Abstract

David Brink, in his book Moral Realism and the Foundations of Ethics, and other writers, have recently offered powerful new arguments for a form of moral realism that sees moral inquiry as being "on par" with scientific inquiry in many important epistemological and metaphysical respects. I call this theory "Assimilative Moral Realism" . AMR is marked by naturalism about moral facts, and by empiricism about moral knowledge. Moral facts are held to be facts about properties that are constituted by, and supervene upon, complexes of ordinary natural and social properties, just as certain scientific facts are facts about the macrophysical properties that are realized by certain microphysical bases. Moral beliefs are held to be justifiable in the way certain scientific beliefs ostensibly are: via certain explanatory inferences and indirect empirical tests of moral principles. Further, traditional semantical objections to naturalistic moral realism can be avoided, it is supposed, because semantical tests for the existence of moral facts can plausibly be rejected, just as they can be in the scientific case. In this thesis, I make a case for rejecting this assimilative theory. I provide a number of technical reasons to suppose that moral principles cannot be tested empirically using the Duhemian method advocated by AMR theorists. I also give a critique of their main arguments for the relevance and indispensability of assumptions about moral facts to explanations of moral judgments and other items, and conclude that their arguments establish neither the explanatory relevance nor necessity of moral assumptions. Finally, in place of the AMR theory of supervening moral properties, I offer an account of moral supervenience claims--an account that is itself compatible with AMR only if there are sound moral explanations of certain nonmoral facts