Thursday, March 18, 1999
at 4:00 PM
Francis Scott Key Building, Room 1117

Sober, Lewontin and Wimsatt argue against genic selectionism by noting that the contribution an individual gene makes to fitness depends upon itsgenetic context and then arguing that selection can only occur for properties which exhibit a context-independent contribution to fitness. This principle follows from a general principle central to most accounts of probabilistic causality, called a contextual unanimity condition. According to this condition, a property can only be a cause of some effect if it exhibits a context-independent contribution to the probability of that effect. In this talk I shall argue that Sober has advocated the right position, but for the wrong reasons. Units of selection cannot be defined in terms of context-independent contributions to fitness because causes cannot be defined in terms of context-independent increases in the probability of their effects. To determine at what level selection is occurring, one must, instead of looking at the conditional probabilities, examine the biological mechanisms which generate these probabilities. What the unit of selection is in a particular case will depend upon the structure of the mechanism of phenotypic expression. A proper understanding of these mechanisms will explain both how genes and gene complexes can give rise to probabilistic relationships that approximate contextually unanimous causal contributions and the conditions under which these approximations fail.

Stuart Glennan received his B.A. in mathematics and philosophy from Yale University in 1985 and his Ph.D. in philosophy from the University of Chicago in 1992. Currently, he is an associate professor of philosophy at Butler University. Glennan's research has chiefly been concerned with developing an account of causation which, in addition to addressing traditional metaphysical and epistemological problems, is informed by and illuminates actual patterns of scientific inference and explanation. His articles on causality, chance and related issues have appeared in a variety of journals, including Philosophy of Science, Nous and Erkenntnis. He is also author of a forthcoming textbook in symbolic logic, Elements of Deductive Inference.

Spring 1999