EPR and Non-Locality - For and Against

[bhobba]

Hi

As most people I think know I do not think that QM requires non-locality of any kind. The reason is it is a limiting case of QFT which since it combines SR and QM it can not violate the assumptions it is built on. Specifically we have the cluster decomposition property:
https://www.physicsforums.com/threads/cluster-decomposition-in-qft.547574/

But I recently came across an interesting paper on it, and a rebuttal:
https://arxiv.org/abs/1703.11003
https://arxiv.org/abs/1705.01356

I agree with Stephen Boughn, but I think it's nice to get arguments for and against in one place for perusal and comment.

Thanks
Bill

 

[PeterDonis]

I do not think that QM requires non-locality of any kind.

"Non-locality" is an ordinary language term and does not have a single well-defined technical meaning. Depending on which well-defined technical meaning you choose, you can make it either true or false that QM does not require non-locality.

For example, if you choose to have "non-locality" mean "violates the cluster decomposition property", then it is true that QM does not require non-locality.

But if you choose to have "non-locality" mean "violates the Bell inequalities", then it is false that QM does not require non-locality.

I find such arguments over definitions of terms pointless. I would rather drop terms like "non-locality" that can be twisted to make either side of an argument right or wrong, and focus on the actual physics, properties like cluster decomposition or Bell inequality violation.

 

[PeterDonis]

I agree with Stephen Boughn

Unfortunately, just reading the abstract of his paper I am already not impressed:

"The magic of entangled quantum states has little to do with entanglementand everything to do with superposition"

I strongly disagree; entanglement and superposition are very different things (for one thing, superposition is basis dependent and entanglement is not) and conflating them, in my experience, causes far more confusion than it solves. (I can probably dig up a number of threads right here on PF that illustrate that.)

 

[PeroK]

Hi

As most people I think know I do not think that QM requires non-locality of any kind. The reason is it is a limiting case of QFT which since it combines SR and QM it can not violate the assumptions it is built on. Specifically we have the cluster decomposition property:
https://www.physicsforums.com/threads/cluster-decomposition-in-qft.547574/

But I recently came across an interesting paper on it, and a rebuttal:
https://arxiv.org/abs/1703.11003
https://arxiv.org/abs/1705.01356

I agree with Stephen Boughn, but I think it's nice to get arguments for and against in one place for perusal and comment.

Thanks
Bill

There's also

"... quantum mechanics and by inference nature herself are nonlocal in the sense that a measurement on a system by an observer at one location has an immediate effect on a distant "entangled" system ..."

That's at best sloppy language.

 

[vanhees71]

It's not sloppy language, it's plain wrong. Standard relativistic QFT is local (fulfills microcausality) by construction but still there is the inseparability described by entanglement, i.e., correlations between the outcome of measurements for observables on parts of a quantum system with the (local!) measurements at the partial systems at far distances with measurement-outcome events space-like separated (and thus NOT causally connected).

 

[DrChinese]

As PeterDonis says, violation of a Bell Inequality essentially requires Quantum Nonlocality. The interpretations handle this in different ways.

@bhobba: You say that you don't know many that believe in Quantum Nonlocality, but most authors of papers I read on entanglement seem to believe it, and most say so explicitly as if it is generally accepted. I don't recall ANY peer-reviewed published paper on entanglement within QFT as saying QFT is local. Almost any experimental paper will indicate it is a disproof of locality. I appreciate that QFT claims to be local by construction, and yet it is obvious from Bell tests that an observer here influences an outcome there. Sorry, it's not just "correlations". Especially in the cases where the "correlation" is 100%.

EDIT: In case I wasn't clear, experimental considerations require quantum non-locality. And by using the term "quantum non-locality", I am referring to the kind of non-locality required by Bell in which either realism or locality (or both) are rejected.

 

[Morbert]

I appreciate that QFT claims to be local by construction, and yet it is obvious from Bell tests that an observer here influences an outcome there. Sorry, it's not just "correlations". Especially in the cases where the "correlation" is 100%.

I'm probably treading old ground but if we have two entangled spacelike-separated systems A and B, none of the expectation values of observables of A are modified by any measurements of observables of B. In what sense does an observer of properties of B influence outcomes of experiments on A?

 

[bhobba]

I am referring to the kind of non-locality required by Bell in which either realism or locality (or both) are rejected.

I can't recall saying most people do not believe in QM non-locality. A lot of QM papers/books I read say, as if it is a proven fact, that QM is non-local. I do not agree with that, and never have. The issue is in an entangled system (of two particles) you can't say you even have two particles for locality to be an issue. You only have two particles once the entanglement is broken. Whats going on in between is anybody's guess. Since you can't say what's going on my position would be more correctly stated as you can't really know.

Thanks
Bill

 

[bhobba]

I'm probably treading old ground but if we have two entangled spacelike-separated systems A and B, none of the expectation values of observables of A are modified by any measurements of observables of B. In what sense does an observer of properties of B influence outcomes of experiments on A?

In the sense they are correlated in Bell type experiments. I take the view correlation does not imply influences. It may of course, but that's all part of the ongoing debate about this stuff. We do know that counterfactual definiteness and locality are ruled out.

Thanks
Bill

 

[bhobba]

I appreciate that QFT claims to be local by construction,

I think it obeys SR by construction. SR does not rule out non-locality providing it can't be used to send information so clocks can be synced.

Thanks
Bill

 

[bhobba]

It's not sloppy language, it's plain wrong.

I think so too - and the Author was disagreeing with it.

But I have to say Peter has a point - this can easily degenerate into semantics. That is something I had not considered before, and I think in some of the things I have written about it I have done just that. There was a recent twitter discussion including a Field's medalist no less about if 2+2 = 5. There was aguments that it depends on what assumptions you make eg the base you are working in. But when you got right down to it, it was really just semantics. By logical construction (eg Peano's axioms ) it must be true . When you apply it you may find from experiment it is wrong - but then you have applied it incorrectly. The discussion went back and forth but in the end it was totally useless - as the Fields medalist admitted later. It's a trap even the best of us can fall for. It's actually unanswerable because it depends on semantics. Of course I believe 2+2=4 - however if you want to argue the point it ends up up in semantic pointlessness.

Thanks
Bill

 

[PeterDonis]

By logical construction (eg Peano's axioms ) it must be true .

That word "true" is another much abused word. What does it actually mean in this connection? As you appear to realize, it does not mean that every system in the real world obeys it. As your use of the phrase "by logical construction" makes clear, what "true" actually means in this connection is "true in any semantic model of the logical axioms in question". So finding an experiment in the real world which does not obey 2 + 2 = 4 does not mean 2 + 2 = 4 is "wrong". It means that real-world situation is not a semantic model of the Peano axioms. But other real world situations may be, and in those real-world situations, 2 + 2 = 4 is indeed true.

Of course I believe 2+2=4

If by this you mean you believe that in any semantic model of the Peano axioms, 2 + 2 = 4 will be true, yes, I agree with you. But I don't think it's "semantic pointlessness" to insist on being clear about what exactly you mean when you say "I believe 2 + 2 = 4". I think it's a requirement for clear thinking in general.

To bring this digression back on topic, consider what Bell's Theorem actually says, in logical terms. In logical terms, it says that any semantic model of a certain set of axioms (which he defines mathematically in his paper) must have a certain property, but actual quantum systems violate that property. That means actual quantum systems cannot be a semantic model of that particular set of axioms. Whether that means actual quantum systems must have "nonlocality" depends, as I've said, on what you take the term "nonlocality" to mean--but the fact that actual quantum systems cannot be a semantic model of a particular set of axioms is true regardless of anyone's choice of words.

 

[bhobba]

That word "true" is another much abused word. n general.

As usual excellent post Peter. I always learn a lot here, even when I 'goofed' like I did in this post.

I could post up in the math sub-forum that 2+2 = 5 podcast, but it gets philosophically off topic for this forum in parts.

Thanks
Bill

 

[vanhees71]

As PeterDonis says, violation of a Bell Inequality essentially requires Quantum Nonlocality. The interpretations handle this in different ways.

Sure, but nonlocality in the sense of inseparability and not in the sense of nonlocality of interactions since standard relativistic QFT (including the standard model describing all phenomena correctly today though heavily tested in the hope to find beyond-the-standard-model physics) are local (microcausal) QFTs. This is a mathematical property and independent of interpretations.

 

[vanhees71]

I think so too - and the Author was disagreeing with it.

But I have to say Peter has a point - this can easily degenerate into semantics. That is something I had not considered before, and I think in some of the things I have written about it I have done just that. There was a recent twitter discussion including a Field's medalist no less about if 2+2 = 5. There was aguments that it depends on what assumptions you make eg the base you are working in. But when you got right down to it, it was really just semantics. By logical construction (eg Peano's axioms ) it must be true . When you apply it you may find from experiment it is wrong - but then you have applied it incorrectly. The discussion went back and forth but in the end it was totally useless - as the Fields medalist admitted later. It's a trap even the best of us can fall for. It's actually unanswerable because it depends on semantics. Of course I believe 2+2=4 - however if you want to argue the point it ends up up in semantic pointlessness.

Thanks
Bill

Well you can call what was called "4" for millenia all of a sudden "5" if you wish, but that's nothing than confusion and totally useless. I doubt that such a proposal would be successful and it's not desireable to rename age-old names of things (as the confusion about milliards and billions in different countries and in different times demonstrates).

It's another issue concerning to try to be precise when necessary: In my opinion to call the long-range correlations described by quantum entanglement "non-locality", though anybody having gone through a QM 1 lecture should know what's meant by this term. The confusion goes back to the unfortunate EPR paper, which Einstein rightfully didn't like too much. I think it's much more to the point to use Einstein's notion of "Inseparability" (Inseparabilität). One should note that Einstein not only was an ingenious physicist but also a master in scientific prose, formulating everything "as simple as possible but not simpler", at least when he wrote the papers himself ;-)).

 

[Demystifier]

As most people I think know I do not think that QM requires non-locality of any kind. The reason is it is a limiting case of QFT which since it combines SR and QM it can not violate the assumptions it is built on. Specifically we have the cluster decomposition property:
https://www.physicsforums.com/threads/cluster-decomposition-in-qft.547574/

From the fact that QM does not require non-locality of a specific kind, it does not follow that it does not require non-locality of any kind.

If you deny non-locality, then what does Bell theorem really prove, in your opinion?

 

[bhobba]

If you deny non-locality, then what does Bell theorem really prove, in your opinion?

It puts a limit on interpretations and how you can look at it.

Thanks
Bill

 

[Demystifier]

It puts a limit on interpretations and how you can look at it.

Sure, but what that limit is, specifically?

 

[DrChinese]

I can't recall saying most people do not believe in QM non-locality.

My bad! Now that I re-read your OP, it says instead that you are that person, not most people. Apologies, I completely misread that.

 

[DrChinese]

The issue is in an entangled system (of two particles) you can't say you even have two particles for locality to be an issue. You only have two particles once the entanglement is broken.

I accept that in a Bell test, the entangled system cannot be considered as two separate photons (or other particles). However, a system with such spatial extent as is used in Bell test (say 2 photons) cannot be considered as being "localized" either. In fact, I am not sure it even makes sense to say a single photon system is "localized" as in many cases, a single photon can be considered to have a very large* spatial extent.

And in either case, we are witnessing quantum nonlocality. I can't imagine any definition of such systems that imply otherwise. We can have systems that are many kilometers in size, and still have perfect correlations in Bell tests with observers Alice and Bob.

So I would certainly say that if you retreat to the position that there is no quantum nonlocality as Alice and Bob perform their respective measurements, and that it is the system instead that has spatial (and/or temporal) extent: you haven't eliminated the quantum nonlocality, you have simply moved it.

And further: Alice can take the portion of an entangled system she receives... and teleport that entanglement FTL to another system as far distant as she likes. And again we can perform Bell tests on the newly enlarged entangled system, and get perfect correlations. That's a pretty good trick if there is no quantum nonlocality.*Perhaps even as large as the observable universe?

 

[Minnesota Joe]

I can't recall saying most people do not believe in QM non-locality. A lot of QM papers/books I read say, as if it is a proven fact, that QM is non-local. I do not agree with that, and never have. The issue is in an entangled system (of two particles) you can't say you even have two particles for locality to be an issue. You only have two particles once the entanglement is broken. Whats going on in between is anybody's guess. Since you can't say what's going on my position would be more correctly stated as you can't really know.

Are you willing to you elaborate on this? In EPR-Bell (GHZ, Hardy) examples the experimenters themselves can have space-like separation. What seems important is that there is an entangled system for the space-like separated experimenters to measure, without saying anything about whether it is really one particle or really two particles.

ETA: I think Dr. Chinese is making a similar point.

 

[DrChinese]

I'm probably treading old ground but if we have two entangled spacelike-separated systems A and B, none of the expectation values of observables of A are modified by any measurements of observables of B. In what sense does an observer of properties of B influence outcomes of experiments on A?

Haha, yes, we have covered this before. In this forum, and you participated in at least some of it (see your post #33).

https://www.physicsforums.com/threa...rpretation-of-quantum-mechanics.989890/page-2

You ask "in what sense"? Per Steven Weinberg, Lectures on Quantum Mechanics, 12.1 Paradoxes of Entanglement (talking about EPR-B):

"There is a troubling weirdness about quantum mechanics. Perhaps its weirdest feature is entanglement, the need to describe even systems that extend over macroscopic distances in ways that are inconsistent with classical ideas. ... Of course, according to present ideas a measurement in one subsystem does change the state vector for a distant isolated subsystem - it just doesn't change the density matrix."

Of course, you interpreted another statement by Weinberg as implying QM is local (or something that I couldn't agree as saying QM is local). I'd love to an actual quote by Weinberg where he denies quantum non-locality, or otherwise denies the obvious implications of Bell. Since the quote above explicitly says:

According to present ideas a measurement in one subsystem does change the state vector for a distant isolated subsystem...

Of course, to be fair: there is no apparent direction of the change referred to. It could be in either direction.

 

[bhobba]

Are you willing to you elaborate on this? In EPR-Bell (GHZ, Hardy) examples the experimenters themselves can have space-like separation.

The experimenters are irrelevant. The particles are entangled - you can't say for sure they exist as separate particles for the idea of locality to even make sense.

Here is the math. Let's say we have two systems that can be in state |a> and |b>. If system 1 is in state |a> and system 2 in state |b> we write this as |a>|b>. Conversely if system 1 is in state |b> and system 2 in state |a> we write it as |b>|a>. By the principle of superposition a possible state of both systems is 1/√2 (|a>|b> + |b>|a>). In such a state what state each system is in, or even if there is still two systems, is unclear - you can only speak with certainty of the combined system 1 and 2. If they have no individuality the idea of signals passing between the two systems does not even make sense. However one can devise an observable to find the state system 1 is in and hence system 2. You can hypothesise - certainly - but Bells Theorem put bounds on that hypothesising. If you hypothesise that system 1 and system 2 still keep their individuality then Bell says some kind of non-local influence must go between the separate systems when you observe it. If you want to keep locality you can't consider them as separate systems - but as a single system. Then you face the issue of how, if say you find system 1 in state |a> then how does system 2 know to be in state |b>. The answer is they are correlated - you have deliberately devised the experiment so that such will always be the case. It is a 'strange' characteristic of QM that the statistics of such a correlation is different from the statistics of classical correlation - which is the essence of Bell's Theorem. An explanation is at the foundations of QM it is a generalised probability theory so there is nothing strange about the idea it would have different statistics. Another explanation I have put forward comes from the Cluster Decomposition property which in order to make sense of you must preclude correlations. But as Peter correctly points it is just semantics on what you mean by non-locality.

Thanks
Bill

 

[wle]

"Non-locality" is an ordinary language term and does not have a single well-defined technical meaning. Depending on which well-defined technical meaning you choose, you can make it either true or false that QM does not require non-locality.

I think this pretty succinctly hits the nail on the head: "local" by itself is a word that people learn by association from other people, seeing examples of things called "local" and others "nonlocal". So arguing about which box quantum physics should be put into is not interesting by itself. The more interesting question is why or what underlying issues might lead you to ask this kind of question in the first place, which could be heavily dependent on how you think about quantum physics.

Bell had reasons for caring about this which you can reasonably agree or disagree with but which I think make a lot of sense if you look at things from his point of view. From some of his writing on the subject, it's clear he was dissatisfied with the standard textbook formulation of quantum physics, particularly the measurement problem, and thought that a hidden-variable model like Bohmian mechanics was the most natural way to resolve this. However, Bohmian mechanics is highly nonlocal and only works for nonrelativistic quantum physics, and Bell viewed the nonlocality of Bohmian mechanics as an impediment to getting it to work for quantum field theory in a way that respects relativity.

If you see things this way then investigating the locality or nonlocality of quantum physics, including getting opinionated about what is a "right" (i.e., relevant) or "wrong" definition of locality makes sense. On the other hand, if you don't have the same perspective as Bell (you think quantum physics is already fine the way it is, or issues with it would be better resolved in a different way than with a hidden-variable model) then it shouldn't make too much difference whether quantum physics is labelled "nonlocal" or not.

 

[Morbert]

Of course, you interpreted another statement by Weinberg as implying QM is local (or something that I couldn't agree as saying QM is local). I'd love to an actual quote by Weinberg where he denies quantum non-locality, or otherwise denies the obvious implications of Bell.

Just to be clear: I don't think Weinberg implies QM is local. I just don't think anything he says challenges the position that QM does not invoke any nonlocal action if the state is interpreted epistemically.

 

[DrChinese]

Since this is the interpretations forum: if you want strict adherence to Einsteinian locality, why not simply adopt one of those interpretations that feature retrocausal effects, or time symmetry, or similar (including the acausal Relational Blockworld). In those, the contexts/effects of the future in one way or another lead to quantum nonlocality even though no influence exceeds c at any time.

See how nicely I slipped my sales pitch in?

 

[DrChinese]

Just to be clear: I don't think Weinberg implies QM is local. I just don't think anything he says challenges the position that QM does not invoke any nonlocal action if the state is interpreted epistemically.

Not picking on you specifically with these comments... but I assume we are allowed a bit more latitude here.

I cannot tell you how many papers I've seen that use hand waving to say that Bell's Theorem is flawed in this way or that. Bell can't be circumvented either just because someone says the state is epistemic rather than ontic. Is there non-local action? I have severe doubts about that too. But thousands of experiments show there is not local realism at the base. And that means there IS quantum nonlocality (since that's the usual definition). It just remains for us to figure out HOW that happens. A system which stretches many kilometers in extent manages to take on a specific value of an observable instantaneously. And that value can be predicted exactly out of a very large set of possible values (assuming we are talking an observable such as momentum).

It's NOT some coincidence as is claimed by some here (mere "correlations"). Selection of a measurement basis by Alice leads to an exact predicted result by far distant Bob on that basis. It matters not whether Bob is measuring part of an entangled system with spatial extent*, or is now measuring a distant independent single particle. If it looks like a duck, waddles like a duck, and quacks like a duck: let's call it a duck. We are witnessing quantum nonlocality.

It should be obvious that NO theory can purport otherwise, regardless of construction. We need to face this challenge head on. I agree that QFT is the best we have. But it is flawed to the extent that it fails to admit quantum nonlocality (which is why I keep talking about models not being reality). I note that even those who believe MWI is correct cannot agree as to whether MWI is local or not (I say it is not of course).

Clearly this is a sticking point for many scientists. And probably explains why Weinberg says "There is a troubling weirdness about quantum mechanics." Apparently, he doesn't think there has been full resolution of the paradox.*like a big balloon that Alice is popping...

 

[PeterDonis]

QFT is the best we have. But it is flawed to the extent that it fails to admit quantum nonlocality

I don't think QFT fails to admit quantum nonlocality in the sense of Bell inequality violations. QFT predicts that the Bell inequalities will be violated under appropriate conditions.

Some physicists like to say that QFT is "local" because, for example, spacelike separated measurement operators commute (the results don't depend on the order in which they are done). But that is simply a different definition of the term "local". And as I pointed out way back in post #2 of this thread, I don't think arguments over definitions of words have much point. The point is the physics, and the physics is that (1) actual quantum systems violate the Bell inequalities, and (2) spacelike separated measurements on actual quantum systems commute. Those are the facts, and the challenge, as you rightly point out, is to figure out how those facts happen.

 

[RUTA]

The point is the physics, and the physics is that (1) actual quantum systems violate the Bell inequalities, and (2) spacelike separated measurements on actual quantum systems commute. Those are the facts, and the challenge, as you rightly point out, is to figure out how those facts happen.

Consider how many human-years have been spent working to “figure out how those facts happen” without resolution. By “how” we mean by “constructive efforts” (per Einstein). It’s interesting that the same is true of time dilation and length contraction in SR, where the physics community long ago decided to accept Einstein’s “principle account” and gave up looking for a “constructive” counterpart. It’s also interesting that Bell state entanglement follows from the same principle account, i.e., the relativity principle (no preferred reference frame). Perhaps the physics community will again eventually accept this principle account and give up looking for a constructive counterpart there, too.

 

[PeterDonis]

By “how” we mean by “constructive efforts” (per Einstein).

I see this Insight was just posted recently; I haven't read it yet, so I'll take the time to do so before trying to respond. I will probably respond in the comment thread for that Insight, which is here for anyone interested:

https://www.physicsforums.com/threa...on-of-the-mysteries-of-modern-physics.992927/

 

[vanhees71]

I accept that in a Bell test, the entangled system cannot be considered as two separate photons (or other particles). However, a system with such spatial extent as is used in Bell test (say 2 photons) cannot be considered as being "localized" either. In fact, I am not sure it even makes sense to say a single photon system is "localized" as in many cases, a single photon can be considered to have a very large* spatial extent.

I think that's the point! If you have two photons in an entangled state they are not localized (despite the fact that even a single photon cannot be strictly localized, because it hasn't even a position) but nevertheless correlated in the sense that you don't have a product of two single-photon states. Nevertheless that's not "non-locality" in a causal sense but what Einstein much more elegantly dubbed "inseparable", and Einstein emphasized that it is inseparability that bothered him most. It's also clear that this is only in tension with causality if you take the collapse of the state when measurements are done on the inseparated parts of the system as a physical process rather than an adaption of our description of the system after the measurement. His suggestion famously were hidden variables, but only with Bell's work this got a physical statement, i.e., a conjecture that could be tested by experiment, and immediately the experimental challenge was taken up by Aspect et al, and QT and inseparability turned out to be correct and not local hidden variable models (apparently to somewhat of a surprise for Bell). The conclusion thus indeed is that nature seams to allow for "inseparable" states in the sense of QT.

And in either case, we are witnessing quantum nonlocality. I can't imagine any definition of such systems that imply otherwise. We can have systems that are many kilometers in size, and still have perfect correlations in Bell tests with observers Alice and Bob.

As I said, the common naming this as nonlocality is confusing in the context with relativistic QFT, where one emphasizes from the very beginning the locality of interactions and microcausality. You can have perfect correlations between inseparable parts of a system at any distance though it is more and more difficult to maintain the entanglement by sufficiently isolating the entire system from the interaction with "the environment" to avoid decoherence.

So I would certainly say that if you retreat to the position that there is no quantum nonlocality as Alice and Bob perform their respective measurements, and that it is the system instead that has spatial (and/or temporal) extent: you haven't eliminated the quantum nonlocality, you have simply moved it.

The "nonlocality" (or rather in my preferred Einsteinian lingo "inseparability") is due to the state preparation rather than the measurement. That's an important point to understand that there's in fact no violation of Einstein locality in Bell tests. As discussed some time ago in these forums, this also holds for realizations of entanglement swapping, which is due to filtering according to local measurements preparing entanglement between parts of a quantum system which never have been interacting themselves.

And further: Alice can take the portion of an entangled system she receives... and teleport that entanglement FTL to another system as far distant as she likes. And again we can perform Bell tests on the newly enlarged entangled system, and get perfect correlations. That's a pretty good trick if there is no quantum nonlocality.*Perhaps even as large as the observable universe?

It's not FTL, because at the instant A does her local measurement it's only her who knows that she has teleported a state. B who does the measurement doesn't know it at this instant but only when A and B share their measurement protocols they can establish the corresponding correlations (on the then enlarged system). That's why it is so important to take the collapse not as a physical process as some Copenhagen-flavored interpretations seem to suggest but simply as an update of the state description, depending on (local) observations.

 

[vanhees71]

Consider how many human-years have been spent working to “figure out how those facts happen” without resolution. By “how” we mean by “constructive efforts” (per Einstein). It’s interesting that the same is true of time dilation and length contraction in SR, where the physics community long ago decided to accept Einstein’s “principle account” and gave up looking for a “constructive” counterpart. It’s also interesting that Bell state entanglement follows from the same principle account, i.e., the relativity principle (no preferred reference frame). Perhaps the physics community will again eventually accept this principle account and give up looking for a constructive counterpart there, too.

It's figured out, how those facts are to be perfectly described, namely by quantum (field) theory, discovered in 1925/26 by Heisenberg, Born, and Jordan.

 

[Morbert]

But thousands of experiments show there is not local realism at the base. And that means there IS quantum nonlocality (since that's the usual definition).

For the purposes of this thread I'd agree. I'll sometimes go to bat for Griffiths's local realist consistent histories but it's not too relevant here. Quantum nonlocality, as described above, is consistent with most mainstream interpretations.

It's NOT some coincidence as is claimed by some here (mere "correlations"). Selection of a measurement basis by Alice leads to an exact predicted result by far distant Bob on that basis. [...] A system which stretches many kilometers in extent manages to take on a specific value of an observable instantaneously.

I think all the subtleties lie in the sentences above.

According to a typical antirealist position, if Alice performs a measurement on this extended system, she will create a record of a measurement outcome in her apparatus, and this record will inform her expectations of measurements Bob might perform on this same extended system.

What she did not create, however, is a quantum property of the extended system corresponding to a specific value of an observable. If we adopt an antirealist position, then we must fully commit: QM does not describe properties of quantum systems, extended or otherwise. Instead it describes frequencies and correlations in experimental records that are produced in response to interactions with quantum systems.

 

[vanhees71]

Exactly. The "collapse" due to Alice's measurement is only that she adapts her description due to its outcome. Then she knows what Bob must find when measuring the corresponding 100% correlated property. E.g., taking the standard example of two photons in the singlet-polarization state then she knows, when she has measured H-polarization on her photon, Bob must measure helicity V-polarization if he measures in the same polarization direction. That's the 100% correlation described by the entanglement.

Of course Bob can also measure the polarization in any other direction. Then Alice can still predict the corresponding conditional probabilities for the outcome of Bob's measurement given here measurement result. What Bob measures for the full ensemble are just unpolarized photons (which is also what Alice measures). Only if Alice and Bob share their measurement protocols they can figure out that their photons were entangled, no matter whether both measure the polarization in the same or in a different direction.

Nothin changes instantaneously in a causal sense on Bob's photon due to Alice's measurement. The collapse is just the adaption of the description of the system after the measurement by Alice. Bob cannot instantaneously update his description, because it needs the information about Alice's outcome, and that Bob can not get instantly or by any faster than light signal. Of course I assume that standard relativistic QFT is correct, which by construction doesn't allow for FTL communication.

 

[martinbn]

If you deny non-locality, then what does Bell theorem really prove, in your opinion?

I might be wrong, but my impression is that no one denies the result of Bell. People only say that non-locality is a poor choice of terminology. Why use it if it is already in use! Because I am cynical, I think the reason is the BM adherents. BM is non-local, so they don't miss any opertunite to say that Bell + experiment have proven that nature is non-local.