In this paper I shall argue that human, as well as animal, behaviour requires an evolutionary explanation, and that this sort of explanation is predicated on a different explanatory logic from that of functional decomposition, conceptual analysis and formal task-description. Evolutionary, or genetic ([Nagel, 1961], or historical [Sloman, 1978] explanation relates what a creature can do to how it is done by examining the stages by which its behaviour emerged through evolution and learning in specific environmental circumstances. Such a description depends on actual occurrences. Its explanatory kinds cannot be discovered by taking a 'design stance'. Functional decomposition and natural selection do not mix, although they have often been combined in quasi-teleological explanations. I examine the logic of genetic and selective processes and argue that decomposition by activity provides our only explanatory access to the level of neurophysiology, making it the 'natural' explanatory level between folk-psychology and descriptions in terms of neurons and action potentials.
This paper is not available online