The Hypothesis of a Non-Epistemology

Translated by Raymond Brassier



1. What does the hypothesis of a non-epistemology signify?

What does the hypothesis of a non-epistemology bring to non-philosophy? Our aim is not to found a new discipline, but to take into account a vast reservoir of conceptual material that accompanies the sciences and put it to a new use. It may be that this does in fact pertain to a discipline –we can discuss this later, along with the question of the constitution of such practices as occasioned by other materials. Our aim is not critical either, since we do not seek to minimize the importance of written works about the sciences, but rather to redistribute such works on the basis of this new use. The use as such is the critique. This is a hypothesis, insofar as it can only be effectuated by way of its progressive realization, it is not something that can be obtained by principle, by surveying all these discourses.

Consequently, a particular region can be tackled only through its material, and non-philosophy will be able to explain how it is using this material.




2. The material of non-epistemology

We will take epistemology as it is, but we will suspend its pretension to be a description of the sciences as well as their objects. This is what we mean by reading it as a material. What do we then notice about epistemology? Science is implicated in it, that much is obvious. But philosophies are also implicated in it, which is less obvious, and one of our hypotheses is that one can only talk about science through concepts that combine various sciences and philosophies. This thesis allows us to see two basic kinds of philosophical discourse about science. There are those that can be characterized as general and ideal representations of science, and which carry out a fundamental classification of the types of knowledge and arrange them hierarchically. We call them ‘philosophies of science’ and they have an ontologico-transcendental thematic. Their goal is to articulate the fundamentals: science, theory, experience, concept, empirical, etc. They bring out the extraordinary nature of the sciences vis-a-vis other forms of knowledge, revealing their exemplary or paradigmatic character with regard to cognition. But as a result –and as certain philosophers have pointed out– they have a tendency to dismiss anything in science that pertains to the ordinary, to the mundane, to facts.

Epistemology compensates for this by proposing descriptions of the sciences, taking into account such issues as scale, connections between disciplines, exchanges between sciences and philosophies, modes of local validation, debates between science and ethics, and practical conditions for the production of science. By staying closer to the local, epistemology elaborates a sort of common sense, of the commonplace, which renders science thinkable for philosophy, whereas the increasing specialization of the two disciplines, and that of the sciences in particular, seemed to render any such links impossible. The thematic of epistemology is principally ontico-regional, and particularly concerned with the constitution of the notion of ‘fact’. It carries out a philosophical and ethical ‘democratization’ of the vast expanse of the sciences, rather than effecting a democratization of philosophy within the sciences –at least at the outset. It inaugurates a new theoretical genre, and manifests itself quite explicitly as a combination, not as an ideal hierarchy.




3) Epistemology has lost its object


Yet epistemology and the philosophy of science share a common assumption, which is that science can be an object for them. This assumption seems so self-evident it is hardly ever addressed. But it is a huge assumption, and one which has oriented reflection on the sciences along very specific paths. It is so huge that it has led epistemology to lose its object. The different stages in this ‘loss’ are as follows:


1. In seeking to specify the commonplace aspect of the sciences, epistemology looked for criteria that would allow it to specify its object. There is an entire tradition of research along these lines, which was given new impetus by the Vienna Circle, Popper, Lakatos, Kuhn, Feyerabend, and which originates in part in the work of Poincaré. From a contemporary perspective, we can see what is valuable in such work but also the impossibility of ever identifying such criteria. Many writers have pointed out this failure, whether from an epistemological perspective, in the name of an ‘ecology of practices’, for example; or from a philosophical perspective, in the name of a neo-pragmatist critique of the notion of ‘hard’ fact, for example. This is to say that, from a theoretical viewpoint, epistemology seemed incapable of grasping the identity of the sciences because of the impossibility of finding a criterion.

2. The expansion of the sciences has reached such a degree of generality that they no longer manifest themselves ‘individually’ as sciences but only through a heterogeneous, kaleidoscopic empirical datum, which can no longer be grasped on the basis of the oppositional relation between theory and experience. At the empirical level, the sciences only appear as combinations of science, technics, technology, worldviews, ethics. They have gradually come to be understood as local by-products of the material and ideal conditions of their own production as effected by technology, which ensures the network of connections between heterogeneous realities. Such a circle exhibits the way in which epistemology’s factual description of the sciences overturns the fundamental conceptual oppositions in the philosophy of science.

3. Armed with such combinations, epistemologists begin to describe science as a social practice among others. Laboratory life becomes better known, but one renounces the attempt to come up with a concept of science distinct from these practices. Science is once more dissolved into social practices that preclude a precise circumscription of their specificity insofar as they claim to provide the rule for the description of science –in a word, they take the place previously occupied by the search for criteria of scientificity.





Thus, the history of epistemology show how science as an object has disappeared, despite the fact that there remain a vast number of different descriptions of it. Its future seems to reside outside of what philosophy called ‘science’.

In our opinion, this disappearance has a cause. It resides in the attempt to grasp ‘science’ as a natural object encapsulating its own causality within itself. This was a sure way of handing it over to philosophy. All that was required for science to be given was a portion of scientific text from any period, plus a fragment of laboratory life, and it would be possible to describe it directly. The notion of ‘paradigm’ merely adds a nuance to this schema. By way of contrast, our assumption is that a direct description of science by philosophy is not possible. There are two types of reason for this. There are philosophical reasons –when dealing with a combination, it is not enough to subtract an essence in order to discover the other one. This gesture is itself philosophical and will ultimately only be capable of giving us another combination, which, in itself, is not uninteresting. But there is also another reason, the one that leads us to talk about a non-epistemology: since we are dealing with combinations, each of them, whatever they may be, can serve as the indirect occasion for the description of science. Such a generalization leads to the idea that a direct image of science is no longer possible, there are only occasions –which are by definition combinations– for indirect descriptions of the scientific and philosophical stances. This indirectness can only be established by granting the notion of a real that is non-constructible, whether it be by science or philosophy. And it will assume a positive value only through the general displacement which allows the epistemological descriptions to be admitted as such in the form of a material. The discourses we can bring to bear on science and philosophy are doubly indirect, both on account of the impossibility of extracting essences, whether through empirical means or philosophical ones; but also, and more fundamentally, because it is important to insist that the real qua real has neither philosophical nor scientific characteristics. Insofar as it is genuinely real, it can only be indifferent to the discourses about it, and it is by virtue of this indifference that these discourses will be occasionally determined according to the philosophical and scientific contents that they encounter in the world. In other words, philosophical critique is not sufficient when it comes to constructing a concept of science. The philosophical description of the sciences produces new philosophical variations occasioned by science as such. In this sense, they have a consistency, but one that is philosophical. Science can only appear as a by-product of philosophy; one that is overrated as knowledge, underrated as thought.




4) How do philosophy and epistemology resist science?



One way of tackling the question of the relative autonomy of science and philosophy is to examine to what extent philosophy and epistemology resist the sciences. We will be able to evaluate their key points of divergence through this resistance...



1. When dealing with science, philosophy invariably runs aground on the issue of science’s opaque relation to the real, which philosophy dissolves into perpetually renascent ambiguities. For example, the move from ‘explanation’, which leads us straight toward a metaphysical explanation of the scientific disciplines, and ‘description’ or ‘prediction’, a conception which tend to bring science close to a pragmatics, gives rise to a faded version of these notions, through which science is caught as though in a chiasmus. Regardless of the precise content ascribed to these notions, they present themselves in such a way that they never entirely match up with one another, and this difference can be used as a symptom. For philosophy, science manifests itself as an explanation, but there is always a ‘remainder’ that cannot be reduced to explanation, a relation to the real that is not exhausted by explanation. Alternately, science manifests itself as description, in which case the remainder is seen in terms of a lack of ‘metaphysics’ or ‘consciousness’ or ‘civilization’. It is usually said that science only explains the ‘how’ of phenomena, not their ‘why’. Nevertheless, the ‘why’ is never entirely evacuated and finds its solution in new ‘hows’. The regular connections between phenomena as described by science oscillate between orders of rank that cannot be strictly separated from each other, but that are not equivalent either. This division comes about through philosophy, it is a symptom of the form of non-positional realist cogito that is proper to science, but which philosophy cannot accept because it a realism that cannot be divided or explained, one that does not conform to the dialectical structures of philosophy, one that is not already constructed by philosophy.



Epistemology will transform these ambiguities into objects by attributing specific, non-exhangeable properties to the models of explanation and prediction.



2. Epistemology runs aground when dealing with science because it cannot admit the possibility of a scientific practice that would not depend on a theory. An empirical science depends on a pure science, which functions as its generality. Epistemology here reproduces the hierarchies proper to philosophy of science at its more local scale. This resistance is crystallized in the notion of ‘fact’, which is a hybrid epistemological creation. The category of ‘fact’ is both a hybrid of science and philosophy and immediately doubled into pre-scientific (brute fact) and scientific (scientific fact), which cannot be separated from one another without denying what Poincaré called ‘the value of science’. The ‘fact’ is the minimal prerequisite for philosophical duality; it is dialectic, i.e. difference reduced to the empirical degree of description. It is not surprising to see the notion of fact currently under attack by neo-pragmatists. ‘Fact’ is a philosophical creation used to understand and ‘democratize scientific realism; it even tries to turn it into a commonplace. Thus, although they do not realize it, it is not science that the neo-pragmatists are criticizing, but its reduction, the fact as a philosophical reduction of science. It is an intra-philosophical debate. Epistemology stumbles on this side of the discussion about the value of facts when confronted with those practices of science that do not put theory and experience into dialectical or differential relation with one another. As a result, epistemology either excludes the practice of scientific modeling and the engineering sciences, or else sees them as derivative. A treatment of science as a multiplicity of aspects unilateralized according to the real (as opposed to bi-lateralized within philosophy) makes it possible to take these other practices into account, along with what is clumsily referred to as pluri-disciplinarity. The fundamental question then becomes that of the relative autonomy of scientific practices. What is required in order to think this relative autonomy is the idea that epistemology does not directly and uniquely describe its intended object. Non-epistemology is a practice of distinctions that had been rendered impossible by the idea that epistemology relates directly to science.




Should we then try to overturn these philosophical positions in favour of what resists them? To do so would be to engage in a form of deconstruction of epistemology by showing that the concept of science is philosophical through and through. This has already been successfully accomplished, in Of Grammatology for instance. Our hypothesis, by way of contrast, is that there is science, there is philosophy, there is a discourse about science and a discourse about philosophy, but according to the real. The basic problem concerns the status we are going to attribute to ‘discourse about’ and the use we put it to in the hypothesis of a non-epistemology.




6) Epistemology as constituting a metadiscourse for identifying science

In philosophy, metadiscourse is not a way of avoiding the thing itself but, on the contrary, the fundamental means of identifying it. In order to identify science and philosophy in discourse, philosophy and epistemology produce metadiscourses and metacategories that are probably necessary for every procedure of identification. How can one identify something within discourses that are always already mixed and combined, without appealing to an apparatus that is at once reflected and reflecting; without those accumulations, condensations, and intensifications which in the analysis of an individual are called consciousness? Epistemology usually starts out by positing disciplines as simply given. It is on the basis of this given that it is able to construct systems of representation and of representations of representations that allow it to identify science.

Non-epistemology does not only apply to science, it is also for epistemology. The question as far as the practice of non-epistemology is concerned, is that of knowing what to do with these metadiscourses and metarepresentations. There is a liberation in the realization that epistemology is only indirectly and partially about science, and that consequently it is not a discourse about a ‘real’ but a series of nested discourses about a discourse. It allows for a new use of epistemological distinctions. Non-epistemology is a projection-without-aim of each of the synthetic elements onto a different dimension; it is the taking into account of epistemology as a mixture, and a set of rules for undoing this mixture. This is its initial stage. First, it ruptures continuities, separates terms, suspends syntheses: this is one way of transforming epistemology into a material, i.e. into a discourse that no longer claims to have a direct bearing on its object. Some of the goals of a non-epistemology are as follows: to free up the use of epistemological discourses; to refuse to submit them to the directions for use imposed by the putative synthesis of its objects; to transform the amphibolies of epistemology into particular objects without merely overturning oppositions.

What are these objects? One of the specific features of non-epistemology is that it operates upon a material that is given as a mixture. In a certain sense, epistemology realizes the mixture such as it is thought by philosophy, and it is insofar as it testifies to this mixed-ness that it is interesting and can finally be taken as an object that very quickly reveals its complexity. Epistemology usually manifests the way in which science relates to itself through the intermediary of the world (the philosophy of science), just as it manifests the way in which the world relates to itself through the intermediary of science (technology) –it moves almost naturally from one system of relation to another, simply according to the rules stipulated by a model (machine, difference, game). Its complexity goes beyond the mere relation between philosophy and science, which it complicates by constructing metarepresentations about each of them. Epistemology is an unstable formation because it makes elements that obey different logics function together, i.e. ‘the (ordinary) world’, ‘science’ (its object), and ‘the real’ (that which its object is about, and which consequently escapes epistemology).




6) Referring metadiscourses back to the identity of the real

By suspending these syntheses, one finds oneself confronted with identities, rather than a procedure of identification for science. ‘The fact’ is no longer a synthesis of a given datum and a theoretical construction; both aspects are given separately, without there being a passage from one to the other, or an opposition between them. We are no longer dealing with a synthetic ‘object’, which is at once given and constructed, but with a real object on one hand, and an object of knowledge on the other, both of which are entirely distinct. The mixture constituted by ‘epistemology’ will provide us with entirely disparate elements about science, philosophy, and the commonplace, along with their possibilities of coincidence or synthesis. Each of these metadiscourses must be seen as identities according to the real. The point is no longer to elaborate a philosophy of the sciences, but a science of the identities of these discourses. ‘Discourse about’ is now no longer a foundation for science, it now becomes the object of another science –this is what the non-Gödelian transformation of the philosophy of mathematics means.

What non-epistemology thereby reveals is an order between these disparate elements, because the ‘transformation’ of identification into identity presupposes a non-epistemological real, which will provide a place for epistemology. This order suggests a new way of generalizing notions or operators neglected by epistemology because of the philosophical interplay between opposites. Obviously, non-epistemology is not the restoration of some sort of ‘pure’ discourse, whether it be about science or philosophy. All discourse is a combination. The only thing that can be opposed to a combination is another combination. A scientific concept can be deconstructed only by opposing another combination of science and philosophy to it. Non-epistemology generalizes this state of affairs, but only by first being a practice that ensures the identity of combinations.




7) What does non-epistemology bring to non-philosophy?

Does non-epistemology bring anything new to non-philosophy? Perhaps not much in terms of principles. Its domain does not modify the practice of non-philosophy considered in its minimal gesture. On the other hand, it opens up new perspectives and new notions through the variation of its vast material:


* It reveals epistemology as a kind of common sense and as a form of knowledge for understanding the connections between disciplines; one that is occasioned by science rather than about it.

* It abandons the search for criteria for science without giving up on the latter.

* It examines the effects of reducing the philosophical and its multiplicities in epistemology as a way of restructuring metadiscourses so as to identify science. To that end, it provisionally re-establishes philosophy and its variations so as to be able to talk about ‘science’.

* It generalizes notions or themes that were not visible in the philosophy of science and epistemology.

* It allows one to import the idea of scientific modeling into philosophy, and of modeling the relations between the philosophy of science and epistemology, as well as the relations between philosophy and the sciences.


These aspects can be interpreted in two different ways. One is philosophical and consists in all the possible varieties of distinction. The other is non-philosophical and describes sciences and philosophies by combining minimalism according to the real and complexity according to effectiveness.

We have already emphasized the idea that epistemology is the formation of a common sense that allows one to move between philosophical, scientific and ethical discourses in spite of their heterogeneity. It is a way of effacing this heterogeneity by means of networks, couplings, chiasms, mixtures, additions, accumulations, etc…This only becomes visible from a perspective that does not confine itself to epistemology alone. This common sense manifests itself as such only if one posits the identity that provides the occasion through which philosophy, science, and epistemology become consistent and relatively autonomous formations.

Non-epistemology acknowledges that there is an essence of science, which consists in identity rather in science as such, just as it acknowledges that this identity becomes manifest through the adoption of a stance rather than through the adoption of criteria. It is a minimal stance, more minimal than those criteria used to describe science, which were too specifically bound to a historical situation, but one that allows one to account for the complexities of contemporary science. Minimalism and complexity allow for a redistribution of the mixtures of philosophy and science. Non-epistemology acknowledges that the search for criteria was not entirely vain, but no longer admits these criteria as such, only as partial and indirect descriptions of the scientific stance on the basis of empirical conjunctures. It then becomes possible to use any epistemological fragment whatsoever; not to enrich the real, but the material that occasions indirect description of the real.



8) A liberation and enrichment of the notions that relate science back to the real


The ‘objects’ of non-epistemology are the already generalized multiplicities of science and philosophy, together with their interaction. These multiplicities are not the result of a metaphysical potentialization, which would go ‘beyond’ known or given cases. The world contains the variations of the possible, which when related back to the real, cannot be distinguished from the world or effectiveness. What changes is the fact that the distinctions, although considered to be valid once posited, no longer operate in terms of a play of oppositions or as a reduced or minimal synthesis of the synthesis of opposites. This brings about a very particular syntax, which consists on one hand in radically separating two distinct terms, and on the other in considering them as identical, instead of constructing metarepresentations that lead to their identification.

This procedure brings about an unexpected enrichment of notions. Specifically, that of ‘hypothesis’, which has always been completely underestimated both in the philosophy of science and in epistemology. We are able to show through the practice of non-epistemology how this notion makes it possible to understand both something of scientific creation and something of philosophical creation, according to another mixture which would allow us to deconstruct and enrich those of epistemology. The philosophical history of the notion of hypothesis could be revisited and we would see that its exposition has never been entirely adequate.

In philosophy, the notion of hypothesis is quite impoverished until Leibniz, then Russell, who is a reader of Leibniz, but who also brings about a complete change in the meaning of ‘possible’. In dictionaries, the hypothesis is seen as an affirmation requiring confirmation in order to change nature and become a law or principle, or invalidation in order to be destroyed. In either case it is cancelled as hypothesis and so is merely a redundant instrument. It is necessary to show that by varying the meaning of hypothesis, and by relating it back to what François Laruelle calls the identity of the last instance, it is possible to give an entirely different function to the notion of hypothesis. Many different paths are possible on the basis of the philosophical material. For example, one could turn the hypothesis into what Russell identified as an irreducible ‘propositional attitude’, and subject it to controlled transformations according to the real.

In science, its use is constant, but obscure and passed over in silence. What we mean by this is that even if a scientific discipline constructs the most exhaustive possible list of its hypotheses, the relation it admits between its hypotheses and what they help describe is essentially obscure.

In epistemology, the notion of hypothesis is conspicuous by its absence (see for example the Dictionnaire d’histoire et de philosophie [Dictionary of History and Philosophy] P.U.F., 1999). In our opinion, this absence is due to the specific character of the philosophy-science mixture in epistemology. One of the effects of the reduction of philosophy in epistemology is the manifestation of science as factual –in conformity with the epistemological conception of the factual as mixture, which we have already described. This effect occludes the notion of hypothesis because the latter modifies the actual.

To add the notion of hypothesis is to break the continuities between an affirmation and its cause, or rather, to transform its cause into something that bears no resemblance to these continuities, or to the principles to which they are supposed to lead. Without hypotheses, there is no multiplicity of theories, disciplines, or philosophies. Without hypotheses, there is no articulation joining them to one another. We are trying to elaborate a concept of hypothesis that is not evanescent, and that has two sources: 1) identity; and 2) the positive condition for the connections between disciplines. The notion of hypothesis thereby gains a positive consistency. We propose to philosophize by hypotheses, but also to practice epistemology by hypothesis. This is the theoretical, and hence material, aspect of non-epistemology.

The notion of hypothesis here is only one case –that of the title of this paper and of the risk it takes both with regard to the epistemological tradition and the concept of science. This work would have to be undertaken for all the notions that recur throughout philosophy and epistemology: experimentation, modeling, simulation, etc….Philosophy and science can be characterized according to identity on the basis of unilateralised aspects, whether these come from the world, science or philosophy. One will then be able to organize a great variety, at once finite and unlimited, of possible importations and exportations of notions and disciplines. The relations between philosophies and sciences will no longer be articulated on the basis of one and the same massive assumption, which is at once too big (philosophy’s image of the sciences) and too particular (because of the particular historical origin of its concept), but will be variously modeled on the basis of different dimensions.

The function of non-epistemology should accompany the enormous development of science in such a way as to reveal –through every possible and effective combination of the identity of science– that it is also that of philosophy and ethics, without turning these into authorities on the real. Why? Because there is necessarily discourse about science and about philosophy. What is required is a practice that provides the theses and notions of epistemology both with its limits and its power. It is a relatively autonomous discipline solely by virtue of the donation of its material. But it is above all a practice that displaces the theses of epistemology into indirect hypotheses about science on the basis of the causality that the real exerts both on it and on philosophy. Epistemology, which has gradually abandoned the concept of science in order to encompass it through technology and social practices, will no longer need to be divided between two incompatible futures. The essence of science, technologies, and social practices will all be able to be used as indirect descriptions of the real, for which science provides the occasion for one of its possible descriptions, alongside those of all the other sub-disciplines that can be imagined. In this practice, non-epistemology manifests the identity of the scientist and the ordinary man, or uses the figure of the scientist as that of a commonplace…Thus, as a theory, it pertains to the material and to effectiveness, while as a practice it pertains to the subject. Non-philosophy introduces order into these distinctions –material, subject, real– in such a way as to preclude attributing to man characteristics drawn from one of these sub-discipline, whether that of the scientist or of the philosopher.