Sociobiology provides a perspective from which much of human behavior seems to make sense, and we expect that its importance will grow in years to come. We believe, however, that its current development is flawed by several widespread misunderstandings. Like other students of nature, sociobiologists are greatly interested in adaptation, particularly as embodied in W.D. Hamilton's theory of inclusive fitness. However, natural selection produces adaptation only when traits' fitnesses do not depend on their frequencies. Yet the fitness effects of social behaviors are generally frequency dependent, and there is, therefore, no reason to expect social behaviors to be adaptive. Neither is there any reason to expect evolution to maximize inclusive fitness. Hamilton's theory has been too widely applied. For similar reasons, evolution can produce mechanisms of cultural transmission that lead to maladaptive behaviors. The circumstances under which adaptive learning mechanisms will evolve are poorly understood. The ethnographic literature is a poor source of data for testing hypotheses because it often fails to distinguish what people do from what they say they do. If, as seems likely, language is used more for manipulation than communication, the ethnographic record will be practically useless. In spite of all this, sociobiological arguments seem to account for a great deal of observed variability in human behavior.