In recent years it has become more and more difficult to distinguish between metaethical cognitivism and non-cognitivism. (Dreier 2004; van Roojen 2004) The distinction most often used to separate the two camps, until recently at least, was their position on the meaningfulness of moral statements and the applicability of truth-predicates to those statements. Cognitivists are said to hold that moral claims are both meaningful and truth-apt, and the early noncognitivists, such as Ayer and Stevenson, certainly denied this. However, in recent years, this distinction has broken down. For example, proponents of the minimalist theory of truth hold that moral claims need not express beliefs in order to be (minimally) truth-apt, and yet some of these proponents still reject the traditional cognitivist analysis of moral language and thought. (Horwich 1993; Blackburn 1984 and 1998)1 Thus, minimalism complicates attempts to draw a clear line between the two views by looking at truth-aptness. Further, the quasi-realist project embraced by the leading non-cognitivists, Allan Gibbard and Simon Blackburn, attempts to show how moral claims can be both meaningful and true. Gibbard goes so far as to claim that we can speak of natural moral properties. (Gibbard 2006) However, Gibbard and Blackburn also claim that even if we can speak of moral truth and meaningful moral sentences, important differences between their position and the cognitivist’s remain. So there must be another way of distinguishing between cognitivism and noncognitivism.