Abstract

Attacks explanatory deductivism, the view that all genuine explanations have the form of a correct deductive argument. The view has strong intuitive appeal to many philosophers. The author offers a defense against the claim that there are no statistical explanations of particular facts. In other words, he shows that statistical explanations of statistical generalizations – in the form that Hempel designated as the deductive–statistical variety – are not the only correct forms of statistical explanation. He exposes a glaring conflict between the deductive–nomological model of explanation and basic causal considerations relevant to explanation.