It's often thought that one's metaethical views are more or less independent of one's first-order moral views. Anti-realists can still value other people's welfare, want to prevent the innocent from suffering, etc. But is this enough? An argument I owe to Helen suggests it may not be. Anti-realists can of course have benevolent preferences, and be disposed to blame people who act malevolently, but there is something they're missing: They can't accommodate the moral datum that other people really matter -- matter simpliciter, "from the point of view of the universe", as opposed to merely mattering to them, personally, in light of their contingent preferences. And, the argument goes, there's something morally disreputable about the more superficial attitude to which (consistent) anti-realists are limited. Positive regard should not be something we choose to bestow upon others; it is something they are owed, in light of the kinds of beings they are. The worry is, in other words, that anti-realists must regard their good will as too... optional. They fail to really see people as mattering in themselves.

Does that seem right? Expressivists and "quasi-realists" seem likely to want to deny it, insisting that they can endorse all the same first-order norms as moral realists. "People do matter, and deserve to be regarded with respect," they will say, by which they mean that they endorse norms of treating people as if they matter and deserve to be regarded with respect. But going through the motions is surely not the same as really believing these things, and it seems plausible that morality (or genuine respect for persons) calls for the latter, over and above the former.