Michael Friedman's recent book represents a valuable contribution to the philosophical understanding of space-time theories. It incorporates the most recent "technical" foundational results, relating these results to more general, non-technical issues. The aims of the book are two-fold. Friedman first focuses on clarifying the content and relations among specific space-time theories. This is certainly a task worth doing. Since the birth of the theories of relativity, there have been many confusions and misconceptions about the concepts of relativity embodied in these theories and their connection with an empiricist philosophy of science. The second goal of Friedman's book is to examine the empiricist arguments of the relationalist and the conventionalist against the independent ontological status of space-time geometry. In the course of this examination, Friedman identifies an important methodological principle, "the principle of unifying power," which, he claims, restricts the proper scope of empiricist arguments and provides the needed defense of realism.
The conceptual framework developed by Friedman enables him to successfully meet the goals described above. If the book has any weaknesses, they are of two types. In some cases, central concepts (e.g., 'indistinguishability group') require more attention to detail in order to eliminate problems of interpretation. In other areas, one wishes for more discussion of the general significance of the concepts (e.g., the 'symmetry group' of a theory), particularly with respect to areas other than the space-time theories discussed. These problems, however, do not detract from the value of the book. In the following discussion, I will sketch and evaluate what I consider to be Friedman's central claims about issues of current interest.
The book begins with an introductory chapter describing the motivating questions. Historically, logical positivism began as a Neo-