Accueil
https://fr.wikipedia.org/wiki/Dominique_Lecourt
"The peculiar history of epistemology in France: an original 'tradition' has arisen there, famous for the few names you will find in this book: Gaston Bachelard, Georges Canguilhem, Michel Foucault. Doubtless it would be hard to imagine works more dissimilar in their respective objects, aims and echoes. If it is indeed possible to compare the work of Foucault with that of Canguilhem because of their common interest in the history of the biological sciences, it has to be admitted that, for his part, Bachelard, who was exclusively attentive to the physics and chemistry of his time, says not a word about them. It should also be added that Canguilhem's strict specialization is opposed to the 'encyclopredism' of Foucault, who talks about linguistics, political economy, etc., just as much as he does about natural history and biology. On the other hand, if Canguilhem in a certain sense takes up the Bachelardian project of 'giving science the philosophy it deserves', i.e., of struggling, in his special domain, against the encroachments of idealistic philosophies of knowledge, it must be admitted that although this polemic is no less constant in Foucault's work, it has there undergone a remarkable displacement, a veritable 'decentring' with respect to the history of the sciences, to bring it to bear on the most general conditions of 'knowledge' (savoir). Lastly, whereas the interest of Bachelardian epistemology was recognized and its originality respected from the very first, the real importance of Canguilhem's works has only been grasped in the last few years, after twenty years of more or less deliberate neglect. As for those of Foucault, they have been the object of a strange misunderstanding, since they have owed a part oft heir rapid celebrity to the echo they aroused in a camp to which it is clear today that they did not belong, that of 'structurahsm'.
For all these reasons, it would be inaccurate to speak of these writers as belonging to an epistemological 'school'. Cangnilhem, after all, is a historian of the sciences, and Foucault would correctly refuse to be described as an 'epistemologist': he calls himself an 'archreologist of knowledge' (savoir). Under these conditions, is it possible to speak in their case, as I propose to do, of an 'epistemological tradition' ? I find this justifiable because of the existence of a common feature which is transmitted to each of these oeuvres, despite their apparent diversity. It would be superficial to seek to explain it by the supposed intellectual 'influeuces' which masters may have exercised on disciples. It is true, of course, that Bachelard was Cangnilhem's 'teacher', and Cauguilhem Foucault's, but we cannot appeal to psychology as a substitute for historical analysis without immediately renouncing any understanding but an anecdotal one. The common feature I have mentioned is more real and more profound, it constitutes their point of agreement and pertains to their common 'position' in philosophy.
To put it briefly, and provisioually formulating it in a negative form, it lies in their radical and deliberate 'non-positivism'. This non-positivism, inaugurated by Bachelard, while it seems to me to form the 'cement' of the tradition uniting my three authors, simultaneously distinguishes this tradition from everything practised elsewhere under the name of 'epistemology'. It counterposes it to another tradition, one which does have its 'schools' and its 'institutions' a tradition dominant today and in which converge investigation conducted in the East and the West, at the University of Yale and in the Moscow Academy of Sciences.
I say that this other tradition, despite the internal distinctions that have to be made in it, is massively 'positivist' in that it always, in one way or another, presents itself as an attempt to elaborate a 'science of science' or -the technocratic variant- a 'science of the organization ofscieutific work'. This is the case with the Auglo-Saxon specialists Bernal and Price who, more than twenty years ago, announced under this title o.f a 'science of science' the emergence of a new discipline and estabhished under this sign an original institution. The insistent references made to it today by the editors of the philosophical journal of the Moscow Academy of Sciences (Voprosy Filosojii) -Kedrov, Mikulinksy and Rodnyi- in order to take up the same project in the same terms, sufficiently prove that they range themselves in the same camp.' Although in a quite different form, this is also the case with the 'logical neo-positivist' current which attempts, on the basis of the concepts of the scientific discipline of mathematical logic, to form the categories of what Reichenbach called 'a scientific philoophy, the philosophy of our time, that of the age of science', which is to be both a 'science of science' and a scientific critique of philosophy.
Whether epistemology is made into a kind of 'cross-roads' at which a cohort of heteroclite disciplines with scientific pretensions come to conjoin their disparate concepts in order to constitute a general theory of science, or whether a determinate science is made responsible for the provision of its categories, the philosophical presupposition of the undertaking is the same, and it is this presupposition that leads me to describe these attempts as 'positivist'. Truly, this common presupposition could find no better expression than the slogan: 'A science of science is possible.'
The claim I have just ventored may cause some surprise: that "the science of science' is mortgaged by a philosophical presupposition. A paradox, since, precisely, both groups claim to have put an end to the 'philosophical usurpation', restoring to science its most legitimate property and its most precious birthright: itself. However, I continue to state that philosophy is not absent from these efforts. Let me add, to complete the paradox, that two philosophical presuppositions are conjoined in it. It is their very conjunction which constitutes the positivism that I am denouncing. One pertains to the unity claimed for the duplicated singular term 'science' ; the other relates to the circle of this duplication within the reflexiveness it attributes to the term with itself: 'science of science'.
The first point is well known today: to speak of science in general as one single entity which might itself take itself as an object is to make use -here: double use- of an ideological notion. It is to suppose that the ensemble of scientific practices can be treated as a homogeneous reality, constituting, at least in principle, the unity of an undifferentiated whole. This 'treatment' or this 'point of view' on science is peculiarly philosophical. It even repeats, beneath a modernist and scientistic exterior, one of philosophy's classical procedures. Better, the classical procedure of idealist philosophy which when it speaks of the sciences, is eager for one thing only: to disengage their common 'essence' so as to be able to speak of 'science' in the singular -and then to take the theses produced thereby as a justification for the elaboration of a theory of knolvledge. Re-read the history of philosophy: it is easy to establish with what remarkable regularity this procedure has been applied, from Plato to Husserl.
Hence the philosophical presupposition whose presence I therefore detect in the positivist epistemologistsis an idealist philosophical presupposition. As such it simultaneously conceals and reveals, inthe symptomatic mode, a reality which we have momentarily glimpsed: the ensemble of scientific practices. More accurately: by attributing to this ensemble the unity of a whole, the presupposition 'resorbs' -imaginarily annihilates- the reality of these practices, which resides in their distinctness -each having its own object, its own theory and its specific experimental protocols- and in their uneven development -each having its particular history. I say that it is the very reality of these practices which is thereby masked, for they do not exist outside the system that they constitute. Now this system, far from unfolding beneath the appearance of calm identity which has been foisted onto it, itself only has reality through the different contradictions entertained between the theoretical disciplines which feature in it. It is the intertwining of its contradictions which gives a form to its history. Thus we now know what is 'concealed' in the last analysis by the idealist philosophical presupposition here in question: it is the actual history of the sciences.
To say that a science of science is possible is, besides to claim that "science" can unveil, by mere reflection on itself, the laws of its constitution, and thus of its functioning as of its formation. It is to claim of the "scientific discourse" that it has the intrinsic -and exceptional- virtue of being able, by itself, without going outside itself, to state the principles of its own theory. In other words, the 'scientific discourse', sovereignly autonomons, is accountable to no one and constructs itself, without let or hindrance, in the pure space of scientificity instituted, laid out and delimited by itself. Without let or hindrance, since every obstacle is in principle always already located, stated and surmounted in the implicit discourse it is constantly conducting with itself -a quiet murmur in its inmost soul that in case of trouble need only be made explicit to illuminat everything. An immediate and decisive consequence follows: if it is the laws of scientific discourse itself which determine the space of its own deployment, if in principle it encounters no let or hindrance in that space, then the completion of knowledge (savoir) -its end and its perfection- is always in principle possible. All that remains in fact is to level out the few, quite formal difficulties that momentarily block It. A question of techmque. Let us translate: there is no real history of the sciences ; time has nothing to do with it. Or rather: time can only intervene in the form of delay or anticipation. The history of 'science' is merely a development, at best: an evolution which guides knowledge from error to truth ; in which all truths are measured against the latest to appear.
Now, once again, we can state that a very old philosophical operation is being repeated in a new form. Did not what are customarily called the 'Great Philosophies' (an appellation à controler) have as their project -and avowed claim- to state the criteria for all real or possible scientificity ? Did they not put themselves, with respect to the existing sciences, in the position here declared of constitutive reflection ? As the point is of some importance, I shall resort to a typical illustration: the philosophy of Hegel. In the discourse it conducts on itself, this philosophy, as is well known, takes as its 'foundation' i.e. as its basis and guarantee, the categories which are stated in that strange book entitled The Science of Logic. The peculiar status of this book in the Hegelian system deserves a special study for itself alone. For our purposes, we shall only consider one small fact: Hegel declares that to it are consigned the categories of the scientificity of science. Witness the following passage in which the author situates the book with respect to the Phenomenology of Mind:
In The Phenomenology ofMind I have set forth the movement of consciousness, from the first crude opposition between Itself and the object, up to absolute knowledge. This process goes through all the forms of the relation of thought to its object, and reaches the Concept of Science as its result. Thus this concept (apart from the fact that it arises within the boundaries of Logic) needs here no justification, having already received its justification in that place.
For us this text has a dual interest. It reveals the position of the content of logic with respect to the existing sciences -those that Hegel calls: 'empirical sciences' ; it is logic that presents the concept of science ; or, better, it is Hegel's book that contains the justfication, the "foundation" for the concept of science. In short, Hegel's Science of Logic is the philosophical science, the true science, the science of the sciences.
Hegel's philosophy presents itself as 'the philosophy of all philosophies'. Doubtless it would not be illegitimate to.take its declarations at their word and see in it in fact, in the position it attributes to itself wish respect to the system of the existing sciences, the truth, having attained self-consciousness, of what previous philosophies were tacitly practising. It would then be seen that the key stone of this philosophy is a 'science of science'. One step further: It is clear that, according to Hegel's own statements, even the writing of this book -The Science of Logic- presupposed that The Phenomenology of Mind was complete, and that therefore history itself was complete In these conditions it is obvious that the project of a 'science of science' is merely the repetition in a new form of the same peculiarly philosophical operation. An operation whose effect fits in well with the one we have registered vis-a-vis the constitution of the unitary ideological notion of 'science': it annihilates the actual reality of the history of the sciences by placing it beneath the sign of a teleology. Now, abstracting from the specific structure taken by this teleology in the Hegelian dialectic, we can say -to return to the terms used at the beginning of this analysis- that the regular effect of the project of a 'science of science' is to reduce the actual history of the sciences to a kind of evolution. In other words: positivism and evolutionism go hand in hand. Or alternatively: evolutionism is the obligatory complement in the history of the sciences to positivism in epistemology.
In these conditions it is hardly surprising that the non-positivism of the epistemological tradition I am discussing begins with and holds to a deliberate rejection of an 'evolutionism'. I shall say that its non-positivism leans on an anti-evolutionism. Today everyone knows the prime term for this rejection in Bachelard's work, the notion of "epistemological 'rupture", is an injunction against every philosophy of history which would like to bring within its jurisdiction the epistemological categories he works on. Nor is it unknown that Georges Canguilhem, proposing to distinguish between the 'beginnings' of a science and its 'origins', relentlessly denouncing every attempt to seek more or less remote 'precursors' for a 'discovery', shares the same concern. But it is surely Michel Foucault who, analysing the notion of 'discontinuity' at the beginning of The Archelology of Knowledge, has best demonstrated the theoretical implications, decisive for the conception that can be held of history, of the anti-evolutionism that all three profess.
I can now give the positive content of the differentia specifica of this epistemological tradition, so far stated in a negative and polemical form. Their non-positivism and their anti-evolutionism pertain to the link they recognize between epistemology and the actual practice of the history of the sciences. A link, or, rather, a unity, whose theory, as we shall see, each has attempted to outline in his own way. However, these attempts seem to me to have failed up to the present. An inevitable failure, no doubt, for this theory belongs in principle to a discipline which they do not recognize: 'historical materialism', the Marxist science of history. For all that, all epistemological problems undergo a revolutionary displacement from the mere fact of their practice of this unity ; and on the other hand, their failure, insofar as it occurs on the basis of a correct practice, is infinitely valuable to us, since it designates to us, caught in its displacement, the site where we have to re-work.
It remains to understand why it was thus possible for this anti-evolutionist non-positivism to arise and be handed down in French philosophy alongside and against other currents more in accord with the dominant -spiritualist and positivist- tendencies of that philosophy. This analysis belongs to a history of philosophy which still has to be written. I can only say that in the last analysis this history would be the history of the unity I have been discussing: a history of its formation, of its inconsistencies, of its failures, of its struggles and of its successes. To this day I only have at my disposal one factual suggestion on this subject ; here it takes the form of a paradox requiring elucidation: it seems that it was really Auguste Comte and his disciples who made this unity possible and inscribed it in French academic institutions, making the history of the sciences a discipline belonging with philosophy. In fact, France since then has been one of the only countries in the world -if not the only one- in which the history of the sciences is practised and taught in philosophy faculties. Thus it is to the founders of positivist philosophy that we should impute the beginnings of a tradition whose major trait is, as we have seen, its non-positivism..." (pp.120-126)
" [From Bachelard to historical materialism]
"Gaston Bachelard was not a Marxist, nor even a materialist ; as a reminder, the passage from Applied Rationalism will be cited where materialism is presented as a 'flat', 'abstract' and 'crude' philosophy; it will be recalled that "discursive" as it may have been, Bachelardian epistemology explicltly wlsed to be idealist ; it wil! be Fointed out, lastly, that the most obvious point of agreement in this apparently contradictory -epistemologlcal and 'poetic' - œuvre is precisely a dynamistic conceptin of thought, which is fundamentally very 'psychologistic'." (p.129)
"From the 'realism' of Meyerson to Sartre's existentialism, no philosophy escaped his sarcasms. This perpetual polemic, far from arising from a personal psychological disposition, has a precise and profound theoretical meaning. It has to be taken seriously. Its principle can be found outside philosophy in the upheavals experienced in the actual history of the sciences at the beginning of tins century: the development of non-Euclidean geometries, the theory of Relativity, the beginnings of Micro-physics, etc. The new disciplines then inaugurated seemed 'unprecedented' to Bachelard ; his first works are but the reflection of this radical novelty: they take note of a rupture in the actual history of the sciences." (p.130)
"[Dans Le nouvel esprit scientifique, il] proposes a novel philosophical category - whose function at this stage is still descriptive: the category of 'No' or 'Non-'. This category was formed by a controlled extension of the negation which had served to think the disconcerting novelty of the non-Euclidean geometries: in the same way, one can speak of 'non-Newtonian' mechanics, 'non-Lavoisierian' chemistry, etc." (p.131)
"The discontinuity he registers in The New Scientific Mind is opposed to the principle that governed the then dominant philosophy of science, that of Emile Meyerson: The Inductive Value of Relativity (1929) is counterposed to La Deduction relativiste (Relativist Deduction), in which Meyerson, with some courage and considerable ingenuity, had undertaken to 'deduce" Einstein from Newton by showing that, at least in its principles, the theory of Relativity was already present 'in germ' in the Principia of 1687. Meyerson, convinced of the identity of the human mind in all its manifestations, held that the discontinuous was an illusory appearance attributable to the philosophical ignorance of scientists. Hence one must state that Bachelardian epistemology, 'starting not from a philosophical 'principle' but from a summons addressed to philosophy by the contemporary re-organization ofthe exact sciences, ran, in the very recognition ofits object, into a set of philosophical theses which were obstacles in its way. This collision was first thought by Bachelard as a delay (retard) of philosophy with respect to science, a delay which was still awaiting its concept." (p.131)
"But the 'encounter' with Meyerson's philosophy had another, still more. decisive consequence: in it Bachelard discovered the solidarity between a thesis concerning the theory of knowledge: 'realism', and a thesis concerning the history of the sciences 'continuism'. In it the historical continuity of learning (savoir) is supported by the homogeneity of the forms of knowledge (connaissance) - ordinary and scientific. An attack on the former could not but ruin the latter: such is the double, historical and epistemological, meaning of the now famous thesis ofthe 'rupture' as it functions in Bachelard's works. It is clear in what sense this epistemology has to to be called 'historical' ; it was so from the very first ; from the
moment it took as its task to draw the philosophical implications of the 'no' which the new scientific mind addressed to previous science. A task which received its formulation and the beginnings of a realization six years later in The Philosophy of No: the question, it says there, is 'to give science the philosophy it deserves'.
Bachelard was to carry out this undertaking during the next ten years: from The Philosophy of No (1940) to Applied Rationalism (1949). In doing so he steadily acquired the conviction that this philosophy' could only be elaborated marginally to all the existing 'theories of knowledge'. Not only 'marginally to' but also against them. What he had defined as the 'delay' of philosophy now seemed to him to be a systematic 'displacement' of every philosophical theory of knowledge with respect to the scientists' actual work. That is why
Applied Rationalism, which can be said to realize, but also to rectify in many points, the programme of The Philosophy of No, opens with several crucial pages in which Bachelard sets out what I propose to call a 'topology' of philosophy. This topology takes the form of a 'spectrum' in which all the types of theory of knowledge appear dispersed around the reality of the work of production of scientific concepts, in which this topology proves to be a 'typology'. Allow me to reproduce this spectrum (on p. 133) in order to analyse it. This spectrum has two very remarkable characteristics: (a) every philosophy -to the extent that it contains as a principal component a theory of knowledge- is defined in it by its place - its specific 'displacement' - with respect to scientific knowledge ; (b) scientific knowledge has the role of an axis, that is to say that by a mere 'fold' around this axis it is possible to make the various typical forms of philosophy coincide. Three conclusions can be drawn from these two characteristics: ,(a) what is given as the content of philosophy is none other than tlte hallucinatory materialization of its peculiar dispersion (teart) from science ; (b) the always possible coincidence of symriletrical and opposed typical forms of philosophy is the index of a fundamental identity: each form is merely the 'inverted' but identical form of an opposite form ; (c) the essence of philosophy can only be determined from the point of view of the axis, i.e., from a non-philosophical point of view.
Idealism
+
Conventionalism
+
Formalism
+
(philosophy ofthe production ofscientific knowledges:
applied rationalism and technical materialism)
+
Positivism
+
Empiricism
+
Realism
This 'identity in inversion' that I have just emphasized is analysed at length in Bachelard's last works. He there remarks that it is supported by a certain number of philosophical couples, each term of which can by turns be dominant or dominated. These categorial couples form the apparent 'content' of the philosophical theories of knowledge. Stated as such in Applied Rationalism, they are as follows: subject-object, abstract-concrete, given-construct, etc.; they are interchangeable and complementary. They all have as correlate and cement the philosophical category of truth which presents itself as the concept of their accord and closes the space of philosophy. Now this last category, implying that the completion of the process of knowledge is always possible in principle or realized, makes inconceivable the actual history of scientific knowledge: its ruptures, its reorganizations, its failures, its contradictions, its risks.... That is why an epistemology which is historical will pay more attention to error, tofailure, to hesitations than to truth: its space will thus be open, and non-systematic." (pp.131-133)
"The philosophy ofthe scientists is contradictory: to use an expression of Bachelard's, it mixes together a 'diurnal philosophy', the clear philosophy of the science, and a 'nocturnal philosophy', the philosophy of the philosophers to which the scientists inevitably turn when they reflect on their own practice. I shall say that the scientist maintains an imaginary relationship to his own practice and that the philosophy of the philosophers plays its part in the constitution of this relationship." (p.134)
"The first concept to be constructed, the one which sustains the whole edifice, is the concept of 'epistemological obstacle'; it designates the effects on the scientist's practice of the imaginary relationship he maintains with this practice. This concept is still famous thanks to the many and often terrifying illustrations Bachelard gave of it on The Formation of the Scientific Mind (1938). It must be said that, preoccupied with what was most urgent, he concerned himself less with its mechanism than with its effects. Or rather with its unique effect ; for, although 'polymorphous', the obstacle functions in a single direction: going back on the 'no', it closes the rupture between ordinary knowledge and scientific knowledge and reestablishes the continuity threatened by the advance of scientific knowledge.
It may arise at the moment of the constitution of the knowledge, in the form of a 'counter-thought', or at a later phase in its development as a 'suspension of thought' (arrêt de pensée). Whatever the case, the obstacle reveals 'a resistance of thought to thought'. More precisely: assuming that scientific thought only advances by its own reorganizations, it can be said that the epistemological obstacle appears whenever -but only then- an existing organization of thought -whether it is already scientific or not- is in danger. Let me add that it appears at the point at which the rupture threatens - which other efforts than those of Bachelard have been able to show to be the site of an 'overdetermination', of an accumulation of contradictions. Localized in its appearance, the obstacle is solidary with a determinate structure of thought which will later appear, by recurrence, as a 'tissue of tenacious errors'. The status Bachelard attributes to the 'philosophy of the philosophers' is now clearer: it is the vehicle and support of the epistemological obstacles, since it is
what structures the relationship of the scientist to his scientific practice. It registers scientific ruptures and reorganizations by inversions in its spectrum ; and, in the closed field of its basic categories, it thereby assimilates the advances of scientific knowledge and lives by their exploitation. It is easy to observe its action, to describe it, even to foresee it: Bachelard, for provocative reasons -in order to shake the philosophers out of their 'torpor'- more than once tried this out. But one question about the epistemological obstacles remains unanswered: once their effects and the mechanism of their intervention in scientific practices have been described, that of their formation remains to be thought. Rather: what necessity is there in the fact that epistemological obstacles are always being formed and reformed ? We shall find that this question reveals the limits of Bachelardian epistemology.
First we must examine the positive categories released by the recognition of the obstacles as such and by the respect for ruptures and re-organizations. We can say that all these categories are ordered according to an unprecedented conception of the dialectic. In Bachelard this category designates the reality of scientific work: the process of mutual adjustment of theory and experiment. Now, given the rejection of every fixed point -through an initial and polemical rejection of the subject-object couplet- this adjustment
has to be thought not as a formal adequation but as a historical process. In a history which implies no certainty, no fate destines theory always to find the means of its realization. The risk is that at a given moment the language of the physicist and that of the mathematician for example, should become contradictory. Philosophy will hasten to see in this a 'crisis' of science. For the mathematicians and physicists, it will be the opportunity for some work: the former will revise their theories, formulate other hypotheses, the latter will
refine their experiments, check their instruments. Neither the former nor the latter will hesitate to reconstruct the edifice of their science 'from top to bottom' if need be. Georges Canguilhem has shown that this category of the dialectic cannot be assimilated to any classical conception of the dialectic: this is readily understandable once the meaning and implications of Bachelard's anti-philosophical polemic have been grasped. The dialectic that emerges is the 'spontaneous' dialectic of scientific practice: it affirms -against philosophical scepticism- the existence of the objects of science and proceeds by a reorganization of learning (savoir) 'from top to bottom'." (pp.135-136)
"Bachelard's polemic against the philosophical theories of knowledge has one last effect: it demands a rectification of the philosophical category of experience by correct appreciation of the function of instruments in the production of scientific concepts. Everyone knows Bachelard's thesis that scientific instruments are 'materialized theories' ; but it is essential to add the following counter-point: the 'materialization' of theory is not for him an acessory phenomenon. On the contrary, he ceaselessly stressed that it is one of the most characteristic features of the contemporary sciences that they are 'artificialist', that they contain as one of their essential components a technique for the production of phenomena ;
what, in parody fashion, he has called a 'phenomeno-technics', and which is the object ofthat new discipline programmatically adopted by Applied Rationalism: 'technical materialism'.
The notion of phenomeno-technics has to be taken seriously: it enables us to understand in what sense the word 'production' is to -be understood : not only the 'theoretical' production of concepts, but indissociably the material production of the object of theoretical labour; of what can no longer be called its 'data' or 'givens' (donnees), but rather its 'material'. The philosophy of the philosophers, when it is confronted with this material intervention of instruments in the production of concepts can only think it as an inessential 'mediation', and leave it to fall within the general and vague category of 'experimental method', a specification 'for the use of the scientist' of the philosophical category of experience. According to Bachelard, this general category, ignoring the specificity of the cases in which instruments function, gives no real knowledge of what takes place in scientific practice. It is unable to explain the new fact that a concept must from now on integrate into its constitution as a concept the technical condition of its realization. Ultimately this is because the notion of 'method' and the philosophical category of experience are solidary with the conceptual couplet 'abstract-concrete'. Now, the practice of phenomeno-technics consists precisely of proceeding- to couplings (couplages) between the abstract and the concrete by the expedient of the setting up of theoretically defined instruments and systems of apparatus according to programmes of rational realization. Thus the 'objects' of science, far from being poor abstractions drawn from the wealth of the concrete, are the theoretically normed and materially ordered products of a labour which endows them with all the wealth of determinations of the concept and with all the sensitivity of experimental specifications. It can thus be said, following Bachelard, that these objects are 'concrete-abstracts'.
The dual status of scientific instruments has the additional interest of revealing -materially- the 'eminently social' character of contemporary science. Bachelard's last epistemological work, on chemistry, Rational Materialism (1953), is particularly instructive in this regard. In it he systematically opposes what the chemist calls a 'substance' to what the philosopher means by substance. Among the essential characteristics of a chemical substance is 'purity'. Here is what Bachelard writes about it:
It can be said that there is no purity without purification. And nothing better proves the eminently social character of contemporary science than the techniques of purification. Indeed, purification processes can only be developed by the utilization of a whole set of reagents whose purity has obtained a kind ofsocial guarantee. Elsewhere, as is well known, Bachelard could write that only 'society can send electricity along a wire'. In short, he reveals rhe need, in order to think the history of a science, to take into account the state
of the 'technological city' (cite technicienne). But that state is quite clearly linked to the history of the techniques of production ; i.e., to the City as such. What type of determination is there between the history of production techniques and that of scientific instruments, between that scientific instruments and that of concepts ? These questions remain unanswered." (pp.137-139)
"In the case of the status of "scientific instruments", what is missing from Bachelard's analysis is a concept that would enable him to think together several histories with different statuses ; in short, the concept of a differential history. When it is a matter of the formation of epistemological obstacles, the same lack is felt, but the consequences, as we shall see, are much more serious. Once the rupture between ordinary knowledge and scientific knowledge has been registered, it is indeed a matter of determining the constitution of that 'tissue of tenacious errors' with which science 'breaks'. At the same time, this is also to account for the necessarily imaginary nature of the relationship the scientist maintains with his own practice.
Bachelard attempted this in two directions, unevenly explored in his work. The first, evoked above all in The Formation of the Scientific Mind, consists of casting doubt on scientific education which, both in its lessons and in its 'practical work', displaces the true interests of science behind the mask of pedagogy: the constant appeal to the images and experiences of everyday life made in physics courses seems to him particularly damaging. The philosophy class is also inscriminated, but for a different although complementary reason: it diffuses an undeserved valorization of "general culture" to the disadvantage of the scientific specializations, which are, however, the only motor of the advance of the sciences. Thus, by means of educational institutions, the conditions of the reproduction of scientific knowledges have their effects on the forms of their production.
The second direction is intended to be more profound, and it is more minutely analysed by Bachelard ; it constitutes the object of the new discipline he believed he had founded: the 'psychoanalysis of objective knowledge'. It must be candidly stated: all that enables Bachelard to think the necessity of the 'epistemological obstacles' is a certain conception of the human soul which roots the 'imaginary relationship' in the imaginary of the images produced by the imagination. The pages -and they are numerous- in which Bachelard describes this inhibitory intervention of images in scientific practice are still famous. But their crucial theoretical status must be understood: they attempt to elaborate a system of concepts which will make it possible to think the intrication of two histories: that of the scientific and that of the non-scientific in the practice of the scientists. Hence it is not surprising that this historical epistemology culminates, in The Rationalist Affectivity of Contemporary Physics[/i (1951), with the project of an epistemological history which is presented as a dual history: a 'ratified history' (or history of the scientific in scientific practice) and a 'lapsed history' (or history of the interventions of the non-scientific in scientific practice).
But if we turn to the realizations of this double history, we find that the only example of a ratified history that he elaborated -in Applied Rationalism- he elaborated vis-a-vis the history of mathematics, whose very special character was stressed by Bachelard himself, following Cavaillès, for mathematics immediately manifests the existence of what he calls a 'logical time', a continuous identity, which is not found as such in the other sciences. As for lapsed history, illustrations of it will be sought in vain: it is thought very precisely as a non-history: as a 'museum of horrors' or an 'unformed magma'. Given this, the conjunction of the two histories and their mutual determination remain obscure.
His resort to the libido of the scientists to explain the constitution of epistmological obstacles here at once reveals all its meaning: it compensates for Bachelard's inability to think the differential history of what I shall call sciences and ideologies. By that very fact, all Baclrelard's epistemological concepts are haunted by psychologism: in the term 'scientific mind' it is the term mind that tends to become dominant ; the notion of scientific labour and, correspondingly, philosophical laziness, take on a subjective connotation ; the application of 'applied rationalism' threatens to capsize into psycho-pedagogy ; obstacles, finally, can be understood as mere difficulties, whereas the scientific city is conceived as an 'inter-subjectivity'. It has to be admitted that many pages of Bachelard are open to such an interpretation.
But if the notion of an 'imaginary relationship' is referred, not to a psychology of the imaginary, but to the scientific concept of ideology as it features in 'historical materialism', in the science of history, and which designates precisely the 'imaginary relationship of men to their material conditions of existence', it is clear what is designated, but not thought, by Bachelard: the necessity, in order to construct the concept of a history of the sciences, to refer it to a theory of ideologies and of their history.
Such a reading really does allow us to assign limits to Bachelard's epistemology since it reveals the 'psychologism' that sustains his 'poetics' as a point of retreat, better as a 'return of philosophy', if it is true that, unable to think the relation scientific non-scientific as a differential history, Bachelard based it on the repetitive permanence of great themes, myths or complexes in an eternal unconscious. In short, paradoxically, he turns to the analogue of a theory of knowledge. Hence it has to be said that Gaston Bachelard's historical epistemology remains a non-philosophy in philosophy. Nevertheless, by its respect for the spontaneous dialectical materialism of scientific practice, it provides us with precious elements for a theory of philosophy and of its history, and on condition that we know how to read it, precisely where it is inconsistent it reveals the ways to supplement it so as to construct a materialist theory of the history of the sciences." (pp.139-141)
-Dominique Lecourt, Marxism and Epistemology. Bachelard, Canguilhem and Foucault, NLB, 1975, 223 pages.
"The peculiar history of epistemology in France: an original 'tradition' has arisen there, famous for the few names you will find in this book: Gaston Bachelard, Georges Canguilhem, Michel Foucault. Doubtless it would be hard to imagine works more dissimilar in their respective objects, aims and echoes. If it is indeed possible to compare the work of Foucault with that of Canguilhem because of their common interest in the history of the biological sciences, it has to be admitted that, for his part, Bachelard, who was exclusively attentive to the physics and chemistry of his time, says not a word about them. It should also be added that Canguilhem's strict specialization is opposed to the 'encyclopredism' of Foucault, who talks about linguistics, political economy, etc., just as much as he does about natural history and biology. On the other hand, if Canguilhem in a certain sense takes up the Bachelardian project of 'giving science the philosophy it deserves', i.e., of struggling, in his special domain, against the encroachments of idealistic philosophies of knowledge, it must be admitted that although this polemic is no less constant in Foucault's work, it has there undergone a remarkable displacement, a veritable 'decentring' with respect to the history of the sciences, to bring it to bear on the most general conditions of 'knowledge' (savoir). Lastly, whereas the interest of Bachelardian epistemology was recognized and its originality respected from the very first, the real importance of Canguilhem's works has only been grasped in the last few years, after twenty years of more or less deliberate neglect. As for those of Foucault, they have been the object of a strange misunderstanding, since they have owed a part oft heir rapid celebrity to the echo they aroused in a camp to which it is clear today that they did not belong, that of 'structurahsm'.
For all these reasons, it would be inaccurate to speak of these writers as belonging to an epistemological 'school'. Cangnilhem, after all, is a historian of the sciences, and Foucault would correctly refuse to be described as an 'epistemologist': he calls himself an 'archreologist of knowledge' (savoir). Under these conditions, is it possible to speak in their case, as I propose to do, of an 'epistemological tradition' ? I find this justifiable because of the existence of a common feature which is transmitted to each of these oeuvres, despite their apparent diversity. It would be superficial to seek to explain it by the supposed intellectual 'influeuces' which masters may have exercised on disciples. It is true, of course, that Bachelard was Cangnilhem's 'teacher', and Cauguilhem Foucault's, but we cannot appeal to psychology as a substitute for historical analysis without immediately renouncing any understanding but an anecdotal one. The common feature I have mentioned is more real and more profound, it constitutes their point of agreement and pertains to their common 'position' in philosophy.
To put it briefly, and provisioually formulating it in a negative form, it lies in their radical and deliberate 'non-positivism'. This non-positivism, inaugurated by Bachelard, while it seems to me to form the 'cement' of the tradition uniting my three authors, simultaneously distinguishes this tradition from everything practised elsewhere under the name of 'epistemology'. It counterposes it to another tradition, one which does have its 'schools' and its 'institutions' a tradition dominant today and in which converge investigation conducted in the East and the West, at the University of Yale and in the Moscow Academy of Sciences.
I say that this other tradition, despite the internal distinctions that have to be made in it, is massively 'positivist' in that it always, in one way or another, presents itself as an attempt to elaborate a 'science of science' or -the technocratic variant- a 'science of the organization ofscieutific work'. This is the case with the Auglo-Saxon specialists Bernal and Price who, more than twenty years ago, announced under this title o.f a 'science of science' the emergence of a new discipline and estabhished under this sign an original institution. The insistent references made to it today by the editors of the philosophical journal of the Moscow Academy of Sciences (Voprosy Filosojii) -Kedrov, Mikulinksy and Rodnyi- in order to take up the same project in the same terms, sufficiently prove that they range themselves in the same camp.' Although in a quite different form, this is also the case with the 'logical neo-positivist' current which attempts, on the basis of the concepts of the scientific discipline of mathematical logic, to form the categories of what Reichenbach called 'a scientific philoophy, the philosophy of our time, that of the age of science', which is to be both a 'science of science' and a scientific critique of philosophy.
Whether epistemology is made into a kind of 'cross-roads' at which a cohort of heteroclite disciplines with scientific pretensions come to conjoin their disparate concepts in order to constitute a general theory of science, or whether a determinate science is made responsible for the provision of its categories, the philosophical presupposition of the undertaking is the same, and it is this presupposition that leads me to describe these attempts as 'positivist'. Truly, this common presupposition could find no better expression than the slogan: 'A science of science is possible.'
The claim I have just ventored may cause some surprise: that "the science of science' is mortgaged by a philosophical presupposition. A paradox, since, precisely, both groups claim to have put an end to the 'philosophical usurpation', restoring to science its most legitimate property and its most precious birthright: itself. However, I continue to state that philosophy is not absent from these efforts. Let me add, to complete the paradox, that two philosophical presuppositions are conjoined in it. It is their very conjunction which constitutes the positivism that I am denouncing. One pertains to the unity claimed for the duplicated singular term 'science' ; the other relates to the circle of this duplication within the reflexiveness it attributes to the term with itself: 'science of science'.
The first point is well known today: to speak of science in general as one single entity which might itself take itself as an object is to make use -here: double use- of an ideological notion. It is to suppose that the ensemble of scientific practices can be treated as a homogeneous reality, constituting, at least in principle, the unity of an undifferentiated whole. This 'treatment' or this 'point of view' on science is peculiarly philosophical. It even repeats, beneath a modernist and scientistic exterior, one of philosophy's classical procedures. Better, the classical procedure of idealist philosophy which when it speaks of the sciences, is eager for one thing only: to disengage their common 'essence' so as to be able to speak of 'science' in the singular -and then to take the theses produced thereby as a justification for the elaboration of a theory of knolvledge. Re-read the history of philosophy: it is easy to establish with what remarkable regularity this procedure has been applied, from Plato to Husserl.
Hence the philosophical presupposition whose presence I therefore detect in the positivist epistemologistsis an idealist philosophical presupposition. As such it simultaneously conceals and reveals, inthe symptomatic mode, a reality which we have momentarily glimpsed: the ensemble of scientific practices. More accurately: by attributing to this ensemble the unity of a whole, the presupposition 'resorbs' -imaginarily annihilates- the reality of these practices, which resides in their distinctness -each having its own object, its own theory and its specific experimental protocols- and in their uneven development -each having its particular history. I say that it is the very reality of these practices which is thereby masked, for they do not exist outside the system that they constitute. Now this system, far from unfolding beneath the appearance of calm identity which has been foisted onto it, itself only has reality through the different contradictions entertained between the theoretical disciplines which feature in it. It is the intertwining of its contradictions which gives a form to its history. Thus we now know what is 'concealed' in the last analysis by the idealist philosophical presupposition here in question: it is the actual history of the sciences.
To say that a science of science is possible is, besides to claim that "science" can unveil, by mere reflection on itself, the laws of its constitution, and thus of its functioning as of its formation. It is to claim of the "scientific discourse" that it has the intrinsic -and exceptional- virtue of being able, by itself, without going outside itself, to state the principles of its own theory. In other words, the 'scientific discourse', sovereignly autonomons, is accountable to no one and constructs itself, without let or hindrance, in the pure space of scientificity instituted, laid out and delimited by itself. Without let or hindrance, since every obstacle is in principle always already located, stated and surmounted in the implicit discourse it is constantly conducting with itself -a quiet murmur in its inmost soul that in case of trouble need only be made explicit to illuminat everything. An immediate and decisive consequence follows: if it is the laws of scientific discourse itself which determine the space of its own deployment, if in principle it encounters no let or hindrance in that space, then the completion of knowledge (savoir) -its end and its perfection- is always in principle possible. All that remains in fact is to level out the few, quite formal difficulties that momentarily block It. A question of techmque. Let us translate: there is no real history of the sciences ; time has nothing to do with it. Or rather: time can only intervene in the form of delay or anticipation. The history of 'science' is merely a development, at best: an evolution which guides knowledge from error to truth ; in which all truths are measured against the latest to appear.
Now, once again, we can state that a very old philosophical operation is being repeated in a new form. Did not what are customarily called the 'Great Philosophies' (an appellation à controler) have as their project -and avowed claim- to state the criteria for all real or possible scientificity ? Did they not put themselves, with respect to the existing sciences, in the position here declared of constitutive reflection ? As the point is of some importance, I shall resort to a typical illustration: the philosophy of Hegel. In the discourse it conducts on itself, this philosophy, as is well known, takes as its 'foundation' i.e. as its basis and guarantee, the categories which are stated in that strange book entitled The Science of Logic. The peculiar status of this book in the Hegelian system deserves a special study for itself alone. For our purposes, we shall only consider one small fact: Hegel declares that to it are consigned the categories of the scientificity of science. Witness the following passage in which the author situates the book with respect to the Phenomenology of Mind:
In The Phenomenology ofMind I have set forth the movement of consciousness, from the first crude opposition between Itself and the object, up to absolute knowledge. This process goes through all the forms of the relation of thought to its object, and reaches the Concept of Science as its result. Thus this concept (apart from the fact that it arises within the boundaries of Logic) needs here no justification, having already received its justification in that place.
For us this text has a dual interest. It reveals the position of the content of logic with respect to the existing sciences -those that Hegel calls: 'empirical sciences' ; it is logic that presents the concept of science ; or, better, it is Hegel's book that contains the justfication, the "foundation" for the concept of science. In short, Hegel's Science of Logic is the philosophical science, the true science, the science of the sciences.
Hegel's philosophy presents itself as 'the philosophy of all philosophies'. Doubtless it would not be illegitimate to.take its declarations at their word and see in it in fact, in the position it attributes to itself wish respect to the system of the existing sciences, the truth, having attained self-consciousness, of what previous philosophies were tacitly practising. It would then be seen that the key stone of this philosophy is a 'science of science'. One step further: It is clear that, according to Hegel's own statements, even the writing of this book -The Science of Logic- presupposed that The Phenomenology of Mind was complete, and that therefore history itself was complete In these conditions it is obvious that the project of a 'science of science' is merely the repetition in a new form of the same peculiarly philosophical operation. An operation whose effect fits in well with the one we have registered vis-a-vis the constitution of the unitary ideological notion of 'science': it annihilates the actual reality of the history of the sciences by placing it beneath the sign of a teleology. Now, abstracting from the specific structure taken by this teleology in the Hegelian dialectic, we can say -to return to the terms used at the beginning of this analysis- that the regular effect of the project of a 'science of science' is to reduce the actual history of the sciences to a kind of evolution. In other words: positivism and evolutionism go hand in hand. Or alternatively: evolutionism is the obligatory complement in the history of the sciences to positivism in epistemology.
In these conditions it is hardly surprising that the non-positivism of the epistemological tradition I am discussing begins with and holds to a deliberate rejection of an 'evolutionism'. I shall say that its non-positivism leans on an anti-evolutionism. Today everyone knows the prime term for this rejection in Bachelard's work, the notion of "epistemological 'rupture", is an injunction against every philosophy of history which would like to bring within its jurisdiction the epistemological categories he works on. Nor is it unknown that Georges Canguilhem, proposing to distinguish between the 'beginnings' of a science and its 'origins', relentlessly denouncing every attempt to seek more or less remote 'precursors' for a 'discovery', shares the same concern. But it is surely Michel Foucault who, analysing the notion of 'discontinuity' at the beginning of The Archelology of Knowledge, has best demonstrated the theoretical implications, decisive for the conception that can be held of history, of the anti-evolutionism that all three profess.
I can now give the positive content of the differentia specifica of this epistemological tradition, so far stated in a negative and polemical form. Their non-positivism and their anti-evolutionism pertain to the link they recognize between epistemology and the actual practice of the history of the sciences. A link, or, rather, a unity, whose theory, as we shall see, each has attempted to outline in his own way. However, these attempts seem to me to have failed up to the present. An inevitable failure, no doubt, for this theory belongs in principle to a discipline which they do not recognize: 'historical materialism', the Marxist science of history. For all that, all epistemological problems undergo a revolutionary displacement from the mere fact of their practice of this unity ; and on the other hand, their failure, insofar as it occurs on the basis of a correct practice, is infinitely valuable to us, since it designates to us, caught in its displacement, the site where we have to re-work.
It remains to understand why it was thus possible for this anti-evolutionist non-positivism to arise and be handed down in French philosophy alongside and against other currents more in accord with the dominant -spiritualist and positivist- tendencies of that philosophy. This analysis belongs to a history of philosophy which still has to be written. I can only say that in the last analysis this history would be the history of the unity I have been discussing: a history of its formation, of its inconsistencies, of its failures, of its struggles and of its successes. To this day I only have at my disposal one factual suggestion on this subject ; here it takes the form of a paradox requiring elucidation: it seems that it was really Auguste Comte and his disciples who made this unity possible and inscribed it in French academic institutions, making the history of the sciences a discipline belonging with philosophy. In fact, France since then has been one of the only countries in the world -if not the only one- in which the history of the sciences is practised and taught in philosophy faculties. Thus it is to the founders of positivist philosophy that we should impute the beginnings of a tradition whose major trait is, as we have seen, its non-positivism..." (pp.120-126)
" [From Bachelard to historical materialism]
"Gaston Bachelard was not a Marxist, nor even a materialist ; as a reminder, the passage from Applied Rationalism will be cited where materialism is presented as a 'flat', 'abstract' and 'crude' philosophy; it will be recalled that "discursive" as it may have been, Bachelardian epistemology explicltly wlsed to be idealist ; it wil! be Fointed out, lastly, that the most obvious point of agreement in this apparently contradictory -epistemologlcal and 'poetic' - œuvre is precisely a dynamistic conceptin of thought, which is fundamentally very 'psychologistic'." (p.129)
"From the 'realism' of Meyerson to Sartre's existentialism, no philosophy escaped his sarcasms. This perpetual polemic, far from arising from a personal psychological disposition, has a precise and profound theoretical meaning. It has to be taken seriously. Its principle can be found outside philosophy in the upheavals experienced in the actual history of the sciences at the beginning of tins century: the development of non-Euclidean geometries, the theory of Relativity, the beginnings of Micro-physics, etc. The new disciplines then inaugurated seemed 'unprecedented' to Bachelard ; his first works are but the reflection of this radical novelty: they take note of a rupture in the actual history of the sciences." (p.130)
"[Dans Le nouvel esprit scientifique, il] proposes a novel philosophical category - whose function at this stage is still descriptive: the category of 'No' or 'Non-'. This category was formed by a controlled extension of the negation which had served to think the disconcerting novelty of the non-Euclidean geometries: in the same way, one can speak of 'non-Newtonian' mechanics, 'non-Lavoisierian' chemistry, etc." (p.131)
"The discontinuity he registers in The New Scientific Mind is opposed to the principle that governed the then dominant philosophy of science, that of Emile Meyerson: The Inductive Value of Relativity (1929) is counterposed to La Deduction relativiste (Relativist Deduction), in which Meyerson, with some courage and considerable ingenuity, had undertaken to 'deduce" Einstein from Newton by showing that, at least in its principles, the theory of Relativity was already present 'in germ' in the Principia of 1687. Meyerson, convinced of the identity of the human mind in all its manifestations, held that the discontinuous was an illusory appearance attributable to the philosophical ignorance of scientists. Hence one must state that Bachelardian epistemology, 'starting not from a philosophical 'principle' but from a summons addressed to philosophy by the contemporary re-organization ofthe exact sciences, ran, in the very recognition ofits object, into a set of philosophical theses which were obstacles in its way. This collision was first thought by Bachelard as a delay (retard) of philosophy with respect to science, a delay which was still awaiting its concept." (p.131)
"But the 'encounter' with Meyerson's philosophy had another, still more. decisive consequence: in it Bachelard discovered the solidarity between a thesis concerning the theory of knowledge: 'realism', and a thesis concerning the history of the sciences 'continuism'. In it the historical continuity of learning (savoir) is supported by the homogeneity of the forms of knowledge (connaissance) - ordinary and scientific. An attack on the former could not but ruin the latter: such is the double, historical and epistemological, meaning of the now famous thesis ofthe 'rupture' as it functions in Bachelard's works. It is clear in what sense this epistemology has to to be called 'historical' ; it was so from the very first ; from the
moment it took as its task to draw the philosophical implications of the 'no' which the new scientific mind addressed to previous science. A task which received its formulation and the beginnings of a realization six years later in The Philosophy of No: the question, it says there, is 'to give science the philosophy it deserves'.
Bachelard was to carry out this undertaking during the next ten years: from The Philosophy of No (1940) to Applied Rationalism (1949). In doing so he steadily acquired the conviction that this philosophy' could only be elaborated marginally to all the existing 'theories of knowledge'. Not only 'marginally to' but also against them. What he had defined as the 'delay' of philosophy now seemed to him to be a systematic 'displacement' of every philosophical theory of knowledge with respect to the scientists' actual work. That is why
Applied Rationalism, which can be said to realize, but also to rectify in many points, the programme of The Philosophy of No, opens with several crucial pages in which Bachelard sets out what I propose to call a 'topology' of philosophy. This topology takes the form of a 'spectrum' in which all the types of theory of knowledge appear dispersed around the reality of the work of production of scientific concepts, in which this topology proves to be a 'typology'. Allow me to reproduce this spectrum (on p. 133) in order to analyse it. This spectrum has two very remarkable characteristics: (a) every philosophy -to the extent that it contains as a principal component a theory of knowledge- is defined in it by its place - its specific 'displacement' - with respect to scientific knowledge ; (b) scientific knowledge has the role of an axis, that is to say that by a mere 'fold' around this axis it is possible to make the various typical forms of philosophy coincide. Three conclusions can be drawn from these two characteristics: ,(a) what is given as the content of philosophy is none other than tlte hallucinatory materialization of its peculiar dispersion (teart) from science ; (b) the always possible coincidence of symriletrical and opposed typical forms of philosophy is the index of a fundamental identity: each form is merely the 'inverted' but identical form of an opposite form ; (c) the essence of philosophy can only be determined from the point of view of the axis, i.e., from a non-philosophical point of view.
Idealism
+
Conventionalism
+
Formalism
+
(philosophy ofthe production ofscientific knowledges:
applied rationalism and technical materialism)
+
Positivism
+
Empiricism
+
Realism
This 'identity in inversion' that I have just emphasized is analysed at length in Bachelard's last works. He there remarks that it is supported by a certain number of philosophical couples, each term of which can by turns be dominant or dominated. These categorial couples form the apparent 'content' of the philosophical theories of knowledge. Stated as such in Applied Rationalism, they are as follows: subject-object, abstract-concrete, given-construct, etc.; they are interchangeable and complementary. They all have as correlate and cement the philosophical category of truth which presents itself as the concept of their accord and closes the space of philosophy. Now this last category, implying that the completion of the process of knowledge is always possible in principle or realized, makes inconceivable the actual history of scientific knowledge: its ruptures, its reorganizations, its failures, its contradictions, its risks.... That is why an epistemology which is historical will pay more attention to error, tofailure, to hesitations than to truth: its space will thus be open, and non-systematic." (pp.131-133)
"The philosophy ofthe scientists is contradictory: to use an expression of Bachelard's, it mixes together a 'diurnal philosophy', the clear philosophy of the science, and a 'nocturnal philosophy', the philosophy of the philosophers to which the scientists inevitably turn when they reflect on their own practice. I shall say that the scientist maintains an imaginary relationship to his own practice and that the philosophy of the philosophers plays its part in the constitution of this relationship." (p.134)
"The first concept to be constructed, the one which sustains the whole edifice, is the concept of 'epistemological obstacle'; it designates the effects on the scientist's practice of the imaginary relationship he maintains with this practice. This concept is still famous thanks to the many and often terrifying illustrations Bachelard gave of it on The Formation of the Scientific Mind (1938). It must be said that, preoccupied with what was most urgent, he concerned himself less with its mechanism than with its effects. Or rather with its unique effect ; for, although 'polymorphous', the obstacle functions in a single direction: going back on the 'no', it closes the rupture between ordinary knowledge and scientific knowledge and reestablishes the continuity threatened by the advance of scientific knowledge.
It may arise at the moment of the constitution of the knowledge, in the form of a 'counter-thought', or at a later phase in its development as a 'suspension of thought' (arrêt de pensée). Whatever the case, the obstacle reveals 'a resistance of thought to thought'. More precisely: assuming that scientific thought only advances by its own reorganizations, it can be said that the epistemological obstacle appears whenever -but only then- an existing organization of thought -whether it is already scientific or not- is in danger. Let me add that it appears at the point at which the rupture threatens - which other efforts than those of Bachelard have been able to show to be the site of an 'overdetermination', of an accumulation of contradictions. Localized in its appearance, the obstacle is solidary with a determinate structure of thought which will later appear, by recurrence, as a 'tissue of tenacious errors'. The status Bachelard attributes to the 'philosophy of the philosophers' is now clearer: it is the vehicle and support of the epistemological obstacles, since it is
what structures the relationship of the scientist to his scientific practice. It registers scientific ruptures and reorganizations by inversions in its spectrum ; and, in the closed field of its basic categories, it thereby assimilates the advances of scientific knowledge and lives by their exploitation. It is easy to observe its action, to describe it, even to foresee it: Bachelard, for provocative reasons -in order to shake the philosophers out of their 'torpor'- more than once tried this out. But one question about the epistemological obstacles remains unanswered: once their effects and the mechanism of their intervention in scientific practices have been described, that of their formation remains to be thought. Rather: what necessity is there in the fact that epistemological obstacles are always being formed and reformed ? We shall find that this question reveals the limits of Bachelardian epistemology.
First we must examine the positive categories released by the recognition of the obstacles as such and by the respect for ruptures and re-organizations. We can say that all these categories are ordered according to an unprecedented conception of the dialectic. In Bachelard this category designates the reality of scientific work: the process of mutual adjustment of theory and experiment. Now, given the rejection of every fixed point -through an initial and polemical rejection of the subject-object couplet- this adjustment
has to be thought not as a formal adequation but as a historical process. In a history which implies no certainty, no fate destines theory always to find the means of its realization. The risk is that at a given moment the language of the physicist and that of the mathematician for example, should become contradictory. Philosophy will hasten to see in this a 'crisis' of science. For the mathematicians and physicists, it will be the opportunity for some work: the former will revise their theories, formulate other hypotheses, the latter will
refine their experiments, check their instruments. Neither the former nor the latter will hesitate to reconstruct the edifice of their science 'from top to bottom' if need be. Georges Canguilhem has shown that this category of the dialectic cannot be assimilated to any classical conception of the dialectic: this is readily understandable once the meaning and implications of Bachelard's anti-philosophical polemic have been grasped. The dialectic that emerges is the 'spontaneous' dialectic of scientific practice: it affirms -against philosophical scepticism- the existence of the objects of science and proceeds by a reorganization of learning (savoir) 'from top to bottom'." (pp.135-136)
"Bachelard's polemic against the philosophical theories of knowledge has one last effect: it demands a rectification of the philosophical category of experience by correct appreciation of the function of instruments in the production of scientific concepts. Everyone knows Bachelard's thesis that scientific instruments are 'materialized theories' ; but it is essential to add the following counter-point: the 'materialization' of theory is not for him an acessory phenomenon. On the contrary, he ceaselessly stressed that it is one of the most characteristic features of the contemporary sciences that they are 'artificialist', that they contain as one of their essential components a technique for the production of phenomena ;
what, in parody fashion, he has called a 'phenomeno-technics', and which is the object ofthat new discipline programmatically adopted by Applied Rationalism: 'technical materialism'.
The notion of phenomeno-technics has to be taken seriously: it enables us to understand in what sense the word 'production' is to -be understood : not only the 'theoretical' production of concepts, but indissociably the material production of the object of theoretical labour; of what can no longer be called its 'data' or 'givens' (donnees), but rather its 'material'. The philosophy of the philosophers, when it is confronted with this material intervention of instruments in the production of concepts can only think it as an inessential 'mediation', and leave it to fall within the general and vague category of 'experimental method', a specification 'for the use of the scientist' of the philosophical category of experience. According to Bachelard, this general category, ignoring the specificity of the cases in which instruments function, gives no real knowledge of what takes place in scientific practice. It is unable to explain the new fact that a concept must from now on integrate into its constitution as a concept the technical condition of its realization. Ultimately this is because the notion of 'method' and the philosophical category of experience are solidary with the conceptual couplet 'abstract-concrete'. Now, the practice of phenomeno-technics consists precisely of proceeding- to couplings (couplages) between the abstract and the concrete by the expedient of the setting up of theoretically defined instruments and systems of apparatus according to programmes of rational realization. Thus the 'objects' of science, far from being poor abstractions drawn from the wealth of the concrete, are the theoretically normed and materially ordered products of a labour which endows them with all the wealth of determinations of the concept and with all the sensitivity of experimental specifications. It can thus be said, following Bachelard, that these objects are 'concrete-abstracts'.
The dual status of scientific instruments has the additional interest of revealing -materially- the 'eminently social' character of contemporary science. Bachelard's last epistemological work, on chemistry, Rational Materialism (1953), is particularly instructive in this regard. In it he systematically opposes what the chemist calls a 'substance' to what the philosopher means by substance. Among the essential characteristics of a chemical substance is 'purity'. Here is what Bachelard writes about it:
It can be said that there is no purity without purification. And nothing better proves the eminently social character of contemporary science than the techniques of purification. Indeed, purification processes can only be developed by the utilization of a whole set of reagents whose purity has obtained a kind ofsocial guarantee. Elsewhere, as is well known, Bachelard could write that only 'society can send electricity along a wire'. In short, he reveals rhe need, in order to think the history of a science, to take into account the state
of the 'technological city' (cite technicienne). But that state is quite clearly linked to the history of the techniques of production ; i.e., to the City as such. What type of determination is there between the history of production techniques and that of scientific instruments, between that scientific instruments and that of concepts ? These questions remain unanswered." (pp.137-139)
"In the case of the status of "scientific instruments", what is missing from Bachelard's analysis is a concept that would enable him to think together several histories with different statuses ; in short, the concept of a differential history. When it is a matter of the formation of epistemological obstacles, the same lack is felt, but the consequences, as we shall see, are much more serious. Once the rupture between ordinary knowledge and scientific knowledge has been registered, it is indeed a matter of determining the constitution of that 'tissue of tenacious errors' with which science 'breaks'. At the same time, this is also to account for the necessarily imaginary nature of the relationship the scientist maintains with his own practice.
Bachelard attempted this in two directions, unevenly explored in his work. The first, evoked above all in The Formation of the Scientific Mind, consists of casting doubt on scientific education which, both in its lessons and in its 'practical work', displaces the true interests of science behind the mask of pedagogy: the constant appeal to the images and experiences of everyday life made in physics courses seems to him particularly damaging. The philosophy class is also inscriminated, but for a different although complementary reason: it diffuses an undeserved valorization of "general culture" to the disadvantage of the scientific specializations, which are, however, the only motor of the advance of the sciences. Thus, by means of educational institutions, the conditions of the reproduction of scientific knowledges have their effects on the forms of their production.
The second direction is intended to be more profound, and it is more minutely analysed by Bachelard ; it constitutes the object of the new discipline he believed he had founded: the 'psychoanalysis of objective knowledge'. It must be candidly stated: all that enables Bachelard to think the necessity of the 'epistemological obstacles' is a certain conception of the human soul which roots the 'imaginary relationship' in the imaginary of the images produced by the imagination. The pages -and they are numerous- in which Bachelard describes this inhibitory intervention of images in scientific practice are still famous. But their crucial theoretical status must be understood: they attempt to elaborate a system of concepts which will make it possible to think the intrication of two histories: that of the scientific and that of the non-scientific in the practice of the scientists. Hence it is not surprising that this historical epistemology culminates, in The Rationalist Affectivity of Contemporary Physics[/i (1951), with the project of an epistemological history which is presented as a dual history: a 'ratified history' (or history of the scientific in scientific practice) and a 'lapsed history' (or history of the interventions of the non-scientific in scientific practice).
But if we turn to the realizations of this double history, we find that the only example of a ratified history that he elaborated -in Applied Rationalism- he elaborated vis-a-vis the history of mathematics, whose very special character was stressed by Bachelard himself, following Cavaillès, for mathematics immediately manifests the existence of what he calls a 'logical time', a continuous identity, which is not found as such in the other sciences. As for lapsed history, illustrations of it will be sought in vain: it is thought very precisely as a non-history: as a 'museum of horrors' or an 'unformed magma'. Given this, the conjunction of the two histories and their mutual determination remain obscure.
His resort to the libido of the scientists to explain the constitution of epistmological obstacles here at once reveals all its meaning: it compensates for Bachelard's inability to think the differential history of what I shall call sciences and ideologies. By that very fact, all Baclrelard's epistemological concepts are haunted by psychologism: in the term 'scientific mind' it is the term mind that tends to become dominant ; the notion of scientific labour and, correspondingly, philosophical laziness, take on a subjective connotation ; the application of 'applied rationalism' threatens to capsize into psycho-pedagogy ; obstacles, finally, can be understood as mere difficulties, whereas the scientific city is conceived as an 'inter-subjectivity'. It has to be admitted that many pages of Bachelard are open to such an interpretation.
But if the notion of an 'imaginary relationship' is referred, not to a psychology of the imaginary, but to the scientific concept of ideology as it features in 'historical materialism', in the science of history, and which designates precisely the 'imaginary relationship of men to their material conditions of existence', it is clear what is designated, but not thought, by Bachelard: the necessity, in order to construct the concept of a history of the sciences, to refer it to a theory of ideologies and of their history.
Such a reading really does allow us to assign limits to Bachelard's epistemology since it reveals the 'psychologism' that sustains his 'poetics' as a point of retreat, better as a 'return of philosophy', if it is true that, unable to think the relation scientific non-scientific as a differential history, Bachelard based it on the repetitive permanence of great themes, myths or complexes in an eternal unconscious. In short, paradoxically, he turns to the analogue of a theory of knowledge. Hence it has to be said that Gaston Bachelard's historical epistemology remains a non-philosophy in philosophy. Nevertheless, by its respect for the spontaneous dialectical materialism of scientific practice, it provides us with precious elements for a theory of philosophy and of its history, and on condition that we know how to read it, precisely where it is inconsistent it reveals the ways to supplement it so as to construct a materialist theory of the history of the sciences." (pp.139-141)
-Dominique Lecourt, Marxism and Epistemology. Bachelard, Canguilhem and Foucault, NLB, 1975, 223 pages.
