Studies in History and Philosophy of Modern Physics

In recent years a growing number of philosophers of science have come to realise that the philosophical foundations of statistical mechanics are not as secure as was once thought. Conceptual problems that had been swept under a carpet of technicalities, or simply ignored, have surfaced, solutions that were accepted are being questioned, and new routes are being explored. Thus, central issues such as the origins of time asymmetry, Maxwell’s demon, and the compatibility of statistical and classical mechanics, are as urgent and intriguing today as they were when statistical mechanics took shape at the end of the nineteenth century. Research in the foundations of the theory of relativity and quantum mechanics has deepened our understanding of these fields. This special issue is motivated by recognition of the need for similar research in the foundations of thermodynamics and statistical mechanics. The analogy between thermodynamics and other fundamental physical theories is itself a matter of controversy. On the one hand, thermodynamics has been recognised as a particularly profound theory, indispensable for our understanding of the physical world. Those who see it in this light are fond of quoting Einstein’s celebrated remark regarding the impression made on him by classical thermodynamics. It was the only physical theory, he said, that he was convinced would never (within the limits of the applicability of its concepts) be overthrown. On the other hand, we frequently hear it said that with the emergence of statistical mechanics, the laws of thermodynamics, and the second law in particular, having been reduced to the law of large numbers, have been explained away. This latter view is often taken to indicate, further, that statistical mechanics is merely a second-order theory, a probabilistic