Abstract

This paper argues that the doctrines of determinism and supervenience, while logically independent, are importantly linked in physical mechanics—and quite interestingly so. For it is possible to formulate classical mechanics in such a way as to take advantage of the existence of mathematical devices that represent the advance of time—and which are such as to inspire confidence in the truth of determinism—in order to prevent violation of supervenience. It is also possible to formulate classical mechanics-and to do so in an observationally equivalent, and thus equally empirically respectable, way—such that violations of supervenience are (on the one hand) routine, and (on the other hand) necessary for achieving complete descriptions of the motions of mechanical systems—necessary, therefore, for achieving a deterministic mechanical theory. Two such formulations—only one of which preserves supervenience universally—will conceive of mechanical law in quite different ways. What's more, they will not admit of being extended to treat thermodynamical questions in the same way. Thus we will find that supervenience is a contingent matter, in the following rather surprising and philosophically interesting way: we cannot in mechanics separate our decisions to conceive of physical law in certain ways from our decisions to treat macroscopic quantities in certain ways.