Given this premiss, and assuming the implausibility of systematic causal overdetermination, it follows that everything that has a physical effect must itself be physical. However, what does 'physical' mean in this context? Defenders of CP face Hempel's Dilemma: if 'physical' means what is recognized by current physics, then CP is likely to be false, for historical form suggests that future research will substantially correct the ontology of current physics; on the other hand, if 'physical' is understood to refer to some future, ideal physics, then CP is irredeemably vague, since we cannot predict what future physics will bring (Hempel 1980). David Spurrett and David Papineau (1999) have suggested that physicalists can respond by equating 'physical' with 'non-mental', thus converting the Completeness of Physics (CP) into the Completeness of the Non-Mental (CNM), and ending up with physicalism as the thesis that everything that has a non-mental effect must itself be non-mental. Carl Gillett and Gene Witmer (2001) argue that this definitional manoeuvre does not help. In particular, they maintain that Hempel's Dilemma bites just as strongly against the 'via negativa' reading of 'phys-