McNaughton and Rawling's anti-reductionist intentions are to be welcomed, but are not well served by their continuing adherence to a neo-Humean notion of the 'descriptive'. Their too-willing acceptance of this notion is reflected in a denial of appropriate dialectical weight to considerations about the way 'pattern' disappears from the domain of value when we try to characterize the constituent features of the latter in non-evaluative terms. The need for a satisfactory account of the immanence of value in nature is real (...) enough, but such an account can be given without recourse to the problematic idea of a class of subvenient 'descriptive' properties. Likewise, McNaughton and Rawling's recognition of 'normative facts' is to be endorsed, but such facts are a more widespread and everyday phenomenon than they would have us think. (shrink)