SEWALL WRIGHT AND EVOLUTIONARY BIOLOGY
William B. Provine. University of Chicago Press, Chicago, 1986. 561 pp., illus. $30.
Sewall Wright has long despaired at being associated with an overly narrow view of the evolutionary process, one that attributes extreme importance to chance factors and neglects the role of natural se lection. His archrival, R.A. Fisher, contributed in no small way to this caricature. Even among those who get Wright's views on evolution straight there are those who minimize the breadth and depth of his researches. This fine biography of Wright goes a long way toward correcting this tendency.
Wright, who is now 96 years old, was much more than just an evolutionary biologist. He was a student of the physiology of gene action, and Provine brings that work to life. He devotes a chapter to another considerable chunk of Wright's work, the development and application of the method of path coefficients, a form of statistical inference.
Although Provine discusses other contributions, the focus of the book is Wright's work in evolutionary biology, the part of his re search that is best known but not necessarily (until now) well understood. In this respect especially, the book is a truly great achievement. More than just an analysis of Wright's evolutionary views, it is also a revealing and comprehensive history of 20th century evolutionary biology and a major contribution to the conceptual foundations of that discipline.
Appropriately, Provine's treatment centers around Wright's views on the role and importance of chance effects in evolution—that is, on the significance of evolution by random drift of gene frequencies. A detailed discussion of Wright's controversies with Fisher on this subject serves as an excellent introduction to the subject. As Provine makes clear, Fisher and Wright disagreed not about the formulation of the so-called synthetic theory of evolution, to which both contributed, but about the significance of chance effects in evolution and the actual sizes of populations in nature.
Wright's views concerning the role and importance of random drift in evolution are more elaborate than just notions about the prevalence of small population sizes. They are embodied in his complex shifting-balance model of evolutionary change, to whose history and conceptual foundations Provine devotes considerable space.
Particularly noteworthy is Provine's analysis of the changing—sometimes misleading, sometimes incoherent—interpretations that have been made of the adaptive landscape diagrams Wright originally drew to help illustrate the model. Provine's critical assessment of adaptive landscapes is not meant to bear on the shifting-balance model.
The model itself is an evolutionary scenario—not a claim about the way evolution always occurs, but an elaborate description of one way in which it may (and presumably sometimes does) occur. In this scenario, a species subdivided into small populations evolves adaptively in response to a particular shifting combination of natural se lection and random drift events. Not every step in the evolution of the species is adaptive, though the overall changes are.
The historical backdrop against which Provine discusses the shifting-balance model is a trend in evolutionary biology dubbed “the hardening of the evolutionary synthesis” by Stephen Jay Gould. According to Gould, the architects of the synthetic theory were originally pluralistic with regard to the importances of various agents of evolution, but during the '40s and '50s they increasingly emphasized natural selection to the exclusion of other agents, like random drift. Gould has documented this change in the writings of G.G. Simpson and Theodosius Dobzhansky. Provine here documents a change in the same direction, if not to the same extent, in the case of Wright.
Provine blames Wright to some extent for the fact that Wright is so often associated with a much narrower view of the importance of random drift than that suggested by the shifting-balance model itself. Provine argues that, considering Wright's early emphasis on the prevalence of nonadaptive species differences, it is understandable that Fisher, among others, should have thought that Wright attributed much more importance to random drift than to natural selection.
Provine includes a superb discussion of the collaboration of Wright and Dobzhansky that paid off for both (though perhaps more so for Dobzhansky) in the series of articles titled “Genetics of Natural Populations.” In many of these articles, Dobzhansky put Wright's views on evolution, and Wright's expertise in experimental design, to use in analyzing and understanding natural population data. The collaboration paid off for Wright, according to Provine, because Dobzhansky popularized his evolutionary views. It is puzzling that, although Dobzhansky made quite clear the importance of natural selection in the shifting-balance model, Wright continued to be associated with an extreme view about the importance of random drift.
Provine has written a challenging and provocative book about a challenging and provocative scientist. Wright himself had a lot to do with the book. It is based on more than a hundred hours of interviews. Moreover, Provine made parts of the manuscript available to Wright for comments and criticisms. According to Provine, unnamed historians of science have raised doubts about the role that he allowed Wright to play in the writing of the book, questioning in particular the author's ability to maintain objectivity under such circumstances. Provine did the historians in question a great favor by not naming them. It is hard to imagine that anyone would avoid the kind of interaction that Provine has so clearly enjoyed and from which he and the rest of us have so clearly benefited.
Beatty is an associate professor in the Department of Ecology and Behavioral Biology and a member of the Program in History of Science and Technology, University of Minnesota, Minneapolis, 55455.
Scientists have identified a species of cockroach never before seen in the U.S., which was spotted last summer on Manhattan’s West Side.