The Historical Epistemology of Bachelard and Canguilhem

Althusser's notion of a symptomatic reading, divested of theoreticist implications, initiates a return to the problem of the production of knowledge from a historical perspective pioneered by French philosopher of science Gaston Bachelard and continued by his successor as the head of the Institut d'Histoire des Sciences at the University of Paris, Georges Canguilhem. There is no need to repeat here the excellent introduction to Bachelard and Canguilhem provided by Dominique Lecourt's Marxism and Epistemology (1975), but we might take a moment to summarize those aspects of their work to which Althusser is most indebted and to make clear the limits of that indebtedness as well, limits that are rather more sharply defined than is generally recognized by those unfamiliar with Lecourt's fine study.^[9]

In a series of books including Le nouvel esprit scientifique (1934; English translation, 1984) and La philosophie du non (1940; English translation, 1964), Bachelard established himself as the first to recognize the importance of historicity to the philosophy of science. Bachelard recognized the theoretical object of the history and philosophy of sciences as a set of historically determinate relations of production of concepts and took the unusual step, for a philosopher, of respecting the autonomy of scientific practice. Thirty years before Kuhn rediscovered certain of his insights, Bachelard put forward the proposition that every particular science produces its own norms of truth at each moment of its history, thereby dissolving the rationalist chains that bound the philosophy of science to what he called the "philosophy of philosophers."

By invalidating the absolute category of Truth, Bachelard denied philosophy the right to tell the truth of the sciences and proposed instead to tell the truth about the Truth of the philosophers. This smaller truth was the fact that the central determination of all philosophy, insofar as it contains a theory of knowledge, is the specific relation of dependency of philosophy on the actual production of science.

Bachelard insisted on the internal autonomy of scientific discourse, the interdependence of the concepts that make up its theoretical structure or problematic, and also the necessity of a history of science in terms of its own internal structures of theoretical production. Innovations and progress in the sciences are neither the result of linear evolution nor the historical accumulation of factual information. Rather, they are achieved by sweeping transformations of existing conceptual frameworks, revolutionary transformations whereby earlier conceptualizations are rejected, replaced, or reformulated by new theoretical constructs. Bachelard rejected philosophies of knowledge and epistemology in favor of a historical epistemology and the concept of scientific practice as a production. Philosophy of science must be separated from the philosophy of philosophers, he maintained, because the latter constitutes a repository of non-scientific images and notions, ranging from idealism to empiricism, which invade scientific discourse and become "epistemological obstacles" to scientific thinking. The only appropriate philosophy of science, Bachelard concludes, is a philosophy of negation, a rejection of the philosophy of philosophers that attempts to construct absolutes from the historical constructs of science.

As Bachelard's epistemology is historical, so the history of science practiced by Canguilhem is epistemological. In works such as The Normal and the Pathological (1943; English translation, 1978) and La formation du concept de réflexe aux XVIIe et XVIIIe siècles (1955), Canguilhem grasped the fact that historical epistemology imposes a new unity on the scientific study of historically specific scientific practices. For Canguilhem, epistemology is no longer understood as a philosophical search for formal guarantees of certainty; rather, it consists in disengaging—discovering and analyzing—the problems posed or evaded, resolved or dissolved by the actual practice of scientists. For Canguilhem, the actual practice of scientists, while relatively autonomous, exists in no reified intellectual sphere; instead, it is firmly inscribed within a historically specific "cultural frame." Therefore, while the progress of knowledge is discontinuous, it is never accidental.

Canguilhem's work in the history of the biological sciences is a

succession of unmaskings (of historical "accidents") and revelations (of historical necessity). To accomplish this task, it is necessary to bring out the specificity of the theoretical object of the sciences and to formulate the historical concept of scientific problems. For example, the history of the concept of reflex motion is obtained by asking what a theory of muscular motion and the action of the nerves must contain for a notion such as that of reflex motion to find in it a sense of truth. Posing the question of the concept in this way enables Canguilhem to discount the traditional paternity attributed to the concept, namely, the mechanistic physiology of Descartes, by demonstrating that reflex motion is unthinkable within the structure of a Cartesian problematic (wherein the motion of the "spirits" from the brain toward the muscle is strictly a one-way movement). The concept becomes conceivable, Canguilhem demonstrates, only with the vitalist theory of Thomas Willis, who by thinking the specificity of life in an integral manner and by assimilating life to light made it possible to think the movement of nerve and muscle as the reflection (analogous to optical laws) of an impulse from the periphery toward the center then back toward its starting point. By this example Canguilhem establishes two methodological principles: first, the history of concepts must be examined in terms of historical problematics and the problems posed and resolved by them; second, the theoretical and practical motives informing the way a science has gone about posing and resolving problems must be demonstrated as part of an adequate historical explanation of scientific practice. For Canguilhem, science takes the form of a struggle or a dispute and the function of the historian is to analyze its phases—not simply to draw up a balance sheet measured by the Truth of the present, but to provide a rational account of the sudden changes of terrain, the historical conjunctures that constitute that history.

Even these brief remarks are sufficient to demonstrate what Lecourt calls the "truly inestimable theoretical debt" that Althusser's reading of Capital owes to the historical epistemology of Bachelard and Canguilhem, a debt Althusser himself refers to as "incalculable" (Althusser and Balibar 1970, 323). However, Althusser also quietly insists that he has "gone beyond" the work of Bachelard and Canguilhem in "certain ways." Although he does not elaborate, it is clear enough that the "certain ways" to which Althusser refers pertain to the historical materialist framework within which he grounds scientific practice. For Althusser, it is not simply a matter of Marxism returning to epistemology from

the standpoint of the history of science in order to enrich its own self-understanding and rectify certain of its own concepts; it is also, and even more important, a matter of establishing historical epistemology and the history of science as a regional field within the general science of history. Because of his historical materialist problematic, Althusser's debt to Bachelard and Canguilhem is critical in nature, in this respect analogous to Marx's dependence on Smith and Ricardo. As Marx was able to interrogate British political economy as to the questions the latter could not answer or even ask, so Althusser forces French historical epistemology beyond the limits of its own self-understanding. Lacking a science of history, in particular a concept of ideology as a social instance, neither Bachelard nor Canguilhem is able to produce an adequate concept of the social nature of scientific practice.

Bachelard attempts to ground science in psychology, referring to his own philosophical practice as a "psychoanalysis of objective knowledge" and defining the scientific imagination as one of the "natural inclinations" of the human soul—inclinations that also include poetic reverie and the "common sense" of immediate experience. Canguilhem uses his brilliant historical investigations of vitalism as a philosophical position pertinent to the development of biology to ground the history of the life sciences in a "philosophy of life," an essentialist and teleological "unity" of the "concept of life" and "life itself" as manifested in the scientific discovery of DNA. In rejecting such psychological and biological explanations of scientific practice in favor of the determination of social structures, Althusser demonstrates convincingly that his concepts of science and philosophy are not reducible to those of Bachelard or Canguilhem, that they do not refer to precisely the same theoretical objects or function in precisely the same way. Althusser's concept of ideology, as Ben Brewster (1971) quite correctly points out, is as invisible to Bachelard's philosophy of science as the concept of surplus value is to Ricardo's political economy. It is worth noting that shortly after Althusser's reformulation of the concept of philosophy in 1967, Canguilhem himself introduced the concept of ideology into his own lectures (see Canguilhem 1988). Althusser neither rejects Bachelard and Canguilhem as "bourgeois" philosophers nor substitutes their problematic for that of Marx; rather, he reads their works in a realist and a materialist manner, accepting and applying those concepts that increase the amplitude of historical materialism while rejecting and replacing those that do not.