Entropy and the Asymmetry of Time

Abstract

Many thermodynamic processes are irreversible: ice melts in warm water but never spontaneously freezes out of warm water, air leaks out of a pressurized bicycle tire but never leaks back in, and so on. The Second Law of Thermodynamics provides a high-level explanation for these temporal asymmetries, but a satisfactory explanation in terms of more fundamental laws has been notoriously elusive. I defend a cosmological explanation, according to which the temporal asymmetry of thermodynamics derives ultimately from an asymmetry in the boundary conditions of the universe--namely, the fact that the distribution of mass was quite different at the beginning of the universe than it will be at the end.

Recently, three formidable objections have been raised against this cosmological explanation of thermodynamics. Two of these objections challenge the idea that the specified boundary conditions of the universe place significant constraints on the behaviors of mundane thermodynamic systems. A third objection calls into question the applicability of certain thermodynamic and statistical mechanical concepts to the universe as a whole. Building upon suggestions due to Hans Reichenbach and (more recently) Huw Price, I show how the specified boundary conditions can indeed constrain thermodynamic behavior in the requisite way; and I argue that when the relationship between the universal boundary conditions and thermodynamic phenomena is properly understood, the third objection can be sidestepped altogether.