What is fundamental?
Reductionists always claim that what is fundamental is what occurs at the shortest length scales and involving the constituent particles of a system, i.e., the microscopic is more fundamental than the macroscopic. However, claims as to what is fundamental is sometimes a matter of personal choice and preference. As pointed out by Alister Rae , work by Ilya Prigogine on the relationship between microscopic dynamics and thermodynamics, can be argued to support the view that the macroscopic can be more fundamental than the microscopic:
“[In a chaotic system] It is … the positions and velocities of the component molecules, that change chaotically whereas the thermodynamic quantities (which are traditionally thought to be derived from the microscopic substructure) are well behaved. These facts led Ilya Prigogine to suggest …. we should consider the thermodynamic quantities to be the primary reality and the allegedly more fundamental description in term of microscopic structure to be secondary.”Such a view is completely the opposite of what a reductionist would claim. In a later post I will give Bob Laughlin's answer to the question, "What is fundamental?"
A. Rae, Quantum Physics: Illusion or Reality? (2nd ed.; Cambridge: Cambridge University Press, 2004), pp. 120-123.
Speak the good word, Ilya! This is very consistent with the view espoused by Jaynes, that statistical mechanics consists of two very distinct sub-problems: 1) the enumeration of the available states of the system and their observables and 2) statistical inference, based most reasonably on the principle of maximum entropy subject to constraint.
ReplyDeleteThe hang-up that most people have re: Jaynes view is that, by placing the maximum entropy postulate front and centre injects 'subjectivity' into the field. I don't agree with this - the apparent subjectivity is not a problem if one is always careful about defining precisely what ensemble one is talking about. Furthermore, as pointed out by Jaynes, taking a hard-line objective point of view doesn't really make anything clearer anyway, since one is then forced into a conceptual pretzel over the need to make everything ergodic when it clearly is not.
It is high time that the full 'statisticality' of quantum mechanics made its way into entry-level curricula. I think that such a change would inject needed rigor and clarity into the field, and point the way to new advances. This would finally bring to completion the intellectual leap made by Gibbs over a century ago.