Abstract

Both alternatives are subjects to serious difficulties. The deductive-nomological model traditionally has had to argue that it can, without violating the concept of a reason for action, incorporate the normative relations of reason and actions into the laws on which it relies. The rational model, on the other hand, traditionally has faced the charge that it fails to explain, in that it fails to show that the action was to be expected. I argue, however, that the most serious charge against each has the same origin: each in fact proves inadequate when faced with a case of inappropriate or irrational action; i.e., action performed contrary to the agent's perceived better reason. ;In light of this problem, a third model is considered: probabilistic explanation. This model has seldom been applied to explanations of action. While it is not argued that the several varieties of this model are adequate in general , I do find that some varieties of probabilistic explanation possess several merits when taken as analyses of explanation by reasons. In particular, they appear able to overcome the problem which most severely afflicts the deductive-nomological and rational analyses. The major difficulty of probabilistic explanation is this: despite its philosophical merits, probabilistic explanation appears to resemble actual explanations of action less than either deductive-nomological or rational analyses of explanation. This, however, is a problem facing all philosophical analyses of explanations by reasons. ;The structure of explanations of human action which explain by citing an agent's reasons for acting has been a subject of considerable debate in the philosophical literature for a number of years. The debate has generally focused on the suitability of two alternative models of explanation. On the deductive-nomological model, reasons for acting are held to be lawfully tied to actions, such that given a statement of the law, and a statement describing the agent's situation , we supposedly may deduce a description of the action, thereby explaining why it occurred. According to the second model, sometimes called rational explanation, to explain by reasons is not to show the action deducible in this way, but is rather to exhibit the rationale or appropriateness of the action. That is, one explains the action by showing it to be justified in light of the agent's reasons for acting