I argue in this essay that the popular “descriptive” approach to jurisprudence can be modeled after attempts at explaining natural phenomena by scientists. I present four assumptions that are underlying this approach to jurisprudence, which are similar to those of natural scientists. I then argue, however, that in the case of jurisprudence (and unlike the natural sciences) these assumptions contradict each other. After presenting my case I respond to several potential replies to my argument. If my arguments are correct, this shows that jurisprudential descriptivism is not just, as some have argued, unimportant, but rather that it is impossible. The suggests that those who claim to offer an account of the “nature” of law are in fact doing something else.