One last bit on Xunzi and language. I think Xunzi holds to neither a realist nor a nominalist metaphysics—I don’t see that this question would have occurred to him either. Rather his view seems really rather straightforward. We should draw pragmatically valuable distinctions between things according to their type (lei), and we do this by looking at things to discover genuine similarities. This much can be said without getting into metaphysics at all. To borrow an example, we can say why it makes little sense to distinguish in everyday language each of the some 100 species of mosquitoes without saying anything about the reality of species. Whether species really exist or are “socially constructed” our eyes tell us mosquitoes are more like each other than other insects and that the differences between two kinds of mosquitoes are unimportant if we’re interested in keeping track of what bites and what doesn’t and the like. To put the point another way, a realist can acknowledge that nature has many joints and that pragmatic considerations are perfectly fine in deciding how fine grained we want to be in our distinctions. Nor must a realist insist that nature’s joints always trump human lines. Dogs may be dogs to Mother Nature, but that hardly means we err in distinguishing hunting dogs from lap dogs. So while I think someone like Kurtis Hagen is right to push a “contructivist” reading of Xunzi, he need not also argue Xunzi is then not a metaphysical realist. Better to say Xunzi had no metaphysics.