Thomson against Moral Explanations

Abstract

One prominent feature of this debate between Harman and Thomson may surprise readers who have followed recent discussions of moral knowledge.' Harman is well-known for having argued that supposed moral facts never seem relevant to the explanation of nonmoral facts.2 He does not repeat that argument here, however; when he mentions moral explanations, he concedes that many appear perfectly reasonable, insisting only that they are of little help in settling deep moral disagreements (168-70). Thomson, on the other hand, although aiming to defend moral knowledge, takes up Harman's earlier skeptical argument, not to challenge it, but in order to agree with its (proximate) conclusion. She argues that moral facts never explain nonmoral ones, and that the moral skeptic must therefore be met on other grounds. Since she also thinks such grounds are available, her conclusion is that moral explanations contribute nothing to moral knowledge, but that that knowledge is secure without them. As a friend of moral explanations who believes that they matter to our moral knowledge, I want to address Thomson's argument. She raises important challenges clearly and forcefully. I offer three replies, the first two including a limited measure of agreement. First, I agree that if the moral explanations Thomson considers were defectivefor the reason she offers, their defectiveness would not threaten our moral knowledge. What limits my agreement, however, is my thinking that what look like moral explanations-even if Thomson is right that they need somehow to be reconstrued-would still play a central role in our acquiring moral knowledge. Second, I believe that Thomson correctly identifies a reason why moral explanations (reconstrued or not) cannot solve the problem of moral skepticism as she presents it. But I believe that this shows a difficulty, not with moral explanations, but with her way of presenting the skeptical problem. Third, I shall argue that moral