Genèse de la théorie cellulaire

or represent aspects of nature or the language-user's own general or circumstantial experience. In his book Professor Bazerman rarely avoids assuming as obvious the distinction between language and that which language is about, so that statements like the following are typical: "Experimental reports attach themselves to the nature that surrounds the text through the representation of the doings, or experiment" (p. 62). It is not hard to see that if one assumes that scientific language is about something else, one is left with little to say about that "something else", for it becomes the choice between two ontological danglers, "experiences" and "things in themselves"-language, apparently, is either about the one or the other. Nor is one left with much to say about language, if it is seen merely as conductive of knowledge rather than integral to it. Yet, on reflection, nobody reading a scientific article today, whether an account of an experiment in laser optics or a theory of turbulent flow around two-dimensional bluff bodies, thinks that the sense of the article (or its primary function) consists in referring to or representing some past or potential state of affairs. (It would be rather like thinking that numerals and operators of arithmetic are about numbers and operations.) Steven Shapin's sense of "virtual witnessing", a style of writing experiments he attributed to Robert Boyle in an article in 1984, brings us closer to a correct conception of the function of language. What makes the reader a virtual witness is that he sees the experiment carried out in front of him in language, he does not see language reaching out to some other realm. (Alternatively, think of language as a calculus with which one does science.) But even though an update on fundamentals would have been welcome, Bazerman does not actually need an attitude on language-world dualism since he intends to argue a very different case, that throughout history scientists have spent their time grappling with the conventions of language and communication. His views emerge piecemeal in a series of well-researched and interesting case studies. In the one on Newton he argues that the Opticks "is far from the spontaneous workings of the creative mind. The book is a hard-won literary achievement forged through some trying literary wars" (p. 124). A large part of what constitutes winning a literary war in science according to Bazerman (and what constituted Boyle's genius according to Shapin), is convincing the relevant community that there is just one language in which the common problems of the subject can be expressed properly, and that it should be preferred over others in use. The one-against-many literary wars are the exception, of course, but through them one may come to see that gradual changes to the literary devices of science, what Bazerman calls "the development of linguistic means for statements that move toward relatively stable meaning and assent among people sharing wide numbers of social variables (even while sharing participation in scientific activity)" (p. 13), betray a constant struggle in the history of science at a level which is easy to miss. For example, his chapter on 'The changing account of scientific doings in the Philosophical Transactions ofthe Royal Society, 1665-1800' should be understood as an analysis at that level. Professor Bazerman, who teaches composition at the City University ofNew York, is a fine writer although his arguments would have benefited had he kept the book to two-thirds of its length. It may still lead a few to revise their ideas about what needs explaining and what constitutes explanation in the history of science.