Pour une histoire des sciences à part entière

Jacques Roger (1920-1990) was one of the leading French historians of science of the twentieth century, noted for the definitive studies of his countryman Buffon. As professor at the Sorbonne and Director of the Ecole des Hautes Etudes en Sciences Sociales, the Centre Internationale de Synthese, and the Centre Alexandre Koyre, Roger was unquestionably a force to be reckoned with amongst French historians of science. According to his publication list, appended to this selection of his writings, he was author of three books, editor of three series, editor of eight scholarly editions, and author of 139 papers, eight dictionary articles, and fifty-eight reviews. The selection of his writings offered here consists of nineteen papers, mostly in their original French, but some translated from English or Italian. There is also a most useful introductory essay on Roger's views on historiography and other matters, by his former student Claude Blanckaert; and an illuminating 'postface' by Jean Gayon, entitled 'De la philosophie biologique dans l'oeuvre historique de Jacques Roger'. An index should have been provided, however. The central landmark for Roger's work was, of course, Buffon, on whom he was the acknowledged authority. It is remarkable indeed how an academic "empire" can be constructed by using one major figure as the focus of one's work and reputation. It has been done by others of course (for example Drake with Galileo, Whiteside with Newton). But in such cases there has nearly always been more to it than that (e.g. Drake's experimentalism, Whiteside's mathematical expertise). For Roger and Buffon, the success seems to have flowed from his use of Buffon as a means to get a grip on the whole of the life sciences (and medicine) and the earth sciences of the eighteenth century, also extending backward and forward in time so as to embrace such Renaissance figures as Jean Femal, and nineteenth-century topics such as eugenics. Moreover, Roger developed general ideas about the way history of science should be written. His ideas on this are stated in his paper 'Pour une histoire historienne des sciences', part of the present collection. The underlying theme of this essay is the establishment of a clear distinction between the work of the "scientist-historian" and the "historian of science". As may be imagined, Roger's sympathies lay with the latter, and he takes the reader through what are today standard arguments for the avoidance of whiggism and historiographic anachronism. In this essay, and others in the anthology, one can clearly see the French tradition of history of science-stemming from the likes of Duhem, Metzger, and Koyr6-firmly underpinning Roger's writing. In fact, his work has a strong "history of ideas" character, though not overtly Lovejoyian in character, being without the "unit idea" doctrine. Even so, it was ideas (in people's heads), and the way in which they developed and influenced others, that were of paramount significance for Roger. Often the ideas that interested him were as much philosophical as scientific, though he did not count himself a philosopher. However, as Blanckaert and Gayon explain in their very useful introduction and postscript, which summarize Roger's thought in a synoptic way that has not, I think, previously been available in the literature, there was a somewhat curious aspect to Roger's crusade against historiographical anachronism. Supposedly there were certain recurring "themes fondamentales" in history of science-perhaps analogous to what Holton