Social kinds and conceptual change: A reply to Haslanger

Sally Haslanger (2006) is concerned with the debate between so-called social constructionists and error theorists about a given category, such as race or gender. For example, social constructionists about race claim that race is socially constructed, that is, the kind or property that unifies all instances of the category is a social feature (not a natural or physical feature, as naturalists about race would hold). On the other hand, error theorists about race claim that the term ‘race’ is an empty term, that is, nothing belongs to this category, since the conditions that something should satisfy in order to fall under ‘race’ are not satisfied by anything. What kind of evidence could we use in order to support one or another of these theories? It seems that this debate is in part semantic: what makes the case that a category is an empty one (and therefore error theory about it holds), as opposed to it being socially constructed, has to do with the meaning of the corresponding expression. In particular, in the case of race, some people have argued that our concept RACE is such that something will fall under it only if it is a natural property that can explain certain features. Arguably, there are no natural properties of human beings that can do the explanatory work that races are supposed to do, and therefore, error theorists have concluded that ‘race’ is an empty term, that is, there are no races (Appiah (1996)). (Some theorists have introduced new terms for a new property that is very similar to that of race and can do part of the explanatory work that races were supposed to do, but it does not have to satisfy all the conditions that races are supposed to satisfy. For instance, Appiah (1996) has introduced the notion of ‘racial identity’ to that effect.) These considerations suggest that if we want to find out whether a certain category is socially constructed, or whether an error theory about it is correct, we have to engage in....