Jaynes' paper “Information Theory and Statistical Mechanics” posits that statistical mechanics is fundamentally an inference theory of many-body mechanical systems. Critiques of this approach focus on its implications for concepts like Maxwell's Demon, which challenges the information-theoretical foundation of entropy. Key references for understanding these critiques include works from core.ac.uk, arxiv.org, and jstor.org. The discussion highlights the necessity of quantum mechanics in addressing Maxwell's Demon, particularly through Szilard's application of information theory to entropy.
PREREQUISITESResearchers, physicists, and students interested in the intersections of information theory, statistical mechanics, and quantum mechanics will benefit from this discussion.
DarMM said:Summary: Seeking references for criticisms of Jaynes's information theoretic approach to Statistical Mechanics.
In his well known paper “Information Theory and Statistical Mechanics” Jaynes attempted to formulated statistical mechanics as "nothing more" than the inference theory of many body mechanical systems. I am looking for critiques of this approach. Also of use would be summaries or reviews of the "current state of debate".
Maxwell's demon is "driven out" by Szilard and precisely by applying the information theoretical approach to the notion of entropy.DEvens said:I have not read it. But that "nothing more" part is kind of worrying. How does he preclude Maxwell's Demon?
If I understand correctly, the usual way to preclude MD comes down to quantum mechanics. MD has to make observations.