Meillassoux's Speculative Materialism

'Everything in the world is without reason, and is thereby capable of actually becoming otherwise without reason... Everything could actually collapse: from trees to stars, from stars to laws, from physical laws to logical laws; and this is 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.' (AF53)

After Finitude: An Essay on the Necessity of Contingency   

Correlationism

 

Meillassoux starts his essay with an appeal: to rehabilitate primary and secondary qualities. Primary and secondary qualities, Meillassoux tells us – and we can surely concur here – are obsolete. A secondary quality, in Meillassoux's account, is a sensation that we derive from an object, i.e. we burn when we touch fire. A primary quality is of course an attribute of an object: its colour, for instance. But of course this breaks down almost immediately, because we realise that colour, an attribute of the object that is supposed to be independent of our perception, or its effect on us, is in fact entirely dependent on our eyesight. Meillassoux, however, wants to restore the possibility of the primary quality.

In a way, this issue of primary and secondary qualities is representative of the entire philosophical consensus that Meillassoux is seeking to overturn, that which he calls correlationism.

'For what decisively discredited the distinction between primary and secondary qualities is the very idea of such a distinction: i.e. the assumption that the 'subjectivation' of sensible properties (the emphasis on their essential link to the presence of a subject) could be restricted to the object's sensible determinations, rather than extended to all its conceivable properties.' (AF2)

One of the ways in which Meillassoux will return to primary qualities is to reassert a renewed form of the Cartesian thesis that:

'[A]ll those aspects of the object that can be formulated in mathematical terms can be meaningfully conceived as properties of the object in itself.' (AF3)

Meillassoux goes to great length to show how precisely a statement such as this is impossible to accept in the current philosophical milieu. A milieu whose warp is precisely that of primacy of the subject:

'We cannot represent the "in itself" without it becoming "for us", or as Hegel amusingly put it, we cannot "creep up on" the object "from behind"...' (AF4)

This philosophical bedrock predates Kant, but is inaugurated in modern form, and its best known sense, by what Meillassoux refers to as the 'transcendental revolution' (AF4) that emerged from Kant's Critique of Pure Reason, which we looked at in our last meeting. It is crucial to note that the correlationism implied by Kant's Critique refutes naïve realism, but also retains a space for objectivity.

'In the Kantian framework, a statement's conformity to the object can no longer be defined in terms of a representation's "adequation" or "resemblance" to an object supposedly subsisting "in itself" since this "in itself" is inaccessible. The difference between an objective representation (such as "the sun heats the stone") and a "merely subjective" representation (such as "the room seems warm to me") is therefore a function of the difference between two types of subjective representation: those that can be universalized, and are thus by right capable of being experienced by everyone, and hence "scientific" and those that cannot be universalized, and hence cannot belong to scientific discourse.' (AF4)

And the former relies on Kant's transcendental subject. Appearances for Kant can be universal and invariant, but they are appearances nonetheless. Kant is not a solipsist.

This entails correlationsim, which Meillassoux defines in After Finitude as such:

'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." (AF5)

And he goes on to define the correlationist circle and the correlationist two step:

'If one calls "the correlationist circle" the argument according to which one cannot think the in-itself without entering into a vicious circle, thereby immediately contradicting oneself, one could call "the correlationist two-step" this other type of reasoning to which philosophers have become so well accustomed [which runs as follows:] it would be naïve to think of the subject and the object as two separately subsisting entities whose relation is only subsequently added to them. On the contrary, the relation is in some sense primary: the world is only world insofar as it appears to me as world [...]' (AF5)

Another quote on correlationism to make it completely clear:

'Correlationsim rests on an argument as simple as it is powerful, and which can be formulated in the following way: No X without givenness of X, and no theory about X without a positing of X. If you speak about something, you speak about something that is given to you, and posited by you. Consequently, the sentence ‘X is’, means: ‘X is the correlate of thinking’ in a Cartesian sense. That is: X is the correlate of an affection, or a perception or a conception, or of any given subjective act. To be is to be a correlate, a term of a correlation.' (Collapse, Volume III, 409)

The correlationist circle involves never being able to grasp an object without a subject, but not being able to grasp the subject independently either, and the two step involves the primacy of the relation: no object exists without a relation to the subject (step one) and this relationship (this 'co-') is then primary. There are two examples that Meillassoux uses to bring this into relief. Firstly, in terms of the vicious circle, the consensus maintains the primacy of the subject 'everything is inside' and yet construes this as a cage, i.e. 'everything is (actually) outside' (AF6). And secondly, in terms of the primacy of relation, Meillassoux's example is Heidegger, the subject-object opposition is not an opposition but an 'originarily constituted […] reciprocal relation'.

Ancestrality

 

After outlining correlationism, Meillassoux then asks the important question, 'Are the following statements meaningless for the correlationist?'

-      the date of the origin of the universe (13.5 billion years ago)

-      the date of the accretion of earth (4.56 billion years ago)

-      the date of the origin of life on earth (3.5 billion years ago)

-      the date of the origin of humankind (Homo habilis, 2 million years ago)

Meillassoux categorizes these statements as 'ancestral'. Ancestrality refers to realities that predate human existence or indeed, any form of life on earth. And the arche-fossil is what gives us an indication of these realities, i.e. the isotopes that allow us to date fossils. Meillassoux holds correlationism to account here: you cannot be a correlationist and accept the ancestral statements. He moves through several ways in which a correlationist might accept this statement: either by positing a God-like or external witness to the formation of the universe, which is reflected in human subjectivity, or by acknowledging these statements only as 'seemingly ancestral' (AF16). In either case it seems that we have to 'cancel' these statements, subordinating them to statements about our sense of the universe.

Let's look more closely at Meillassoux's notion of the ancestral and how it relates to speculation. Any validation of an ancestral statement implies a new demand upon thought:

'To think ancestrality is to think a world without thought – a world without the givenness of the world. It is therefore incumbent upon us to break with the ontological requisite of the moderns, according to which to be is to be a correlate. Our task, by way of contrast, consists in trying to understand how thought is able to access the uncorrelated, which is to say, a world capable of subsisting without being given. But to say this is just to say that we must grasp how thought is able to access an absolute, i.e. a being whose severance (the original meaning of absolutus) and whose separateness from thought is such that it presents itself to us as non-relative to us, and hence is capable of existing whether we exist or not.' (AF28)

Despite this, Meillassoux knows that we cannot, however, become dogmatic realists/pre-critical philosophers. In other words, we cannot throw away the legacy of Kant, we must transform it. And it is in the impasse between Kantian correlationism and dogmatism that Meillassoux will try to construct a philosophy 'after finitude.'

In the second chapter, Meillassoux shows that, in rejecting the critical tradition, we cannot lapse back in any Cartesian argument about primary and secondary qualities, because Cartesian metaphysics is dogmatic, or, in other words, it posits a 'necessary being' that 'must absolutely be because it is the way it is' (AF33). (And here the being is obviously God). This positing of a necessary being leads to principle of sufficient reason, the principle according to which

'[F]or every thing, every fact, and every occurrence, there must be a reason why it is thus and so rather than otherwise.' (AF33)

'For not only does such a principle require that there be a possible explanation for every worldly fact; it also requires that the thought account for the unconditioned totality of beings, as well as for their being thus and so.' (AF33)

It is the combination of necessity and the principle of sufficient reason that makes this into a dogmatic metaphysics. Meillassoux wants to recuperate a philosophy of the absolute, a philosophy that does not deny ancestral statements, but does not want to lapse back into such a dogmatic metaphysics.

And here he starts to foreshadow his alternative: speculative materialism.

'Let us call "speculative" every type of thinking that claims to be able to access some form of absolute, and let us call "metaphysics" every type of thinking that claims to be able to access some form of absolute being, or to access the absolute through the principle of sufficient reason. If all metaphysics is "speculative" by definition, our problem consists in demonstrating, conversely, that not all speculation is metaphysical, and not every absolute is dogmatic – it is possible to envisage an absolutizing thought that would not be absolutist.' (AF34)

Meillassoux wants to throw out dogmatic metaphysics as the alternative to Kantian critique, and recuperate the absolute through a different philosophical regime. And it is this vision of a third system that offers an alternative to both the Cartesian and the Kantian systems that gives Meillassoux's little book its profound significance.

From Correlationism to Speculative Materialism

 

In response to the challenge of the ancestral statement Meillassoux sets out in pursuit of a non-dogmatic, non-metaphysical, but ‘speculative’ absolute (following his definition above).

‘But first we must expound that variety of correlationism which is the most rigorous, as well as the most contemporary. For it is only by confronting the most radical form of the correlation that we will be able to know whether in fact de-absolutisation is the unsurpassable horizon for all philosophy.’ (AF35)

Over the following pages Meillassoux elaborates on five possible positions mapped out according to their adherence-to or deviance-from the two poles of Naive Realism and Absolute Idealism, so that he might identify the exact conditions of this most contemporary, most radical form of correlation:

Dogmatic/Naive Realism - dogmatically asserts the real and our access to it; this is different from scientific realism, as we shall see later (for it wields the power of mathematics). (AF27)

Weak Correlationism - the Kantian position: the in-itself exists, we can think it, but our thoughts about things-in-themselves do not constitute knowledge. (AF30) We can make certain definite claims about the in-itself, i.e., that it follows the law of non-contradiction. '[But] by what miraculous operation is Kantian thought able to get outside of itself in order to verify that what is unthinkable for us is impossible in itself?' (AF35)

Strong Correlationism - the dominant model in continental philosophy; we cannot know things-in-themselves, nor can we even think them, but we cannot by the same token say that they do not exist. (AF35) Furthermore, because we cannot apply norms of the for-us, like the law of non-contradiction, to the in-itself, irrational discourses about the in-itself cannot be discounted. (AF41) Strong Correlationism therefore holds that 'it is unthinkable that the unthinkable be impossible.' (AF41) We simply cannot say anything at all about what lies beyond thought, which means we cannot deny its existence, whether rational or irrational. (cf. The early Wittgenstein, Tractatus Proposition 7, and Heidegger: 'the very fact that there are beings, and that there is a givenness of beings' lies beyond all rational access [What is Metaphysics].) Nevertheless, as heirs of the legacy of Critical philosophy, Strong Correlationists still attempt to uncover a set of 'universal conditions for our relation to the world, whether these be construed as conditions for empirical science, conditions for linguistic communication between individuals, conditions for the perception of the entity, etc.' (AF42) Now, according to Meillassoux, insofar as Strong Correlationism thinks no (rational) statement about the in-itself is possible, it leads directly to fideism. Strong correlationism can be read as an epistemological claim about our finitude, ignorance, and need for skeptical agnosticism.

Very Strong Correlationism - as above, but there are also no universal conditions under which we may describe our relation or non-relation to any rational or irrational in-itself: the correlate is marked by its facticity. 'The Very Strong Correlationist takes it for granted that the de-absolutisation of thought also results in de-universalisation, so that each person becomes a sort of private fideist with a personal worldview impenetrable to reason.' (Harman18) Very Strong Correlationism is the hallmark of postmodern philosophy, dismissing 'every variety of universal as a mystificatory relic of the old metaphysics [these pilosophers] will claim that it is necessary to think the facticity of our relation to the world in terms of a situation that is itself finite'. And thus the correlation in which we find ourselves has no universal features, but is 'anchored in a determinate era of the history of being, or in a form of life harbouring its own language games [cf. the later Wittgenstein], or in a determinate cultural and interpretive community, etc.' (AF43)

Absolute Idealism (having so far argued that all correlationism is a form of idealism in the face of ancestrality, Meillassoux needs to distinguish Absolute Idealism from its closest position, Very Strong Correlationism. Both agree that 'we cannot think a world without an entity [that is] capable of 'thinking' this world in the most general sense', while only Absolute Idealism actually 'abolish[es] any such notion of the in-itself'. (AF37) The in-itself is nonsensical: the arche-fossil in-itself does not exist an absolute sense - in other words, even though Absolute Idealism holds that the in-itself is unthinkable, it still maintains that the absolute is thinkable. (AF38) Whereas correlationism is marked by the correlation's facticity (the Kantian ground, our essential finitude), Absolute Idealism is marked by the correlation's absoluteness. Again 'Kant maintains that we can only describe the a priori forms of knowledge [space, time, and the twelve categories] whereas Hegel insists that it is possible to deduce them.' (AF38) Absolute Idealism professes positive knowledge that the in-itself does not exist. 

Meillassoux has demonstrated that the most rigorous, most contemporary, and most radical form of correlationism is Very Strong Correlationism. But this is not a satisfactory position in itself and this is why:

 

Strong Correlationism and Fideism 

'Fideism is merely the other name for strong correlationism.' (AF48)

Strong Correlationism has engendered fideism and 'the return of the religious', a state of affairs brought on by the fact that all possibility of rational discourse about the in-itself has ended. (AF43).

'[B]y destroying metaphysics, one has effectively rendered it impossible for a particular religion to use a pseudo-rational argumentation against every other religion. But in doing so - and this is the decisive point - one has inadvertently justified belief's claim to be the only means of access to the absolute. Once the absolute has become unthinkable, even atheism, which also targets God's inexistence in the manner of an absolute, is reduced to a mere belief, and hence to a religion, albeit of the nihilist kind.' (AF45-46)

‘Contemporary fanaticism cannot therefore simply be attributed to the resurgence of an archaism that is violently opposed to the achievements of Western critical reason; on the contrary, it is the effect of critical rationality, and this precisely insofar as – this needs to be underlined – this rationality was effectively emancipatory; was effectively, and thankfully, successful in destroying dogmatism.’ (AF47)

The ‘return of the religious’ is the very product of Western critical reason. The fervour with which twentieth century philosophers carried out the 'end of metaphysics' has opened up a vast void of the irrational, unknowable in-itself.

In short: 'the victorious critique of ideologies has been transformed into a renewed argument for blind faith.' (AF49)

So:

- A return to dogmatic or naive realism is out of the question (Meillassoux adheres to Kant's rejection of dogmatic metaphysics and to critical philosophy's requisite of beginning an interrogation with the conditions bearing on human apprehension of the world - see AF27).

- Correlationism as it stands is unacceptable because has been revealed to equate to fideism.

- And Meillassoux rejects Absolute Idealism because he does not accept that one can have positive knowledge that the correlation itself is absolute. Rather, he holds that even if thinking an unthought thing is a completely meaningless notion, this does not mean that unthought things do not exist. 'What is meaningful for humans need not exhaust the totality of what is real.' (Harman 22)

Against this final possibility of absolutising the correlation, Meillassoux chooses the essential facticity of correlationism as the most acceptable starting point, for if he can uncover an absolute other than the absolute of the correlation the problem of fideism will also be solved in the process. So we begin with the strong correlationist maxim that 'it is unthinkable that the unthinkable should be impossible'. (AF41)

Graham Harman has a good little summary of this: 

‘Traditionally, Kantians have been at position 2 and Hegelians and position 5. Meillassoux places various twentieth century figures at positions 3/4, preferring 4 for himself... But this position 4 will soon be transformed into one of the most surprising philosophical positions of our time: Speculative Materialism, which tries to avoid the two extremes of absolute idealism and correlationism. To escape the former it must abstain from hypostatising the correlation of thought and world; to avoid the latter, it must move beyond finitude and think the absolute.’ (Harman 23)

Thus:

'Every materialism that would be speculative, and hence for which absolute reality is an entity without thought, must assert both that thought is not necessary (something can be independently of thought), and that thought can think what there must be when there is no thought.' (AF36)

Facticity and Contingency

The Facticity (that is universalised in strong correlationism) is Kant's legacy. Meillassoux undertakes to move from facticity ('the world is governed by structural invariants that can only be described, not deduced') to contingency ('physical laws remain indifferent as to what happens, whether any entity emerges, subsists, or perishes'). 'If contingency consists in knowing that worldly things could be otherwise, facticity just consists in not knowing why the correlational structure has to be thus.' (AF39) One is positive knowledge, the other, negative knowledge. Meillassoux's strategy is to transform correlationism's supposed ignorance of things-in-themselves into absolute knowledge that things-in-themselves exist without reason, and that they can change at any time for no reason at all, thereby founding an ontology of absolute contingency.

'We are going to back into the thing itself what we mistakenly took to be an incapacity in thought'. (AF53) [i.e. the fact that we can’t discover a reason for why things are a certain way and not another is not a deficiency in thought, it’s a real property of things-in-themselves.

The doctrine of human finitude is generally associated with some hidden reason, impenetrable to human thought. But Meilassoux maintains that this is simply not the case, and that our unfounded allegiance to the principle of sufficient reason has led us astray:

'There is no hidden reason, there is no reason at all: This absence of reason is, and can only be the ultimate property of the entity'. (AF53)

'There is nothing beneath or beyond the manifest gratuitousness of the given - nothing but the limitless and lawless power of its destruction, emergence, or persistence'. (AF63)

There strategy here is to turn German Idealism's move on Kant against German Idealism itself: whereas the Idealists hypostatised the correlation, Meillassoux hypostatises the facticity of the correlation. So how does he do this: How does he move from the negative knowledge of facticity to the positive knowledge of contingency?

Deducing the Facticity of the Correlation

 

By means of a five way dialogue involving a dogmatic believer, a dogmatic atheist, a correlationism, a subjective idealist and, finally, a speculative materialist, Meillassoux sets out to prove that the negative knowledge of facticity is in fact an absolute principle of contingency. By putting pressure on correlationism from two opposing sides, Meillassoux demonstrates that it is a fundamentally inconsistent and unstable position. In the face of certain questions, it either collapses into idealism or finds itself maintaining a different kind of absolute without, seemingly, having realised it. 

A breakdown of the dialogue can be found here.

 

Summary of the Deduction

1. The Atheist and Believer are both derailed by the Kantian point that we cannot have access to reality outside our conditions of access to it. 

2. The Idealist is defeated by the point that the meaninglessness of thinking a world outside thought does not mean a world outside thought is impossible. 

3. The Correlationist can only avoiding slipping into Idealism by insisting that there might be an in-itself different from the for-us and that this 'might' refers to a real in-itself, not just an in-itself for-us (which would be a contradiction). The Correlationist, despite themselves, has to absolutise the facticity of the correlation, by holding that there are absolutely a number of different possibilities for the in-itself. She must maintain positive knowledge of the possibilities beyond the for-us in order to defeat the Idealist, who would just argue that these are only possibilities for thought. 

4. There is no way to avoid either making the correlation or facticity absolute: the Correlationist must either become a Subjective Idealist or a Speculative Materialist.

'We have now identified the faultline that lies at the heart of correlationism; the one through which we can breach its defences.' (AF59)

The Speculative Materialist speculates from within the faciticity of the correlation. However, her speculative act is buoyed up by the absolute possibility that any theory she entertains about the in-itself is potentially absolutely true. The speculative act suddenly attains an unprecedented level of gravity.

What Meillassoux has just done is deduce the absolute status of facticity. While the features of our facticity can only be described, the absoluteness of facticity itself can be deduced. In short, facticity cannot be factical, it must be absolute. This quality is the 'factiality' of facticity.

'From now on, we will use the term "facticality" to describe the speculative essence of facticity, viz., that the facticity of everything cannot be thought of as a fact' [rather, every thing is absolutely factical]. (AF79)

To recap: how is Speculative Materialism Different from Correlationism?

'Instead of saying that the in-itself could actually be anything whatsoever without anyone knowing what, we maintain that the in-itself could be anything whatsoever and that we know this.' (AF65)

The Correlationist 'is incapable of disqualifying any hypothesis about the nature of the absolute'. (AF65) While the Speculative Materialist 'knows two things that the sceptic did not: first, that contingency is necessary, and hence eternal; second, that contingency alone is necessary' (AF65).

 

The Figures of Factiality

'The factial is defined as the very arena for speculation that excludes all metaphysics.' (AF128)

The goal now becomes the identification and analysis of the conditions and necessary features of factiality. These comprise:

- The law of non-contradiction (AF 69-70) 
- There must be something rather than nothing (AF 80)
- Every mathematical statement can be absolutised (AF126)

Hyperchaos

Meillassoux's absolute is anything but rigid, delimiting or tyrannical: in hyperchaos 'nothing would seem to be impossible, not even the unthinkable'. (AF64)

'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 realising 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... It is a Time capable of destroying even becoming itself, by bringing forth, perhaps forever, fixity, stasis, and death.' (AF64)

Importantly, contingency does not necessarily imply flux.

[Here, we skip over a large, important section of After Finitude, in which Meillassoux begins to draw out the consequences of his deduction of factiality. For notes on this part of the book, click here.]

Diachronicity and Mathematics 

Meillassoux claims that dia-chronic statements in the metaphysical open 'a possibility that was rendered meaningful by the very inception of modern science' and this possibility is 'the decentering of thought relative to the world within the process of knowledge'. (AF112)

Dia-chronicity 'provide[s] a general characterisation of all such statements about events that are anterior or ulterior to every terrestrial-relation-to-the-world – the former expressing the temporal hiatus between world and relation-to-the-world that is inherent in the very meaning of such discourse.' (AF112)

How or why do diachronic statements affirm this possibility? Diachronic statements attest to 'a temporal hiatus between being and terrestrial thought'? (AF113) In this sense we need a new epistemological formulation for these types of statements, which Meillassoux identifies in the 'hypothesis'. Whilst he argues that, in a way, ancient myth enacted this temporal hiatus, hypotheses are different, because they can become part of a cognitive process – or rather, they become knowledge, and knowledge that can transform our understanding of reality. With the scientific hypothesis, we comprehend knowledge that does not need a witness - laws that do not need a human witness to exist.

Copernican Revolution/Ptolemaic Counter-Revolution 

In the final chapter, Meillassoux returns to the problem of ancestrality (which he expands into dia-chronicity) in order to posit, as did Descartes, that mathematics is in fact our conduit to the in-itself. 

'This is the enigma which we must confront: mathematics’ ability to discourse about the great outdoors; to discourse about a past where both humanity and life are absent.' (AF26)

'The Galilean-Copernican revolution has no other meaning than that of the paradoxical unveiling of thought’s capacity to think what there is whether thought exists or not.’ (AF116)

Galileo, Descartes and Copernicus all employed a mathematised natural science to theorise a world existing apart from humans, thereby removing humans from the centre of meaning. (AF116) 

‘Yet this is where we encounter a rather disconcerting paradox...: when philosophers refer to the revolution in thought instituted by Kant as “the Copernican revolution”, they refer to a revolution whose meaning is the exact opposite of the one we have just identified.’ (AF117)

In light of this decentralisation of the subject by science, Meillassoux (along with Latour, incidentally), opines that Kant's supposed Copernican Revolution is in many ways a Ptolemaic Counter-Revolution. For, at the precise moment when modern science was trying to give us diachronic knowledge about 'the nature of a world without us' in which 'the truth or falsity of physical law is not established with regard to our own existence', Kant returned humans to the centre of epistemology. (AF114)

Having debunked Kant’s spurious terminology, Meillassoux holds fast to the ‘Copernico-Galilean event’ which ‘institutes the idea of a mathematical knowledge of nature’. (AF124)

Primary Qualities are Mathematisable Qualities

Mathematisation answers the question of how thought can think what exists when there is no thought - 'the most urgent question' posed by science to philosophy. (AF121) And our modern mathematical natural science, which is better than ever before, gives us a cosmos in which 'thought has become able to think a world that can dispense with thought, a world that is essentially unaffected by whether or not anyone thinks it.' (AF116)

It is important to note that, for Meillassoux: 'Every mathematical statement is not necessary true, but absolutely possible'. (AF126) So, that to which mathematics points is not an ideal referent lying outside space and time, but rather to the structure of the given: 

'all those aspects of the given that are mathematically describable can continue to exist regardless of whether or not we are there to convert the latter into something that is given-to or manifested-for.' (AF117)

Graham Harman gives a good breakdown of this:

'In other words, the mathematical is not deeper than the given; primary qualities are not somehow buried beneath human access. Instead, the difference between primary and secondary qualities is one that plays out entirely within the realm of the given. Some aspects of the given can be conceived as existing even if I were to vanish, while others cannot; the former are mathematisable ones, and they count as the primary qualities. [...] That is to say, primary qualities are contingent just like everything else, since they do no belong to some sort of special eternal realm outside the given, but they are also absolute, since they can possibly exist outside human thought, even if in contingent rather than necessary form.' (Harman 52)

Meillassoux writes: 'What is mathematisable cannot be reduced to a correlate of thought' (AF117) - which is to say that the mathematiseable exists in thought, but also exceeds thought, and in this way, is capabale of indexing the real. 

Mathematics is exemplary of this in a way, because mathematical worlds often have no relation to the terrestrial world or human existence: many mathematical objects are either intuitively impossible or cannot be realised or observed in our world.

Now we get to the crucial aspect of Meillassoux's book: his new formulation of the absolute, which rests on the necessity of contingency rather than existence.

'[T]his dia-chronic referent may be considered to be contingent while simultaneously being considered to be absolute: it can be construed as an event, an object, or a processual stability, that need not be shown to be unconditionally necessary, since this would be contrary to our ontology. On the other hand, however, the meaning of the dia-chronic statement about a radioactive decay older than all terrestrial life is only conceivable if it is construed as absolutely indifferent to the thought that envisages it. Accordingly, the absoluteness of that which is mathematizable means: the possibility of factial existence outside thought – and not: the necessity of existence outside thought.' (AF117)

Meillassoux, Quentin. After Finitude: An Essay on the Necessity of Contingency. trans. Ray Brassier. London: Continuum, 2008.
Quentin Meillassoux, 'Speculative Realism,' Collapse III (2007): 409

Harman, Graham. Quentin Meillassoux: Philosophy in the Making. Edinburgh: Edinburgh University Press, 2011.  

Based on a presentation given by Baylee Brits and Amy Ireland in Sydney, February 19th, 2013