Published online by Cambridge University Press: 28 February 2022
Since the mid-1970s, when philosophical scrutiny of biology began to be focused on evolutionary theory, the center of philosophical attention in that field has been what is called the “units of selection” controversy. In particular, two problems have attracted philosophical and, on occasion, biological attention: (i) should natural selection be regarded as capable of operating on groups of individuals as distinct units, or should such a process only be regarded as a special type of selection on the constituent individuals of such groups even when every member of the group, by virtue of its membership of that group, is affected by selection in an identical manner?; and (ii) similarly, should selection be regarded as “ultimately” operating only on individual alleles since, even when the genotype seems to feel the action of selection more directly, any change in the genotypic composition of a population also consists of a change in its allelic composition?
The influence of Richard Lewontin on this paper should be obvious. My skepticism about the value of variance-based analyses in theoretical population genetics arose from conversation with John Maynard Smith. Discussions with J. F. Crow. T. Nagylaki and, in particular, W.C. Wimsatt have been useful. This analysis is part of a larger project of attempting to give a comprehensive account of the structure of contemporary evolutionary theory, with full attention to the technical details and complexities of the models and strategies employed in that field. I apologize for several references to that incomplete work but these are needed to situate the arguments being offered here in the context of past work on the units of selection. Please note that the basic—and technical—argument of this paper is fully elaborated here and does not rely on that work.