Moral realism can be defined as "a claim that moral judgments are independent of certain mental states. ''1 Moral realism is usually linked to cognitivism. "For a sentence in ethics to be true is ... for that sentence to correspond to the objective 'moral facts, '''2 where a moral fact is, at the least, derived from the belief that the injunctions, prohibitions, and values incorporated in our moral code have a standing, an authority, independent of our society's requiring adherence to them. 3 More strongly, moral realism is sometimes identified with ethical absolutism, that is, "that there are some moral principles which are true for or binding on or valid for all men. ''4
Moral realism, particularly in the form of moral absolutism, has not always enjoyed widespread popularity. Interestingly, recently there have been several attempts to make a case for moral realism from a less absolutist perspective. One such attempt draws on a conventionalist interpretation of the later Wittgenstein, where "consensus [is] a necessary as well as perhaps sufficient foundation for the notion of correctness. ''5 What is real and true, according to this reading of Wittgenstein, is what is stipulatively agreed upon by a community to be true or real. Similarly, it is argued, through consensus a community and its social practices set up standards for moral judgments, and these standards, being independent of individual moral judgments, become regarded as objective and real for that community. 6
In contrast to this kind of approach I shall argue that we cannot defend a viable form of moral realism from the perspective of linguistic conventionalism. This is because assertability conditions cannot replace truth conditions as valid criteria for the truth of a moral judgment, and because * Presidential Address, American Society for Value Inquiry, 1987.