New experimental proof of wave-function collapse?

[vanhees71]

What is unclear concerning quantum statistics? The H theorem is most naturally derived from detailed balance which follows from the unitarity of the S matrix, i.e., the (generalized) optical theorem, which is at the heart of quantum-many body theory. Ironically, it's much harder to do classical than quantum statistical physics. Even if you try to do everything in terms of classical theory, you need to introduce quantum ideas to make everything clear. Although thermodynamics and statistical physics survived the quantum revolution best, many "clouds on the horizon of classical physics" were solved by the discovery of quantum physics and triggered its development. One must not forget that quantum theory started with Planck's solution of the black-body radiation problem, a typical statistical-physics problem, and Einstein's idea about "wave-particle duality" (although obsolete now) came from his analysis of this solution.

 

[kith]

What is unclear concerning quantum statistics?

There isn't something unclear about what you probably would call the physics. The last couple of posts were just concerned with how the Copenhagen interpretation fits in with what I wrote in post #79. This could be a starting point for another fundamental discussion about interpretations but I don't want to lead such a discussion right now.

 

[Demystifier]

The H theorem is most naturally derived from detailed balance which follows from the unitarity of the S matrix, i.e., the (generalized) optical theorem, which is at the heart of quantum-many body theory.

That sounds like the Weinberg's proof of the H-theorem, in
S. Weinberg, The Quantum Theory of Fields vol I, Sec. 3.6, pages 150-151.
Can you explain why at the left hand side of Eq. (3.6.19) we have dt and not d(-t)? The sign of t should not matter in a T-invariant theory. On the other hand, with d(-t) in Eq. (3.6.19) we would eventually "derive" that entropy decreases with time, contrary to what we wanted to obtain.

My point is, you cannot really derive the H-theorem without assuming some form on time asymmetry from the beginning.

 

[Derek Potter]

No, what I said was that being unquestionably right was a criterion for teaching it to undergraduates.

Well, it was a light-hearted comment but if you want to be serious about it, perhaps you can say which version of CI you regard as unquestionably right. CI tends to be an umbrella for all sorts of interpretations including Heisenberg fuzziness which inspired Schrödinger's Cat. However as far as I know, CI always has some sort of randomness built into it, whether as a projection postulate or a slightly simpler wavefunction collapse. MW manages without any such thing. So it would seem that CI actually has redundant hypotheses making it pretty unlikely to be right at all, let alone unquestionably so.

 

[vanhees71]

That sounds like the Weinberg's proof of the H-theorem, in
S. Weinberg, The Quantum Theory of Fields vol I, Sec. 3.6, pages 150-151.
Can you explain why at the left hand side of Eq. (3.6.19) we have dt and not d(-t)? The sign of t should not matter in a T-invariant theory. On the other hand, with d(-t) in Eq. (3.6.19) we would eventually "derive" that entropy decreases with time, contrary to what we wanted to obtain.

My point is, you cannot really derive the H-theorem without assuming some form on time asymmetry from the beginning.

This is precisely the one and only correct proof of the detailed-balance relation for the most general case. You don't need time-reversal or parity invariance at all. I also don't understand your question concerning dt vs d(-t), because there's no time integral involved in (3.6.19). You just integrate (3.6.19) over $\mathrm{d} \alpha$. Then both integrals are equal, and thus $\int \mathrm{d} \alpha P_{\alpha}$ time-independent.

However, your final statement is correct: Of course in deriving transition-matrix elements you assume a directnesses of time, the socalled "causality time arrow". The H theorem just proves that this "causality time arrow" is the same as the "thermodynamical time arrow", defined as the direction of time, where entropy doesn't decrease.

Of course, a system in thermal equilibrium doesn't admit the determination of any time direction, because it forgot any history. In other words, if you make a movie from a equilibrium system, you cannot tell whether you show it forward or backward running, as long as you look at the macroscopic state only, i.e., averaged/coarsegrained quantities.

 

[LaserMind]

Note that EPR, Bell's inequality and entanglement don't demonstrate "nonlocality" (though this is the common word for it) so much as it confirms the initial "superposition of states" as predicted by quantum mechanics. In other words, the initial state of the photons are not polarized in a particular direction, the initial spin of the fermions are not in some specific x-y-z direction. The "nonlocalitiy" has to do with those states being 100% correlated antisymmetrically, as required by standard quantum mechanics.

Like others in this thread, I'm not seeing anything that looks like "proof of wave function collapse". It's called "proof of existing quantum theory." There is an unfortunate tendency in physics to conceive of the math as being the reality. The math is the description of the reality, the quantitative language we use to communicate about the reality, subject to experimental verification.

Or to use an analogy from the Matrix, the quote of "There is no spoon." There is no wavefunction. There are phenomena that we measure that are described by math we call "wavefunctions", which aptly predict our measurements. The notion that you can "prove" that a mathematical construct has objective material behavior (collapsing or otherwise) is absurd.

 

[LaserMind]

The wave function collapse is a value exchange - a numerical event!

 

[Dale]

The OP is long gone and now the discussion is just going in circles. Thread closed.