108 SOCIAL SCIENTIST The idea that science develops not only by breaks with ordinary experience but also by breaks with previous scientific theories, was developed by Bachelard two or three decades prior to similar ideas developed by Kuhn and Feyerabend. As a corollary, Canguilhem's critique of Bachelard and Kuhn of reducing paradigm choice to the level of social psychology is very similar to Kuhn's criticism by Popper and Lakatos. To students of history and the philosophy of science, who are familiar only with the Anglo-Saxon tradition, these two volumes would be of immense value. The book edited by Redondi provides a representative selection from the writings of Cendorcet (1743-1794), Auguste Comte (1798-1857), Pierre Duhem (1861-1916), Paul Tannery (1843-1904), Alexandre Koyre (1892-1964), Gaston Bachelard (1884-1962), Georges Canguilhem (1904- ) and Michel Foucault (1926-1984). The book provides a fair of the French writings on the subject over the last three centuries but suffers by its very nature as an edited volume in that the selections of scores of pages from major thinkers are not likely to provide any meaningful explication of their positions. Divided into three main section^, positivism, history of scientific thought and historical epistemology, the volume has a brief introduction by Pietro Redondi which locates the French tradition in relation to the Anglo-Saxon tradition and provides the context for the division of various authors in the different sections. ' Gary Cutting's book places Foucault's work 'in the context of recent French history ami philosophy of science, particularly the work of Gaston Bachelard and Georges Canguilhem.' This contextualisation of Foucault complementing his construe as a social critic provides tremendous insights into the process of achieving the Enlightenment ideal of using reason for human liberation. Foucault who shifts the terms of the debate from what constitutes rational to why some body of thought is considered rational, develops a technique 'for revealing how a discipline has developed norms of validity and objectivity.' According to Gutting Foucault's project of history of thought has a twofold purpose: 'to show how 'particular domains of knowledge have constrained human freedom and to provide the intellectual resources for overcoming these constraints.' Pursuing this line of argument Foucault arrives at the conclusion that 'modem social sciences are inextricably intferwoven with techniques of social control and their very constitution as knowledge depends essentially on mechanisms of power.' The Bachelard-Canguilhem-Foucault network provides useful insights for an elaboration of materialist epistemology, despite none of them being a professed materialist. Foucault's project of showing the contingent nature of what present themselves as necessary and a priori limits on knowledge and a similar idea in Bachelard of episte-mological obstacles or more specifically the idea that philosophical a priories derive from our inability (or non-willingness) to think beyond the categories of current science, provide a framework which could