I EPR and Non-Locality - For and Against

[RUTA]

I didn't bring up "acausal"; you did in your post #91. So I can't really respond to this. All I said was:

There MUST be some channel of some kind of communication, influence, selection or otherwise between Alice's setting or particle, and Bob's setting or particle. That being the mystery "channel" we seek to solve.

I have no idea what that is or looks like, and don't pretend I do (like some people). But how you can read that as me asserting anything more than the obvious, as embodied by the words of Weinberg (or Zeilinger or a hundred others who can state the mystery better than I) - I don't follow your reasoning. The combination of perfect correlations and violation of Bell Inequalities in hundreds of experiments point to the same mystery: quantum nonlocality exists and remains unexplained*.*Beyond the predictions of QM themselves. I am unaware of any predictions QFT makes on entanglement that go beyond QM, but that probably is a reflection of my ignorance on QFT... which is why I am participating here.

You guys post much faster than I can keep up with, but let me jump in here and use this post to make my point again. The mystery is, as DrChinese states, the existence of spacelike separated correlations that violate Bell’s inequality to 8 sigma (that’s the largest I’ve seen anyway) as predicted by QFT. Bell stated famously that ”correlations cry out for explanation.” While QFT predicts these Bell-inequality-violating correlations, QFT is just a mathematical formalism. So, unless you accept QFT’s mathematical description of the correlations as “explanatory,” QFT doesn’t solve the mystery, it only predicts it. Here is a nice quote from Fuchs in 2016:

Compare [quantum mechanics] to one of our other great physical theories, special relativity. One could make the statement of it in terms of some very crisp and clear physical principles: The speed of light is constant in all inertial frames, and the laws of physics are the same in all inertial frames. And it struck me that if we couldn't take the structure of quantum theory and change it from this very overt mathematical speak -- something that didn't look to have much physical content at all, in a way that anyone could identify with some kind of physical principle -- if we couldn't turn that into something like this, then the debate would go on forever and ever. And it seemed like a worthwhile exercise to try to reduce the mathematical structure of quantum mechanics to some crisp physical statements.

And Smolin in 2020:

So, my conclusion is that we need to back off from our models, postpone conjectures about constituents, and begin again by talking about principles.

Both of these quotes evidence the frustration of the foundations community with a debate that shows no sign of ever ending.

 

[PeterDonis]

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

If Weinberg were posting here, I'd ask him the same question I have already asked you, and which you have refused to answer: how can the state vector of the other subsystem "change" when it didn't even have a well-defined state vector before the measurement? (And, for extra credit, why should we care about the state vector rather than the density matrix, which everyone agrees does not change?)

This is what I mean about appealing to authority. You just read Weinberg's words and repeat them as though they were unchallengeable. I don't accept anything anyone says as unchallengeable. I apply my own intelligence to what they say to see what I think. And if I have a question about what is said, I ask it. Of course it's up to you whether you answer the question I ask, or even pay attention to it. But if you won't answer it, then I have no reason to change my opinion. (I might not even if you do answer it, but there's only one way to find that out.)

If you assert that QFT resolves the mystery

I have said no such thing. I have not said anything resolves any mystery. Others might have, but I haven't.

I have simply asked what seem to me to be obvious questions about assertions that others have made. The reason I ask is to see if anyone will answer them, and if so, what their answers are.

 

[PeterDonis]

Here is a nice quote from Fuchs in 2016

And Smolin in 2020

Can you give links? (If they were already given earlier in this thread, a pointer or at least a rough idea of the post number would be fine.)

 

[RUTA]

Can you give links? (If they were already given earlier in this thread, a pointer or at least a rough idea of the post number would be fine.)

Sorry, you can find those quotes with citations in this Insight.

 

[PeterDonis]

Sorry, you can find those quotes with citations in this Insight.

Ok, thanks!

 

[vanhees71]

I guess your definition of indistinguishable is different than mine. Experimentally, we can have entangled photons that have never interacted in the past, as they are from different sources. They are clearly identifiable, and can have markers (such as their wavelength) that identify them uniquely. They can be passed through filters to confirm that uniqueness.

If QFT is local and realistic, as has been said a hundred times, it's wrong. So tell me exactly where you consider it non-realistic? Non-realistic meaning there is observer dependent reality.

Photons are indistinguishable bosons. That's a very well defined mathematical property and has nothing to do with any interpretational issues.

I don't know, what "realistic" means. It's an undefined philosophical expression. QFT is by construction a local relativistic QFT. Local means that

(a) the Heisenberg-picture field operators transform under proper orthochronous Poincare transformations as their classical analogues. The transformation property is determined by the unitary representation of this group the corresponding field operators refer to.

(b) the Hamilton density commutes with any operator representing a local observable if the arguments of these are space-like separated

Further one assumes

(c) the Hamilton is bounded from below. By convention the ground-state energy (vacuum energy) is choosen to be 0.

From this the spin-statistics theorem follows, i.e., you have to quantize the field theory either as bosons (if the spin of the field is integer) or fermions (if the spin of the field is half-integer). This implies indistinguishability of the quanta represented by these fields.

There is no observer-dependent reality. States of quantum systems are well-defined, and I don't know, what entanglement swapping, which you seem to refer to once more, has to do with all this. There's no problem to describe entanglement swapping within standard local relativistic QFT.

Once more the claim that the outcome of local measurements of B's photon would instantaneously be changed by the outcome of local measuremnts by A on a photon pair prepared in a Bell state, cannot be made within local relativistic QFTs. What B measures are unpolarized photons. What A measures are unpolarized photons. There's no dependence of this result on the A's or B's choice which observable they measure (e.g., they can measure the polarization of the photons in arbitrary directions). Only if they compare measurement protocols one finds the correlations between A's and B's outcome of measurements as predicted by QFT for the photons prepared in this entangled state. This implies the violation of Bell's inequality and related properties of this kind. There's not tension between Einstein causality and locality in the sense of relativistic local QFTs whatsoever. Inseparability is implied by any kind of quantum theory, including local relativistic QFTs and no contradiction with locality either.

I call this "realistic", because it's in accordance with all known very accurate experiments on this issue. For me there's no other notion of "reality" which makes sense as a scientific rather than a philosophical notion.

 

[vanhees71]

The physics community considers this an open problem (to the extent they think about it at all).

A small very specialized part of the philosophy-of-physics community considers this as an open problem. Practitioners of standard relativistic QFT either don't care about this apparent problem or, if they are interested in it, don't see at as a problem, because relativistic QFT describes very well all observations. Particularly experimentalists in the quantum optics and/or AMO community I know, have a very clear scientific probabilistic-only interpretation of QT. I don't know anybody who claims there is a contradiction between the usual notion of locality of relativistic QFTs (see my previous posting) and experiments.

 

[Sunil]

I take the view correlation does not imply influences.

Yes, this is a message which has to be explained and repeated many times in everyday life and usual applications of statistics that there exist other explanations for correlations. Namely, common causes.

But in the case of the Bell inequalities, those common cause explanations are excluded by the theorem, common causes cannot lead to violations of the Bell inequalities. So, in this case correlation does imply causal influences.

 

[Sunil]

I don't know, what "realistic" means. It's an undefined philosophical expression. QFT is by construction a local relativistic QFT.

The notion of realism is precisely defined in the EPR paper (the EPR criterion of reality) as well as in various presentations of the theorem. It is what is necessary to obtain the formula

$E(AB|a,b) = \int A(a,b,\lambda) B(a,b,\lambda) \rho(a,b,\lambda) d \lambda$

Then you have to exclude superdeterminism to get

$E(AB|a,b) = \int A(a,b,\lambda) B(a,b,\lambda) \rho(\lambda) d \lambda$

and Einstein causality in a stronger form than what holds in QFT to obtain

$E(AB|a,b) = \int A(a,\lambda) B(b,\lambda) \rho(\lambda) d \lambda$

There is Stone's theorem that for every Boolean algebra there exists a space $$\Lambda$$ so that it can be embedded into the Boolean algebra of subsets of this space, so the existence of this space of possible states of reality is a triviality too. But, trivial or not, the requirement that

$E(AB|a,b) = \int A(a,b,\lambda) B(a,b,\lambda) \rho(a,b,\lambda) d \lambda$

has to follow from realism is precise enough.

Once more the claim that the outcome of local measurements of B's photon would instantaneously be changed by the outcome of local measuremnts by A on a photon pair prepared in a Bell state, cannot be made within local relativistic QFTs.

That QFT is not realistic is nothing in favor of QFT. Whatever, realistic interpretations of QFT exist too.

I call this "realistic", because it's in accordance with all known very accurate experiments on this issue. For me there's no other notion of "reality" which makes sense as a scientific rather than a philosophical notion.

You are not free to use an already established scientific notion for something completely different.

 

[PeterDonis]

and Einstein causality in a stronger form than what holds in QFT to obtain

What follows here is the same equation as the one just above it, so I think you have a typo somewhere. Perhaps you meant the equation after this sentence to read as follows?

$E(AB|a,b) = \int A(a,\lambda) B(b,\lambda) \rho(\lambda) d \lambda$

 

[Sunil]

If Weinberg were posting here, I'd ask him the same question I have already asked you, and which you have refused to answer: how can the state vector of the other subsystem "change" when it didn't even have a well-defined state vector before the measurement? (And, for extra credit, why should we care about the state vector rather than the density matrix, which everyone agrees does not change?)

If it didn't even have a state vector before, but has one after, the description of the state of the system has, obviously, changed. From "state without state vector, described by a non-pure density matrix" to "state with state vector". So this is at best a minor quibble about words, which can be used only if someone has sloppily applied "state vector" to that part of an entangled state.

(And, for extra credit, why should we care about the state vector rather than the density matrix, which everyone agrees does not change?)

Because we are interested to find out whatever can be found out. And the density matrix, of course, changes if it takes into account the additional information measured at the place far away. So, it is unchanged only for those who don't have the complete information.

What follows here is the same equation as the one just above it, so I think you have a typo somewhere. Perhaps you meant the equation after this sentence to read as follows?

$E(AB|a,b) = \int A(a,\lambda) B(b,\lambda) \rho(\lambda) d \lambda$

Of course. Thanks, corrected.

 

[Sunil]

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).

Retrocausality would be an exceptional, extraordinary hypothesis. But for an extraordinary claim one would need extraordinary evidence.

There is no such extraordinary evidence, given that what is known allows for a simple and straightforward causal explanation using a hidden preferred frame.

Non-locality of such preferred frame explanations is not a serious objection, given that non-local theories may appear as limits of local theories when the limiting speed goes to infinity, and against local theories with a hidden preferred frame with limiting speed greater than c there is nothing to object in principle. Moreover, we have enough strong hints that GR is no fundamental (it has singularities, it is non-renormalizable, thus, good for an effective QFT but not more), so that there is also no base for assuming that symmetries of GR should be really fundamental, instead of being approximations.

 

[PeterDonis]

the description of the state of the system has, obviously, changed

This is the same problem. If the system didn't even have a well-defined state vector before, then if "description" means "state vector", it didn't have a well-defined "description" before the measurement. "The description has changed" is not a valid description of this situation.

From "state without state vector, described by a non-pure density matrix" to "state with state vector".

No, that's not correct. As has already been pointed out, the density matrix of Bob's particle does not change when Alice makes her measurement. It only changes when Bob makes his measurement.

this is at best a minor quibble about words

No, it isn't. See above.

the density matrix, of course, changes if it takes into account the additional information measured at the place far away

Bob can only take that information into account when he learns it. He doesn't learn it when Alice makes her measurement. He only learns it when the information about Alice's result reaches him, at the speed of light (or slower, depending on how the information is transmitted).

it is unchanged only for those who don't have the complete information.

Which includes Bob. (And if we consider Alice's knowledge about Bob's measurement result, we get the same issue.)

 

[Sunil]

This is the same problem. If the system didn't even have a well-defined state vector before, then if "description" means "state vector", it didn't have a well-defined "description" before the measurement. "The description has changed" is not a valid description of this situation.

If the system doesn't even have a well-defined state vector, the "description" is the density matrix.

No, that's not correct. As has already been pointed out, the density matrix of Bob's particle does not change when Alice makes her measurement. It only changes when Bob makes his measurement.

No. Once Alice has made the measurement, the state of Bob's part of the system has a state vector, which is identified by the measurement made by Alice and its result.

Bob can only take that information into account when he learns it.

So what? The state of a system is not the state of the mind of Bob, but what is known about the preparation in the whole universe.

Which includes Bob. (And if we consider Alice's knowledge about Bob's measurement result, we get the same issue.)

But in QT we consider the whole state, which includes the knowledge about the preparation procedure available at a given moment of time in the whole universe.

 

[Lynch101]

The only people I have ever heard talk about QFT as if it is "problem solved" are a few people here.
...
There are no interpretations of QFT that solve Weinberg's "weirdness" [what I call "mystery"]. As best I can tell, it is silent; and further, I don't see where it adds anything past what earlier QM does. The physics community considers this an open problem (to the extent they think about it at all).
...
If you assert that QFT resolves the mystery, then please say so outright

I've read through the whole thread and I was going to start a separate thread with questions about what I've read here, but I thought that might be redundant. If it would be better to start a new thread that's no problem, I can do that, but I'll post here for the time being. I'll outline my understanding and maybe someone can - if they are so inclined - point out my error.

It reads to me as though you and others are talking past each other somewhat. It reads as though two different questions are being addressed:
@DrChinese: Is nature non-local?
Others: Is QM non-local?

The argument from the others appears to be: it cannot be stated that QM is non-local because QFT is, by construction, a local theory. Alice and Bob both make local measurements and can only share information at or slower than, c. The observed correlations are precisely those predicted by QFT, therefore it cannot be said that QM necessitates non-locality.

@DrChinese, you seem to be speaking to the explanation of those correlations which, as you have said, QFT does not provide.

To me, this reads as though it goes back to the question of completeness. Is QFT a complete description of nature? If the mathematics of QFT are taken to be nothing more than a computational tool, which make predictions about the outcomes of experiments, then it would seem not to be complete. That doesn't mean that it is incomplete in the manner that EPR suggested, but in the more general sense of not describing the system prior to its interaction with a measurement device.

Lee Smolin puts it better than I can.

Edited to shorten quote:

despite the successes of quantum field theory, many physicists, beginning with Einstein, have wanted to go beyond it to a deeper theory that gives a complete description of each individual experiment--which, as we have seen, no quantum theory does. Their searches have consistently found an irreconcilable conflict between quantum physics and special relativity.

As long as we’re just checking the predictions of quantum mechanics at the level of statistics, we don’t have to ask how the correlations were actually established. It is only when we seek to describe how information is transmitted within each entangled pair that we need a notion of instantaneous communication.

 

[bhobba]

But in the case of the Bell inequalities, those common cause explanations are excluded by the theorem, common causes cannot lead to violations of the Bell inequalities. So, in this case correlation does imply causal influences.

It proves classical correlations do not apply unless you have non local influences. But QM is a generalised probability model where the statistics of classical correlations do not hold, nor would you a-priori expect them to. It does not imply there must be non-local influences - simply that it is not standard probability theory, but a generalisation of it that does not always have the same properties.

Thanks
Bill

 

[RUTA]

To me, this reads as though it goes back to the question of completeness. Is QFT a complete description of nature? If the mathematics of QFT are taken to be nothing more than a computational tool, which make predictions about the outcomes of experiments, then it would seem not to be complete. That doesn't mean that it is incomplete in the manner that EPR suggested, but in the more general sense of not describing the system prior to its interaction with a measurement device.

Lee Smolin puts it better than I can.

Edited to shorten quote:
despite the successes of quantum field theory, many physicists, beginning with Einstein, have wanted to go beyond it to a deeper theory that gives a complete description of each individual experiment--which, as we have seen, no quantum theory does. Their searches have consistently found an irreconcilable conflict between quantum physics and special relativity.

As long as we’re just checking the predictions of quantum mechanics at the level of statistics, we don’t have to ask how the correlations were actually established. It is only when we seek to describe how information is transmitted within each entangled pair that we need a notion of instantaneous communication.

It looks like you're referring to two distinct objections to QT based on Bell state entanglement -- locality and completeness -- as in this Mermin article in Physics Today. I argue against both objections in this Insight.

 

[PeterDonis]

Once Alice has made the measurement, the state of Bob's part of the system has a state vector, which is identified by the measurement made by Alice and its result.

Perhaps this is true in your preferred interpretation. But you cannot just help yourself to the claim that it is true, period.

The state of a system is not the state of the mind of Bob, but what is known about the preparation in the whole universe.

Same response as above.

in QT we consider the whole state, which includes the knowledge about the preparation procedure available at a given moment of time in the whole universe.

Same response as above. Also note that this viewpoint can't possibly be valid in a relativistic model, since there is no such thing as a unique "given moment of time in the whole universe" in a relativistic model.

 

[Lynch101]

It looks like you're referring to two distinct objections to QT based on Bell state entanglement -- locality and completeness -- as in this Mermin article in Physics Today. I argue against both objections in this Insight.

Thanks RUTA, I had read your insight article and as much as I understood it, it was very interesting, but I don't quite understand it to the level of seeing how it refutes the challenge of incompleteness.

I'm talking about a specific attribute of "anti-realist" interpretations of the mathematics, which says that the mathematics is simply a tool which makes predictions about the outcomes of experiments. If this is the case, then the mathematics does not describe the system prior to its interaction with the measurement device, which would render it an incomplete description of nature.

 

[Sunil]

Perhaps this is true in your preferred interpretation. But you cannot just help yourself to the claim that it is true, period.

No, this is true in the minimal interpretation, and the minimal interpretation is not my preferred interpretation. What is true in the minimal interpretation is true in QT.

The minimal interpretation does not contain relations between particular measurements and observers localized somewhere. We have measurements, defined by operators, and a measurement leads to a change of the state. There is no measurement process, we have a state before the measurement $|\psi\rangle = \sum_\alpha c_\alpha| \psi_\alpha\rangle$, then a miracle the measurement happens, and after this we have a state after the measurement, which is, with probability $|c_\alpha|^2$, the state $|\psi_\alpha\rangle$. The minimal interpretation does not contain any notion of space, which could be used to define something like a local measurement, or a measurement made only in the environment of Alice but not of Bob.

There are also no different notions of time in the minimal interpretation, there is one time, and the change of the state during a measurement happens immediately, without any delay, and without any specification how it happens.

Also note that this viewpoint can't possibly be valid in a relativistic model, since there is no such thing as a unique "given moment of time in the whole universe" in a relativistic model.

It can be valid in a relativistic model with a hidden preferred frame. So, it is only in conflict with a particular interpretation of relativity, which forbids, for some metaphysical reasons, a hidden preferred frame. Such a conflict with relativistic metaphysics does not change QT in its minimal interpretation.

 

[RUTA]

It can be valid in a relativistic model with a hidden preferred frame. So, it is only in conflict with a particular interpretation of relativity, which forbids, for some metaphysical reasons, a hidden preferred frame. Such a conflict with relativistic metaphysics does not change QT in its minimal interpretation.

That is my understanding, too. Relativity of simultaneity in SR does not rule out a preferred frame. Nor does SR rule out superluminal causation. Indeed, some particle theorists have postulated the existence of tachyons. One reason people claim that SR is incompatible with superluminal causation is that they demand causes precede effects and subscribe to no preferred reference frame. Those two assumptions/preferences then rule out a causal connection between spacelike separated events.

 

[RUTA]

Thanks RUTA, I had read your insight article and as much as I understood it, it was very interesting, but I don't quite understand it to the level of seeing how it refutes the challenge of incompleteness.

I'm talking about a specific attribute of "anti-realist" interpretations of the mathematics, which says that the mathematics is simply a tool which makes predictions about the outcomes of experiments. If this is the case, then the mathematics does not describe the system prior to its interaction with the measurement device, which would render it an incomplete description of nature.

The argument is simply that if you believe time dilation and length contraction are "completely" explained by the light postulate (NPRF applied to the speed of light c), then you should believe that violations of the Bell inequality to the Tsirelson bound are "completely" explained by "conservation per NPRF" (NPRF applied to Planck's constant h). Both the light postulate and the Bell state correlations are based in NPRF, they just involve two different constants, c and h. So, clearly, this argument fails for those who still seek "constructive efforts" to explain time dilation and length contraction

The power of principle explanation is summed up nicely by the quote I just today added to the Insight from Pauli (reference in the Insight):

Understanding nature surely means taking a close look at its connections, being certain of its inner workings. Such knowledge cannot be gained by understanding an isolated phenomenon or a single group of phenomena, even if one discovers some order in them. It comes from the recognition that a wealth of experimental facts are interconnected and can therefore be reduced to a common principle. In that case, certainty rests precisely on this wealth of facts. The danger of making mistakes is the smaller, the richer and more complex the phenomena are, and the simpler is the common principle to which they can all be brought back. ... The ability to predict is often the consequence of understanding, of having the right concepts, but is not identical with 'understanding'.

But, this is just personal preference.

 

[Lynch101]

The argument is simply that if you believe time dilation and length contraction are "completely" explained by the light postulate (NPRF applied to the speed of light c), then you should believe that violations of the Bell inequality to the Tsirelson bound are "completely" explained by "conservation per NPRF" (NPRF applied to Planck's constant h). Both the light postulate and the Bell state correlations are based in NPRF, they just involve two different constants, c and h. So, clearly, this argument fails for those who still seek "constructive efforts" to explain time dilation and length contraction

Ah, OK. I had understood that point in the insights article then. I just wasn't sure if there was something I was missing.

Leaving relativity aside for the moment, I don't think that point addresses the fundamental issue relating to anti-realist interpretations of QFT, that if the mathematics of QFT only describe/predict the [classical level] observable outcomes of experiments then, by necessity, this would render it an incomplete description of nature because it doesn't describe the system prior to its interaction with the measurement device.

The power of principle explanation is summed up nicely by the quote I just today added to the Insight from Pauli (reference in the Insight):

Einstein's thoughts on the difference between constructivist and principled explanations are also worth noting. Emphasis is my own.

When we say that we have succeeded in understanding a group of natural processes, we invariably mean that a constructive theory has been found which covers the processes in question.
...
The advantages of the constructive theory are completeness, adaptability, and clearness, those of the principle theory are logical perfection and security of the foundations. The theory of relativity belongs to the latter class.

 

[PeterDonis]

this is true in the minimal interpretation

No, it isn't. The minimal interpretation does not say Bob's description of his particle changes when Alice makes her measurement. It only says it changes when Bob makes his measurement.

What you are calling "the minimal interpretation" is your personal preference, not the minimal interpretation.

It can be valid in a relativistic model with a hidden preferred frame.

That's certainly not the minimal interpretation either. The minimal interpretation of QFT has no preferred frame, since that would be additional structure added that does not affect any predictions and therefore gets scraped right off by Occam's razor.

 

[RUTA]

Leaving relativity aside for the moment, I don't think that point addresses the fundamental issue relating to anti-realist interpretations of QFT, that if the mathematics of QFT only describe/predict the [classical level] observable outcomes of experiments then, by necessity, this would render it an incomplete description of nature because it doesn't describe the system prior to its interaction with the measurement device.

Einstein's thoughts on the difference between constructivist and principled explanations are also worth noting. Emphasis is my own.

Yes, those who still advocate for a constructive account of SR provide many Einstein quotes showing his support for a "completion" of SR, i.e., a constructive account of SR. And, yes, there is little doubt that in the EPR paper he is advocating for a constructive account of Bell state entanglement, i.e., it is "incomplete" without a constructive account. The question is, what would Einstein say today about the situation?

Today, we know that the same principle (NPRF) accounts for time dilation and length contraction in SR and Bell state entanglement and the Tsirelson bound in QM. And, 65 years after his death, we still have no accepted constructive accounts of these "mysteries." Keep in mind he knew nothing of the Bell inequality and its experimental violation by the Bell spin states.

Personally, I advocate for the explanatory power of principle explanation only because of my view that physics is concerned with providing constraints on experience, as explained in this 2020 Entropy paper. Since that deals with consciousness studies, i.e., it's beyond physics, I don't expound on that in PF. Essentially, this redefines what it means to provide a constructive account, so it is not "anti-realist".

 

[bhobba]

No, this is true in the minimal interpretation,

What interpretation is that? I know a few I would describe as minimalist eg Bayesian and Ensemble. I would not call Copenhagen minimalist since there are various versions. I maintain discussion of issues like the papers I started this thread with is best done using a stated interpretation so it does not devolve into a quagmire of unidentified assumptions.

The minimal interpretation does not contain any notion of space,

Hmmm. How do you take a partial derivative wrt to x,y and z as per the Schrodinger's equation without the concept of space? In fact it would play havoc with Wigner's Theorem.

So, it is only in conflict with a particular interpretation of relativity, which forbids, for some metaphysical reasons, a hidden preferred frame. Such a conflict with relativistic metaphysics does not change QT in its minimal interpretation.

The modern view of relativity as space time geometry does not forbid a preferred frame - it merely says it is superfluous as far as the laws of physics are concerned. It's in the statement of the POR which is the laws of physics are the same in any inertial frame. If there was a preferred frame then its 'wind' would be frame dependant. If the POR is true, this 'wind' can not affect the laws of physics. If it has no affect on that why have it in the first place ie it is superfluous. Those that want a preferred frame seem to want it for philosophical not physical reasons.

Thanks
Bill

 

[PeterDonis]

What interpretation is that?

PF's consensus presentation of the minimal interpretation is here:

https://www.physicsforums.com/insights/the-7-basic-rules-of-quantum-mechanics/

Postulate 7 is the one most relevant to this discussion. It needs to be read carefully, particularly the phrase "successive, non-destructive projective measurements"; Alice's and Bob's spacelike separated measurements in the scenario under discussion are not "successive, non-destructive projective measurements" under this postulate.

 

[Lynch101]

The question is, what would Einstein say today about the situation?

It would certainly be interesting to find out.

65 years after his death, we still have no accepted constructive accounts of these "mysteries."

And there might be a fundamental limit on how far we can probe nature, meaning that such a complete description of nature, as explains these "mysteries", simply is not possible.

Essentially, this redefines what it means to provide a constructive account, so it is not "anti-realist".

Apologies, I keep coming back to this point but I think it is key to the question of completeness - and this only really applies to "anti-realist"/instrumentalist interpretations of QFT: if the mathematics is only a computational tool which calculates predictions for the observable outcomes of experiments then it does not, by necessity, describe the state prior to its interaction with the measurement device. If does not describe the system prior to this, then it cannot be a complete description of nature.

I may not be understanding your position, however. Are you saying that the interpretation you are advocating does not treat the mathematics purely instrumentally?

 

[bhobba]

I'm talking about a specific attribute of "anti-realist" interpretations of the mathematics, which says that the mathematics is simply a tool which makes predictions about the outcomes of experiments.

I would describe that as realist. I would describe anti-realist as a view like Wittgenstein's where he thinks it is about nothing. Turing took him to task on that one - saying bridges would fall over etc if it wasn't true. But either way this is not a subject we discuss as it really is philosophy. We simply think of mathematics as part of models. It s the models that are true or false - what the math is we leave open.

Thanks
Bill

 

[bhobba]

PF's consensus presentation of the minimal interpretation is here:

I should have remembered that. Sticking to that certainly makes things easier to discuss.

Thanks
Bill

 

[Lynch101]

I would describe that as realist. I would describe anti-realist as a view like Wittgenstein's where he thinks it is about nothing. Turing took him to task on that one - saying bridges would fall over etc if it wasn't true. But either way this is not a subject we discuss as it really is philosophy. We simply think of mathematics as part of models. It s the models that are true or false - what the math is we leave open.

Thanks
Bill

Thanks Bill. I'm using the term "anti-realist" in the sense it was explained in this thread on anti-realist interpretations of QM. As it was explained there, the "anti-realist" interpretation is not anti-realist in the strict philosophical sense of being juxtaposed with realism, rather it says that the mathematical elements of the theory don't correspond to elements of reality. This might be contrasted with naiive instrumentalism which is agnostic on the question of whether or not the mathematics represents an underlying ontology.

In this sense of "anti-realism", the mathematics of QFT are taken to be nothing but a tool to calculate the predictions for the outcomes of experiments. That is, they only tell us the probability of a particle registering on a classical level measurement device. Such anti-realist interpretations would, by necessity, be incomplete descriptions of nature because they don't describe the system prior to its interaction with the measurement device.

To my untrained eye, it appears that this is where some people are talking at cross purposes in this thread. On one hand, there are those who are arguing that QM does not necessitate non-locality because QFT is a local theory, by construction. It doesn't appear as though this particular point is being disputed.

The issue appears to arise when people call for an explanation of the correlations observed in experiments. This, to my mind, appears to be a slightly different question which is being conflated with the question, does QM necessitate non-locality? A more accurate question might be, does nature necessitate non-locality?

It appears that QM doesn't necessitate non-locality because QFT is local, but QFT doesn't appear to be a complete description of nature. Any attempt at a more complete description of nature appears to necessitate non-locality.

At least, that is how I have read the arguments in this thread.

 

[PeterDonis]

It appears that QM doesn't necessitate non-locality because QFT is local

As has already been pointed out multiple times (the first time was me, way, way back in post #2 of this thread--it would be really nice if people would read it before blithely using the term "nonlocality"), this depends on what you mean by "nonlocality".

QFT says that spacelike separated measurements commute; it is "local" in this sense.

QFT also predicts violations of the Bell inequalities; it is "nonlocal" in this sense.

Any attempt at a more complete description of nature appears to necessitate non-locality.

Again, it depends on what you mean by "nonlocality". See above.

Also, both of the statements that QFT agrees with, above, are experimental facts; so any more complete description of nature must still agree with them.

 

[Lynch101]

As has already been pointed out multiple times (the first time was me, way, way back in post #2 of this thread--it would be really nice if people would read it before blithely using the term "nonlocality"), this depends on what you mean by "nonlocality".

QFT says that spacelike separated measurements commute; it is "local" in this sense.

QFT also predicts violations of the Bell inequalities; it is "nonlocal" in this sense.

Again, it depends on what you mean by "nonlocality". See above.

From my reading of it, it isn't simply the predicted violations of Bell inequalities that render the theory non-local, it is the attempted explanation of a mechanism to account for the observed correlations which appears to necessitate non-locality, in the sense that a measurement performed in one location has an instantaneous effect in a spatially separated location.

"Anti-realist" interpretations of QFT don't seem to attempt to explain how or why the observed correlations deviate from the predictions of classical physics, they simply predict that they will. But, such anti-realist interpretations appear to be incomplete descriptions of nature. The attempted explanations, as opposed to the simple predictions, of the observed correlations appear to require some form of FTL influences - but not of the sort that can be used to transmit signals.

Again, this is where it appears to me that people are talking past each other. While QM/QFT might not necessitate non-locality, of the sort @DrChinese appears to be advocating, nature herself might.

Lee Smolin talks about this in Time Reborn (p.142)

As long as we’re just checking the predictions of quantum mechanics at the level of statistics, we don’t have to ask how the correlations were actually established. It is only when we seek to describe how information is transmitted within each entangled pair that we need a notion of instantaneous communication. It’s only when we seek to go beyond the statistical predictions of quantum theory to a hidden-variables theory that we come into conflict with the relativity of simultaneity.

This means giving up the relativity of simultaneity and embracing its opposite: that there is a preferred global notion of time. Remarkably, this does not require overthrowing relativity theory; it turns out that a reformulation of it is enough. The heart of the resolution is a new and deeper way of understanding general relativity theory which reveals a new conception of real time.

 

[RUTA]

Apologies, I keep coming back to this point but I think it is key to the question of completeness - and this only really applies to "anti-realist"/instrumentalist interpretations of QFT: if the mathematics is only a computational tool which calculates predictions for the observable outcomes of experiments then it does not, by necessity, describe the state prior to its interaction with the measurement device. If does not describe the system prior to this, then it cannot be a complete description of nature.

I may not be understanding your position, however. Are you saying that the interpretation you are advocating does not treat the mathematics purely instrumentally?

Once you understand reality per neutral monism (as in our Entropy paper) with physics providing the constraints on experience, fundamental explanation is not rooted in causal mechanisms. Certainly some principles/constraints have corresponding causal mechanisms, e.g., Fermat's principle has Snell's law, but since causal mechanisms are not fundamental, it's perfectly ok that some principles/constraints do not have corresponding causal mechanisms. Consequently, time dilation and length contraction are explained fundamentally by NPRF, just like Bell state entanglement and the Tsirelson bound, without any corresponding causal mechanisms (no ether, no superluminal signals, etc.). Once you have the most fundamental principles/constraints on experience, you have a complete understanding of reality according to this ontology. That's not instrumentalism, because we have provided an ontology. It's just an ontology rooted in principles/constraints rather than causal mechanisms. We wrote an entire book arguing for this type of ontology (Beyond the Dynamical Universe), but the Entropy paper is a sufficient summary

 

[RUTA]

To summarize again, Bell state entanglement does not entail that QT is incomplete or inconsistent with SR. As we argue in our book, papers, and Insights, the many "mysteries" of modern physics don't indicate a need to "fix" it. While modern physics is certainly not finished, what exists of it now can be seen as beautifully comprehensive and coherent with the appropriate view of physics.

 

[PeterDonis]

it isn't simply the predicted violations of Bell inequalities that render the theory non-local

Did you even read my post? Did you go back and read post #2 in the thread?

The term "non-local" does not have a single unique meaning. It means different things to different people.

"Violates the Bell inequalities" is one of the multiple meanings this term has been used to denote. So to say that violations of the Bell inequalities don't render the theory non-local is meaningless; it could be true or false, depending on what you decide "non-local" means.

So the best thing to do is to not use the term at all in this discussion.

 

[Jimster41]

QFT says that spacelike separated measurements commute; it is "local" in this sense.

QFT also predicts violations of the Bell inequalities; it is "nonlocal" in this sense.

Do like all of them commute? If so, is it pair-wise, n-wise, all-n-wise? Or only ones Alice and Bob set up real careful?

 

[PeterDonis]

Do like all of them commute?

In QFT, any measurement operators at spacelike separated events commute.

 

[Morbert]

"Anti-realist" interpretations of QFT don't seem to attempt to explain how or why the observed correlations deviate from the predictions of classical physics, they simply predict that they will. But, such anti-realist interpretations appear to be incomplete descriptions of nature. The attempted explanations, as opposed to the simple predictions, of the observed correlations appear to require some form of FTL influences - but not of the sort that can be used to transmit signals.

Lee Smolin talks about this in Time Reborn (p.142)

It is only when we seek to describe how information is transmitted within each entangled pair that we need a notion of instantaneous communication.

Anti-realists would not grant Smolin the implicit assumption he is making. He is assuming the system consists of a pair of spacelike separated divisible objects. If this is what the system really is, then quantum theory is incomplete in the sense you describe. But that is not necessarily what the system really is.

Asher Peres makes a subtle point in his book "Quantum Theory: Methods and Concepts": Quantum theory is universal insofar as it can describe the physics of any system, but it does not offer a single closed description of every system. By this, he means the application of quantum theory to any system (e.g. one the scientist wishes to observe) is only meaningful in the context of an ancillary "exophysical" system to which the theory is not applied (e.g. the scientist). If this is the character of physical descriptions of nature, then quantum theory is complete insofar as there is no physics it does not account for. If you insist on a closed ontic account of nature, then quantum theory is incomplete. But you should also justify this insistence.

 

[Lynch101]

Did you even read my post? Did you go back and read post #2 in the thread?

The term "non-local" does not have a single unique meaning. It means different things to different people.

I did, but I thought it was pretty clear which meaning @DrChinese and Lee Smolin were referring to, and therefore the meaning I was referencing.

"Violates the Bell inequalities" is one of the multiple meanings this term has been used to denote. So to say that violations of the Bell inequalities don't render the theory non-local is meaningless; it could be true or false, depending on what you decide "non-local" means.

There is a pretty obvious connection between the definition of non-local as involving instantaneous (or FTL) influences and the definition of non-local as "Violates the Bell inequalities". The former is the attempt to explain the latter and/or the latter is evidence of the former. This might lead us to suspect that there is a slight misappropriation of the term "non-local" when trying to define it as "Violates the Bell inequalities".

To try to elaborate: We have the experimental evidence of violations of Bell inequalities and know that these go against the predictions of classical physics. If we then try to explain how these violations of Bell inequalities occur and the answer we arrive at is: there must be some sort of instantaneous (or FTL) influence occurring, whereby a measurement in location A has an instantaneous effect at location B (which is spatially separated). This means that actions, such as measurements, don't simply exert influence in their locality at a maximum speed of c. This instantaneous (or FTL) influence we refer to as non-local - to contrast it with the idea that actions can only exert influence in their locality at a maximum speed of c.

To then suggest that "non-locality" is just another name for the violations of Bell inequalities seems to conflate the effect with the cause or the observed phenomenon with the attempt to explain how the observed phenomenon is possible, given that it goes against classical physics.

Do the other definitions of "non-local" share a similar connection?

So the best thing to do is to not use the term at all in this discussion.

Or, it might be worth exploring which term was first in use, which one offers the most explanatory power, and whether or not there are connections between the different definitions, as above, which can be delineated.

 

[Lynch101]

Once you understand reality per neutral monism (as in our Entropy paper) with physics providing the constraints on experience, fundamental explanation is not rooted in causal mechanisms. Certainly some principles/constraints have corresponding causal mechanisms, e.g., Fermat's principle has Snell's law, but since causal mechanisms are not fundamental, it's perfectly ok that some principles/constraints do not have corresponding causal mechanisms. Consequently, time dilation and length contraction are explained fundamentally by NPRF, just like Bell state entanglement and the Tsirelson bound, without any corresponding causal mechanisms (no ether, no superluminal signals, etc.). Once you have the most fundamental principles/constraints on experience, you have a complete understanding of reality according to this ontology. That's not instrumentalism, because we have provided an ontology. It's just an ontology rooted in principles/constraints rather than causal mechanisms. We wrote an entire book arguing for this type of ontology (Beyond the Dynamical Universe), but the Entropy paper is a sufficient summary

Thanks RUTA, I am interested in the idea of monism, so I'll certainly start by checking out the paper.

We might be talking at cross purposes here because I am only talking about anti-realist/instrumental interpretations of QM i.e. those interpretations which say that the mathematics ONLY allows us to predict the observable outcomes of experiments i.e. the probability that a particle will register at a particular measurement device. These interpretations would be incomplete descriptions of nature for the reasons mentioned, they don't describe the system prior to its interaction with the measurement device.

Does the interpretation you are proposing describe the system prior to measurement?

 

[Lynch101]

Anti-realists would not grant Smolin the implicit assumption he is making. He is assuming the system consists of a pair of spacelike separated divisible objects. If this is what the system really is, then quantum theory is incomplete in the sense you describe. But that is not necessarily what the system really is.

But, if we don't grant Smolin his assumptions and we are left with the mathematical framework which only predicts the probability of a particle registering on a particular measurement device, then we are left without a description of the system prior to its interaction with the instrument, which would mean we have an incomplete description of nature.

Are there attempted explanations of the observed correlations which don't necessitate non-locality (in the FTL sense)? I've encountered the idea in multiple places that, in order to explain the violations of Bell inequalities we must give up one or more of the following:
1) Locality (in the FTL sense)
2) Realism (in the strict philsophical sense)
3) Local realism
4) Free Will

Is that accurate?

By this, he means the application of quantum theory to any system (e.g. one the scientist wishes to observe) is only meaningful in the context of an ancillary "exophysical" system to which the theory is not applied (e.g. the scientist). If this is the character of physical descriptions of nature, then quantum theory is complete insofar as there is no physics it does not account for. If you insist on a closed ontic account of nature, then quantum theory is incomplete. But you should also justify this insistence.

Apologies, there might be a nuance in there that I'm not quite picking up on because I don't quite follow how it would address the point I've been making.

I'm not sure if I'm insisting on a closed ontic account of nature, or not, genuinely. I just know that the point I am making is fairly rudimentary: if a theory [or interpretation of that theory] only makes predictions about what will be observed when a system interacts with a measurement device, then it doesn't describe the system prior to that interaction. As a result of not describing the system prior to its interaction with the measurement device, it could not be considered a complete description of nature.

To me, that would seem like a reasonable justification (well, if it were formulated to be a justification that is.)

 

[Morbert]

But, if we don't grant Smolin his assumptions and we are left with the mathematical framework which only predicts the probability of a particle registering on a particular measurement device, then we are left without a description of the system prior to its interaction with the instrument, which would mean we have an incomplete description of nature.

I'm not sure if I'm insisting on a closed ontic account of nature, or not, genuinely. I just know that the point I am making is fairly rudimentary: if a theory [or interpretation of that theory] only makes predictions about what will be observed when a system interacts with a measurement device, then it doesn't describe the system prior to that interaction. As a result of not describing the system prior to its interaction with the measurement device, it could not be considered a complete description of nature.

The anti-realist position maintains that quantum theory produces a description of the physical system both prior to measurement and after measurement, and both of these descriptions are on equal footing, physically and ontologically. The spin of a particle is equally real or not real whether or not the particle is described with the state $|\psi\rangle = |\uparrow\rangle+|\downarrow\rangle$ or with the state $|\psi\rangle = |\uparrow\rangle$, and hence the collapse of a wavefunction after measurement does not render the spin of a particle any more or less real than it was before measurement. Should these descriptions be in terms of a primitive ontology, or in terms of records that can be generated in ancillary systems that might interact with the system? Smolin would insist on the former. But it's not a very compelling insistence.

Are there attempted explanations of the observed correlations which don't necessitate non-locality (in the FTL sense)? I've encountered the idea in multiple places that, in order to explain the violations of Bell inequalities we must give up one or more of the following:
1) Locality (in the FTL sense)
2) Realism (in the strict philsophical sense)
3) Local realism
4) Free Will

Is that accurate?

If we

i) Reject the premise that a quantity refers to an element of reality if we can predict with absolute certainty the experimental outcome associated with that quantity
ii) Interpret quantities in terms of experimental records and their likelihoods/expectations, and hence do not attempt to simultaneously model complementary quantities/mutually exclusive experiments

then the correlations follow naturally from quantum theory without any violation of the principle of local causes.

[edited first paragraph to make it clearer]

 

[RUTA]

We might be talking at cross purposes here because I am only talking about anti-realist/instrumental interpretations of QM i.e. those interpretations which say that the mathematics ONLY allows us to predict the observable outcomes of experiments i.e. the probability that a particle will register at a particular measurement device. These interpretations would be incomplete descriptions of nature for the reasons mentioned, they don't describe the system prior to its interaction with the measurement device.

Does the interpretation you are proposing describe the system prior to measurement?

Ah, yes, I see where what I've said so far does not answer your question per se. You need to read Section 3 of the Entropy paper, specifically Section 3.2.1. Quantum-Classical Contextuality. Therein you will see that the only objects of physical reality that exist are those that are interacting with everything else in physical reality. Since the QM systems you are talking about are not interacting with anything until they are measured (they're "screened off"), there is nothing in physical reality corresponding to them and consequently nothing to describe prior to interaction with the measurement device.

 

[PeterDonis]

I thought it was pretty clear which meaning @DrChinese and Lee Smolin were referring to, and therefore the meaning I was referencing.

Nobody is arguing about whether QM/QFT "requires nonlocality" in the specific sense of violating the Bell inequalities. Everybody agrees that it does. So a statement like this...

It appears that QM doesn't necessitate non-locality because QFT is local, but QFT doesn't appear to be a complete description of nature. Any attempt at a more complete description of nature appears to necessitate non-locality.

...is nonsense if by "non-locality" you meant "violation of Bell inequalities". You don't need a "more complete description of nature" to predict violations of the Bell inequalities. QM/QFT already do that just fine. So you could not possibly have been clear in your own mind about what you meant by "non-locality", since if you had been you would never have made this nonsensical statement.

You see why I keep saying it would be better if people would just not use the term "non-locality" at all? People confuse not just others, but even themselves, with it.

 

[PeterDonis]

There is a pretty obvious connection between the definition of non-local as involving instantaneous (or FTL) influences and the definition of non-local as "Violates the Bell inequalities". The former is the attempt to explain the latter and/or the latter is evidence of the former. This might lead us to suspect that there is a slight misappropriation of the term "non-local" when trying to define it as "Violates the Bell inequalities".

To try to elaborate: We have the experimental evidence of violations of Bell inequalities and know that these go against the predictions of classical physics. If we then try to explain how these violations of Bell inequalities occur and the answer we arrive at is: there must be some sort of instantaneous (or FTL) influence occurring, whereby a measurement in location A has an instantaneous effect at location B (which is spatially separated). This means that actions, such as measurements, don't simply exert influence in their locality at a maximum speed of c. This instantaneous (or FTL) influence we refer to as non-local - to contrast it with the idea that actions can only exert influence in their locality at a maximum speed of c.

To then suggest that "non-locality" is just another name for the violations of Bell inequalities seems to conflate the effect with the cause or the observed phenomenon with the attempt to explain how the observed phenomenon is possible, given that it goes against classical physics.

Now you're contradicting yourself. First you say that you are using "non-locality" the way @DrChinese and Lee Smolin meant it. The way they meant it is "violations of the Bell inequalities".

Then you argue that that very same meaning of "non-locality" doesn't make sense and shouldn't be used.

Wouldn't it be better to just drop the term "non-locality" altogether? All it's doing is confusing you and making you contradict yourself.

 

[PeterDonis]

it might be worth exploring which term was first in use, which one offers the most explanatory power, and whether or not there are connections between the different definitions, as above, which can be delineated.

Or we could just drop the term "non-locality" altogether, since it is not helping the situation, and just talk about, for example, what connections there might be between, say, violations of the Bell inequalities and the fact that spacelike separated measurements commute. Wouldn't that be clearer and less confusing?

 

[Lynch101]

Or we could just drop the term "non-locality" altogether, since it is not helping the situation, and just talk about, for example, what connections there might be between, say, violations of the Bell inequalities and the fact that spacelike separated measurements commute. Wouldn't that be clearer and less confusing?

I'm inclined to think that simply dropping the redundant definition of the term would be easier.

EDIT: It would also clear up the confusion that it creates.

 

[PeterDonis]

I'm inclined to think that simply dropping the redundant definition of the term would be easier.

Good luck with that, since there is not general agreement on which definition is "redundant".

 

[PeterDonis]

simply dropping the redundant definition

Also, as I pointed out in post #146, you're using the term yourself in contradictory ways, so it's not clear what definition you think is "redundant".