Platonic Laws of Nature

David Armstrong accepted the following three theses: universals are immanent, laws are relations between universals, and laws govern. Taken together, they form an attractive position, for they promise to explain regularities in nature-one of the most important desiderata for a theory of laws and properties-while remaining compatible with naturalism. However, I argue that the three theses are incompatible. The basic idea is that each thesis makes an explanatory claim, but the three claims can be shown to run in a problematic circle. I then consider which thesis we ought to reject (hint: see the title) and suggest some general lessons for the metaphysics of laws.