Individuality and Selection

Evolutionary theory is currently undergoing a period of rapid development, but in the process several problems have cropped up that are proving to be infuriatingly difficult to resolve-e.g. the presence of so much genetic heterogeneity in natural populations, the prevalence of sexual forms of reproduction in the face of an apparent 50% cost of meiosis, and the difficulty of explaining how selection can operate at higher levels of orga­ nization. In their most recent publications, the leading theoretical biologists of our day seem to have all but given up hope of making further progress (28, 34, 58). Comparable stalemates in the history of science have tended to result from everyone concerned taking for granted something so funda­ mental that no one in their right mind would question it. In the present case, I think two assumptions are at fault: (a) the view that genes and organisms are "individuals" while popUlations and species are "classes," and (b) our traditional way of organizing phenomena into a hierarchy of genes, cells, organisms, kinship groups, populations, species, and ecosystems or commu­ nities. In his classic paper on units of selection, Lewontin (27) accepts the traditional organizational hierarchy and asks at what level selection can occur. His answer is that it takes place primarily at the lower levels and becomes rarer and more problematic at the higher levels. However, some­ thing peculiar happens as we follow Lewontin up the traditional hierarchy: We pass from such commonsense individuals as genes and organisms, through such borderline cases as colonies, to such commonsense groups as populations and species. It would be truly amazing if a single process could operate on entities as different as individuals and groups. At least some of