Many of the commentaries cohere around two major points of criticism. The first is that we have omitted discussion of the mechanisms that are assumed to operate at levels 2, 3, and 4. Campbell, Cloak, Dewsbury, Eckberg, Mundinger, Pulliam, Richerson & Boyd, Slobodkin, Simon, Williams, and Wahlsten all make comments that bear on this point. The second point is that we have omitted discussion of the fact that "organisms change the environment by their activities" (Lewontin) and thereby modify the selection pressures that act on them. Apart from Lewontin, this point has been raised in different forms by Barkow, Chase, Hull, Perzigian, Ruyle, and less directly by Slobodkin. Stemming from these two basic omissions there is a third general criticism. Numerous commentators have pointed out that we have failed to discuss sufficiently either the extent or nature of the constraints that may or may not be operating on supraordinate evolutionary processes as a function of their nested relationships to subordinate evolutionary processes within our scheme (Barkow, Campbell, Cloak, Chase, Lenington, Lewontin, Perzigian, Peterson, Pulliam, Richerson & Boyd, Simon, Slobodkin, and Williams). We start by acknowledging these three basic omissions. We agree that they exist and that they are serious.

Classifying is a fundamental operation in the acquisition of knowledge. Taxonomic theory can help students of cognition, evolutionary psychology, ethology, anatomy, and sociobiology to avoid serious mistakes, both practical and theoretical. More positively, it helps in generating hypotheses useful to a wide range of disciplines. Composite wholes, such as species and societies, are “individuals” in the logical sense, and should not be treated as if they were classes. A group of analogous features is a natural kind, but a group of homologous features is not. Imposing hypotheses justified only on the basis of nominalist, realist, phenomenalist, or conceptualist metaphysics upon the neurophysiology of organisms or upon the causes of behavior exemplifies the “psychologist's fallacy” of William James. Levels should be distinguished from their members and from classes of levels, and the ontological status of entities ranked at levels should be made clear. It is important not to confuse such categories as substance and process with one another. Several genetical terms, such as “gene” and chromosome,” are even more equivocal than has been realized. Discussions about units of selection, behavior, and thinking suffer from the ambiguity of a “unit of” an entity. An important source of misunderstanding about natural selection is the habit of treating it as an “agent”: in an important sense, natural selection does not “act” at all.