The history of the philosophy of science is now getting more attention than it previously received, and rightly so. One encouraging sign is the burgeoning society, HOPOS, which is now entering its second decade. This is an organization devoted to exploring the routes that have taken the field of the philosophy of science—in all its diversity—to where it is today. The present work by an American now teaching at the University of Montpelier, which traces the origins of philosophy of science in France and its subsequent influence, is a worthy contribution to this effort.

Since I have been told that Brenner's title is a bit off-putting, I begin by saying what he does not try to do in his book. This is not a chauvinist tract. Brenner does not seek to show, à la manière d'un chauvin, that all of the salient features of contemporary philosophy of science have French roots. Rather, he argues that several lines of French thought at the end of nineteenth century and the beginning of the twentieth were influential in helping to define philosophy of science as a discipline and in carefully addressing, as well as formulating, some of its central questions—questions that are still being debated a century later. What he tries to do, in short, is to describe some of the ways that French scientists and philosophers of the time viewed the nature and methods of science, why their reflections were influential in their own time—both in France and elsewhere—and why they continue to exert their influence on contemporary philosophy of science. In this, I think he succeeds.

Brenner's treatment of these matters is wide-ranging, learned, and often very subtle. In a brief review, I can mention only some of the claims that he defends. To begin, he thinks that the end of the nineteenth century and the beginning of the twentieth was an important formative period in philosophy of science, in France and elsewhere, this because science was changing so radically (4). Here he is surely right. He is also right to argue that some authors were more receptive to the scientific novelty than others—Henri Poincaré being among the former and Pierre Duhem among the latter. Poincaré was initially attracted to Duhem's énergétique—a phenomenological theory based on continuous magnitudes—but he later embraced the discontinuities and causal realism implied by the electron and early quantum theories (80–81).

Second, Brenner thinks that important elements in our current understanding of the field of philosophy of science were provided by French thinkers, notably by Poincaré and Duhem, whom he regards, rightly, as having exerted the most influence on German- and English-speaking readers (67). He focuses on two of these elements in his book: the way (and extent to which) the doing of science is a matter of convention, of “choix libre,” and the question of experimental method—how it goes and what it shows. Brenner does not think that conventionalism, the emergence of which he traces to independent works by Poincaré and Duhem in the early 1890's (51), has ever received the attention it deserves. Members of the Vienna circle—Schlick and Carnap, for example—have recently been the subjects of detailed studies, and there are also recent reassessments of logical positivism; but “We have to admit,” he writes, “that conventionalism has not been the object of a comparable study” (9). This is a serious lacuna, Brenner argues, since “With [the advent of] conventionalism we have a decisive stage in the formation of an autonomous discipline [i.e., philosophy of science]. Henceforth, epistemology [read philosophy of science] is grounded in the rational analysis of the structure of scientific knowledge and in the historical study of its process of development” (36). A lot of Brenner's book is thus devoted to what conventionalism meant to its originators, how its various meanings likely got transmitted, via Gaston Milhaud, Éduard Le Roy, and Abel Rey, for example, and how it continues to influence contemporary philosophy of science. One nice example is Brenner's discussion of Rudolf Carnap (128–137), which argues that—despite what he said—Carnap was more indebted to Poincaré than to Russell (137).

Third, Brenner argues—convincingly, I think—that the origin of the so-called “Duhem-Quine” thesis is a complex matter. Poincaré originally argued that a specific experiment, this bearing on two competing theories in optics, was not crucial; but it was Duhem who generalized Poincaré's conclusion and elevated it to the methodological position we now call “holism” with regard to experimental testing: “An hypothesis does not face the tribunal of experience by itself, but only in conjunction with many other hypotheses.” Indeed, as he shows, Poincaré continued to speak of “decisive” experiments and urged scientists to make hypotheses one at a time, so that they will know just what has been refuted by an experimental result (55–57).

Fourth, this is because Poincaré and Duhem were engaged in a debate, mostly indirect, about the nature of science and its proper methodology. Brenner argues that Poincaré was an inductivist who based scientific practice on probabilities, but that Duhem was a deductivist who took holism seriously (68, 72, 93). So, while both of them embraced the idea that hypotheses in geometry and physics are conventions, conventionalism developed along different paths—one originated by Poincaré and the other by Duhem (86–95). Poincaré's version probably had the most influence on the Vienna Circle—on Philip Frank, Hans Hahn, Otto Neurath, and later Carnap—with its avowed interest in formulating an inductive logic of confirmation (108, 116–122) In fact, says Brenner, “We can regard Reichenbach and Carnap as the inheritors of Poincaré” (88). The more explicitly historical approach of Duhem is continued by most French thinkers—Gaston Bachelard and Émile Meyerson, for example—but, Brenner argues, this was not revived (or, better, reinvented) in Anglo-American philosophy of science until much later—by, among others, Thomas Kuhn (3–4, 167, 173, 178–179). An important exception, which he also discusses (109–116), is the complex interaction between Duhem and Ernst Mach.

Two comments: Brenner thinks that contemporary philosophy of science only came into it own when it showed, via Poincaré, Duhem, and others, that it was an autonomous discipline. This occurred gradually in France at the end of the nineteenth century, since many philosophers then still thought of philosophy of science as only an adjunct to a metaphysics that could reconcile science with human freedom. One finds that way of thinking in, for example, the writings of Jules Lachelier and Émile Boutroux. Most of the emerging French philosophy of science, by contrast, generally eschewed metaphysics and advocated a “positive” method in science. Duhem certainly did, and so did Poincaré. But this did not mean that French conventionalists were opposed to metaphysics. Far from it. Brenner argues that they were opposed to scientism, the idea that science could explain everything (125), and not to the idea that science, and so philosophy of science, had to be reconciled with more basic human concerns, the concern to preserve freedom being the most basic. This is one of the reasons that conventionalism was attractive: it left room for freedom. Indeed, it stressed it. As Brenner writes, “The emphasis [in conventionalism] is placed on the irreducible role of decision in science” (205). Second, by the end of the 19^th century, French philosophy had already departed from the old positivism of Auguste Comte; but, at the same time, most of it still held fast to the Comtian idea that our knowledge—in particular, our scientific knowledge—is historically conditioned, so that philosophy of science, rightly construed, is therefore an historically based effort to understand both the nature and the structure of scientific knowing. A century later, this might appear obvious; but at the time French thinkers who promoted this idea were usually swimming against the tide, and would be for some time. To illustrate how complex and uncertain the situation was, Brenner discusses Karl Popper, an Austrian, who was indebted to both Poincaré and Duhem (137–150).

Finally, Brenner offers a useful analysis of the virtues and limitations of both logically oriented and historically oriented conceptions of the philosophy of science by tracing the impact of Poincaré and Duhem on a variety of later thinkers—French, German, Austrian, and American (153–188). While he eventually sides with the “historicists”—history “does not come in simply to complete the logical analysis; the elements of scientific theories are known in their temporal dimension,” so that we may rightly refer to “the irreducible historicity of scientific knowledge” (207)—he does his best to give the “logicists” their due.

I return to my opening remark that Brenner's book is best characterized as a contribution to the history of the philosophy of science. As such, it may not appeal to some readers of this journal. If so, that is unfortunate, since I think that we cannot really understand what we are concerned with now in the philosophy of science without having tried to understand how the discipline has evolved. Brenner provides valuable, if not always uncontroversial, assistance in doing that. Moreover, he does so in a style that is admirably clear and direct. Intentionally so, I would add: He is not an author who thinks that he must be obscure in order to appear profound.