Foucault on Bachelard


Bachelard and Philosophy of Education
James Marshall
University of Auckland

Gaston Bachelard was born 27 June 1884 at Bar-sur-Aube, Champaigne, where his parents owned a small paper and tabacconist shop. He received his secondary education in Bar-sur-Aube, and served in the first World War (including 38 months in the trenches and the award of Le Croix de Guerre). Essentially he was an outsider to the academy, for he had taught himself chemistry whilst working as a postman, before teaching in secondary colleges at Bar-sur-Aube. Studying slowly he was awarded his doctorate from the Sorbonne in 1927. It is clear that the achievements in particle physics had both excited and influenced him. He was to say: ‘one decade in our epoch is equal to centuries in earlier epochs’ (quoted in Lecourt, 1975, p.33). Having turned to philosophy of science, he was to teach for 10 years in the Faculté des Lettres de Dijon before becoming a Professor at the Sorbonne in 1940. In May 1960 he was made an Officier de la Legion d’Honneur and, dying on 16 October 1962, he was to be interred at his Bar-sur-Aube on 19 October.

Bachelard was recognised as both philosopher and poet, but it is only his work as a philosopher which will be considered here. He is important not only for his work in the area of philosophy of science, but also for his influence upon a number of French philosophers,. Jean Cavailles, shot as a member of the Resistance to the German occupation in WW II, was an influence upon Bachelard and upon this French stream of philosophy. Within his general philosophy this entry will concentrate on those features which were to later influence French poststructuralism.

Bachelard’s anti-positivist philosophy of science was developed in a number of texts between 1927 and 1953, anticipating some of the conclusions of Thomas Kuhn, though he was not to exert any direct influence on Anglo-Saxon philosophy of science. For Bachelard science was an autonomous activity, dependent only upon itself for its norms and practices – in spite of the intrusions of philosophers. His work involved a constant polemic against philosophers because, according to him philosophy wanted to cover up, hide and occlude ‘the real historical conditions of the production of scientific knowledges’ (Lecourt,1975: 27). Indeed the idea that the history of science could be a fruitful source, even a source, for logical analyses was quite alien to analytic philosophy of science (Tiles, 1984: 3). Also he believed that physical theories were not divorced from metaphysical commitments, though they might claim to be so. Bachelard then uncovers the ‘unthought’ of philosophical discourse about science in a ‘recovering’, which also opens up the notions of the ‘unconscious’ and the ‘unthought’ and the further possibility of the ‘recovery’ of these by a ‘psycho-analysis of reason’. (Traces of this notion of a ‘psychoanalysis of reason’ are to be found in the early work of Michel Foucault).

But Bachelard’s scientific reputation depends also, in part, upon his studies of poetic language, day-dreaming, and phenomenology, and their application to episodes in the history of science. Thus poetic imagination, as involving an affective engagement with ‘things’, was seen by him as a condition of scientific productivity. But this insight is not hived off into the deflated notion of a ‘logic of discovery’ as in the philosophy of Sir Karl Popper, for Bachelard include it in his philosophy of science. Thus Kuhn’s important Structure of Scientific Revolutions (1962) may have been ‘old hat’ for French philosophers of science.

According to Gary Gutting (1989) the account of scientific change and, accordingly, scientific progress was the focus of Bachelard’s philosophy of science. Bachelard held that reason was best exemplified in science, and this was best understood, not by studying science now, but by studying its history (in philosophy of education, on the importance of the history of science, see here the Althusserian inspired work of Michael Matthews). Reason was best discovered in science by discovering how reason was used in practice, in addressing particular problems (cf. John Dewey [1938] The Logic of Inquiry), as opposed to studying abstract theories. The history of science showed, he argued, how abstract philosophical and metaphysical theories in science were gradually overturned through the practical work of scientists.

Bachelard found discontinuities in the history and development of science that were not, in principle, capable of logical reconstruction, prior to Kuhn, with whose work there are some similarities. But whereas Kuhn was concerned with theories, or paradigms, Bachelard’s concern is with concepts. The key concept here for Bachelard is that of rupture, which has four epistemological aspects or categories. These he terms breaks, obstacles, profiles and acts (Gutting, 1989).

Breaks are concerned with (i) how science breaks away from common sense in formulating its concepts (see also Stephan Körner [1966] Experience and Theory; an essay in the philosophy of science), and with (ii), breaks between scientific concepts. An example of (i) would be the shift from common sense notions of intelligence to IQ conceived as intelligence, whereas an example of (ii) would be the changes that have occurred historically in the scientific concepts of the atom. Breaks, in turn, are opposed by obstacles, which can be construed as residues from earlier concepts which, especially if they were important in the past, tend to block changes to new concepts. Here common sense may be a major obstacle, operating as it does as an implicit assumption of the way the world is. If they operate at an implicit level then Bachelard was to propose a set of techniques to bring these assumptions to the ‘surface’.

An epistemological profile is an analysis of a given person’s understanding of a particular scientific concept. This analysis would show where an individual stood in relation to the historical development of a concept or theory, and the extent to which residues were retained or breaks maximised. In recent liberal political philosophy a profile would show whether a person was a liberal or neo-liberal according to how underlying concepts such as individualism, freedom and equality were conceived. Residues are countered by breaks but this requires on behalf of the scientist the Bachelardian notion of an act. Acts are leaps but not just a leap in any old direction at all, but one in the direction of progress.

Cavaillès and Bachelard rejected the idea that scientific progress was to be determined by collating in a jigsaw type pattern the established truths of science. For Bachelard, from his work on the imagination and reverie, epistemological acts which produced these breaks were guided not by normal science but by poetry and art. (See also the philosophical work of distinguished biochemist Sir Peter Medawar [1966], Induction and Intuition in Scientific Thought). Foucault was to declare concerning Bachelard in 1954: “no one has better understood the dynamic work of the imagination”. Indeed Foucault’s early work on Binswanger (also a Bachelardian) cites Bachelard against Sartre on the topic of the imagination. Bachelard, in turn, was very complimentary to Foucault on his Folie et Deraison.

For Bachelard scientific progress depends therefore upon epistemological breaks. According to him these breaks were quite distinctive – eg, the break from Newtonian to Einsteinean physics, was seen as very distinctive. Science therefore cannot be seen as progressing linearly or necessarily rationally (cf. Kuhn). But progress is conceived and measured from the present status of science, and history of science will provide a history of rationality. But this notion of progress is questionable, because Bachelard accepts that the science of the present is not only an advance (as opposed to a mere change), but that the science of the present should be invoked to both understand and assess the achievements of past science. This modern assessment is to be based then upon how past scientific endeavours contribute to the ‘developed’ science of the present. What is sought are the effects of the past upon the present. There are of course some questions which historians would raise about the appropriateness of such moves for doing history: for example, Phillipe Ariès’ Centuries of Childhood (1962) was criticised for his projection of a 2Oth. century concept of childhood backwards to medieval times.

Canguilhelm is to temper Bachelard’s concept of the (distinctive) epistemological break and the influence upon Foucault is more general than specific.

The implications for education from Bachelard’s work include at least: the account of rationality; the importance of teaching the history of science and not merely science as it is now; the implications for those research methodologies which appeal to ‘a’ philosophy of science; the importance of the imagination (in his sense) for rational thought; and the notion of conceptual discontinuity in scientific thought.


Bachelard, Gaston. (1928). Essai sur la Connaisance Approchée. Paris: Vrin.

Bachelard, Gaston (1940). La philosophie de Non: Essai d’une philosophie de nouvel esprit scientifique. Paris: Presse Universitaires de France. Transl G.C. Waterstone, New York: Orion Press, 1969.

Bachelard, Gaston (1953). Le Nouvel Esprit Rationnel. Paris: Presses Universitaires de France.

Bachelard, Gaston (1960). La Poetiqué de la Rêverie. Paris: Presses Universitairs de France. Transl. D.Russell, New York: Orion Press, 1969.

Ginestier, Paul (1987). Pour Connaître Bachelard. Paris: Bordas.

Gutting, Gary (1989). Michel Foucault’s Archaeology of Scientific Reason. Cambridge: Cambridge University Press.

Lecourt, Dominique (1975) Marxism and Epistemology: Bachelard, Canguilhelm, Foucault. Transl. Ben Brewster. London: New Left Books. (Originally published as The Historical Epistemology of Gaston Bachelard. Paris: Vrin, 1969).

Margolin, Jean-Claude (1974). Bachelard. Paris: Seuil.

Tiles, Mary (1984). Gaston Bachelard: science and objectivity. Cambridge: Cambridge University Press.

Encyclopedia of Philosophy of Education


Leave a Reply

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>