ABSTRACT
Our initial survey of the problems of induction and explanation is now complete. We have considered some of the forms these problems take, some of the reasons they are so difficult to solve, and some of the weaknesses that various attempts to solve them suffer. In the last chapter, I also offered something more constructive, by attempting an improved version of the causal model of explanation. Up to now, however, we have treated inference and explanation in near mutual isolation, a separation that reflects most of the literature on these subjects. Although the discussion of inference in chapter 1 construed the task of describing our practices as itself an explanatory inquiry, the attempt to specify the black box mechanism that takes us from evidence to inference and so explains why we make the inferences we do, none of the models of inference we considered explicitly invoked explanatory relations between evidence and conclusion. Similarly, in the discussion of explanation in chapters 2 and 3, inferential considerations played a role in only one of the models, the reason model. That model uses an inferential notion to account for explanation, by claiming that we explain a phenomenon by giving some reason to believe that the phenomenon occurs, and it was found to be unacceptable for inferential reasons, since it does not allow for self-evidencing explanations, such as the explanation of the tracks in the snow or of the red-shift of the galaxy, virtuous explanations where the phenomenon that is explained nevertheless provides an essential part of the reason for believing that the explanation is correct.
