In their recent “Moral Fictionalism vs. The Rest,” (Australasian J. Phil. 83 (Sept. 2005), 307-330), Daniel Nolan, Greg Restall, and Caroline West do a very illuminating job of characterizing moral fictionalism, laying out its advantages, and opening up some challenges facing the view.
With standard forms of non-realism, moral fictionalism [MF] holds that all of the claims in first-order moral discourse, discourse offering moral judgments and the like (what the authors call “positive” moral claims), are false. First-order moral claims are false because, on the surface anyway, they purport to describe some moral reality with too much ontological baggage. But, for MF, these are useful fictions, so, contrary to the eliminativism of, say, error theory, we should keep realism-committed, erroneous moral claims.
MF is one of those have-your-cake-and-eat-it-too positions: get all the ways in which moral realism can accommodate ordinary moral discourse without committing to realism’s ontology. But MF is not, I think, as advantageous as Nolan, Restall, and West think it is.