Long-distance correlation, information, and faster-than-light interaction

[ddd123]

fermionic indistinguishability in white dwarfs.

Ironically, Ken G, when I think of EPR I can sort of understand what you're saying, but the white dwarf example is holding me back. Actually, are you sure it is a valid example? If the information of an emptied state on one side doesn't travel at least light-like to the other side so that you can fill it, you risk having an overfilled star for some inertial frames (i.e. in some frames the emptying occurs after the filling).

 

[Ken G]

Ironically, Ken G, when I think of EPR I can sort of understand what you're saying, but the white dwarf example is holding me back. Actually, are you sure it is a valid example? If the information of an emptied state on one side doesn't travel at least light-like to the other side so that you can fill it, you risk having an overfilled star for some inertial frames (i.e. in some frames the emptying occurs after the filling).

What I mean about white dwarfs is that if you look at interactions between particles, say you want to understand the heat conduction, you find that electrons deep in the Fermi sea don't scatter at all-- because the state they would have to scatter into is already "occupied." That's the language we use to talk about what is happening, but it's not a very good language, because it is deeply steeped in a form of local realism that can lead us astray in other applications. We imagine that star is full of different electrons, each with their own momentum state, and we say that the one electron can't go into a momentum state where there is already another one, but the whole reason that can't happen is because the electrons don't have identities like that! So the language is internally inconsistent, though common and somewhat innocuous if not taken too literally.

What is actually true is that if you look at the combined wavefunction of all the electrons, there simply are not unique individual electron states-- you can decompose into a concept of individual electron states in a host of different ways, akin to choosing a different basis for a single-particle wavefunction. So it's just not true that there is "already an electron in that momentum state", that is merely one way to translate the combined wavefunction into language that sounds like it kind of makes sense, but should not be taken literally. For example, it should not be taken so literally as to imagine that when we try to scatter an electron and find we cannot, somehow that "other electron" that is "already in that momentum state" produces some "nonlocal influence" that "prevents" our electron from scattering. None of that language is supported by the quantum mechanics that is determined by the total wavefunction of all the electrons, which does not distinguish any individual electrons at all. The experiment that is trying to scatter an electron could determine properties like the momentum of whatever electron is being culled out in that way, and only then is there "that electron" with "that momentum", but no experiment is doing that for any "other individual electron", so we shouldn't even talk about each "other individual electron" like it was a real thing there. When you avoid that, you avoid the whole concept of "influences" between electrons, you simply don't think of the system as being comprised of individual independent electrons-- it is a whole system that contains some number of electrons, none of which are distinguished and none of which have individual momenta.

 

[Mentz114]

[Mentor's note: Split off from https://www.physicsforums.com/threads/an-abstract-long-distance-correlation-experiment.852684/page-6 because this subthread was discussing possible mechanisms behind the phenomenon]

My take on these:

This entry has been deleted by me and sent as a private message.

 

[DrChinese]

... Caution: All the wisdom says that one cannot exceed the classical limit with a simulation, so don't be too hopeful.

Mentz114,

I don't think we should be covering this subject in this thread (or any thread actually). The CHSH formula is subject to the same objections I had above for the listed page. There is not going to be a breakthrough here, you and I both know Bell is a limitation. If someone want to make assertions that violate generally accepted science, those should be published elsewhere rather than debated here.

The simulation is fine as a basic tutorial. It does not make any unusual claims. It is only what is being said here that I have an issue with.

 

[Mentz114]

Mentz114,

I don't think we should be covering this subject in this thread (or any thread actually). The CHSH formula is subject to the same objections I had above for the listed page. There is not going to be a breakthrough here, you and I both know Bell is a limitation. If someone want to make assertions that violate generally accepted science, those should be published elsewhere rather than debated here.

The simulation is fine as a basic tutorial. It does not make any unusual claims. It is only what is being said here that I have an issue with.

I agree. What I posted should have been a private message.

 

[DrChinese]

I agree. What I posted should have been a private message.

I think that makes more sense to follow up on. Thanks.

-DrC

 

[ddd123]

What I mean about white dwarfs is that if you look at interactions between particles, say you want to understand the heat conduction, you find that electrons deep in the Fermi sea don't scatter at all-- because the state they would have to scatter into is already "occupied." That's the language we use to talk about what is happening, but it's not a very good language, because it is deeply steeped in a form of local realism that can lead us astray in other applications. We imagine that star is full of different electrons, each with their own momentum state, and we say that the one electron can't go into a momentum state where there is already another one, but the whole reason that can't happen is because the electrons don't have identities like that! So the language is internally inconsistent, though common and somewhat innocuous if not taken too literally.

What is actually true is that if you look at the combined wavefunction of all the electrons, there simply are not unique individual electron states-- you can decompose into a concept of individual electron states in a host of different ways, akin to choosing a different basis for a single-particle wavefunction. So it's just not true that there is "already an electron in that momentum state", that is merely one way to translate the combined wavefunction into language that sounds like it kind of makes sense, but should not be taken literally. For example, it should not be taken so literally as to imagine that when we try to scatter an electron and find we cannot, somehow that "other electron" that is "already in that momentum state" produces some "nonlocal influence" that "prevents" our electron from scattering. None of that language is supported by the quantum mechanics that is determined by the total wavefunction of all the electrons, which does not distinguish any individual electrons at all. The experiment that is trying to scatter an electron could determine properties like the momentum of whatever electron is being culled out in that way, and only then is there "that electron" with "that momentum", but no experiment is doing that for any "other individual electron", so we shouldn't even talk about each "other individual electron" like it was a real thing there. When you avoid that, you avoid the whole concept of "influences" between electrons, you simply don't think of the system as being comprised of individual independent electrons-- it is a whole system that contains some number of electrons, none of which are distinguished and none of which have individual momenta.

Of course I agree with all this. But what would you say about my issue specifically? Surely, the information that a state has been emptied must travel timelike or lightlike to avoid Pauli principle violation for some inertial frames, in case someone wants to fill that state on the other side of the star. But then aren't we falling back to localized parts for a star that should be a holistic object?

 

[Ken G]

Surely, the information that a state has been emptied must travel timelike or lightlike to avoid Pauli principle violation for some inertial frames, in case someone wants to fill that state on the other side of the star. But then aren't we falling back to localized parts for a star that should be a holistic object?

It sounds like you are asking, if a cosmic ray enters a white dwarf and knocks an electron clear out from deep in the Fermi sea, what is the wavefunction of the white dwarf? If the interaction is very quick, the cosmic ray could establish very accurately the energy of the electron that emerges, so let's say the electron and cosmic ray both come out with a very definite energy. Then at first it will look a wavefunction with a missing momentum state, much like an atom with a hole deep in one of its energy levels. But the location from where the electron came from would be uncertain, perhaps anywhere in the star. The wavefunction will then evolve via the Schroedinger equation, which has some interaction terms that will eventually fill that hole and release some energy, perhaps a photon. But it will take a long time for that transition to happen, I would guess at least the light travel time across the star, but perhaps much longer. I expect the situation would be like ionizing an electron from a deep shell in an atom-- the light crossing time is much shorter than the timescale for a transition to actually occur, so the issue doesn't even come up.