A purely metaphysical formulation of physicalism is surprisingly elusive. One popular slogan is, 'There is nothing over and above the physical'. Problems with this arise on two fronts. First, it is difficult to explain what makes a property 'physical' without appealing to the methodology of physics or to particular ways in which properties are known. This obviously introduces epistemic features into the core of a metaphysical issue. Second, it is difficult to cash out 'over-and-aboveness' in a way that is rigorous, (...) metaphysically pure and extensionally apt for the purposes of the debate. In this paper I will touch on the first problem, but I wish to focus on the second. In particular, I will focus on the claim that supervenience theses cannot define physicalism because they allow classical emergentist dualism through the physicalist door [Horgan 1993; Kim 1998; Wilson 2005]. I will argue that when the relevant supervenience thesis is metaphysical, emergentism is excluded. Against recent arguments to the contrary, I maintain that this is the case even given necessitarianism about natural laws [Wilson 2005]. I will argue that a necessitarian with emergentist sympathies will be forced either into a type of quasi-panpsychism, where our basic physical properties contain the illicit seeds of mentality at their core, or she will be forced to admit that emergence laws are not necessary after all. Either way, the combination of necessitarianism and emergentism does not provide a counterexample to supervenience physicalism. (shrink)