The rather unstable historical configuration which came to be known as the 'French epistemological tradition' seems to owe much of its specificity to a distinctive anti-Bergsonian stand. It is nonetheless possible, as we shall argue, to trace a number of Bergsonian patterns in the basic assumptions of what was sometimes identified as an 'anti-positivist' view of the history of ideas. Thus the notion of problem for instance, a fundamental category in Bergson's philosophy, underlies several themes and procedures that seem essential to the self-definition of an 'anti-positivist' epistemology. The idea that problems, even more than theories and concepts, are the genetic element in the development of thought, keeps coming back in Bergson's works, suggesting the stronger claim that his philosophy, more than any other in the beginning of the Twentieth-century, may provide the philosophical grounds for a non-positivist conception of problems themselves.
What follows is probably closer to an ideal reconstruction or thought experiment than to a historical investigation in the traditional sense. For the claim is not really a historical one. If one sets aside Merleau-Ponty, Canguilhem and the Ecole des Annales, it is difficult to find any direct acknowledgement of Bergson's influence upon the actors involved in this rather entangled plot. Quite the contrary, Bergson generally played the role of a scape-goat or obtrusive mentor, and the overthrowing of his ideas proved a most efficient rallying theme for the generation that reached philosophical maturity in the thirties or in the immediate post-war period. Accordingly, the idea of an unavowed Bergsonian heritage expressing itself in a thread of thinkers running from Bachelard to Foucault seems at best counter-intuitive or far-fetched. Challenging the obviousness of what appears as a major topos in the history of French philosophy requires a preliminary examination of the philosophical implications of the so-called 'anti-positivist' epistemological tradition. In order to understand exactly in what sense it is supposed to part with Bergson's theory and practice, it is useful to start from Foucault's famous statement about two conflicting orientations in Twentieth-century French philosophy.
Two Traditions
In an homage to Georges Canguilhem which was originally meant to serve as a preface to the English translation of The Normal and the Pathological, Michel Foucault observed that 'a dividing line' ran through many philosophical and ideological oppositions in the French post-war intellectual landscape 'the one that separates a philosophy of experience, of meaning, of the subject, and a philosophy of knowledge, of rationality, and of the concept.
On one side, a filiation which is that of Jean-Paul Sartre and Maurice Merleau-Ponty ; and then another, which is that of Jean Cavaillès, Gaston Bachelard, Alexandre Koyré, and Canguilhem.' (Foucault 1998466). In the original version of this text, written in 1978, Foucault was content with underlining 'two strains that remained profoundly heterogeneous' (Foucault 1994-1430). He traced back this cleavage to an original discrepancy regarding the interpretation of Husserl's Cartesian Meditations (1929, published in 1931). Bluntly stated, there is Sartre on the one hand, with The Transcendence of the Ego, and Cavaillès on the other, with his two theses on the axiomatic method and set theory. Curiously, in the 1985 revised version of the same paper, published in French in the Revue de Métaphysique et de Morale (probably the last article Foucault published during his lifetime), the genealogy of this tension within French philosophy leads Foucault much further down the timeline, as if the divergent readings of phenomenology were only the symptom of a deeper fracture 'Doubtless this cleavage comes from afar, and one could trace it back through the nineteenth century Henri Bergson and Henri Poincaré, Jules Lachelier and Louis Couturat, Pierre Maine de Biran and Auguste Comte.' (Foucault 1994-2764). It is through this series of oppositions, which Foucault somewhat dramatically describes as deeply-rooted ('well established') as early as the turn of the century, that phenomenology itself is thought to have been filtered and eventually 'admitted' into France.
It is striking that in spite of the rhetorical power of Foucault's assemblage (a genuine historical 'montage', as it is), it is not quite clear, when one thinks of it, why Bergson should naturally fit in the first category, that of the philosophers 'of experience, of meaning, of the subject'. At any rate, this piece of philosophical doxa accounts for an important aspect of the prevalent perception of 'Bergsonism' - a perception that often obscures the exact nature of Bergson's philosophical achievement.
However, one may favour an opposite reading of the dividing lines of this philosophical Kampfplatz. It is not stretching things too far to argue that Bergson in fact provided a genuine philosophy of the concept, although not exactly the kind that Cavaillès had in mind when he said that 'it is not a philosophy of consciousness, but a philosophy of the concept that will provide a doctrine of science.' (Cavaillès 196078). Cavaillès' 'dialectics' amounts to an immanent dynamism, a developmental process of concepts or 'empiricism of thought in action'. Bergson finds the 'generating necessity' of science in 'an activity,' and that of philosophy in a 'superior empiricism,' but such an activity cannot easily be identified either with the traditional activity of judgement and its ensuing doctrine of faculties (in an intellectualist vein still perceptible in Bachelard, a true heir of Brunschvicg in this respect), or with the activity of a transcendental ego (in a phenomenological way). In this respect it does not openly contradict Cavaillès' 'dialectics' (1). Bergson's emphasis is on the particular operations at play in scientific knowledge in general scientific understanding parcels out, arrests, quantifies and measures, whereas reality, as it is given to immediate intuition, is an unfolding, continuous and essentially qualitative process. This is so because however disinterested it may have become today, science remains in close relation with the demands of our action. Yet by the same token, science can be described to a certain extent without the need of a subject of science - as an activity without a subject.
Scientific theories, with their apparatus of forms and principles, are still expressions of life itself in its ability to use intelligence for its own purpose. In this regard, Foucault's statement about the two traditions is perhaps most confusing in the opposition it wishes to stress between Bergson and Poincaré. The former's view of science as rooted in practice is at times strikingly close to the latter's so-called conventionalism, and it is no wonder that Leroy entered the French philosophical scene by conflating Bergsonian philosophy with Poincarean epistemology. Whereas Poincaré firmly protested against what he considered as a dubious manœuvre (see The Value of Science, published in 1905), Bergson did not mind being forced into an alliance with what some described as a 'skeptical' trend in the philosophy of science (Boutroux, Duhem, Milhaud, etc. ; see 'La Philosophie française' in Bergson 19721176-1179).
On the whole, it seems that Bergson cuts across the categories which his name (rather than his actual philosophy) helps devising. As a consequence, what Foucault says about Husserl could very well apply to Bergson himself he may be more aptly described as the point of bifurcation from which two very different traditions emerge and start diverging. Sartre's first encounter with philosophy, as he recalls, was the reading of Time and Free Will - it is ironical, yet not surprising, that despite his harsh criticism of Bergson's views about consciousness and freedom, he should find himself in his company under the heading of 'philosophers of the subject'. Surely, Cavaillès, Bachelard (who wrote The Dialectics of Duration in 1936), and above all Canguilhem (who published in 1943 a commentary of the third chapter of Creative Evolution), had all read Bergson too, though more or less carefully. Yet, as early as the 1930s, following the publication of Politzer's pamphlet, 'Bergsonism' had more or less become the target of almost every innovative philosophical trend in France. The story is well known, and not worth recounting again. It remains to be seen if the motivations underlying the rejection of Bergson are homogeneous. Tracing back diverging readings of Bergson himself (Merleau-Ponty being a much more ambiguous example than Sartre in this regard (2)), would probably reveal cleavages as profound (although probably less apparent) as those Foucault identified in the history of Husserl's reception. Without going that far, it is interesting to bring the case of Foucault into focus. His negative identification of Bergson as an inverted forerunner of the line of thinkers to which he feels endebted, involves a retrospective scheme which tells us a great deal about himself and the nature of his intellectual formation in the heydays of existentialism, when phenomenology (or at least a certain interpretation of the phenomenological project) was the dominant paradigm. It is doubtful that it tells much about Bergson, who is merely presented through the lenses of the post-war situation. The rough opposition drawn by Foucault naturally owes much to the special circumstances of the 'homage,' a genre which easily lends itself to a certain kind of 'monumental' historiography (as Nietzsche would have it).
The text that mentions Bergson was written for a special issue of the Revue de Métaphysique et de Morale devoted to Canguilhem. It is a tribute, and every tribute needs its 'villains'. It is important to bear in mind similar and equally famous instances of this academic ritual tributes paid by Canguilhem to Cavaillès, by Bachelard to Brunschvicg, or by Merleau-Ponty to Bergson, etc.
On other occasions, Bergson is presented by Foucault as being chiefly responsible for a pervasive tendency of contemporary philosophers to downplay space in favour of time in 1978, during an interview with Watanabe, he refers to 'a kind of latent Bergsonism which dominated French philosophy' (Foucault 1994-1576), cautiously qualifying his initial statement with a rather ambiguous remark 'I say Bergsonism, I'm not saying this was the actual Bergson, far from it. There was a certain privilege of all temporal analyses over space, which was held as something dead and inert.' In fact, it turns out that what Foucault wishes to oppose is rather something like an existentialist or vulgarly marxist over-valorization of historicity and historical consciousness at the expense of the 'reactionary' categories of spatiality. The reference to 'a Bergsonian valorization of time' does not only function as a philosophical cliché. It may in fact best be explained from an auto-biographical perspective. It is with the academic primacy of Bergsonian time (or Bergsonian primacy of time) that Foucault himself had to struggle in the fifties and sixties, as he tried to foster a new form of investigation invested in the constructions of space - sites, boundaries, thresholds, where power inscribes its marks. Ironically, however, similarly sweeping accounts of the influence of Bergson on twentieth century French thought lead other commentators to radically divergent claims Martin Jay, for example, traces back the 'denigration of vision' in French philosophy to the Bergsonian critique of spatialization as a specific mechanism of intellectual and social control, a critique which is supposed to bear its effects everywhere, from Luckàcs (whose concept of reification owes much to 'spatialization') and the Marxist tradition, up to Foucault himself (Jay 1993191-208, 430).
This interpretative flip-flop hints at the fact that such totalizing views of Bergson's position in the century are most often put to work for ideological purposes. In the most recent Deleuzian literature, Bergson is often fancied as the denied father of French postmodern philosophy (Douglass 1992). Once again Bergson becomes an overarching figure casting its shadow on the whole philosophical stage, overdetermining every philosophical division. These 'grandes manoeuvres' may well revive Bergsonian studies, but as far as the precise understanding of Bergson is concerned, they do not provide much philosophical insight. Whether he is believed to be heralding what is now known as a philosophy of difference (thus providing the groundwork for the metaphysics of postmodernism), or whether one holds his epistemological insights as a foreshadowing of the contemporary scientific image (in modern physics, psychology, neurosciences, biology, theories of human information-processing, etc.), in every case Bergson functions as a dubious reference point - an origin. Pete Gunter at least warns us against the shortcomings of such 'retrospective' readings, of which it is quite certain that Bergson would have disapproved 'it is no easy matter to specify what amount of influence on, or degree of prophetic anticipation of, the course of scientific thought would constitute a 'verification' of his philosophical methods.' (Gunter 196936)
'Positivism'
Bergson's metaphysical stance involved maintaining as a guiding principle the possibility of an intuitive, immediate grasp of reality, undercutting the interplay of concepts and symbols to which common-sense and science have accustomed us. A rather paradoxical consequence of this central claim is that he was repeatedly charged with being a hidden 'positivist.' Politzer criticized Bergson's notion of duration as substantial, and so did Bachelard when he indicted duration as a massive, positive plenum. Sartre reproached him for confusing the self-transparency of consciousness with the opaque presence of a thing. Bergson, a 'positivist' The accusation may well be misplaced, and yet the concern for immediacy, combined with the praise of a 'positive metaphysics' (in the line of Ravaisson) and a 'superior empiricism,' for all their ambiguity ('metaphysics' and 'superior' being of course essential qualifications here), made Bergson seem like a good candidate for the label. As a matter of fact, some of Bergson's critics have been described as the proponents of a form of 'radical and deliberate non-positivism' (Lecourt 19727), though in a sense that brings another twist to the very notion of positivism. It is Canguilhem who first remarked that Bachelard, having realized that science reformed itself through epistemological breaks, gradually formed 'a conception of the relations between science and the history of science that in itself constituted such a break a non-positivist conception.' (Canguilhem 1970186 ; see Canguilhem 199431-35).
The trademark 'positivism' was recently endorsed by Pete Gunter in relation to Bergson. This clever move was chiefly motivated by the desire to blur the lines drawn by the dominantly positivist Anglo-American epistemology, but it is rather unfortunate from the standpoint of the French context, considering the negative connotations of the term. Of course, describing Bergson as a 'positivist' because 'intuitionism must run the guantlet of observation and experiment,' because the statements drawn from our intuitions definitely satisfy the requirement that, to be factually meaningful, it 'must be possible that experience render them probable - or improbable' (Gunter 19879), is partly a matter of words. If our definition of 'positivism' is modeled after a weakened or more 'pragmatic, biocentric' (Gunter 19876) version of the verificationist principle upheld by the Vienna Circle, it may well be argued that Bergson was indeed a 'positivist' in his own way, at least to the extent that he 'limited factual statements to empirical statements and empirically meaningful statements to those capable of being empirically verified or diversified.' (Gunter 19873).
But this reallocation of Bergson's epistemological legacy within the philosophical scene is misleading in yet another, more important respect. For if 'anti-positivism' is indeed the rallying point of the kind of epistemology that developed around Bachelard and Canguilhem, and whose effects can still be traced in Althusser, Foucault and Bourdieu, it does not chiefly imply a reaction against the given or the immediate (even if it often takes this form in the most psychologically-driven, least philosophically interesting writings of Bachelard). Its philosophical grounds are not primarily a reaction against the tyranny of empirically verifiable statements, the verdict of experience. What anti-positivism fundamentally protests against is a static, non-problematic view of concepts themselves, and the plainly linear view of the development of science that is derived from such a view. The uses of 'positivism' in the French epistemological tradition are naturally far from homogeneous. Just as the uses of 'Bergsonism,' they are certainly loaded with many prejudices and prone to many misunderstandings. Nonetheless, if we were tempted to indulge in the kind of monumental genealogy sketched by Foucault, we would probably have to admit that the overarching figure of the whole scene is Auguste Comte, with his positivist conception of the history of the sciences. And yet, paradoxical as it may seem, in the specific sense of 'positivism' that is now under scrutiny, Auguste Comte would actually count as an anti-positivist. For as Pierre Macherey reminds us, 'he creates the concept of a philosophical history of sciences a history of problems and concepts, rather than a history of solutions and 'theories,' in other words a rational and reasoned history, rather than an empirical, descriptive history...' (Macherey 198995). And he adds in a footnote 'this view has greatly inspired the historical epistemology of G.Bachelard and G.Canguilhem.' As a matter of fact, Bachelard has never kept secret his admiration for Comte 'there is no scientific culture, he says, that does not fulfill the obligations of positivism' (Bachelard 1949104). This rather unexpected filiation wouldn't make sense if we forgot that Comte was precisely not a 'positivist' regarding concepts, even if his definition of the scientific method as a mixture of observation and induction leaves us with a rather weak concept of science.
Despite the ambivalent character of their Comtian origin, the uses of 'positivism' seem to converge in the definition provided by Canguilhem as a preliminary to the understanding of the true task of the historian of sciences 'positivism, a philosophy of history based on a generalization of the notion that theory ineluctably succeeds theory as the true supplants the false.' (Canguilhem 199442).
What is remarkable in Canguilhem's definition is first of all that the dividing line cuts through different conceptions of history, and not primarily different metaphysics or even epistemological stances. 'Positivism' does not imply an a-historical vision of things (the one that would take the universal or natural character of the given for granted) ; it is itself a particular view of history, a 'philosophy of history.' The distinctive mark of this philosophy is that it presents concepts as the more or less adequate replica of verified facts, supplying truths according to which theories must be judged and eventually vindicated or overthrown.
What an 'Anti-Positivist' Epistemology Means
By contrast, the meaning of an anti-positivist epistemology appears more clearly. It is a stance that must first be evaluated according to its strategic effects on certain interpretative and constructive practices. Canguilhem writes 'To take as one's object of inquiry nothing other than sources, inventions, influences, priorities, simultaneities, and successions is at bottom to fail to distinguish between science and other aspects of culture.' (Canguilhem 19883). It is interesting that this indictment of positivist historiographical practices should be couched in quasi-Bergsonian terms.
'Sources' and 'influences' point to the 'retrospective illusion' ; 'simultaneities' and 'successions' are reminiscent of the spatial conception of the historical timeline. Canguilhem says elsewhere that 'A history of results can never be anything more than a chronicle. The history of science concerns an axiological activity, the search for truth. This axiological activity appears only at the level of questions, methods and concepts, and nowhere else.' (Canguilhem 199430). How this bears on the question of historicity itself, and how 'the history of the relation of intelligence to truth generates its own sense of time' (Canguilhem 199431), is another question, but it is a Bergsonian question as well.
It seems that we are now in a position to formulate the general principle of anti-positivism it consists in the belief that epistemology is not concerned with facts (neither scientific facts, nor historical facts uncovered by the history of science), but with concepts. This means that its objects are always projects, and in the case of the history of sciences, 'the object of historical discourse is, in effect, the historicity of scientific discourse.' (Canguilhem 199426).
Yet again, the real problem with positivism is not so much that it strives at some kind of immediate contact with the real it is rather that this attitude implies a very naive idea of what a concept is, in general. Hence the problem is not so much to replace facts by concepts as the proper objects of inquiry, but to reach an adequate understanding of the formation and functioning of concepts in the first place. Whether the substitute for positivism is found in dialectics or intuition, what is being criticized is always an abstract view of concepts which considers them apart from their theoretical setting, the network or system of notions to which they belong, their vital connections in the web of thought (Canguilhem 199450-51).
In his study on the formation of the concept of reflex movement in the Seventeenth and Eighteenth century, Canguilhem emphasizes the necessity of studying conceptual filiations rather than the succession of theories (Canguilhem 1994181). Bachelard develops a similar line of argument concerning Fizeau's experiment we do not know what we are talking about before 'the object of knowledge is replaced in a problematic, situated in a discursive process of instruction.' (Bachelard 194955). On this construal, the object becomes 'more than a historical fact, more than a fact resulting from observation it solves a problem.' (Bachelard 194953)
Briefly stated, the kind of 'positivism' that is being discarded by the French epistemological tradition under consideration is essentially one which considers concepts in isolation from their variation in a problematic configuration, one which proves incapable of engaging in what Canguilhem calls the 'working of a concept' ('Dialectique et philosophie du Non chez Gaston Bachelard,' in Canguilhem 1970206).
Anti-positivism thus reverts the ordinary direction of thought instead of going from theories (and facts) to concepts, it goes from concepts to theories (and problems), because to define a concept is to formulate a problem. Through the succession of theories, one must realize that a problem endures, even within the solutions devised for solving it. Problems must be tracked, identified, properly recast and posed, even where scientists and thinkers themselves were not in a position to do so, or believed they could do without it by simply stating the solutions. To quote Bachelard once more Above all one must kow how to state problems. Whatever one may say, in scientific life, problems do not arise by themselves. It is precisely this sense of problems which is the distinctive mark of a genuine scientific mind. For a scientific mind, every piece of knowledge is an answer to a question. If there is no question, there cannot be any scientific knowledge. Nothing is granted. Nothing is given. Everything is constructed (Bachelard 193814).
And elsewhere 'Scientific research does not need the bravado of universal doubt, but the constitution of a problematic. It takes its departure in a problem, even if this problem is badly stated.' (Bachelard 194951). Needless to say, according to Bachelard, this sense of problems is a requisit for the history of sciences, and a fortiori for any rigorous philosophy of science. These considerations bring us one step further in the understanding of what an 'anti-postivist' epistemological stance should imply. For it seems now that the proper object of an anti-positivist epistemology is not so much facts, theories or even concepts themselves. Rigorously speaking, 'the perpetual revision of contents by deeper investigation and by erasure' is only a symptom of the constant recasting of a problem in various theoretical fields. So the real objects of an anti-positivist epistemology are problems, along with the conditions under which problems are formulated, posed, and sometimes solved. The positivist image of truth is overturned only when one realizes that thought does not primarily strive at knowing what reality really is, but deals with its own problems as immanent, genetic functions.
Bergson's Project of a 'History of Problems'
The difficulty, once again, is that if this is a correct characterization of anti-positivism, then anti-positivism seems to be the rallying point of both Bergsonism and its opponents. This sounds like an odd thing to say only because we have been trained to take for granted the narrative of the history of philosophy designed by some of its actors. As early as 1899, in his evaluation of Bergson's application to the Collège de France, Théodule Ribot was already aware that Bergsonian philosophy offered a rather unconventional perspective on the history of ideas, and of scientific ideas in particular 'Mr. Bergson, if he obtained the chair of modern philosophy, would substitute for the classical method of the history of systems that of the history of problems for example the history of theories of induction, from Bacon to now, or of the idea of matter from Galileo to the modern times.' (quoted in Soulez and Worms 199791-92). To substitute for the history of systems (and of solutions) the history of problems (and of concepts) there is no better definition, indeed, of a non-positivist historical epistemology. Bergson was finally elected at the Collège de France at the turn of the century. The titles of his courses confirm Ribot's remark Sketches for a history of the notion of time in its relation to systems (1902-1903), The evolution of the theories of memory (1903-1904), Study of the evolution of the problem of freedom (1904-1905)....
Two claims remain to be examined first, that Bergson's practice of the history of ideas and problems actually conformed to the anti-positivist principles ; second and more interestingly, that Bergson provided the philosophical underpinning of an anti-positivist view, not only of concepts or knowledge in general, but of problems themselves.
Bergson's History of Philosophy
How does Bergson concretely work towards the aim of a history of problems As far as the history of philosophy is concerned, an interesting example is found in Bergson's lessons, at a time where he was still teaching at the Lycée Henri IV. In a chapter compiled by Henri Hude on the theories of the soul (Bergson 1995247ff.), Bergson draws the outlines of what may be called a problematic introduction to the history of systems. These are notes from Bergson's pupils, which the professor, of course, never intended for publication. But the controversy over the philosophical relevance of these sources need not concern us here, as we are only interested in their value as samples of the general method which governed Bergson's pedagogy. The example is simple enough to be stated roughly. Examining theories of the soul in Descartes, Malebranche, Spinoza and Leibniz, Bergson manages to isolate a structural bipolarity, or rather the entanglement of two different and sometimes opposite tendencies the concern for individuality, and the need for a satisfying conceptual account of causality. The situation is rather unstable, and Bergson identifies the reason for the constant shift from one kind of system to the other. It has to do with the very question of causality. The problem of the conciliation of causality with individuality takes a different form depending on the notion of causality philosophers wish to adopt. In Leibniz for example, 'parallelism' sacrifices freedom and consequently individuality, in order to gain a better, less confused understanding of the communication of substances. The formulation of the problem avoids the obscure ideas of 'influence,' or 'reciprocal action'. While acknowledging this Leibnizian 'tour de force,' Bergson claims that another solution could have been explored, and the mind-body problem posed anew, if instead of relying on a general and a priori notion of causality, one had set out to examine the 'lines of fact' suggested by nature (as he would later say) in order to challenge the assumption that causal relations necessarily occur between terms of the same kind. Bergson's suggestion is that homogeneous causal relations can be (and should be) dropped in favor of 'relations of derivation between heterogeneous terms' (Bergson 1995251). This summary (or anticipation) of Matter and Memory is interesting in that it provides a very striking instance of what it means, in practice, to replace a history of theories with a history of problems. Theories of the soul are indeed considered as so many systems of variation of a deeper, more relevant problem, which needs to be extracted and stated in each specific case the problem of causal relations between substances or realities.
What really matters is exposing the problem that corresponds to a given concept or set of concepts. In an interesting text on the question of how philosophers should write, Bergson explains that if the same philosophical term ('reason,' for instance, or 'soul') sometimes stands for different concepts and solutions, it is probably because it indicates the persistence of the same problem 'indeed, a term such as 'reason' does not so much refer to a thing as to a problem' (Bergson 19976). Thus the history of the concept of the soul, conceived as a history of the theories of the soul or of the mind-body relation, is replaced by the history of the transformations of a typically Seventeenth century problematic configuration binding together the notion of causality and the requirement of individuality. Similar reconstructions of particular problematics can be found in Bergson's later works. They tend to qualify the idea that Bergson's practice of the history of philosophy is condemned to chose between two equally unacceptable alternatives on the one hand a sweeping but schematic account of past philosophies that tends to reduce each of them to a mere reflection of the state of scientific thought, on the other hand an intuitive approach that strives to grasp the simple act from which every philosophy originates, at the risk of losing the sense of how problems actually arise and get solved.
Gueroult admired Bergson's forceful genetic reconstruction of the illusion that made past philosophers fail to recognize the reality of time and change, but he could not be satisfied by an approach that directly aimed at the backbone of systems, instead of exploring the details of their articulations. The result, he believed, was a seductive 'phenomenology of metaphysics' (Gueroult 196019), unfortunately closer to a 'historical novel' (ibid., 22) than a genuine analysis of the evolution of systems. He pointed out that this overly philosophical view of the history of philosophy was in fact balanced by Bergson's simultaneous insistence on the necessity of an intuitive grasp of the absolutely unique and simple generating intuition refracted in every facet a philosopher's system (ibid., 24-26). The first approach is too broad ; the second runs the risk of losing sight of the problematic nature of systems in favour of mere matters of expression. Gueroult argues that the organic elements which constitute the conditions of the stating of a problem are overlooked as Bergson aims directly at the intuitive core of a 'vision'. However, a third possibility can be discerned in the process described in 'L'Effort intellectuel' (ibid., 31) Bergson, according to Gueroult, is at his best when he tries to unpack the meaning of a problematic situation by adjusting the focus, moving back and forth from a given system of concepts to the dynamic scheme that engendered it, recapturing a problem and stating it in a way that suggests the different threads one may follow in an attempt to solve it. Such a process probably meets Bergson's purpose in studying the history of philosophy to get a clearer view of 'the unity of the problem underlying the multiplicity of the solutions.' (Bergson 19977).
History of Science
As far as the history of science is concerned, it is important to remark first that Bergson conceived it as a 'history of systems' that needed to be entirely recast in terms of the philosophical problem of the apprehension of change. The most interesting Bergsonian text in this regard can be found in the fourth chapter of Creative Evolution. Gueroult's criticism of the kind of global, all-encompassing approach to the history of ideas that he thought typical of Bergsonism, would naturally apply here. As a matter of fact, in Gueroult's view, the reason why Bergson is finally drawn to a history of metaphysics that lumps together all philosophical systems in a general classification of historical 'mishaps,' is that he first chose to present the history of philosophy as the mere 'epiphenomenon' (Gueroult 196021) or 'reflection' (ibid., 22) of the history of science. The evolution from Greek metaphysics to classical and Kantian philosophy, on such a view, is only the by-product of the move from Aristotle's physics to Galileo's. Yet, conversely, the evolution from Greek science to modern science is retraced according to a certain number of structural patterns which reveal at once the successive transformations of certain scientific problems, and the persistence of the same philosophical problem, specified in different ways.
Modern science dates from the day when mobility was set up as an independent reality. It dates from the day when Galileo, setting a ball rolling down an inclined plane, firmly resolved to study this movement from top to bottom for itself, in itself, instead of seeking its principle in the concepts of high and low, two immobilities by which Aristotle believed he could adequately explain mobility' (Bergson 199954).
That is the real problem behind the scientific paradigm shift. The alleged difference between a science that applies concepts (as the Greeks did) and a science that seeks laws (constant relations between variable magnitudes), is of secondary importance in this respect.
Now, how did the astronomical problem present itself to Kepler The question was, knowing the respective positions of the planets at a given moment, how to calculate their positions at any other moment. So the same question presented itself, henceforth, for every material system. [...] Let us conclude, then, that our science is not only distinguished from ancient science in this, that it seeks laws, nor even in this, that its laws set forth relations between magnitudes we must add that the magnitude to which we wish to be able to relate all others is time, and that modern science must be defined pre-eminently by its aspiration to take time as an independent variable. But with what time has it to do ' (Bergson 1944364-365).
Thus Bergson displaces a rather conventional philosophical problem (concepts versus laws), substituting for it a new one (time as an independent variable), which he considers more apt to characterize a new range of scientific problems (Kepler's). The same kind of operation is systematically carried out in relation to different issues. It is, obviously, an essentially problematic method.
Bachelard reproached Bergson for his outdated image of science. This criticism was first formulated by the mathematician Emile Borel in 1908 (Castelli 1998278). Whether it is justified or not, it seems to miss the point, for Bergson was not so much interested in the history of sciences for its own sake (the 'evolution' of scientific theories, with their positive contents and methods), as in the historicity of their problems. It is of course typical of Bergson's method that he should bring out the historicity of scientific problems by showing their relevance to certain philosophical problems. Problems nevertheless imply by necessity a particular kind of historical method. Bergson believes it is essential 'to confer some importance on the (often contingent) order in which problems were stated. In this respect, science would be contingent, in part at least, and the history of science would be part and parcel of science itself.' ('Discussion à la Société Française de Philosophie, 18 décembre 1902,' in Bergson 1972568). Thus Bergson repeatedly criticized any teaching of science that would not simultaneously deal with the history of science, and he strongly supported the reinstallation of a program in the history of science at the Collège de France in 1919 (Castelli 1998279). His view of the history of science is remarkable in at least two respects he approaches it as a history of scientific problems (for example, the problems of classical mechanics, or the problems raised and solved by mechanistic conceptions of the living), and in relation to certain philosophical questions (the problem of change, the problem of creation).
An Anti-Positivist View of Problems
Bergson was in effect committed to a history of problems, to an essentially problematizing practice of the history of ideas. Reaching the intuitive core of conceptual constructions did not insiduously imply falling back on the kind of positivist view which the French epistemological tradition so strongly condemned. It is tempting to go one step further and acknowledge that Bergson actually provided what has not been properly stated so far, namely, an anti-positivist notion of problems themselves. For the anti-positivist emphasis on the priority of problems over propositions and solutions (facts and theories) would be of little consequence if it did not transform the very notion of problem. And yet it is a mark of the philosophical limitations of the anti-positivist tradition that it has not devoted much effort to the conceptual clarification of the problem as such. As we shall try to argue, Bergson does precisely this, and in even clearer terms than Bachelard or Canguilhem (3).
Anti-positivism, for its own part, always menaces to degenerate into a positivist view of problems as 'historically given,' or merely deduced from the examination of possible 'moves' in a given situation, i.e. what it is possible to say within the limits of a certain system of propositions and concepts. Structuralism had to struggle with a positivist conception of structures, and it is no accident that what Deleuze targeted in The Logic of Sense ultimately came down to a certain misrepresentation of the nature of problems themselves a structure is indeed nothing but an 'objective' system of problem positions.
Now how does Bergson manage to lay the foundations of an anti-positivist conception of problems In a letter to Floris Delattre, he writes 'I consider an amateur in philosophy, the one who accepts the terms of an ordinary problem as they come, and holds the problem as definitively posed, merely choosing between apparent solutions which necessarily precede [préexistent] preexist his choice. This is how Butler rejects Darwin's solution in favour of Lamarck's. But philosophizing for real should mean at once creating the position of a problem and creating the solution. [...] I consider an amateur the one who chooses between ready-made solutions, as one decides to register as a member of this or that political party. But I consider a philosopher the one who creates the necessarily unique solution of the problem which he has posed anew by the very effort of trying to solve it' ('Lettre à Floris Delattre, 1935' in Bergson 19721528).
These remarks provide the simplest definition of what is implied in a positivist view of problems. Positivism resurfaces whenever the problem is pictured as the mere disjunction of ready-made solutions, on the model of questions and answers. In such a view, problems are considered as given, and yet as truly secondary to their solutions, because they are entirely designed after them, like neutralized doubles of supposedly pre-existent propositions which may or must serve as responses. On the contrary, posing a problem anew implies creating the terms according to which it is to be stated, and finding the solution in the same stroke.
The passage of Difference and Repetition in which Deleuze denounces the process whereby problems are neutralized whenever they are conceived on the model of interrogations, merely rephrases Bergson problems and questions must no longer be traced from the corresponding propositions which serve, or can serve, as responses. We know the agent of this illusion ; it is interrogation which, within the framework of a community, dismembers problems and questions, and reconstitutes them in accordance with the propositions of the common empirical consciousness - in other words, according to the probable truth of a simple doxa (Deleuze 1994157) (4).
The positivist view of problems eventually comes down to the illusion that pictures the problem as a phantomatic double of its solutions. This illusion itself can be explained by reference to our tendency to give too much importance to the activity of problem-solving, at the expense of problem-stating. As Bergson says in the Creative Mind 'The truth is that in philosophy and even elsewhere it is a question of finding the problem and hence of posing it even more than of solving it. For a speculative problem is solved as soon as it is well posed' (Bergson 193451).
Disagreement
Bergson, whether we like it or not, shares a common concern with Bachelard, Koyré, Canguilhem, Foucault, Althusser (and Deleuze, which goes without saying) a concern for what has been called a history of problems. Not only does Bergson's actual conceptual practice confirm this, but his method (or theory of philosophical practice) provides all the underpinnings for a non-positivist view of problems. He can be considered as the true purveyor of the theory of problems implicit in the anti-positivists' epistemological practice.
This fact was not always acknowledged by those who reproached Bergson for his rigid conception of science, or who classified him as a 'philosopher of experience'. But common problematics paradoxically manifest themselves most clearly through this kind of disagreement (Rancière 1999 xii), or differential of interpretations. A common problematic need not be stated as such, it does not require that any common problem be stated. In a truly anti-positivist manner, we may say that it does not need to be stated by both parties in the same way, or even stated at all.
The real discrepancy is of course not between anti-intellectualism and intellectualism, irrationality and rationality, not even experience and knowledge, subject and concept. It runs deeper, which means that it cannot be grasped anywhere else than within philosophical practices themselves - not in the types of 'philosophies' that these rather 'ideological' designations try to capture, but in the working of certain operations and procedures, and in the interpretations that allow them to function, sometimes on condition that they become invisible.
Now where do we stand There is no superior viewpoint. We must install ourselves within the interpretative differential, where contrasting interpretations and procedures play off each other. After opening a pair of pliers to their widest possible angle (Foucault's scenography), we squeeze them back to their pressure point (a common anti-positivist stance). Somewhere along the way, we find that interpretations start diverging it is the point where differences are at once minimal and maximal, the point of contradiction where contradictories oscillate around a center, the smallest angle. It is there that we have a chance to find what separates Bergson from the French epistemological tradition, in their greatest proximity.
An easy way out, as we have seen, could be found in emphasizing that Bergson is not a historian of science, nor an epistemologist, nor even a philosopher of science. He is a metaphysician, and a metaphysician of a specific kind a metaphysician of intuition. He thinks that a direct apprehension of reality is possible when one thinks sub specie durationis, in duration, as Merleau-Ponty reminds us. This, indeed, is distinctive of Bergson - it suffices to set him aside from the 'philosophers of the concept,' those who are interested in the immanent dynamics of concepts, what Cavaillès refers to a 'dialectics.' But it does not quite help us. It is still too general an answer. One must ask what does it mean, as far as problems are concerned, to craft a philosophy of intuition It is not enough to merely recast concepts and systems in order to understand to which problems they answer, or even to realize that scientists and philosophers create unprecedented concepts to pose and solve problems of their own devising. It is a matter of understanding problems themselves in terms of the possibility of this superior form of knowledge which 'intuition' stands for. This is where the debunking of false problems reveals itself as an essential component of Bergson's overall strategy.
False Problems
If we retrace our steps, the trajectory appears clearly
1) the first point consisted in showing that the anti-positivist stance gives priority to concepts over facts ;
2) a deeper move was then identified that of giving priority to problems over concepts and theories ;
3) this reversal of priority was grounded in a renewed conception of problems themselves ;
4) the last step consists in showing that it is essential to any consistent anti-positivist view of problems that one should be able to account for the possibility of pseudo-problems. And this is where Bergson's philosophical strategy parts with the anti-positivist epistemological tradition (to reverse the chronological order of things).
Bergson does not simply insists on the necessity of assessing problems before the solutions that theories and systems provide. He develops, as Gilles Deleuze rightly emphasized, a genuine method of intuition whose first rule can be stated as follows 'Apply the test of true and false to problems themselves. Condemn false problems and reconcile truth and creation at the level of problems.' (Deleuze 199115).
There is an art of inventing problem positions that is assumed, yet never taken up, in the works of the 'philosophers of the concept'. There is a question they cannot really ask, because they lack a transcendental, truly genetic conception of problem conditions, and this question is 'how can this constitutive power which resides in the problem be reconciled with a norm of the true ' As Deleuze goes on 'While it is relatively easy to define the true and the false in relation to solutions whose problems have already been stated, it seems much more difficult to say in what the true and the false consist when applied to the process of stating problems.' (ibid., 16).
It will not do 'to define the truth or falsity of a problem by the possibility or impossibility of its being solved.' (ibid., 17). Insoluble or intractable problems are often held as pseudo-problems. But this, of course, runs contrary to the anti-positivist principle, for it reverts the proper order of things by putting solutions (or their absence) first, thus resorting to a merely extrinsic determination of problems. Yet how are we to reach the 'intrinsic determination of the false in problems' (ibid.) For Bachelard and Canguilhem, for Althusser and even Foucault, there are no genuinely insoluble problems (remember Marx 'Humanity only sets itself problems that it is capable of solving'). But there are indeed no false problems either there are only unstatable problems, the effect of which can be characterized negatively as lacks, lapses, blanks, gaps, blindspots - and positively as symptoms (Althusser 196827).
This picture of things does not offer any grips for a genetic account of false or badly stated problems. At best, general principles are invoked, to the effect that epistemological problems should be 'corrected,' yet always in the negative mode which consists in reaffirming the priority of problems over solutions. Against the ideological question and its 'necessarily closed space,' Althusser stresses the need 'to open up, in some other place, a new space - the space required by a proper position of the problem, which does not presuppose its solution.' (ibid., 63). But to carry out such a plan to the end, up to the transcendental conditions of the problem, is precisely what a purely critical theory cannot achieve. It does not make room for an art of stating problems in their singularity, it does not teach how to debunk or undo specific false problems. It is either too empirical or too abstract. Sooner or later, it is bound to fall back on a positivist view of problems, since the only criteria at hand are solutions themselves. One can only trace back problems, or rather infer their existence on the basis of the concepts that are supposed to express and distort them. There is no method for probing problems themselves, no sounding in the depths of problems, to speak like Bergson. Accordingly, there is no genuine criticism of problems, only a criticism of solutions. There is no truly intrinsic determination of false and true in problems themselves. Invention at the level of problem positions remains something unexplicable. The only authorized moves consist in demystifying, unveiling and laying bare a problem that was never posed. But such a problem, by definition, can never be false, it is at best unstatable (or at least was so). And it can never be properly said to be invented either philosophy's role is merely to acknowledge its presence or absence.
Bergson, by contrast, is not satisfied with uncovering the problem '[...] I mean that its solution exists [as soon as it is well posed], although it may remain hidden and, so to speak, covered up the only thing left to do is to uncover it. But stating the problem is not simply uncovering, it is inventing. Discovery, or uncovering, has to do with what already exists, actually or virtually ; it was therefore certain to happen sooner or later. Invention gives being to what did not exist ; it might never have happened' (Bergson 193453).
The critical stance can only uncover, it is bound to remain uncreative. If false problems are ever debunked, they cannot be explained - why this false problem rather than that one how is it that we have begged the question by projecting the answer in the question why precisely this question, and this answer why now Curiously enough, it is often the alleged partisans of the concept which display the most naive and psychologizing view of problems. Bachelard's 'dialectics,' for example, often sounds like an elaboration of the Brunschvcgian 'activity' of reason, conceived on the model of judgement. The criticism of problems amounts to drawing boundaries, refusing, overthrowing. 'It is fruitless, says Bachelard, to put a false problem at the origin of a real one, it is even absurd to compare alchemy with nuclear physics. [...] Science has nothing to gain from false continuities, where what is at stake is frank dialectics.' (Bachelard 1953104). At best then, false problems boil down to problems that could not be solved when they were formulated, problems that needed to be dropped and replaced by the right ones by the means of an 'epistemological break'.
Spelling out all the implications of the Bergsonian critique of false problems would require a systematic investigation of the interplay of intuition and intelligence in the operations of problem-making dismantling and debunking, stating and posing, answering and solving. Bergson's own philosophical practice provides glimpses of what an inventive art of problem positions, as opposed to the mechanical techniques of problem-solving, would look like. In the case of metaphysical false problems (nothingness, disorder), it is as if intelligence, triggered by intuition, was folded back unto itself and forced to undo its own constructions, thus providing the 'intellectual counterpart of the intellectualist illusion,' to quote from La Pensée et le Mouvant (Bergson 193469 ; During 2000). But only intuition can expose the wrong gesture or false movement that caused the problem to be badly posed in the first place.
Conclusion
Carved out, so to speak, according to its natural articulations, the disagreement between Bergson and the 'philosophers of the concept' has taken a new shape. Only Bergson has managed to extend the anti-positivist view of the positivity or priority of problems to a conception of their intrinsic determination. This move, which Deleuze singled out as the main characteristic of the method of intuition, necessarily involves probing problems according to a norm of the true and the false, for this is really all that is intrinsic about them. It also explains why Bergson cannot be captured in Foucault's net he is neither a philosopher of experience, nor a philosopher of the concept, he is a philosopher of the concept as experience whose taste for more fluid and flexible concepts does not quite match the more familiar ways of 'frank dialectics'.
NOTES
(1) Cavaillès' opposition runs between 'dialectics' and 'activity,' the latter being modelled after a mainstream neo-kantian philosophy.
(2) See 'Bergson se faisant' ('Bergson in the making'), his tribute to Bergson, pronounced in 1959, with its dialectic of 'Bergson' and 'Bergsonism'.
(3) A careful study of Canguilhem's 'Concept and Life' (Canguilhem, 1970335-364) would qualify this statement. This text gives a very useful clue to the understanding of Bergson's own theory of problems. Problems are described as an expression of life itself it is in fact life which poses problems about life, even if man alone is able to raise truly false problems.
(4) Also compare Deleuze 1994158 and Bergson 193451. See During 200167-69.
BIBLIOGRAPHY
Louis Althusser (1968). Lire le Capital, t. I, Paris, Maspéro.
Gaston Bachelard (1938). La Formation de l'esprit scientifique, Paris, Vrin.
Gaston Bachelard (1949). Le Rationalisme appliqué, Paris, PUF.
Gaston Bachelard (1953). Le Matérialisme rationnel, Paris, PUF.
Henri Bergson (1934). 'L'Intuition philosophique,' in La Pensée et le Mouvant, Paris, PUF.
Henri Bergson (1944). Creative Evolution, trans. A.Mitchell, New York, The Modern Library.
Henri Bergson (1946). The Creative Mind, trans. M.Andison, Westport CT, Greenwood Press.
Henri Bergson (1972). Mélanges, Paris, PUF.
Henri Bergson (1995). Cours, vol. III, Henri Hude éd., Paris, PUF.
Henri Bergson (1997). 'Comment doivent écrire les philosophes ,' Letter to Constant Bourquin, 1924, in Philosophie, nº54, 1997.
Henri Bergson (1999). An Introduction to Metaphysics, trans. T.E.Hulme, Indianapolis, Hackett Publishing Company.
Georges Canguilhem (1970). Etudes d'histoire et de philosophie des sciences, Paris, Vrin.
Georges Canguilhem (1988). Ideology and Rationality in the History of the Life Sciences, Cambridge, MA, MIT Press.
Georges Canguilhem (1994). A Vital Rationalist, Selected Writings from Georges Canguilhem, F.Delaporte ed., trans. A.Goldhammer, New York, Zone Books.
Milic Capek (1959). 'La théorie biologique de la connaissance chez Bergson et sa signification actuelle,' in Revue de Métaphysique et de Morale, April 1959, pp.194-211.
Marie Cariou (1995). Bachelard et Bergson, Paris, PUF.
Enrico Castelli Gattinara (1998). Les Inquiétudes de la raison épistémologie et histoire en France dans l'entre-deux-guerres, Paris, Vrin.
Jean Cavaillès (1960). Sur la logique et la théorie de la science, Paris, PUF.
Gilles Deleuze (1973). 'A quoi reconnaît-on le structuralisme ,' in François Châtelet éd., La Philosophie au XXème siècle, t. IV, Paris, Hachette.
Gilles Deleuze (1991). Bergsonism, trans. H.Tomlinson and B.Habberjam, New York, Zone Books.
Gilles Deleuze (1994). DifferenceandRepetition, trans.P.Patton, ColumbiaUniversityPress, New York.
Elie During (2000). 'Fantômes de problèmes,' in 'Dossier Bergson,' in Magazine Littéraire, nº386, april2000, p.39-42.
Elie During (2001). 'Deleuze and Nietzsche On Frivolous Propositions and Related Matters,' in Pli. The Warwick Journal of Philosophy, vol.11, 2001, p.62-78.
Paul Douglass (1992). 'Deleuze's Bergson Bergson redux,' in F.Burwick, P.Douglass eds., The Crisis in Modernism. Bergson and the Vitalist Controversy, Cambridge, Cambridge University Press.
Michel Foucault (1994-1). Dits et Ecrits, t.III, Paris, Gallimard.
Michel Foucault (1994-2). Dits et Ecrits, t.IV, Paris, Gallimard.
Michel Foucault (1998). Aesthetics, Method, and Epistemology, James D.Faubion ed., New York, The New Press.
Henri Gouhier (1949). 'Bergson et l'histoire des idées,' in Revue Internationale de Philosophie, October 1949, nº10.
Martial Gueroult (1960). 'Bergson en face des philosophes,' in Les Etudes Bergsoniennes, vol.VI, Paris, PUF.
Pete Y.A.Gunter ed. (1969). Bergson and the Evolution of Physics, Knoxville, The University of Tennessee Press.
P.A.Y. Gunter (1987). 'The Dialectic of Intuition and Intellect Fruitfulness as a Criterion,' in P.A.Y. Gunter and A.C.Papanicolaou eds., Bergson and Modern Thought Toward a Unified Science, London, Harwood Academic Publishers.
Martin Jay (1993). Downcast Eyes, The Denigration of Vision in XXth c. French Thought, Berkeley, University of California Press.
Dominique Lecourt (1972). Pour une critique de l'épistémologie, Paris, Maspéro.
Dominique Lecourt (1975). Marxism and Epistemology Bachelard, Canguilhem, Foucault, London, Humanity Press.
Pierre Macherey (1989), Comte La philosophie et les sciences, Paris, PUF.
