary (mathematically measurable) qualities purged of any merely sensory, subject-dependent secondary qualities. But pure mathematics is arguably the supreme example of absolutely subject-dependent thought – that is, a thought that proceeds without reference to any sort of objective reality ‘outside’ it. No one denies that every mathematical measurement is ‘indifferent’ to the thing it measures. But leaving aside the question of why an abstract, mathematized description of an object should be any less mind-dependent or anthropocentric than a sensual or experiential description, there is no eliding the fundamental difference between pure number and an applied measurement. The idea that the meaning of the statement ‘the universe was formed 13.5 billion years ago’ might be independent of the mind that thinks it only makes sense if you disregard the quaintly parochial unit of measurement involved (along with the meaning of words like ‘ago’, to say nothing of the meaning of meaning tout court). As a matter of course, every unit of measurement, from the length of a metre to the time required for a planet to orbit around a star, exists at a fundamental distance from the domain of number as such. If Meillassoux was to carry through the argument of ‘ancestrality’ to its logical conclusion, he would have to acknowledge that it would eliminate not only all reference to secondary qualities like colour and texture but also all conventional primary qualities like length or mass or date as well. What might then be known of an ‘arche-fossil’ (i.e. a thing considered independently of whatever is given of it, including its material extension) would have to be expressed in terms of pure numbers alone, rather than dates or measurements. Whatever else such (neo-Pythagorean?) knowledge amounts to, it has no obvious relation to the sorts of realities that empirical science tries to describe, including realities older than the evolution of life.

After Finitude is a beautifully written and seductively argued book. It offers a welcome critique of the ambient ‘necessitarian’ world-view, that pensée unique which tells us ‘there is no alternative’, and which underlies both the listless political apathy and the deflating humility of so much contemporary philosophy and critical theory. In the rationalist tradition of the Enlightenment and of ideology-critique, Meillassoux launches a principled assault on every ‘superstitious’ presumption that existing social situations should be accepted as natural or inevitable. His insistence that such situations are actually a matter of uncaused contingency, however, offers us little grip on the means of their material transformation. The current fascination with his work, in some quarters, may be a symptom of impatience with a more traditional conception of social and political change – not that we might abruptly be other than we are, but that we might engage with the processes whereby we have become what we are, and might now begin to become otherwise. A critique of metaphysical necessity and an appeal to transfinite mathematics will not provide, on their own, the basis upon which we might renew a transformative materialism.

back