Today the Ethics Group had a guest speaker, Rob Shaver from the University of Manitoba, discussing the non-naturalist position in early twentieth century analytic ethics, so I dropped in to listen. It was a very good paper; although it took a historical approach to the subject, so I might be biased. Essentially the argument was that two common objections against non-naturalism, the extravagance objection and the cheapness objection, aren't really much for a non-naturalist to worry about.
Non-naturalism, which in the twentieth century is associated with the names Moore, Broad, and Ewing, is the position that ethical or moral discourse cannot be reduced to purely descriptive discourse; and a great deal of Shaver's argument was concerned with emphasizing that this doesn't require non-naturalists to believe that there are any 'moral properties' in objects themselves - all it requires them to say is that there are two ways of describing objects, a non-naturalist way (involving moral imperatives, values, what have you) and a naturalist way, and the former can't be reduced to the latter. This eliminates the extravagance objection, which holds that non-naturalists are committed to a weird moral ontology (moral objects, moral properties) in addition to the ordinary ontology (the sort of objects described by science and common sense), because non-naturalists actually aren't committed to such a thing. Some of them (Moore) perhaps flirted with it, but in Moore's case the appearance of it may just be that he didn't properly distinguish concepts and properties, and other non-naturalists pretty much deny it. Likewise, the cheapness objection, which holds that non-naturalism doesn't really contribute much, doesn't really hinge on anything particularly non-naturalistic (indeed, the same objection can be, and has been, leveled against most meta-ethical positions, including the naturalistic ones).
As I said, a good paper.