In a typical introductory ethics course, moral relativism is presented only to be exposed as contradictory or confused. Ethical relativists are accused of inconsistency when they attempt to use relativism as a reason for advocating universal tolerance. They are said to be confused when they think that cultural diversity indicates genuine moral disagreement, or that moral disagreement is itself proof that truth is relative. It is safe to say that ethical relativism in its "vulgar and unregenerate forms" has little or no philosophical credibility. 1
Recently, however, Gilbert Harman has defended a version of moral relativism which is not vulnerable to these criticisms. 2 His thesis is that there is a class of moral judgments which he terms "inner judgments". Inner judgments, Harman claims, make sense only in relation to an implicit agreement among a group of people about their relations with one another. Harman tells us that his thesis about inner judgments is a "soberly logical thesis", 3 and he defends it in two ways: first, he provides linguistic evidence for it, and secondly, he thinks that it follows from a Humean conception of motivation, which he assumes. I hope to show that the argument from the linguistic evidence fails, that the Humean view does not have the implications Harman thinks it has, and that therefore we have no reason to accept Harman's version of relativism and, in particular, the specific conclusions he believes it warrants.
Harman's thesis is that inner judgments make sense only in relation to an implicit agreement among a group of people. Inner judgments include the judgment that someone ought or ought not to do something or that it was right or wrong of him to do it. They do not include judgments in which we say, for example, that someone is evil, a traitor or an enemy, or in which we say that an institution is unjust. Harman's claim is that inner judgments