Published online by Cambridge University Press: 01 April 2022
I discuss a broad critique of the classical approach to the foundations of statistical mechanics (SM) offered by N. S. Krylov. He claims that the classical approach is in principle incapable of providing the foundations for interpreting the “laws” of statistical physics. Most intriguing are his arguments against adopting a de facto attitude towards the problem of irreversibility. I argue that the best way to understand his critique is as setting the stage for a positive theory which treats SM as a theory in its own right, involving a completely different conception of a system's state. As the orthodox approach treats SM as an extension of the classical or quantum theories (one which deals with large systems), Krylov is advocating a major break with the traditional view of statistical physics.
I would like to thank Larry Sklar for valuable advice and comments on earlier versions of this work. I have also benefited from discussions with Anne Bezuidenhout, Gary Ebbs, Steve Jacobson, Jim Joyce, Jim Maffie, Joe Mendola, Peter Railton, and Bill Schmitz.