Frequentist statistical mechanics

[DrDu]

When I learned statistical mechanics, it followed the lines of the maximum entropy principle from information theory as laid out by Jaynes which can also be seen as a Bayesian statistical theory.
I wonder whether there exist some orthodox frequentistic interpretation of statistical mechanics. Any references would be great!

 

[nonequilibrium]

Yes. So you got your ensemble. Jaynes says every member of the ensemble is equally probable due to ignorance. Another interpretation is that of those using the "ergodic hypothesis" where one postulates that a specific system will in due time (on the order of poincaré recurrence time, i.e. VERY FRIGGIN LONG) pass every point on phase phase arbitrarily close. Since every member of the ensemble corresponds with one phase space point, in the frequentist definition of probability every member of the ensemble becomes equally probable. Jaynes criticized this by saying that the recurrence time is absurdishly long and that when you make a measurement, it's on the order of seconds, not age-of-the-universe. Of course there are also many critiques on the bayesian interpretation. As far as I know, it hasn't been conclusively settled (?)