Abstract
In this paper a conception of rationality is developed which bears on three important issues in the social sciences -- the status of the principle of rationality, the criteria for rational actions, and the nature of rational explanations. It is argued that the principle of rationality should be interpreted as a methodological principle and is valuable only inasmuch as it leads to true hypotheses about human action. Definitions of rational beliefs, rational means, and rational ends are provided. These definitions provide criteria for rational actions which better accord with our intuitive judgments about what should count as rational action than the criteria given in traditional accounts. Finally, the nature of rational explanations is specified and arguments for their usefulness and respectability presented although these explanations do not necessarily fit the standard covering law model.