Confused by nonlocal models and relativity

[vanhees71]

No, the problem, as we have already discussed ad nauseam in previous threads, is that different sources in the literature use "locality" to mean different things. All of those things are perfectly well-defined; they're just different. Which means that in any discussion of this general topic, one needs to say which meaning of "locality" one is using. It does not mean you can discount someone else's arguments because they are using "locality" with one of the other meanings besides the one you prefer.

It's confusing, however. In the quoted book chapter by Pan and Zeilinger they use "locality"/"nonloclity" in two different meanings. At least they are kind enough to label the one as "Einstein locality" (and this is the usual meaning of "locality" in the relativistic-QFT literature) and as "quantum non-locality" (which simply means that Bell's inequalities or other properties of local-hidden HV models are violated by QT via entanglement). I didn't "discount" anybody's argument. I only tried to clarify this confusion once more!

 

[martinbn]

It's confusing, however. In the quoted book chapter by Pan and Zeilinger they use "locality"/"nonloclity" in two different meanings. At least they are kind enough to label the one as "Einstein locality" (and this is the usual meaning of "locality" in the relativistic-QFT literature) and as "quantum non-locality" (which simply means that Bell's inequalities or other properties of local-hidden HV models are violated by QT via entanglement). I didn't "discount" anybody's argument. I only tried to clarify this confusion once more!

But what is Einstein locality? Some say that it is violated and that signal locality is preserved, so they are not the same. So, what is the einstein locallity then?

 

[gentzen]

Maybe Bell's famous paper on the definitions of the words used in discussing these issues can help clarify some things:
https://www.informationphilosopher.com/solutions/scientists/bell/Against_Measurement.pdf

I was recently reminded of the paper in a lecture by Tim Maudlin.

I also recently reread parts of that paper too, especially section "The quantum mechanics of N G van Kampen". The reason was that vanhees71 had mentioned van Kampen as a great no-nonsense quantum "explainer". I was fascinated by the fact that Bell didn't get at all where van Kampen was trying to go. And then he tried various ridiculous "reconstructions" of what van Kampen might have thought, and where he supposedly made mistakes.

That reminded me that even the great Bell could sometimes be wrong, at least when it comes to interpret the writings of other people.

 

[A. Neumaier]

We can't speak of separate particles when they are entangled. Yet we talk of particles moving away from each other. There seem to be inherent semantic difficulties.

The particle terminology survived from the times where entanglement experiments were science fiction, and is no longer appropriate except in interpreting scattering experiments.

All semantic difficulties are absent if one thinks of particles not as little balls but as excitations of fields. Excitations can be entangled, as we can see already for water wavelets, a kind of classical analogue. They are nonlocal pieces of a single entity.

 

[A. Neumaier]

"quantum non-locality"

As there is nothing quantum about Bell's argument, this should rather be named ''classical hidden-variable non-locality", or "Bell non-locality".

 

[gentzen]

The particle terminology survived drom the times where entanglement experiments were science fiction, and is no longer appropriate except in interpreting scattering experiments.

All semantic difficulties are absent if one thinks of particles not as liitle balls but as excitations of fields. Excitations can be entangled, as we can see already for water wavelets.

Whenever vanhees71 dreams of changing terminology, I think to myself that we don't even manage to get rid of that seriously misleading "particle terminology". For an electron or an ion, it is good enough, but for a photon, or a phonon, or other excitations, it is badly off, both from quantum mechanics, but also from any "effective" description of the physical situation.

Of course, there is no better word from normal daily experience, but one could have at least modified the word "particle" slightly, to make it clear that we are talking about something completely different. For example, in German one could use "Teilchon" instead of "Teilchen", in French "corpuscole" instead of "corpuscule" and "particole" instead of "particule". In English, well probably parton could have been nice, but it doesn't follow the pattern for other languages, and is already used for something else. Maybe particon, which could also be transfer to French as corpuscon and particon (or maybe not, French has its own "feeling").

 

[gentzen]

As there is nothing quantum about Bell's argument, this should rather be named ''classical hidden-variable non-locality", or "Bell non-locality".

well, "classical hidden-variable non-locality" doesn't sound good, and will never gain general acceptance. But "Bell nonlocality" is nice, especially because "Einstein locality" is already in use.

 

[gentzen]

but even Mermin‘s very modest attempt to change qubit to qbit failed

 

[Fra]

Post #13 in this thread I think put the finger on the key... rather than the various definitions

the physical theory must explain the quantum correlation between the measurement outcomes.
In other words, the problem of understanding the quantum entanglement is not because the time order of the space-like events is meaningless in SR but because any observer will find the experiment outcomes always violates BI (Bell`s Inequality). And this can even happen in a non-moving frame where suppose that there are two observers each is attached to the measurement device and each claim that his device made the first measurement. In this setup, BI is still violated not because that there is no significance to talk about the order of the measurements but because something else is missing, whether to call it non-locality or retro-causality,,,etc.

IMO, some confusion of the locality concepts is that I think in it's simple non-technical form it usually means that an object X is only affected by it's immedate surroundings. Then is of course made more precise with special relativity etc.

But the problem I see is that as it's not clear the localisation of a quantum state means in the first place, because the position of the "quantum state" is hardly the same as the "position" in the wavefunction!

Especially in CI or the statistical of ensemble interpretation. I am inclined to conceptually think of the quantum state, as localized to the observer. But in the CI interpretation, this is the whole classical environment, which is those delocalized to start with, this is a problem from the beginning. So what is the sense in saying that the "observer" os only affected by it's immediate surroundings? what IS the "surrounding" the the QM observer?

/Fredrik

 

[vanhees71]

Whenever vanhees71 dreams of changing terminology, I think to myself that we don't even manage to get rid of that seriously misleading "particle terminology". For an electron or an ion, it is good enough, but for a photon, or a phonon, or other excitations, it is badly off, both from quantum mechanics, but also from any "effective" description of the physical situation.

True, but here the harm is less severe, because a photon is a photon in almost all physics text, i.e., a 1-quantum Fock state of the (asymptotically) free electromagnetic field.

Of course, there is no better word from normal daily experience, but one could have at least modified the word "particle" slightly, to make it clear that we are talking about something completely different. For example, in German one could use "Teilchon" instead of "Teilchen", in French "corpuscole" instead of "corpuscule" and "particole" instead of "particule". In English, well probably parton could have been nice, but it doesn't follow the pattern for other languages, and is already used for something else. Maybe particon, which could also be transfer to French as corpuscon and particon (or maybe not, French has its own "feeling").

Well, I guess there's little chance to change this.

The problem with quantum foundations is that it's full of confusing terminology due to too much philosophy involved. It is really hard to discuss it, because it's subtle by itself and in addition, because you have always to explain again when using standard words from physics like locality (in physics, at least in my community in connection with QT what's meant is that interactions are described in realtivistic local QFT obeying the microcausality constraint on local observables, i.e., self-adjoint operators that represent local observables must commute at space-like separation of their argument).

 

[vanhees71]

I also recently reread parts of that paper too, especially section "The quantum mechanics of N G van Kampen". The reason was that vanhees71 had mentioned van Kampen as a great no-nonsense quantum "explainer". I was fascinated by the fact that Bell didn't get at all where van Kampen was trying to go. And then he tried various ridiculous "reconstructions" of what van Kampen might have thought, and where he supposedly made mistakes.

That reminded me that even the great Bell could sometimes be wrong, at least when it comes to interpret the writings of other people.

For me Bell is great for the discovery of anomalies and the physics core of his work on quantum foundations. On the other hand for me he's very confusing in his language. A measurement is not allowed to be called measurement anymore and observables don't exist anymore but are relabeled as "beables" with obscure meaning. I always feel reminded of Einstein's advice concerning theoretical physicists: "Don't listen to their words, look at their deeds".

 

[gentzen]

A measurement is not allowed to be called measurement anymore

indeed, a bit unfortunate

and observables don't exist anymore but are relabeled as "beables" with obscure meaning.

Well, "beables" are something completely different than "observables". If you try to use them as a sort of observables, then everything will look very obscure, and you might end-up even believing that you disproved that Bohmian mechanics makes the same predictions as normal QM.

 

[A. Neumaier]

Of course, there is no better word from normal daily experience

'field' is a better word from normal daily experience than 'particles', except in scattering experiments.

a photon is a photon in almost all physics text, i.e., a 1-quantum Fock state of the (asymptotically) free electromagnetic field.

But ''2 photons'' is ambiguous since it may be two single photon states or one entangled photon state.

 

[vanhees71]

I never could figure out, what Bell intended to mean by creating this playful world "beable".

 

[vanhees71]

'field' is a better word from normal daily experience than 'particles', except in scattering experiments.

But ''2 photons'' is ambiguous since it may be two single photon states or one entangled photon state.

There are of course many $n$-photon states. It's just an eigenstate of the photon-number operator, which is highly degenerated.

 

[A. Neumaier]

I never could figure out, what Bell intended to mean by creating this playful world "beable".

Anything that exists objectively, whether measured or not.

 

[vanhees71]

So it's an observable after all! It's of course interpretation-dependent, but an observable is something I can observe at least in principle, and only such a thing can exist objectively.

 

[gentzen]

So it's an observable after all! It's of course interpretation-dependent, but an observable is something I can observe at least in principle, and only such a thing can exist objectively.

No, it is not an observable. The Bohmians know pretty well that their trajectories are not directly observable, and that this is important for avoiding various paradoxes, or "mispredictions".

 

[Morbert]

So it's an observable after all! It's of course interpretation-dependent, but an observable is something I can observe at least in principle, and only such a thing can exist objectively.

We can observe pointer states, and so if values of observables defined on the state space of a microscopic system are interpreted as statements about macroscopic systems used to test the microscopic system, then there is an objectivity there. But if a value of an observable is interpreted as a statement about a property inherent in the microscopic system itself, then it is hard to interpret them objectively.

 

[A. Neumaier]

So it's an observable after all!

Not necessarily. It could be a particle, a spectrum, or a cross section....

It's of course interpretation-dependent, but an observable is something I can observe at least in principle, and only such a thing can exist objectively.

Does the momentum of a particle exist if it is not oberved? If yes, momentum is a beable, if no, it is not.

 

[martinbn]

Anything that exists objectively, whether measured or not.

I don't think so. My understanding is that a beable is part of the theory that corresponds to something real.

 

[A. Neumaier]

There are of course many $n$-photon states. It's just an eigenstate of the photon-number operator, which is highly degenerated.

A Bell-entangled 2-photon state with the photons far apart is very particle-unlike since it behaves like a unity. Thus this non-local object should not be called two particles!

 

[vanhees71]

Not necessarily. It could be be a particle, a spectrum, or a cross section....

Does the momentum of a particle exist if it is not oberved? If yes, momentum is a beable, if no, it is not.

Of course. In both non-relativistic QM and relativistic QFT momentum is a basic observable, which exists because you have to realize some (ray) representation of the Galilei or proper orthochronous Poincare group, respectively and thus there must be momentum observables described by the generators of spatial translations. So in this sense momentum should be a beable, because it's defined as an objective property and something that can be observed/measured. But then "beable" is indeed synonymous with "observable". Of course an observable doesn't need to take determined values before measurement. All that can be said, given the state the system is prepared in, are the probabilities for the outcome of measurements of the observable.

 

[A. Neumaier]

I don't think so. My understanding is that a beable is part of the theory that corresponds to something real.

What is the difference to what I said? 'something real' is the same as 'what exists objectively'. The examples I gave

It could be a particle, a spectrum, or a cross section

illustrate this.

 

[vanhees71]

A Bell-entangled 2-photon state with the photons far apart is very particle-unlike since it behaves like a unity. Thus this non-local object should not be called two particles!

I've my problems with the word "particle" particularly for photons anyway, but why shouldn't this be a proper two-photon state? After all it's an eigenstate of the photon-number operator, which itself is, of course, not a local operator.

 

[A. Neumaier]

Of course. In both non-relativistic QM and relativistic QFT momentum is a basic observable, which exists because you have to realize some (ray) representation of the Galilei or proper orthochronous Poincare group, respectively and thus there must be momentum observables described by the generators of spatial translations. So in this sense momentum should be a beable,

Then a particle has at every point on its trajectory a well-defined momentum, and by the same argument a well-defined position. But this contradicts the uncertainty relation. Thus your argumentation cannot be valid.

an observable doesn't need to take determined values before measurement.

But then the particle has neither momentum nor position before measurement, hence neither is a beable in Bell's sense.

 

[A. Neumaier]

I've my problems with the word "particle" particularly for photons anyway, but why shouldn't this be a proper two-photon state? After all it's an eigenstate of the photon-number operator, which itself is, of course, not a local operator.

I would say it is ''the e/m field in a 2-photon state'', but it is not ''two photons''.

 

[martinbn]

What is the difference to what I said? 'something real' is the same as 'what exists objectively'. The examples I gave

illustrate this.

Yes, but I think it is meant to be the counter part in the theory not the objectively existing thing itself. I might be wrong of course.

 

[vanhees71]

I would say is ''the e/m field in a 2-photon state'', but it is not ''two photons''.

That's of course the much more precise formulation.

 

[vanhees71]

Then a particle has at every point on its trajectory a well-defined momentum, and by the same argument a well-defined position. But this contradicts the uncertainty relation. Thus your argumentation cannot be valid.

No, why? An observable usually doesn't take a determined value. Nevertheless the observable always exists as an objective property of the system in the sense it can (at least in principle) always be measured. The outcome is of course random with the probability for the outcome given by usual generalized Born's rule.

But then the particle has neither momentum nor position before measurement, hence neither is a beable in Bell's sense.

But then "beable" is a notion within a ficticious deterministic theory and not within QT, i.e., if you mean by "objective" that all observables must always take determined values, but as far as I understand Bel wanted to express something within quantum theory with his word "beable".

Our discussion again just demonstrates that this word is not sharply defined at all. Everybody seems to have an individual understanding of its meaning, and thus it's the opposite of a clear scientific definition. As too easily, again philosophical speculations like that lead to confusion rather than clarification of the "physical meaning" of QT.

 

[gentzen]

but as far as I understand Bel wanted to express something within quantum theory with his word "beable".

In Bohmian mechanics, you want to distinguish between the "status" of the wavefunction and the "status" of the particle trajectories. Here, the wavefunction is not a "beable", but the particle trajectories are.

In GRW, once again you don't want to grant the wavefunction a "status" too different from "nomological" (law-like). So you use something else, like the "mass density," or the "flashes" to which you can assign the "status" of a "beable".

 

[A. Neumaier]

Yes, but I think it is meant to be the counter part in the theory not the objectively existing thing itself. I might be wrong of course.

Yes, but these are often silently identified in discussions, and even Bell does it....

 

[vanhees71]

In Bohmian mechanics, you want to distinguish between the "status" of the wavefunction and the "status" of the particle trajectories. Here, the wavefunction is not a "beable", but the particle trajectories are.

These socalled trajectories are fictitious. At least they cannot be observed in any way.

In GRW, once again you don't want to grant the wavefunction a "status" too different from "nomological" (law-like). So you use something else, like the "mass density," or the "flashes" to which you can assign the "status" of a "beable".

This would mean that you interpret "beable" in a much more restricted sense as "obervable" than the standard definition of "observable" in QT, i.e., only the macroscopic "pointer readings" on a measurement device are "beables", but I thought somehow Bell wanted to get rid of this notion of "observables" in the sense of something that has to be measured...

 

[PeterDonis]

These socalled trajectories are fictitious.

Not in Bohmian mechanics. In that interpretation, they are real.

At least they cannot be observed in any way.

Which illustrates why Bell invented the term "beable" instead of just using "observable": that in some interpretations (Bohmian was, IIRC, one of Bell's favorites, at least for illustration), the two are not the same--there are beables that are not observable, such as particle positions in Bohmian mechanics.

I thought somehow Bell wanted to get rid of this notion of "observables" in the sense of something that has to be measured

I don't think Bell wanted to get rid of this notion; he just wanted to distinguish it from the notion of "beable". See above.

 

[vanhees71]

But then "beable" is something outside of the natural sciences, referring to someting that cannot be observed and thus not tested by observations.

This is as if you'd call the electromagnetic potentials in some arbitrary gauge a "beable", although it's never observable. What is observable is the electromagnetic field derived from these potentials. So shouldn't rather this field be the "beable" and not the potentials?