It is widely agreed that properties play causal roles: they capture the causal powers of things. But do properties have their causal roles essentially? David Lewis did not think so. He adhered to the doctrine of quidditism, namely the doctrine that the identity of properties is primitive and that they can trade their causal roles. Quidditism is controversial. But Lewis did not see why he should want to reject it. He knew that he could avoid quidditism on the cheap by (...) treating individuals and properties alike in rejecting transworld multilocation of properties and endorsing a counterpart theory for properties. But he did not see why he should want to do so. In this article, I argue that Lewis should have wanted to endorse a counterpart theory for properties in order to reject quidditism. My argument concerns resemblance relations among properties. Another constitutive role of properties is that they capture objective resemblances between their instances. The premises of my argument are intuitive claims about resemblances among some properties that Lewis held on Humean grounds. (shrink)