In 2008, Grove proposed that there is a binary distinction between taxonic and nontaxonic latent variables, although causal structures which do not produce sharp category boundaries have long been recognized by Meehl and others. I argue that this position is incoherent. As soon as one admits causal mechanisms which do not yield dichotomous outcomes and acknowledges data structures which have intermediate category membership, the taxon concept becomes nontaxonic. This position does not imply that determining something to be taxonic is arbitrary but suggests that one should not make strong inferences about causation (e.g., specific etiology) on the basis of taxometric research findings.