Does a type specimen necessarily or contingently belong to its species?

In a recent article, Alex Levine raises a paradox. It appears that, given some relatively uncontroversial premises about how a species term comes to refer to its species, a type specimen belongs necessarily and contingently to its species. According to Levine, this problem arises if species are individuals rather than natural kinds. I argue that the problem can be generalized: the problem also arises if species are kinds and type specimens are paradigmatic members used to baptize names for species. Indeed, the same problem arises with respect to kinds like gold and the samples used to ground names for them. After arguing that the paradox arises whether or not species are individuals, I attempt to show how the paradox can be resolved. Levine’s argument that a type specimen belongs necessarily to its species is specious. The appeal of the argument stems from a failure to distinguish between two different modal statements concerning type specimens, one de dicto and the other de re. Type specimens belong contingently to their respective species. Even so, they can be known a priori to belong to them: hence, that a particular type specimen belongs to its species is an example of contingent a priori knowledge. In an interesting recent article, Alex Levine (2001: 332-6) raises a paradox. It seems that, given some relatively uncontroversial premises about how a species term comes to refer to its species, a type specimen belongs necessarily and contingently to its species. According to Levine, this problem arises if species are individuals like Mt. Rushmore rather than kinds like gold. I will argue that the problem can be generalized: the problem also arises if species are kinds and type specimens are paradigmatic members used to baptize names for species. Indeed, the same problem arises with respect to kinds like gold and the samples used to ground names for them. These samples serve the same role that type specimens do for species, at least if standard views about kind term reference are right. Because the problem is more general than Levine indicates, it is also more threatening. It calls for attention. Fortunately, I will argue, the paradox can be resolved. Levine’s argument that a type specimen belongs necessarily to its species is specious. The appeal of the argument stems from a failure to distinguish between two different modal claims concerning type specimens. Type specimens belong contingently to their respective species, though they can be known a priori to belong to them.