Brassier
14/2/07
09:20
Page 15
COLLAPSE II
The Enigma of Realism: On Quentin
Meillassoux’s After Finitude1
Ray Brassier
1. THE ARCHE-F OSSIL
Quentin Meillassoux has recently proposed a
compelling diagnosis of what is most problematic in postKantian philosophy’s relationship to the natural sciences.2
The former founders on the enigma of the ‘arche-fossil’. A
fossil is a material bearing the traces of pre-historic life, but
an ‘arche-fossil’ is a material indicating traces of ‘ancestral’
phenomena anterior even to the emergence of life. It
provides the material basis for experiments yielding
estimates of ancestral phenomena – such as, for instance,
the radioactive isotope whose rate of decay provides an
index of the age of rock samples, or the starlight whose
luminescence provides an index of the age of distant stars.
Natural science produces ancestral statements, such as, for
example, that the universe is roughly 13.7 billion years old,
that the earth formed roughly 4.5 billion years ago, that life
1. This is a heavily edited version of a chapter from the author’s Nihil Unbound:
Enlightenment and Extinction (Basingstoke: Palgrave Macmillan, forthcoming 2007).
2. Après la Finitude: Essai sur la necessité de la contingence (Paris: Seuil, 2006). English
translation After Finitude (trans. R. Brassier) (London: Continuum, forthcoming
2008).
15
Brassier
14/2/07
09:20
Page 16
COLLAPSE II
developed on earth approximately 3.5 billion years ago,
and that the earliest ancestors of the genus Homo emerged
about 2 million years ago.3 Yet it is also generating an ever
increasing number of ‘descendent’ statements, such as that
the Milky Way will collide with the Andromeda galaxy in
3 billion years, or that the earth will be incinerated by the
sun 4 billions years hence, or that all the stars in the
universe will stop shining in 100 trillion years, or that
eventually, one trillion, trillion, trillion years from now, all
matter in the cosmos will disintegrate into unbound
elementary particles. Philosophers should be more
astonished by such statements than they seem to be, for
they present a serious problem for post-Kantian
philosophy. Yet strangely, the latter seems to remain
entirely oblivious to it. The claim that these statements are
philosophically enigmatic has nothing to do with qualms
about the methods of measurement involved, or with
issues of empirical accuracy, or any other misgivings about
scientific methodology. They are enigmatic because of the
startling philosophical implications harboured by their
literal meaning. For the latter seems to point to something
which violates the basic conditions of conceptual intelligibility stipulated by post-Kantian philosophy. In order to
understand why this is so, we need to try to sketch the
latter.
For all their various differences, post-Kantian
philosophers can be said to share one fundamental
conviction: that the idea of a world-in-itself, subsisting
3. ‘Billion’ and ‘trillion’ will be used throughout following their now internationally
accepted US usage, as meaning a thousand million and a million million
respectively.
16
Brassier
14/2/07
09:20
Page 17
Brassier – Enigma of Realism
independently of our relation to it, is an absurdity.
Objective reality must be transcendentally guaranteed,
whether by pure consciousness, intersubjective consensus,
or a community of rational agents; without such
guarantors, it is a metaphysical chimera. Or for those who
scorn what they mockingly dismiss as the ‘antiquated’
Cartesian vocabulary of ‘representationalism’, ‘subject/
object dualism’, and epistemology more generally, it is our
pre-theoretical relation to the world, whether characterized
as Dasein or ‘Life’, which provides the ontological precondition for the intelligibility of the scientific claims listed
above. No wonder, then, that post-Kantian philosophers
routinely patronize these and other scientific assertions
about the world as impoverished abstractions whose
meaning supervenes on this more fundamental subrepresentational or pre-theoretical relation to phenomena.
For these philosophers, it is this relation to the world –
Dasein, Existence, Life – which provides the originary
condition of manifestation for all phenomena, including
those ancestral phenomena featured in the statements
above. Thus if the idea of a world-in-itself, of a realm of
phenomena subsisting independently of our relation to it,
is intelligible at all, it can only be intelligible as something
in-itself or independent ‘for-us’. This is the reigning doxa of
post-metaphysical philosophy: what is fundamental is
neither a hypostasized substance, nor the reified subject,
but rather the relation between un-objectifiable thinking
and un-representable being, the primordial reciprocity or
‘co-propriation’ of logos and physis which at once unites and
distinguishes the terms which it relates. This premium on
relationality in post-metaphysical philosophy – whose
17
Brassier
14/2/07
09:20
Page 18
COLLAPSE II
telling symptom is the preoccupation with ‘difference’ – has
become an orthodoxy which is all the more insidious for
being constantly touted as a profound innovation.4
Meillassoux has given it a name: ‘correlationism’.
Correlationism affirms the indissoluble primacy of the
relation between thought and its correlate over the
metaphysical hypostatization or representationalist
reification of either term of the relation. Correlationism is
subtle: it never denies that our thoughts or utterances aim
at or intend mind-independent or language-independent
realities; it merely stipulates that this apparently
independent dimension remains internally related to
thought and language. Thus contemporary correlationism
dismisses the problematic of scepticism, and of epistemology more generally, as an antiquated Cartesian hang-up:
there is supposedly no problem about how we are able to
adequately represent reality, since we are ‘always already’
outside ourselves and immersed in or engaging with the
world (and indeed, this particular platitude is constantly
touted as the great Heideggerian–Wittgensteinian insight).
4. Graham Harman has elaborated a profound critique of this tendency in contemporary philosophy, seeing in it an avatar of a generalized anti-realism. Whether the
relation in question is the epistemological relation between mind and world, the phenomenological relation between noesis and noema, the ekstatic relation between Sein
and Dasein, the prehensive relation between event-objects, or the processual relation
between matter and memory, Harman argues that this premium on relationality
occludes the discontinuous reality of objects in favour of their reciprocal idealizations. Harman’s startlingly original interpretation of Heidegger provides the point of
departure for his complete re-orientation of phenomenology away from the primacy
of the human relation to things and toward things themselves considered independently of their relation to humans or each other. Accordingly, the fundamental task
for this ‘object-oriented philosophy’ consists in explaining how autonomous objects
can ever interact with each other, and to that end Harman has developed a particularly ingenious theory of ‘vicarious causation’ – See Harman’s contribution to the
present volume, 171-205.
18
Brassier
14/2/07
09:20
Page 19
Brassier – Enigma of Realism
Note that correlationism need not privilege ‘thinking’ or
‘consciousness’ as the key relation – it can just as easily
replace it with ‘being-in-the-world’, ‘perception’,
‘sensibility’, ‘intuition’, ‘affect’, or even ‘flesh’. Indeed, all of
these terms have featured in the specifically phenomenological varieties of correlationism.5
But the arche-fossil presents a quandary for the
correlationist. For how is the correlationist to make sense
of science’s ancestral claims? Correlationism insists that
there can be no cognizable reality independently of our
relation to reality; no phenomena without some transcendental operator – such as life or consciousness or Dasein –
generating the conditions of manifestation through which
phenomena manifest themselves. In the absence of this
originary relation and these transcendental conditions of
manifestation, nothing can be manifest, apprehended,
thought or known. Thus, the correlationist will continue,
not even the phenomena described by the sciences are
5. The writings of Husserl and Heidegger are littered with paradigmatic expressions
of the correlationist credo. Here are just two examples:
The existence of Nature cannot be the condition for the existence of consciousness, since Nature itself turns out to be a correlate of consciousness:
Nature is only as being constituted in regular concatenations of consciousness. (Edmund Husserl, Ideas Pertaining to a Pure Phenomenology and to a
Phenomenological Philosophy. Book One. Tr. F. Kersten, Dordrecht, The
Netherlands: Kluwer, 1982: 116).
[S]trictly speaking we cannot say: there was a time when there were no
human beings. At every time, there were and are and will be human
beings, because time temporalizes itself only as long as there are human
beings. There is no time in which there were no human beings, not
because there are human beings from all eternity, but because time is not
eternity, and time always temporalizes itself only at one time, as human,
historical Dasein. (Martin Heidegger, Introduction to Metaphysics, tr. G. Fried
and R. Polt, New Haven and London: Yale University Press, 2000: 88-9).
19
Brassier
14/2/07
09:20
Page 20
COLLAPSE II
possible independently of the relation through which
phenomena become manifest. Moreover, the correlationist
will add, it is precisely the transcendental nature of the
correlation as sine qua non for cognition that obviates the
possibility of empirical idealism. Thus, contra Berkeley,
Kant maintains that known things are not dependent upon
being perceived precisely because known things are representations and representations are generated via transcendental syntheses of categorial form and sensible material.
Synthesis is rooted in pure apperception, which yields the
transcendental form of the object as its necessary correlate
and guarantor of objectivity. The transcendental object is
not cognizable, since it provides the form of objectivity
which subsumes all cognizable objects; all of which must be
linked to one another within the chains of causation
encompassed by the unity of possible experience and circumscribed by the reciprocal poles of transcendental
subject and transcendental object. Yet the arche-fossil
indexes a reality which does not fall between these poles
and which refuses to be integrated into the web of possible
experience linking all cognizable objects to one another,
because it occurred in a time anterior to the possibility of
experience. Thus the arche-fossil points to a cognizable
reality which is not given in the transcendental object of
possible experience. This is a possibility which Kant
explicitly denies:
Thus we can say that the real things of past time are given
in the transcendental object of experience; but they are
objects for me and real in past time only in so far as I represent to myself (either by the light of history or by the
guiding clues of a series of causes and effects) that a
20
Brassier
14/2/07
09:20
Page 21
Brassier – Enigma of Realism
regressive series of possible perceptions in accordance
with empirical laws, in a word, that the course of the
world, conducts us to a past time-series as the condition
of the present time – a series which, however, can be
represented as actual not in itself but only in the connection of a possible experience. Accordingly, all events which
have taken place in the immense periods that have preceded my own
existence really mean nothing but the possibility of extending the
chain of experience from the present perception back to the conditions
which determine this perception in respect of time.6
For Kant, then, the ancestral time of the arche-fossil
cannot be represented as existing in itself but only as
connected to a possible experience. But we cannot
represent to ourselves any regressive series of possible
perceptions in accordance with empirical laws capable of
conducting us from our present perceptions to the
ancestral time indexed by the arche-fossil. It is strictly
impossible to prolong the chain of experience from our
contemporary perception of the radioactive isotope to the
time of the accretion of the earth indexed by its radiation,
because the totality of the temporal series coextensive with
possible experience itself emerged out of that geological
time wherein there simply was no perception. We cannot
extend the chain of possible perceptions back prior to the
emergence of nervous systems, which provide the material
conditions for the possibility of perceptual experience.
Thus it is precisely the necessity of an originary
correlation, whether between knower and known, or Sein
and Dasein, that science’s ancestral statements flatly
6. Critique of Pure Reason, tr. N. K. Smith (London: MacMillan 1929), A495; emphasis
added.
21
Brassier
14/2/07
09:20
Page 22
COLLAPSE II
contradict. For in flagrant disregard of those transcendental conditions which are supposed to be necessary for every
manifestation, they describe occurrences anterior to the
emergence of life, and objects existing independently of
any relation to thought. Similarly, science’s descendent
statements refer to events occurring after the extinction of
life and the annihilation of thought. But how can such
statements be true if correlationism is sound? For not only
do they designate events occurring quite independently of
the existence of life and thought; they inscribe the transcendental conditions of manifestation themselves within a
merely empirical timeline. How can the relation to reality
embodied in life or thought be characterized as transcendentally necessary (sine qua non) for the possibility of spatiotemporal manifestation when science unequivocally
states that life and thought, and hence this fundamental
relation, have a determinate beginning and end in spacetime? Don’t science’s ancestral and descendent statements
strongly imply that those ontologically generative
conditions of spatiotemporal manifestation privileged by
correlationists – Dasein, life, consciousness, and so on – are
themselves merely spatiotemporal occurrences like any
other?
2. THE CORRELATIONIST RESPONSE
Confronted by Meillassoux’s argument from the archefossil, partisans of correlationism have not been slow in
mounting a counter-offensive. In a supplement to the
forthcoming English translation of Après la finitude,
Meillassoux recapitulates the two most frequently voiced
objections and responds to both. The correlationist defence
22
Brassier
14/2/07
09:20
Page 23
Brassier – Enigma of Realism
is two-tiered. In the first stage, Meillassoux is accused of
inflating an un-observed phenomenon into a negation of
correlation, when in fact it is merely a lacuna in
correlation. In the second stage, Meillassoux is deemed
guilty of naively conflating the empirical and the transcendental. We will consider each of these objections, as well as
Meillassoux’s responses to them, in turn.
In the first stage, the correlationist contends that, far
from being novel and challenging, the argument from the
arche-fossil is merely a restatement of a hackneyed and
rather feeble objection to transcendental idealism. Thus,
the correlationist continues, the arche-fossil is simply an
example of a phenomenon which went un-perceived. But
un-perceived phenomena occur all the time and it is
excessively naive to think they suffice to undermine the
transcendental status of the correlation. In this regard, the
temporal distance which separates us from the ancestral
phenomenon is no different in kind from the spatial
distance which separates us from contemporaneous but
unobserved events occurring elsewhere in the universe.
Thus the fact that there was no-one around 4.5 billion
years ago to perceive the accretion of the earth is no more
significant than the fact that there is currently no-one 25
million, million miles away perceiving events on the
surface of Alpha Centauri. Moreover, the notion of
‘distance’ is an inherently ambiguous and unreliable
indicator of the limits of perception: technology allows us
to perceive objects extraordinarily far away in space and
time, while myriad occurrences close at hand routinely go
unperceived. In this regard, instances of spatiotemporal
extremity are no different in kind from other banal
23
Brassier
14/2/07
09:20
Page 24
COLLAPSE II
instances of un-witnessed or un-perceived phenomena,
such as the fact that we are never aware of everything
going on inside our own bodies. Thus the arche-fossil is
just another example of an un-perceived phenomenon and,
as with all other examples of un-perceived phenomena, it
merely exemplifies the inherently lacunary nature of manifestation – the fact that no phenomenon is ever exhaustively or absolutely apprehended by perception or consciousness. Far from denying this, both Kant and Husserl
emphasized the intrinsically limited and finite nature of
human cognition. Thus for Kant sensible intuition is
incapable of exhaustively apprehending the infinite
complexity of a datum of sensation. Similarly for Husserl,
intentionality proceeds by adumbrations which never
exhaust all the dimensions of the phenomenon. But the fact
that every phenomenon harbours an un-apprehended
remainder in no way undermines the constitutive status of
transcendental consciousness. All that it shows is that manifestation is inherently lacunary and that the non-manifest
inheres in every manifestation. A counterfactual suffices to
establish the persistence of transcendental constitution
even in cases of lacunary manifestation such as the archefossil. Thus the contingent fact that no-one was there to
witness the accretion of the earth is ultimately of no
importance; for had there been a witness, they would have
perceived the phenomenon of accretion unfolding in
conformity with the laws of geology and physics which are
transcendentally guaranteed by the correlation. Ultimately,
the correlationist concludes, the argument from the archefossil fails to challenge correlationism because it has simply
confused a contingent lacuna in manifestation with the
24
Brassier
14/2/07
09:20
Page 25
Brassier – Enigma of Realism
necessary absence of manifestation.
Against this initial line of defence, Meillassoux insists
that the arche-fossil cannot be reduced to an example of
the un-perceived because the temporal anteriority involved
in the notion of ancestrality remains irreducible to any
notion of temporal ‘distance’ concomitant with correlational manifestation. To reduce the arche-fossil to an unwitnessed or un-perceived occurrence is to beg the
question because it is to continue to assume that there is
always a correlation in terms of which to measure gaps or
lacunae within manifestation. But the arche-fossil is not
merely a non-manifest gap or lacuna in manifestation; it is
the lacuna of manifestation tout court. For the anteriority
indexed by the ancestral phenomenon does not point to an
earlier time within manifestation; it indexes a time anterior to
the time of manifestation in its entirety; and it does so according
to a sense of ‘anteriority’ which cannot be reduced to the
past of manifestation because it indicates a time wherein
manifestation – along with its past, present, and future
dimensions – originally emerged. Thus, Meillassoux
contends, the ‘ancestral’ cannot be reduced to the ‘ancient’.
There are always greater or lesser degrees of ‘ancientness’
depending on whatever temporal metric one happens to
choose. ‘Ancientness’ remains a function of a relation
between past and present which is entirely circumscribed
by the conditions of manifestation and in this sense any
past, no matter how ‘ancient’, remains synchronous with
the correlational present. In equating temporal remove
with spatial distance, the correlationist objection outlined
above continues to assume this underlying synchronicity.
But ancestrality indexes a radical ‘diachronicity’ which
25
Brassier
14/2/07
09:20
Page 26
COLLAPSE II
cannot be correlated with the present because it belongs to
the time wherein the conditions of correlation between
past, present, and future passed from inexistence into
existence. Accordingly, ancestrality harbours a temporal
diachronicity which remains incommensurable with any
chronological measure that would ensure a reciprocity
between the past, present, and future dimensions of the
correlation.
Meillassoux detects in this initial correlationist response
a subterfuge which consists in substituting a lacuna in and
for manifestation – a lacuna that is contemporaneous with
constituting consciousness, as is always the case with the
un-perceived – for a lacuna of manifestation as such; one
which cannot be synchronized with constituting consciousness (or whatever other transcendental operator happens
to be invoked). The correlationist’s sleight-of-hand here
consists in reducing the arche-fossil – which is non-manifest
insofar as it occurs prior to the emergence of conditions of
manifestation – to the un-perceived, which is merely a
measurable gap or absence within the extant conditions of
manifestation. However, Meillassoux insists, the archefossil is neither a lacunary manifestation nor a temporal
reality internal to manifestation (internal to the
correlation); for it points to the temporal reality in which
manifestation itself first came into existence, and wherein it
will ultimately sink back into inexistence. Consequently,
Meillassoux concludes, it is a serious misunderstanding to
think that a counterfactual suffices to reintegrate the archefossil within the correlation, for the diachronicity it indexes
cannot be synchronized with any correlational present.
Having failed to rebuff the argument from the arche26
Brassier
14/2/07
09:20
Page 27
Brassier – Enigma of Realism
fossil with this initial line of defence, the correlationist
adopts a second strategy. This consists in contesting the
claim that ancestrality indexes a temporal dimension
within which correlational temporality itself passes into
and out of being. For such an assertion betrays a
fundamental confusion between the transcendental level at
which the conditions of correlation obtain and the empirical
level at which the organisms and/or material entities which
support those conditions exist. The latter are indeed spatiotemporal objects like any other, emerging and perishing
within physical space-time; but the former provide the
conditions of objectivation without which scientific
knowledge of spatiotemporal objects – and hence of the
arche-fossil itself – would not be possible. Though these
conditions are physically instantiated by specific material
objects – i.e. human organisms – they cannot be said to
exist in the same manner, and hence they cannot be said to
pass into or out of existence on pain of paralogism. Thus,
the correlationist continues, the claim that the conditions of
objectivation emerged in space-time is an absurd
paralogism because it treats transcendental conditions as
though they were objects alongside other objects. But the
transcendental conditions of spatiotemporal objectivation
do not exist spatiotemporally. This is not to say that they
are eternal, for this would be to hypostatize them once
again and to attribute another kind of objective existence
to them, albeit in a transcendent or supernatural register.
They are neither transcendent nor supernatural – they are
the logical preconditions for ascriptions of existence, rather
than objectively existing entities. As conditions for the
scientific cognition of empirical reality – of which the
27
Brassier
14/2/07
09:20
Page 28
COLLAPSE II
arche-fossil is a prime example – they cannot themselves be
scientifically objectified without engendering absurd
paradoxes. The claim that ancestral time encompasses the
birth and death of transcendental subjectivity is precisely
such a paradox, but one which dissolves once the
confusion from which it has arisen has been diagnosed.
Yet for Meillassoux, the initial plausibility of this
response masks its underlying inadequacy, for it relies on
an unacknowledged equivocation. We are told that
transcendental subjectivity cannot be objectified, and hence
that it neither emerges nor perishes in space-time; but also
that it is neither immortal nor eternal, in the manner of a
transcendent metaphysical principle. Indeed this is
precisely what distinguishes transcendental subjectivity in
its purported finitude from any metaphysical hypostatization of the principle of subjectivity which would render it
equivalent to an infinitely enduring substance. But as finite,
transcendental subjectivity is indissociable from the
determinate set of material conditions which provide its
empirical support. Thus Husserl insists on the necessary
parallelism which renders the transcendental indissociable
from the empirical. Indeed, it is this necessary parallelism
which distinguishes transcendental subjectivity from its
metaphysical substantialization. Accordingly, though
transcendental subjectivity is merely instantiated in the
minds of physical organisms, it cannot subsist independently of those minds and the organisms which support them.
Although it does not exist in space and time, it has no other
kind of existence apart from the spatiotemporal existence
of the physical bodies in which it is instantiated. And it is
precisely insofar as it is anchored in the finite minds of
28
Brassier
14/2/07
09:20
Page 29
Brassier – Enigma of Realism
bounded physical organisms with limited sensory and
intellectual capacities that human reason is not infinite. But
if transcendental subjectivity is necessarily instantiated in
the spatiotemporal existence of physical organisms, then it
is not quite accurate to claim that it can be entirely
divorced from objectively existing bodies. Indeed, in the
wake of Heidegger’s critique of the ‘worldless’ or
disembodied subject of classical transcendentalism, postHeideggerean philosophy can be said to have engaged in
an increasing ‘corporealization’ of the transcendental.
Merleau-Ponty is probably the most prominent (though
certainly not the only) advocate of the quasi-transcendental status of embodiment. Accordingly, although transcendental subjectivity may not be reducible to objectively
existing bodies, neither can it be divorced from them, for
the existence of bodies – and a fortiori of language, society,
history, culture, etc. – provides the conditions of instantiation for the transcendental (i.e. the ‘always already’). Thus,
Meillassoux concludes, while it is perfectly plausible to
insist that the correlation provides the transcendental
condition for knowledge of spatiotemporal existence, it is
also necessary to point out that the time in which the
bodies that provide the conditions of instantiation for the
correlation emerge and perish is also the time which
determines the conditions of instantiation of the transcendental. But the ancestral time which determines the
conditions of instantiation of the transcendental cannot be
encompassed within the time that is co-extensive with the
correlation, because it is the time within which those
corporeal conditions upon which the correlation depends
pass into and out of existence. Where such conditions of
29
Brassier
14/2/07
09:20
Page 30
COLLAPSE II
instantiation are absent, so is the correlation. Thus the
ancestral time indexed by the arche-fossil is simply the time
of the inexistence of the correlation. This ancestral time is
indexed by objective phenomena such as the arche-fossil;
but its existence does not depend upon those conditions of
objectivation upon which knowledge of the arche-fossil
depends, because it determines those conditions of instantiation which determine conditions of objectivation.
3. ANCESTRALITY AND C HRONOLOGY
Meillassoux’s responses to his correlationist critics are
as trenchant as they are resourceful and they undoubtedly
constitute a significant addition to his already weighty case
against correlationism. However, they also invite a number
of critical observations. First, it is not at all clear how
Meillassoux’s distinction between ancestrality and spatiotemporal distance can be squared with what twentiethcentury physics has taught us concerning the fundamental
indissociability of time and space, as enshrined in the
Einstein-Minkowski conception of four-dimensional spacetime. ‘Anteriority’ and ‘posteriority’ are inherently
relational terms which can only be rendered intelligible
from within a spatiotemporal frame of reference. In this
regard, Meillassoux’s insistence on the irreconcilable
disjunction between a lacuna in manifestation and the
lacuna of manifestation continues to rely on an appeal to
the scalar incommensurability between the anthropomorphic time privileged by correlationism and the cosmological time within which the former is nested. This
incommensurability is attributed to the fundamental
asymmetry between cosmological and anthropomorphic
30
Brassier
14/2/07
09:20
Page 31
Brassier – Enigma of Realism
time: whereas the former is presumed to encompass the
beginning and end of the latter, the reverse is assumed not
to be the case. However, Meillassoux conducts his case
against correlationism in a logical rather than empirical
register – indeed, we shall see below how this leads him to
reiterate the dualism of thought and extension – yet the
asymmetry to which he appeals here is precisely a function
of empirical fact, and as Meillassoux himself acknowledges,7 there is no a priori reason why the existence of
mind, and hence of the correlation, could not happen to be
coextensive with the existence of the universe. Indeed, this
is precisely the claim of Hegelianism, which construes
mind or Geist as a self-relating negativity already inherent
in material reality. Accordingly, the transcendence which
Meillassoux ascribes to ancestral time as that which exists
independently of correlation continues to rely upon an
appeal to chronology: it is the (empirical) fact that cosmological time preceded anthropomorphic time and will
presumably succeed it which is invoked in the account of the
asymmetry between the two. In light of this implicit appeal
to chronology in Meillassoux’s claim that the arche-fossil
indexes the absence of manifestation, rather than any
hiatus within it, it is difficult to see how the temporal
anteriority which he ascribes to the ancestral realm could
ever be understood wholly independently of the spatiotemporal framework in terms of which cosmology coordinates
relations between past, present, and future events. A
simple change in the framework which determines
chronology would suffice to dissolve the alleged incommensurability between ancestral and anthropomorphic
7. Cf. Après la finitude, 161.
31
Brassier
14/2/07
09:20
Page 32
COLLAPSE II
time, thereby bridging the conceptual abyss which is
supposed to separate anteriority from spatiotemporal
distance.
The conclusion to be drawn is the following: as long as
the autonomy of the in-itself is construed in terms of a
merely chronological discrepancy between cosmological
and anthropomorphic time, it will always be possible for
the correlationist to convert the supposedly absolute
anteriority attributed to the ancestral realm into an
anteriority which is merely ‘for us’, not ‘in itself’. By
tethering his challenge to correlationism to the spatiotemporal framework favoured by contemporary cosmology,
Meillassoux mortgages the autonomy of the in-itself to
chronology. The only hope for securing the unequivocal
independence of the ‘an sich’ must lie in prizing it free from
chronology as well as phenomenology. This would entail a
conception of objectivity which excludes chronological
relationality as much as phenomenological intentionality.
Spatiotemporal relations should be construed as a function
of objective reality; rather than objective reality construed
as a function of spatiotemporal relations. By insisting on
driving a wedge between ancestral time and spatiotemporal distance, Meillassoux inadvertently reiterates the
privileging of time over space which is so symptomatic of
idealism and unwittingly endorses his opponents’ claim
that all non-ancestral reality can be un-problematically
accounted for by the correlation. Thus the trenchancy of
Meillassoux’s rejoinders above actually masks a significant
concession to correlationism. For surely it is not just
ancestral phenomena which challenge the latter, but simply
the reality described by the modern natural sciences tout
32
Brassier
14/2/07
09:20
Page 33
Brassier – Enigma of Realism
court. According to the latter, we are surrounded by
processes going on quite independently of any relationship
we may happen to have to them: thus plate tectonics, thermonuclear fusion, and galactic expansion (not to mention
undiscovered oil reserves or unknown insect species) are as
much autonomous, human-independent realities as the
accretion of the earth. The fact that these processes are
contemporaneous with the existence of consciousness,
while the accretion of the earth preceded it, is quite
irrelevant. To maintain the contrary, and insist that it is
only the ancestral dimension that transcends correlational
constitution, is to imply that the emergence of consciousness marks some sort of fundamental ontological rupture,
shattering the autonomy and consistency of reality, such
that once consciousness has emerged on the scene, nothing
can pursue an independent existence any more. The
danger is that in privileging the arche-fossil as sole
paradigm of a mind-independent reality, Meillassoux is
ceding too much ground to the correlationism he wishes to
destroy.8
3. THE P RINCIPLE OF FACTUALITY
Meillassoux distinguishes between two varieties of correlationism: weak correlationism, which claims that we can
think noumena even though we cannot know them, and
strong correlationism, which claims that we cannot even
think them. Weak correlationism, exemplified by Kant,
insists on the finitude of reason and the conditional nature
of our access to being. The conditions for knowledge (the
8. I am indebted to Graham Harman, Robin Mackay, and Damian Veal for all these
critical points.
33
Brassier
14/2/07
09:20
Page 34
COLLAPSE II
categories and forms of intuition) apply only to the
phenomenal realm, not to things in-themselves. Thus the
cognitive structures governing the phenomenal realm are
not necessary features of things-in-themselves. We cannot
know why space and time are the only two forms of
intuition or why there are twelve rather than eleven or
thirteen categories. There is no sufficient reason capable of
accounting for such a fact. In this sense, and this sense
alone, these transcendental structures are contingent. But
Hegel will point out that Kant has already overstepped the
boundary between the knowable and the unknowable in
presuming to know that the structure of things-inthemselves differs from the structure of phenomena.
Accordingly, Hegel will proceed to re-inject that which is
transcendentally constitutive of the ‘for us’ back into the
‘in-itself’. Thus in Hegel’s absolute idealism thinking
grounds its own access to being once more and rediscovers
its intrinsic infinitude. Where Kant’s weak correlationism
emphasises the uncircumventable contingency inherent in
the correlation between thinking and being, Hegelianism
absolutizes the correlation and thereby insists on the
necessary isomorphy between the structure of thinking and
that of being. In this regard, strong correlationism, which
encompasses everything from phenomenology to
pragmatism, can be understood as a critical rejoinder to the
Hegelian absolutization of correlation. Though strong correlationism also jettisons the thing-in-itself, it retains the
Kantian premium on the ineluctable contingency of the
correlation, which Heidegger famously radicalizes in the
notion of ‘facticity’. Thus strong correlationism, as
exemplified by figures such as Heidegger and Foucault,
34
Brassier
14/2/07
09:20
Page 35
Brassier – Enigma of Realism
insists – contra Hegel – that the contingency of correlation
cannot be rationalised or grounded in reason. This is the
anti-metaphysical import of Heidegger’s epochal ‘history
of being’ or of Foucault’s ‘archaeologies of knowledge’.
Accordingly, if we are to break with correlationism, we
must re-legitimate the possibility of thinking the thing-initself, yet do so without either absolutizing correlation or
resorting to the Principle of Sufficient Reason.
In a remarkable tour de force, Meillassoux shows how
what is most powerful in strong correlationism can be used
to overcome it from within. And what is most powerful in
it is precisely its insistence on the facticity of correlation.
For on what basis does strong correlationism reject the
Hegelian rehabilitation of the Principle of Sufficient
Reason – the claim that contradiction is the ground of
being – and the ensuing isomorphy between thinking and
being? It does so by insisting on the facticity or nonnecessity of correlation against its Hegelian absolutization
– thought’s access to being is extrinsically conditioned by
non-conceptual factors, which cannot be rationalised or
reincorporated within the concept, not even in the form of
dialectical contradiction. Thus, in order to emphasize the
primacy of facticity against the speculative temptation to
absolutize correlation, strong correlationism must insist
that everything is without reason – even correlation itself.
Against Hegel’s speculative idealism, which seeks to show
how the correlation can demonstrate its own necessity by
grounding itself, thereby becoming absolutely necessary or
causa sui, strong correlationism must maintain that such
self-grounding is impossible by demonstrating that the
correlation cannot know itself to be necessary. For though
35
Brassier
14/2/07
09:20
Page 36
COLLAPSE II
we can claim that an empirical phenomenon is necessary or
contingent in conformity with the transcendental principles
governing the possibility of knowledge, we cannot know
whether these principles themselves are either necessary or
contingent, for we have nothing to compare them to. This
argument proceeds on the basis of a distinction between
contingency, which is under the jurisdiction of knowledge,
and facticity, which is not. Contingency is empirical and
pertains to phenomena: a phenomenon is contingent if it
can come into or out of existence without violating the
principles of cognition that govern phenomena. Facticity is
transcendental and pertains to our cognitive relation to
phenomena, and hence to the principles of knowledge
themselves, concerning which it makes no sense to say
either that they are necessary or that they are contingent,
since we have no other principles to compare them to.
Against absolute idealism then, strong correlationism
insists that to affirm the necessity of the correlation is to
contravene the norms of knowledge. Yet in so doing, it
violates its own stricture: for in order to claim that the
correlation is not necessary, it has no choice but to affirm
its contingency.
Accordingly, strong correlationism is obliged to
contravene its own distinction between what is knowable
and what is unknowable in order to protect it; it must
assert the contingency of correlation in order to contradict
the idealist’s assertion of its necessity. But to affirm the
contingency of correlation is also to assert the necessity of
facticity and hence to overstep the boundary between what
can be known – contingency – and what cannot be known
– facticity – in the very movement that is supposed to
36
Brassier
14/2/07
09:20
Page 37
Brassier – Enigma of Realism
reassert its inviolability. For in order to maintain the
contingency of correlation and stave off absolute idealism,
strong correlationism must insist on the necessity of its
facticity – but it cannot do so without knowing something
which, by its own lights, it is not supposed to know. Thus
it finds itself confronted with the following dilemma: it
cannot de-absolutize facticity without absolutizing the
correlation; yet it cannot de-absolutize the correlation
without absolutizing facticity. But to absolutize facticity is
to assert the unconditional necessity of its contingency and
hence to assert that it is possible to think something that
exists independently of thought’s relation to it:
contingency as such. In absolutizing facticity, correlationism subverts the empirical-transcendental divide separating
knowable contingency from unknowable facticity even as
it strives to maintain it; but it is thereby forced to
acknowledge that what it took to be a negative characteristic of our relation to things – viz., that we cannot know
whether the principles of cognition are necessary or
contingent – is in fact a positive characteristic of things-inthemselves.
It is worthwhile pausing here to underline the decisive
distinction between the idealist and realist variants of the
speculative overcoming of correlationism. Speculative
idealism claims that the in-itself is not some transcendent
object standing ‘outside’ the correlation, but is rather
nothing other than the correlation as such. Thus it
converts relationality per se into a thing-in-itself or absolute:
the dialectician claims that we overcome the metaphysical
reification of the in-itself when we realize that what we took
to be merely for-us is in fact in-itself. Correlation is
37
Brassier
14/2/07
09:20
Page 38
COLLAPSE II
absolutized when it becomes in-itself for-itself. But this
involves transforming correlation into a metaphysically
necessary entity or causa sui. By way of contrast,
Meillassoux’s speculative materialism asserts that the only
way to preserve the in-itself from its idealist incorporation
into the for-us without reifying it metaphysically is by
realizing that what is in-itself is the contingency of the for-us,
not its necessity. Thus, when facticity is absolutized, it is
the contingency or groundlessness of the for-us (the
correlation) which becomes in-itself or necessary precisely
insofar as its contingency is not something which is merely
for-us. Speculative materialism asserts that, in order to
maintain our ignorance of the necessity of correlation, we
have to know that its contingency is necessary. In other
words, if we can never know the necessity of anything, this
is not because necessity is unknowable but because we
know that only contingency necessarily exists. What is
absolute is the fact that everything is necessarily contingent
or ‘without-reason’.
Consequently, when forced to pursue the ultimate
consequences of its own premises, correlationism is obliged
to turn our ignorance concerning the necessity or
contingency of our knowledge of phenomena into a
thinkable property of things-in-themselves. The result, as
Meillassoux puts it, is that ‘[t]he absolute is the absolute
impossibility of a necessary being’.9 This is Meillassoux’s
‘principle of factuality’ and though it might seem
exceedingly slight, its implications are far from trivial. For
it imposes significant constraints upon thought. If a
necessary being is conceptually impossible then the only
9 Après la finitude, 82.
38
Brassier
14/2/07
09:20
Page 39
Brassier – Enigma of Realism
absolute is the real possibility of the completely arbitrary
and radically unpredictable transformation of all things
from one moment to the next. It is important not to
confuse this with familiar Heraclitean or Nietzschean
paeans to absolute becoming, for the latter merely
substitutes the metaphysical necessity of perpetual differentiation for the metaphysical necessity of perpetual identity.
To affirm the metaphysical primacy of becoming is to
claim that it is impossible for things not to change;
impossible for things to stay the same; and ergo to claim
that it is necessary for things to keep changing. The flux of
ceaseless becoming is thereby conceived as ineluctable and
metaphysically necessary as unchanging stasis. But metaphysical necessity, whether it be that of perpetual flux or of
permanent fixity, is precisely what the principle of absolute
contingency rules out. The necessity of contingency,
Meillassoux maintains, implies an ‘absolute time’ which
can interrupt the flux of becoming with the same arbitrary
capriciousness as it can scramble the fixity of being.
Absolute time is tantamount to a ‘hyper-chaos’ for which
nothing is impossible, unless it be the production of a
necessary being. It is a contingency which usurps every
possible order, including the order of disorder or the
constancy of inconstancy. It is all-powerful; but an absolute
power which is ‘without norms, blind, and devoid of all the
other divine perfections […] It is a power possessing
neither goodness nor wisdom […] a time capable of
destroying becoming itself by bringing forth, perhaps
forever, fixity, stasis, and death’.10
10. Ibid., 88.
39
Brassier
14/2/07
09:20
Page 40
COLLAPSE II
4. THE PARADOX OF ABSOLUTE CONTINGENCY
In a move that effectively sidesteps the entire
problematic of representation, Meillassoux boldly declares
his intention to reinstate intellectual intuition:
[W]e must project unreason into the thing itself, and
discover in our grasp of facticity the veritable intellectual
intuition of the absolute. ‘Intuition’, since it is well and
truly in [à même] what is that we discover a contingency
with no bounds other than itself; ‘intellectual’, since this
contingency is nothing visible, nothing perceptible in the
thing: only thought can access it as it accesses the Chaos
which underlies the apparent continuities of
phenomena.11
The deployment of this presumably non-metaphysical
variety of intellectual intuition circumvents Kant’s critical
distinction between knowable phenomena and
unknowable things-in-themselves – between reality as we
relate to it through representation and reality as it is
independently of our representational relation to it – and
rehabilitates the distinction between primary and
secondary qualities; the former being mathematically
intuitable features of things-in-themselves; the latter being
phenomenological features of our relation to things.12 This
reinstatement of intellectual intuition is of a piece with
Meillassoux’s overturning of Kant’s critical delimitation of
the possibilities of reason. Intellectual intuition now
provides us with direct access to a realm of pure possibility
coextensive with absolute time. Kant displaced the
metaphysical hypostatization of logical possibility by
11. Ibid., 111.
12. Ibid., 28.
40
Brassier
14/2/07
09:20
Page 41
Brassier – Enigma of Realism
subordinating the latter to a domain of real possibility
circumscribed by reason’s relation to sensibility. Time qua
form of transcendental synthesis grounds the structure of
possibility.13 But Meillassoux’s absolutization of
contingency effectively absolutizes the a priori realm of pure
logical possibility and untethers the domain of mathematical intelligibility from sensibility. This severing of the
possible from the sensible is underwritten by the chaotic
structure of absolute time. Where the bounds of real
possibility remain circumscribed by the correlational a
priori, intellectual intuition uncovers a realm of absolute
possibility whose only constraint is non-contradiction.
Moreover, where real possibility is subsumed by time as
form of transcendental subjectivity, absolute possibility
indexes a time no longer anchored in the coherence of a
subjective relation to reality or in the correlation between
thinking and being. Thus the intellectual intuition of
absolute possibility underwrites the ‘diachronicity’ of
thinking and being; a diachronicity which for Meillassoux
is implicit in the ancestral dimension of being uncovered by
modern science. In ratifying the diachronicity of thinking
and being, modern science exposes thought’s contingency
for being: although thought needs being, being does not
need thought.
The question, then, is whether Meillassoux’s
reinstatement of intellectual intuition may not compromise
the very asymmetry which he takes to be science’s
speculative import. Similarly, it may be that the Galilean
hypothesis harbours ramifications concerning the
13. This is the upshot of Heidegger’s reinterpretation of Kant in Kant and the Problem
of Metaphysics, tr. R. Taft (Indianapolis: Indiana University Press, 1990).
41
Brassier
14/2/07
09:20
Page 42
COLLAPSE II
mathematization of thinking which also vitiate
Meillassoux’s appeal to intellectual intuition. To consider
these questions, we must examine the distinction which
Meillassoux invokes in order to stave off idealism. This is
the distinction between the reality of the ancestral
phenomenon and the ideality of the ancestral statement. It is
on the basis of this distinction that Meillassoux, like
Badiou, seeks to distance himself from the Pythagorean
thesis according to which being is mathematical:
[W]e will maintain that, for their part, the statements
bearing on the ancestral phenomenon which can be mathematically formulated designate effective properties of the
event in question (its date, its duration, its extension),
even though no observer was present to experience it
directly. Accordingly, we will maintain a Cartesian thesis
about matter, but not, let us underline this, a Pythagorean
one: we shall not claim that the being of the ancestral
phenomenon is intrinsically mathematical, or that the
numbers and equations deployed in ancestral statements
exist in themselves. For it would then be necessary to
maintain that the ancestral phenomenon is a reality every
bit as ideal as that of a number or an equation. Generally
speaking, statements are ideal insofar as they possess a
signifying reality; but their eventual referents are not necessarily ideal (the cat on the mat is real, though the statement ‘The cat is on the mat’ is ideal.) In this regard, we
will say that the referents of ancestral statements bearing
on dates, volumes, etc. existed 4.56 billion years ago as
described by these statements – but not these statements
themselves, which are contemporaneous with us.14
14. Après la finitude, 28-9.
42
Brassier
14/2/07
09:20
Page 43
Brassier – Enigma of Realism
This distinction between the reality of the ancestral
phenomenon and the ideality of the ancestral statement is
necessary in order to maintain the ontological disjunction
between the correlational present and the ancestral past –
precisely the diachronicity which correlationism cannot
countenance. Nevertheless, if Meillassoux evokes such a
distinction, he cannot sequester it on the side of being
alone, for it must pertain to thinking as well as to being.
Thus this secondary disjunction between real and ideal
subdivides both poles of the primary disjunction between
thinking and being: thought possesses a real and an ideal
aspect, just as being possesses real and ideal features.
Clearly, the diachronicity harboured by the arche-fossil can
only be indexed by a disjunction between the ideality of
the ancestral statement and the reality of the ancestral
phenomenon which promises to prove irreducible to the
neighbouring distinctions between the real and ideal aspect
of thought and the real and ideal features of being, for both
of these remain entirely encompassed by the correlation
between thinking and being. For the point of Meillassoux’s
distinction between physical reality and discursive ideality
is to discount the idealist claim that the reality of the
phenomenon is exhausted by its mathematical idealization
in the statement. Although the reality of the ancestral
phenomenon can be mathematically encoded, it must
transcend this mathematical inscription, otherwise
Meillassoux finds himself endorsing Pythagoreanism. And
as Meillassoux well knows, the latter provides no bulwark
against correlationism, since it effectively renders being
isomorphic with mathematical ideality. The point seems to
be that the reality of the ancestral phenomenon must be
43
Brassier
14/2/07
09:20
Page 44
COLLAPSE II
independent of its mathematical intellection – being does
not depend upon the existence of mathematics. But
Meillassoux’s problem consists in identifying a speculative
guarantor for this disjunction between reality and ideality
which would be entirely independent of the evidence
provided by the mathematical idealization of the ancestral
phenomenon in the ancestral statement. To rely upon the
latter would be to render this speculative disjunction supervenient upon the procedures of post-critical epistemology
and thus to find oneself confronted by the injunction to
verify or otherwise justify it within the ambit of the correlationist circle.
Thus the question confronting Meillassoux’s
speculative materialism is: under what conditions would
this secondary disjunction between the real and the ideal
be intellectually intuitable without reinstating a correlation
at the level of the primary disjunction between thinking
and being? To render the distinction between the reality of
the phenomenon and the ideality of the statement
dependent upon intellectual intuition is to leave it entirely
encompassed by one pole of the primary disjunction, i.e.
thought, and hence to recapitulate the correlationist circle.
For just as we cannot maintain that this primary
disjunction is intellectually intuitable without reinscribing
being within the ideal pole of the secondary disjunction,
similarly, we cannot maintain that the secondary
disjunction is encoded in the ancestral statement without
reincorporating the real within the noetic pole of the
primary disjunction. How, then, are we to guarantee the
disjunction between real and ideal independently of the
intelligible ideality of science’s ancestral claims? For the
44
Brassier
14/2/07
09:20
Page 45
Brassier – Enigma of Realism
ideality of the latter cannot be a guarantor of the reality of
the former. Moreover, intellectual intuition subsumes both
poles of the secondary disjunction within one pole of the
primary disjunction.
Consequently, Meillassoux is forced into the difficult
position of attempting to reconcile the claim that being is
not inherently mathematical with the claim that being is
intrinsically accessible to intellectual intuition. He cannot
maintain that being is mathematical without lapsing into
Pythagorean idealism; but this relapse into
Pythagoreanism is precluded only at the cost of the
idealism which renders being the correlate of intellectual
intuition. The problem lies in trying to square the GalileanCartesian hypothesis that being is mathematizable with an
insistence on the speculative disjunction whereby being is
held to subsist independently of its mathematical intuitability. Part of the difficulty resides in the fact that although
Meillassoux presumably discounts metaphysical and phenomenological conceptions of being, whether as necessary
substance or eidetic presence, since both are encompassed
within the correlationist circle, he has not provided us with
a non-metaphysical and non-phenomenological alternative
– such as we find, for example, in Badiou’s subtractive
conception of the void.15 Like Badiou, Meillassoux recuses
the Kantian formulation of the problematic of access while
striving to uphold the authority of scientific rationality.
However, unlike Badiou, he does not characterize ontology
as a situation within which the presentation of being is subtractively inscribed in such a way as to obviate any
straightforwardly metaphysical or phenomenological
15. Cf. Being and Event, tr. O. Feltham (London: Continuum, 2006).
45
Brassier
14/2/07
09:20
Page 46
COLLAPSE II
correlation between thought and being.
But as a result he must explain why – given that
science teaches us that intellection is in no way an ineliminable feature of reality but merely a contingent by-product
of evolutionary history, and given that for Meillassoux
himself reality can be neither inherently mathematical nor
necessarily intelligible – being should be susceptible to intellectual intuition. In this regard, it is worth noting that one
of the more significant ramifications of the GalileanCartesian hypothesis about the mathematizability of nature
consists in the recent endeavour to deploy the resources of
mathematical modelization in order to develop a science of
cognition. Admittedly, the latter is still in its infancy; nevertheless, its maturation promises to obviate the Cartesian
dualism of thought and extension – and perhaps also the
residues of the latter which subsist in Meillassoux’s own
brand of speculative materialism – while conceding
nothing to correlationism. The diachronic disjunction
between thinking and being is not the only speculative
implication harboured by modern science; the
development of a science of cognition implies that we,
unlike Descartes and Kant, can no longer presume to
exempt thought from the reality to which it provides
access, or continue to attribute an exceptional status to it.
If thought can no longer be presumed to exempt itself
from the reality which it thinks, and if the real can no
longer be directly mapped onto being, or the ideal directly
mapped onto thought, then thinking itself must be reintegrated into any speculative enquiry into the nature of
reality. Thus the central question raised by Meillassoux’s
speculative materialism becomes: Does the principle of
46
Brassier
14/2/07
09:20
Page 47
Brassier – Enigma of Realism
factuality, which states that ‘everything that exists is
necessarily contingent’, include itself in its designation of
‘everything’? Like Badiou, Meillassoux sees Cantor as
having definitively pulverized the concept of ‘totality’, so
that the latter is now devoid of ontological pertinence. But
we do not have to assume a spurious totalization of
existence to enquire whether the thought that everything is
necessarily contingent is itself necessarily contingent. On
the contrary, all that we assume is that thinking is just a
contingent fact like any other. What we should refuse,
however, is the claim that it is necessary to exempt the
thought that ‘everything is necessarily contingent’ from the
existential ‘fact’ that everything is contingent on the
grounds that a transcendental abyss separates thinking
from being. Once the recourse to this transcendental
divide has been ruled out, we are obliged to consider what
follows if the principle refers to itself. More precisely, we
must consider whether the truth of the principle, and a
fortiori Meillassoux’s speculative overcoming of correlationism, entails its self-reference. Here we have to distinguish
between the contingency of the existence of the thought,
which does not generate paradox, and the contingency of
the truth of the thought, which does. Two distinct possibilities can be envisaged depending on whether the thought
does or does not refer to itself. First let us consider what
follows if it does refer to itself. Then if the thought exists,
it must be contingent. But if it is contingent then its
negation could equally exist: ‘Not everything is necessarily
contingent’. But in order for the thought to exclude the
possibility of the truth of its negation, then its truth must
be necessary, which means that the thought must exist
47
Brassier
14/2/07
09:20
Page 48
COLLAPSE II
necessarily. But if it exists necessarily then not everything
that exists is necessarily contingent; there is at least one
thing which is not, i.e. the thought itself. Thus if the
thought refers to itself it necessitates the existence of its
own negation; but in order to deny the possible truth of its
negation it has to affirm its own necessary truth, and hence
contradict itself once more. What if the thought does not
refer to itself? Then there is something which is necessary,
but which is not included under the rubric of existence.
Reality is ‘not-all’ because the thought that ‘everything is
necessarily contingent’ is an ideality which exempts itself
from the reality which it designates. But then not only does
this very exemption become necessary for the intelligible
ideality of the thought that ‘everything is
necessarily contingent’, but the intelligibility of reality
understood as the necessary existence of contingency
becomes dependent upon the coherence of a thought
whose exemption from reality is necessary in order for
reality to be thought as necessarily contingent. Thus the
attempt to exempt the ideal from the real threatens to reinstantiate the correlationist circle once more. Lastly, let us
consider the possibility that the necessary contingency of
existence does not depend on the truth of the thought
‘everything is necessarily contingent’. If everything is
necessarily contingent regardless of the truth of the
thought ‘everything is necessarily contingent’, then
everything could be necessarily contingent even if we had
no way of thinking the truth of that thought coherently.
But this is to re-introduce the possibility of a radical
discrepancy between the coherence of thinking and the
way the world is in-itself – any irrational hypothesis about
48
Brassier
14/2/07
09:20
Page 49
Brassier – Enigma of Realism
the latter become possible and strong correlationism looms
once again.
Whatever the shortcomings attendant upon their lack
of formal stringency, these conjectures seem to point to a
fundamental dilemma confronting Meillassoux’s project. If
he accepts – as we believe he must – that thinking is part
of being as the second fundamental speculative implication
of scientific rationality after that of diachronicity, then the
universal scope of the principle of factuality generates a
paradox whereby it seems to contradict itself: the claim
that everything is necessarily contingent is only true if this
thought exists necessarily. Alternatively, if Meillassoux
decides to uphold the exceptional status of thinking vis-àvis being then he seems to compromise his insistence on
diachronicity, for the intelligible reality of contingent being
is rendered dependent upon the ideal coherence of the
principle of factuality. Indeed, the appeal to intellectual
intuition in the formulation of the principle already seems
to assume some sort of reciprocity between thinking and
being
As one might expect, both these criticisms – viz., that
intellectual intuition reestablishes a correlation between
thought and being and that the principle of factuality
engenders a paradox – have elicited typically acute
responses from Meillassoux. In a personal communication,
Meillassoux has explained why he believes he can parry
both objections. For Meillassoux, the principle of factuality
is designed to satisfy two requirements. First, the
fundamental rationalist requirement that reality be perfectly
amenable to conceptual comprehension. This is a rebuttal
of the prototypical religious notion that existence harbors
49
Brassier
14/2/07
09:20
Page 50
COLLAPSE II
some sort of transcendent mystery forever refractory to
intellection. Second, the basic materialist requirement that
being, though perfectly intelligible, remain irreducible to
thought. Meillassoux insists that the claim that everything
that is is necessarily contingent satisfies both criteria. In his
own words:
Being is thought without-remainder insofar as it is without-reason; and the being that is thought in this way is
conceived as exceeding thought on all sides because it
shows itself to be capable of producing and destroying
thought as well as every other sort of entity. As a factual
act produced by an equally factual thinking being, the
intellectual intuition of facticity is perfectly susceptible to
destruction, but not that which, albeit only for an instant,
it will have thought as the eternal truth which legitimates
its name, viz., that it is itself perishable just like everything
else that exists. […] Thus, it is on account of its capacity
for a-rational emergence that being exceeds on all sides
whatever thought is able to describe of its factual production; nevertheless, it contains nothing unfathomable for
thought because being’s excess over thought just indicates
that reason is forever absent from being, not some eternally enigmatic power.16
These remarks already prefigure Meillassoux’s recusal
of the second objection, viz., that if applied to itself, the
principle of factuality becomes contradictory. Meillassoux
maintains that the paradox can be averted by carefully distinguishing the referent of the principle from its (factual)
existence. Thus, though the latter is indeed contingent, and
hence as liable to be as not to be, the former is strictly
16. Personal Communication, 9/8/2006.
50
Brassier
14/2/07
09:20
Page 51
Brassier – Enigma of Realism
necessary, and indeed it is the eternal necessity of the
principle’s referent that guarantees the perpetual
contingency of the principle’s existence:
One may then say that the principle as something that is
thought in reality is factual, and hence contingent. But
what is not contingent is the referent of this principle; viz.,
facticity as such insofar as it is necessary. And it is
because this facticity is necessary that the principle, insofar as it is – in fact – proffered and insofar as it will be or
will have been thought by some singular entity – no matter when or under what circumstances – it is for this reason that the principle will always be true the moment it is
posited or thought. What is contingent is that the principle, as a meaningful statement, is actually thought; but
what is not contingent is that it is true insofar as it is – as
a matter of fact – thought in a time and place – no matter
when or where. Consequently there is no paradox so
long as the principle’s domain of application is precisely
restricted to entities in their being.17
The crucial operative distinction here is that between
the necessity of contingency qua referent of thought and
the contingency of the (factual) existence of the thought
that everything is necessarily contingent. The question
then is: How does Meillassoux propose to account for this
separation between the contingent existence of thought
and the necessary existence of its referent? Clearly, this
separation is intended to safeguard the coherence of the
principle, as well as the materialist primacy of the real over
the ideal, by ensuring a strict differentiation between
thought and reality. But given that, for Meillassoux,
17. Ibid.
51
Brassier
14/2/07
09:20
Page 52
COLLAPSE II
thought’s purchase on reality is guaranteed by intellectual
intuition, it follows that it must also be the latter which
accounts for this distinction between thought and referent.
Accordingly, it would seem that it is in and through the
intellectual intuition of absolute contingency that the
contingency of the thought is separated from the necessity
of its referent. Everything then hinges on how Meillassoux
understands the term ‘intellectual intuition’.
Clearly, he cannot be using the term in its Kantian
acceptation, since, for Kant, intellectual intuition actively
create its own object, unlike sensible intuition, which
passively receives an independently existing object.
According to Kant, only the intuitive understanding of an
‘archetypical’ intellect (intellectus archetypus) unburdened by
sensibility – such as God’s – possesses this power to
produce its object; for our discursive understanding,
mediated as it is by sensibility, it is the synthesis of concept
and intuition which yields the cognitive relation between
thought and its object. Meillassoux clearly rejects Kant’s
representationalist account of the relation between mind
and world, just as he must refuse phenomenology’s appeal
to an intentional correlation between thought and referent.
Yet it is far from evident what plausible theory of intellectual intuition could simultaneously ensure the scission
between the contingency of thought and the necessity of its
referent – which Meillassoux takes to be sufficient to stave
off contradiction – while circumventing representational
and intentional correlation as well as abjuring the archetypical intellect’s production of its object (since the claim that
intellection creates its object is clearly incompatible with
any commitment to materialism). Though Meillassoux
52
Brassier
14/2/07
09:20
Page 53
Brassier – Enigma of Realism
insists that the paradox of absolute contingency can be
obviated by restricting the principle’s domain of reference
to ‘entities in their being’, he does not explain how he
proposes to enforce this rigid demarcation between the
principle’s contingently effectuated intension and what he
deems to be its ‘eternally’ necessary extension.
‘Reference’, of course, is intimately related to ‘truth’, but
though Meillassoux claims that the truth of the principle is
guaranteed by its ontological referent, this connection is
anything but semantically transparent, since the extension
of the expression ‘absolute contingency’ is no more
perspicuous than that of the term ‘being’. The customary
prerequisite for realist conceptions of truth is an extratheoretical account of the relation between intension and
extension, but Meillassoux’s attempt to construe the latter
in terms of intellectual intuition makes it exceedingly
difficult to see how it could ever be anything other than
intra-theoretical.18 Indeed, it is unclear how the referent
‘absolute contingency’ could ever be rendered intelligible
in anything other than a purely conceptual register.
Consequently, Meillassoux presents us with a case in
which the determination of extension, or ‘truth’, remains
entirely dependent upon a conceptually stipulated
intension, or ‘sense’ – the referent ‘absolute contingency’ is
exclusively determined by the sense of the contingently
existing thought ‘everything that is, is absolutely
contingent’. But if the only way to ensure the separation
between the (contingently existing) ideality of meaning
and the (necessarily existing) reality of the referent is by
18. Cf. Hilary Putnam, ‘The Meaning of ‘Meaning’’ in Mind, Language, and Reality:
Philosophical Papers Volume 2 (Cambridge: Cambridge University Press, 1975), 236.
53
Brassier
14/2/07
09:20
Page 54
COLLAPSE II
making conceptuality constitutive of objectivity, then the
absolutization of the non-correlational referent is won at
the price of an absolutization of conceptual sense which
violates the materialist requirement that being not be
reducible to thought. Far from reconciling rationalism with
materialism, the principle of factuality, at least in this
version, continues to subordinate extra-conceptual reality
to a concept of absolute contingency.
Although Meillassoux’s speculative overcoming of
correlationism strives to deploy the latter’s strongest
weapons against it – as we saw with the principle of
factuality itself – the distinction between the real and the
ideal is part of the correlationist legacy which cannot be
mobilized against it without first undergoing decontamination. For correlationism secures the transcendental divide
between the real and the ideal only at the cost of turning
being into the correlate of thought. Meillassoux is right to
insist that it is necessary to pass through correlationism in
order to overcome it, and in this regard we should follow
his recommendation and find a way of deploying the
distinction between real and ideal against correlationism
itself. But precisely here a fundamental speculative problem
reveals itself, namely: Can we think the diachronic
disjunction between real and ideal while obviating any
recourse to a transcendental divide between thinking and
being?
54