Moral realism holds that our moral judgments are "about the world" in much the way our other judgments are. Antirealists deny this, holding that in making moral judgments we are projecting our own attitudes onto things, or expressing these attitudes, or the like. As usually stated the issue between these two views is thus a linguistic one, about the meaning of moral judgments. It is not the parallel metaphysical or ontological issue of whether there are any "moral facts" or "objective values." Thus, one might, like J. L. Mackie, hold that while moral judgments are indeed intended to be "about the world" (as realism holds), there are in fact no objective values and hence anyone who makes moral judgments is committed to a claim which is false.' In the realism/antirealism debate, realism seems to have a strong initial advantage since, as Simon Blackburn says, "Nobody denies that the surface phenomena of language-the fact that we use moral predicates, and apply truth or falsity to the judgments we make when we use them-pose a problem for projectivism."2 The task for an antirealist then will be to push through what Blackburn calls "the quasi-realist project," which is to show that the features of our moral language which appear to commit us to the idea that there are objective moral truths and the like are really neutral as between realism and antirealism, that is, that these features can be explained adequately by the view that in making moral judgments we merely express our attitudes toward things. One such feature of moral language which has been thought to commit us to realism is the use of modus ponens in reasoning about moral issues. The problem for antirealists is that there are obvious examples of valid reasoning with sentences using moral terms (i.e., examples which all sides agree are valid) but which, on the antirealist theory, do not have or at least do not seem to have the correct logical form to allow us to say that it is, for example, modus ponens, which justifies them, that is,