American Scientist

The Logic of Medicine. Second Edition. Edmond A. Murphy. xiv + 511 pp. Johns Hopkins University Press, 1997. $25.95.

Edmond Murphy states that his book is not supposed to be a "logic of medicine in the strict sense," but rather an "analysis of ideas in medicine"; it is a work, not in logic, but in "the underpinning of medicine as an academic subject in its own right: its scope, methods and criteria." Nor does it cover medical ethics. Rather it examines an under-worked area of the study of medicine, namely, the fundamental concepts and methods medicine uses to formulate its theories and for evaluating and interpreting its data. The book, then, is something like a treatise in logic and the philosophy of science tailored to peculiarities of medical theory and methodology. This is necessary, in the author's opinion, because if one simply tries to apply other theories to medicine, "there is a serious risk that the development of the [theory of medicine] will eventually be shaped by what the outside forces have to offer rather than by what the basic discipline needs." (This edition has substantial revisions that include a new chapter on modeling, making it 25 percent larger than the first edition, published in 1976.)

The author divides his subject into medical ontology and medical epistemology. The first concentrates on clarifying and explaining important concepts that concern the clinician, such as disease and normality and the methodological notions of accuracy, precision and causation. The second focuses on assessing data, making inferences from them and making clinical judgments concerning disease and prognosis. The author seems to be at his best when discussing statistically definable concepts and illustrating them with medical examples. His discussions of bias and confounding usefully stimulate one to think about various errors scientists and clinicians are prone to and thus promotes critical thinking.

Although I found much interesting and good in this book, I don't think it succeeds as the first steps toward a theory of medicine. The problem is that it consists mainly of discussions of examples, and although these discussions were often interesting and informative, the book usually leaves one to one's own devices to glean the morals. Often I found myself using the chapter and section titles to guess at the point of the author's examples. This left me uncertain as to the theoretical issues he was trying to raise and how he proposed to deal with them. It is hard to assess a theory (or the underpinnings of one) when one cannot identify it with any surety. I also found the book's treatment of matters of logic and philosophy of science to be quite weak.

I would not go so far as to say that logic qua academic subject is part of the theoretical basis for a theory of medicine in the way that, say, it is for biology and chemistry. But as a logician I must protest against the distorted view that the book presents of this field. One would hardly guess from the book's impoverished picture of the subject that it was responsible for the development of computers, and that various parts of logical theory are now central to both computer science and several approaches to computer programming. I think the book would have been of higher quality without this material, and deleting it would not have hindered developing the book's more central ideas.—Michael D. Resnik, Philosophy, University of North Carolina at Chapel Hill