Entanglement and FTL signaling in professional scientific literature

[martinbn]

Well, IMO, that is precisely what @DrChinese is saying!

Then it is all lost in translation (from maths to English) because it seems to me to be the opposite.

 

[Cthugha]

Well, IMO, that is precisely what @DrChinese is saying!

I cannot follow. Maybe that is a question of terminology? Nonlocal correlations are pretty exactly the thing @vanhees71 proposes. Every serious full quantum field theory has these built in automatically. Mermin once wrote the hilarious tongue-in-cheek quote:
"My complete answer to the late 19th century question “what is electrodynamics trying to tell us” would simply be this: Fields in empty space have physical reality; the medium that supports them does not.

Having thus removed the mystery from electrodynamics, let me immediately do the same
for quantum mechanics: Correlations have physical reality; that which they correlate does not."

Nonlocal correlations just go along these lines. Consider any down conversion process, where one photon is converted into two. The individual photon energies are not an element of reality. They are not fixed. Their correlations are an element of reality. This means only that the sum of their energies is fixed due to energy conservation. Now you just find that the measurement of course respects this and any measurement result will show nonlocal correlations that result in the sum of both photon energies summing up to the energy before. The same goes for any conserved quantity, such as particle spin, linear momentum or whatever.

All what is relevant is that this property is not an element of reality for the individual particles, but the correlations between them are. This is the minimal assumption required to reproduce Bell-type experiments in a theory. However, of course such a theory is silent on how these non-local correlations are enforced. This is a question of interpretation. "Non-local correlations" themselves, however, are rather a technical term.

However, I fully understand if someone considers this quite technical and minimal take as not really satisfactory from a philosophical or ontological point of view.

 

[PeterDonis]

Well, IMO, that is precisely what @DrChinese is saying!

Consider:

the synchronization of the distant end components clearly cannot occur without "spooky action at a distance" as demonstrated by the cited experiment.

The authors do not intend to demonstrate spooky action at a distance in this paper. The authors aim to debunk this idea.

Seems pretty opposite to me.

 

[PeroK]

Consider:
Seems pretty opposite to me.

Perhaps it's best that @DrChinese confrms whether he thinks the cited paper debunks or supports superluminal "influences".

 

[PeterDonis]

Perhaps it's best that @DrChinese confrms whether he thinks the cited paper debunks or supports superluminal "influences".

@DrChinese is the one who cited the paper; what I quoted from him is from his post citing the paper. So I think we already have his opinion.

 

[vanhees71]

I never said there was a causal signal faster than 10000 c. I would assume you already knew of this experiment by a top team, but assuming not:

https://arxiv.org/abs/0808.3316
"Testing spooky action at a distance"

"In science, one observes correlations and invents theoretical models that describe them. In all sciences, besides quantum physics, all correlations are described by either of two mechanisms. Either a first event influences a second one by sending some information encoded in bosons or molecules or other physical carriers, depending on the particular science. Or the correlated events have some common causes in their common past. Interestingly, quantum physics predicts an entirely different kind of cause for some correlations, named entanglement. This new kind of cause reveals itself, e.g., in correlations that violate Bell inequalities (hence cannot be described by common causes) between space-like separated events (hence cannot be described by classical communication). Einstein branded it as spooky action at a distance.

"A real spooky action at a distance would require a faster than light influence defined in some hypothetical universally privileged reference frame. Here we put stringent experimental bounds on the speed of all such hypothetical influences. We performed a Bell test during more than 24 hours between two villages separated by 18 km and approximately east-west oriented, with the source located precisely in the middle. We continuously observed 2-photon interferences well above the Bell inequality threshold. Taking advantage of the Earth’s rotation, the configuration of our experiment allowed us to determine, for any hypothetically privileged frame, a lower bound for the speed of this spooky influence. For instance, if such a privileged reference frame exists and is such that the Earth’s speed in this frame is less than 10−3 that of the speed of light, then the speed of this spooky influence would have to exceed that of light by at least 4 orders of magnitude."

As I mentioned, this experiment demonstrates that the selection of a measurement basis at one place (call that Alice) affects the outcome at a distant portion of the entangled system (call that Bob). The speed of the change cannot be less than 10,000 c per the experiment, and the experiment does NOT specify the time direction of the "spooky action at a distance". There is no known signal, nor can any causal element be deduced from this entanglement experiment (other than the fact that the subsystems Alice and Bob synchronize much faster than c would allow).

We are now over 60 replies into a thread started just a few hours ago. In case anyone forgot, the OP's question was: "are there any suggestions that entanglement might imply some sort of faster than light signaling between the entangled particles?" I said YES, and this experiment confirms those suggestions (subject to the caveats I provided). Keep in mind that before a measurement, the full entangled system cannot be said to be 2 individual systems (although there is spatial extent). However, the synchronization of the distant end components clearly cannot occur without "spooky action at a distance" as demonstrated by the cited experiment.

I didn't know this experiment, but indeed the paper does in no way contradict my point of view. The "nonlocality" they are referring to is rather meaning the long-ranged correlations due to entanglement. This has nothing to do with faster-than-light signal propagation and a contradiction to microcausality, i.e., the mathematically precise meaning of "locality" in relativistic QFT.

Microcausality implies that entanglement can NOT be used for some sort of "faster-than-light signaling between the entangled particles", and indeed it implies the cluster decomposition principle which rules this out. The authors explicitly say so in the 2nd column of the first page of their paper (published version):

In both of these analyses, the hypo-
thetical superluminal influence was termed the speed of quantum
information to stress that it is not classical signalling. We shall use
this terminology, but we emphasize that this is only the speed of a
hypothetical influence and that our result casts very serious doubts
on its existence.
[\QUOTE]

 

[Lord Jestocost]

... superluminal "influences"

To my mind, “superluminal influence” is an unfortunate term because of its unphysical connotations. From “Quantum entanglement” by Ryszard Horodecki, Pawel Horodecki, Michal Horodecki and Karol Horodecki (Rev. Mod. Phys. 81, 865 – Published 17 June 2009):

“Nature on its fundamental level offers us a new kind of statistical non-message-bearing correlations, which are encoded in the quantum description of states of compound systems via entanglement. They are ‘nonlocal’²² in the sense, that

i) They cannot be described by a LHVM.
ii) They are nonsignaling as the local measurements performed on spatially separated systems cannot be used to transmit messages.

Generally speaking, quantum compound systems can reveal holistic nonsignaling effects even if their subsystems are spatially separated by macroscopic distances. In this sense quantum formalism offers holistic description of Nature (Primas, 1983), where in a non-trivial way the system is more than a combination of its subsystems.

²² The term nonlocality is somewhat misleading. In fact there is a breaking of conjunction of locality and counterfactuality.”

[Bold by LJ]

 

[PeroK]

Words fail us!

 

[DrChinese]

My citation (see below) is one of many that demonstrate quantum nonlocality, that the Bell test synchronization between Alice's measurements and distant Bob's measurements occur FTL. We all "know" that such nonlocality "must" be instantaneous - there is no theoretical upper limit after all - but Bell tests don't normally attempt to quantify that speed or set a lower limit. The cited experiment does just that, a lower limit of 10,000 c given certain assumptions (and you don't have to agree with those).

Obviously: Alice's selection of a measurement basis causes (by assumption only) distant Bob's later measurement to be synchronized with Alice's results. This occurs FTL, possibly instantaneously, and further the causal direction is ambiguous. Put a different way: the relative timing of Alice's and Bob's measurements in a Bell test are irrelevant (this is consistent with any interpretation). Whatever influence occurs to make this happen, is labeled "quantum nonlocality" in the more modern usage of the term, and this is generally accepted as well.

What is not generally accepted any longer is the term "spooky action at a distance", but it still appears in the literature and bothers a lot of folks. To me, this phrase means the same as "quantum nonlocality" because it describes a key element of entanglement. In the cited article, they gave a DIFFERENT definition of "spooky action at a distance" than mine.

Hopefully nothing above is disagreeable to anyone, or any interpretational view.

-------------------------------

https://arxiv.org/abs/0808.3316

The experiment itself was not attempting to set a lower limit for the speed of quantum nonlocality per se, rather they attempt to set a lower limit for the speed of "quantum nonlocality assuming there is a preferred reference frame in the universe". They define this as "spooky action at a distance" but I don't believe Einstein used this term with the idea there might exist a preferred reference frame in the universe.

So that is really the source of the issue many of you are having with my citation, and I wasn't clear enough about my use of the citation to prevent your confusion and comments. I was primarily using the reference because it is the only one I have seen that specifically attempts to show the minimum speed of quantum nonlocality using a Bell test. Most Bell tests demonstrating violation of strict Einsteinian locality simply use c as a lower limit, and stop there (figuring any entanglement-related influence shown as exceeding c accomplishes the goal).

The Salart et al experiment goes much farther, although its objective was different: they wanted to show that Bell tests can rule out the existence of a "hypothetical universally privileged reference frame" given certain constraints ("that the Earth's speed in this frame is less than 10^-3 that of the speed of light"). Some Bohmian type interpretations, for example, posit such. Their experimental conclusion was that there is no such preferred reference frame, and since they equated this to "spooky action at a distance": they reject that (as several of you point out). They didn't reject that there are quantum nonlocal influences, just that they couldn't exist with a preferred reference frame.

Although they don't state this, I believe their experiment also points to a lower bound for the "speed of quantum information" (I would use the word "nonlocality" instead of "information"). From the paper: "The violation of the Bell inequality at all times of the day allows one to calculate the lower bound for the speed of quantum information for any reference frame."

So while apologizing for the confusion I created by using this citation, I would stand by my main point: Bell tests demonstrate that quantum nonlocality is real, specifically that synchronization between Alice's measurements and distant Bob's measurements occur FTL (>c). Call it whatever you want: an influence, action at a distance, quantum information transfer, speed of entanglement... it's real and generally accepted by the community.

To paraphrase @PeroK: sometimes words fail me.

[An additional note: I don't think their formal conclusion would be considered as strong today as when it was originally published in 2008. There is evidence from some CMBR studies that Earth might have a velocity relative to some cosmic reference frames that are in the neighborhood of 10^-3 c (or greater), which would then violate one of the assumptions for their specific conclusion.]

 

[vanhees71]

But the conclusion of the cited paper concerning such a superluminal "quantum speed" is that "that our result casts very serious doubts on its existence." I don't know, why you insist on claiming the opposite. At least you should be clear that this is your opinion, which is opposite to the authors' conclusion about their experiment!

 

[vanhees71]

BTW: In the meantime I found a paper, which precisely makes the connection between the "long-ranged correlations" (often dubbed "nonlocality", which in my opinion is a highly confusion imprecision in language) and microcausality (which implements indeed "locality" in the usual meaning of the word within relativistic QFT):

Molotkov, S.N., Nazin, S.S. The Einstein-Podolsky-Rosen effect and causality. Jetp Lett. 70, 54–60 (1999). https://doi.org/10.1134/1.568129

 

[PeroK]

Although they don't state this, I believe their experiment also points to a lower bound for the "speed of quantum information" (I would use the word "nonlocality" instead of "information"). From the paper: "The violation of the Bell inequality at all times of the day allows one to calculate the lower bound for the speed of quantum information for any reference frame."

My understanding is that this lower limit is for a hypothetical (and highly questionable) quantum information mechanism. This would be in contrast to what I perceive as the default position:

“Nature on its fundamental level offers us a new kind of statistical non-message-bearing correlations, which are encoded in the quantum description of states of compound systems via entanglement.

And, in this default position on entanglement there is no component of the theory where the term "speed" or "FTL" would be appropriate. There is simply no sense in which anything (however you describe it) is being exchanged between entangled particles.

 

[PeterDonis]

The cited experiment does just that, a lower limit of 10,000 c

The general form of these claims doesn't make sense to me. If the measurement events are spacelike separated, the FTL "speed" of anything between them is frame dependent. I can make it 10,000 c or 100,000 c or instantaneous by choosing an appropriate frame, but I can also make it, say, 10 c or 1.1 c or 1.00000000000000000001 c by choosing an appropriate frame. There is no invariant that I can see that would correspond to the "speed" being talked about.

 

[gentzen]

but I can also make it, say, 10 c or 1.1 c or 1.00000000000000000001 c by choosing an appropriate frame

This is where their assumption

given certain constraints ("that the Earth's speed in this frame is less than 10^-3 that of the speed of light")

comes in, to allow then to get "a lower limit of 10,000 c".

 

[PeroK]

This is where their assumption

comes in, to allow then to get "a lower limit of 10,000 c".

And that's another questionable aspect of looking for a signaling mechanism: the need for a preferred, universal frame.

 

[PeterDonis]

This is where their assumption

comes in, to allow then to get "a lower limit of 10,000 c".

In other words, they are only considering a small subset of all possible reference frames. Which leads to the obvious next question: why should I care about whatever they think they're showing, since it's frame dependent and so has no physical meaning?

 

[PeroK]

In other words, they are only considering a small subset of all possible reference frames. Which leads to the obvious next question: why should I care about whatever they think they're showing, since it's frame dependent and so has no physical meaning?

Because the implication is that relativity is wrong!

 

[DrChinese]

The general form of these claims doesn't make sense to me. If the measurement events are spacelike separated, the FTL "speed" of anything between them is frame dependent. I can make it 10,000 c or 100,000 c or instantaneous by choosing an appropriate frame, but I can also make it, say, 10 c or 1.1 c or 1.00000000000000000001 c by choosing an appropriate frame. There is no invariant that I can see that would correspond to the "speed" being talked about.

I think the paper, being from a top team (and assuming you are questioning that), can defend its conclusions better than I. They provide their thinking on reference frames, and specifically a hypothetical preferred reference frame.

On the other hand, I think you have acknowledged my main point: the "speed" of whatever is happening is greater than c (and most probably instantaneous regardless of frame or even type of frame). When Alice and Bob measure in the same reference frame, there certainly no frame dependence on the result. Your answer on the lower limit on "speed of quantum nonlocality" is dependent on how close together the measurements occur, as well as distance separating.

If you could perform a Bell test in a lab where Alice and Bob (in the same inertial reference frame) are measuring 30 meters apart, and can clock their respective measurements within a 100 femtosecond window of each other: By my back of napkin computation the synchronization between the 2 observations would be on the order of:

30 meters at c = 300,000,000 meters/second... is traveled in 10^-7 seconds (= 10^8 femtoseconds).
But the synch occurred in only 100 femtoseconds (at most - could be much less).

That's 1,000,000 c in that reference frame. I don't believe you would get any significantly different results regardless of whether you boosted one relative to the other. (Not sure how you would/could measure that anyway, since the objective would be to start and stop the timings as closely as possible together.)

 

[PeterDonis]

Because the implication is that relativity is wrong!

Why is that the implication? Why isn't the implication that their method of analysis is wrong since it attributes physical meaning to frame-dependent quantities?

 

[PeroK]

Why is that the implication? Why isn't the implication that their method of analysis is wrong since it attributes physical meaning to frame-dependent quantities?

As I understand, they decided to analyse the idea of a quantum information signaling and make the requisite assumptions and see what conclusions they could draw. And, this led to at least a certain scepticism that the mechanism is viable. And possibly the researchers felt they had debunked the whole idea.

 

[PeterDonis]

they decided to analyse the idea of a quantum information signaling and make the requisite assumptions and see what conclusions they could draw

"The requisite assumptions" can't in themselves be inconsistent with relativity because there is a relativistic theory, QFT, that correctly predicts the correlations. The assumption of a preferred frame does not seem to me to be "requisite" to explain the correlations; it seems to me to be an additional assumption driven only by the particular method of analysis chosen. The fact that that additional assumption is inconsistent with relativity seems to me to be far more likely to be an issue with the assumption and the method of analysis that drove it than with relativity itself, particularly in view of the fact that, as above, we can correctly predict the correlations with a relativistic theory.

 

[DrChinese]

"The requisite assumptions" can't in themselves be inconsistent with relativity because there is a relativistic theory, QFT, that correctly predicts the correlations.

You assume that all explicitly nonlocal/nonrelativistic quantum theories/interpretations are wrong (of which Bohmian Mechanics is just one). I'm sure you are aware of some others (such as GRW). Those generally make identical predictions for Bell tests as does QFT (which has many other advantages),

That was their point, without them saying so directly. IF there exists a universal preferred reference frame, AND the Earth's motion relative to it is within a threshold, THEN: a Bell test like the one they ran will in fact show differences in some directions (which are not predicted by garden variety QM). It doesn't, so one should rule those out by experiment. Of course, defenders of those alternate candidate theories will attempt to explain away the experiment or further modify the theory...

 

[PeterDonis]

You assume that all explicitly nonlocal/nonrelativistic quantum theories/interpretations are wrong (of which Bohmian Mechanics is just one).

No, I'm not assuming that. It's perfectly possible to have an "explicitly nonlocal/nonlrelativistic interpretation" that still acknowledges and accounts for the fact that, in domains where relativistic effects are significant, our best current experimental data says that those relativistic effects are present. For example, the version of Bohmian Mechanics that @Demystifier has published is explicitly nonlocal and nonrelativistic, but it still has Lorentz invariance as an emergent symmetry that is expected to be observed in all of our experimental tests to date.

The particular paper in question, though, is "explicitly nonlocal and nonrelativistic" in a different way: the method of analysis it is using appears to me to explicitly deny Lorentz invariance even as an "emergent" symmetry valid in the domain we have experimentally tested. And I would like to see some explanation of why they adopted that method of analysis, when it seems open to the obvious objection that we have observed Lorentz invariance as a valid symmetry in all of our experimental tests to date, which certainly include relative speeds larger than the limits the paper is using, so we should not expect to see preferred frame effects in this domain. (This question becomes even more acute, btw, if the paper's conclusion is, as it seems to be, that the version of "quantum signaling" they are analyzing is questionable or debunked.)

 

[PeterDonis]

Those generally make identical predictions for Bell tests as does QFT (which has many other advantages),

But if that's the case, how can any such interpretation possibly be ruled out by an experiment that confirms those same predictions for Bell tests?

IF there exists a universal preferred reference frame, AND the Earth's motion relative to it is within a threshold, THEN: a Bell test like the one they ran will in fact show differences in some directions (which are not predicted by garden variety QM).

But if there is any such interpretation, it would make predictions for Bell tests different from the predictions of standard QM/QFT. But you said above (in what I quoted) that these interpretations make identical predictions for those tests to standard QM/QFT. So what, exactly, is being ruled out by this experiment?

 

[DrChinese]

But if that's the case, how can any such interpretation possibly be ruled out by an experiment that confirms those same predictions for Bell tests?But if there is any such interpretation, it would make predictions for Bell tests different from the predictions of standard QM/QFT. But you said above (in what I quoted) that these interpretations make identical predictions for those tests to standard QM/QFT. So what, exactly, is being ruled out by this experiment?

You missed the modifying word "generally". The purpose of the cited paper was to demonstrate a little known consequence of some of the outlier candidate/competing theories. Thus showing that there are some predictions which are not identical.

Clearly, the race is on (and has been on) to discover any possible theoretical consequence of the various interpretations/candidate quantum theories, no matter how small, and see if an experiment can be formulated to distinguish. No different than in the study of dark matter vs MOND vs everything else. Each accounts for many of the basics alike, but some have discrepancies which are coming to light.

 

[PeterDonis]

The purpose of the cited paper was to demonstrate a little known consequence of some of the outlier candidate/competing theories. Thus showing that there are some predictions which are not identical.

How have proponents of those theories/interpretations responded?

 

[vanhees71]

You assume that all explicitly nonlocal/nonrelativistic quantum theories/interpretations are wrong (of which Bohmian Mechanics is just one). I'm sure you are aware of some others (such as GRW). Those generally make identical predictions for Bell tests as does QFT (which has many other advantages),

Of course, within non-relativistic QM there is no such apparent tension between causality and the assumption that there's instantaneous influence between far-distant objects, and thus there's nothing to be debated in this direction within non-relativistic QM and even interpretational extensions of it like Bohmian mechanics. GRW is another case, because that's a new theory beyond QM. There's not the slightest hint for its necessity though.

I don't know, how often I have to repeat it: By construction microcausal relativistic QFT exclude causal connections between space-like separated events. In our context "events" are the registration of two photons at far distant places, and if these registration events are space-like separated, the observed correlations due to entanglement cannot be caused by the measurement on one photon influencing the measurement of the other. The only conclusion is that the correlations are due to the state preparation, and that's indeed the standard interpretation of the entire formalism: The entanglement is due to the preparation of the two photons (in this case by parametric down conversion). It describes long-ranged correlations but not faster-than-light signalling, and I think that's also the interpretation the authors of the said paper have about their results, and indeed, for space-like separated events it doesn't even make sense to talk about some "signal propagation" or "causal order".

This is very similar to the fact that phase velocities as well as group velocities in classical field theory (e.g., electrodynamics) can be and in fact are larger than $c$ without indicating any violation of relativistic causality. This is known since 1907!

That was their point, without them saying so directly. IF there exists a universal preferred reference frame, AND the Earth's motion relative to it is within a threshold, THEN: a Bell test like the one they ran will in fact show differences in some directions (which are not predicted by garden variety QM). It doesn't, so one should rule those out by experiment. Of course, defenders of those alternate candidate theories will attempt to explain away the experiment or further modify the theory...

So, why have you claimed the opposite before?

 

[Morbert]

Obviously: Alice's selection of a measurement basis causes (by assumption only) distant Bob's later measurement to be synchronized with Alice's results. This occurs FTL, possibly instantaneously, and further the causal direction is ambiguous. Put a different way: the relative timing of Alice's and Bob's measurements in a Bell test are irrelevant (this is consistent with any interpretation). Whatever influence occurs to make this happen, is labeled "quantum nonlocality" in the more modern usage of the term, and this is generally accepted as well.

Surely this is interpretation dependent and therefore not obvious. Many interpretations would not frame quantum correlations as evidence of nonlocal influences.
https://www.webofstories.com/play/murray.gell-mann/165

 

[vanhees71]

If by nonlocal influences you mean faster-than-light causal influences then any interpretation which assumes them is in plain contradiction with the mathematical facts of microcausal relativistic QFTs!

 

[Demystifier]

If by nonlocal influences you mean faster-than-light causal influences then any interpretation which assumes them is in plain contradiction with the mathematical facts of microcausal relativistic QFTs!

This statement cannot be found in professional scientific literature, which this thread is supposed to be about.

 

[martinbn]

This statement cannot be found in professional scientific literature, which this thread is supposed to be about.

I cannot cite one now, but I am sure I have seen published papers that state or suggest that the causal structure as described by relativity is incorrect. My impression is that is most Bohmians even if they don't say it out loud.

 

[Demystifier]

I cannot cite one now, but I am sure I have seen published papers that state or suggest that the causal structure as described by relativity is incorrect. My impression is that is most Bohmians even if they don't say it out loud.

Sure, I have written it myself, but such papers emphasize that it is not in contradiction with "the mathematical facts of microcausal relativistic QFTs". The idea of such papers is that relativistic QFT is right but incomplete. Completeness is often an additional tacit assumption in standard QFT literature, but completeness is not a mathematical fact.

 

[vanhees71]

This statement cannot be found in professional scientific literature, which this thread is supposed to be about.

Of course it can. The microcausality condition is precisely imposed to avoid faster-than-light signal propagation.

 

[Demystifier]

The microcausality condition is precisely imposed to avoid faster-than-light signal propagation.

That's true, but signal is the key word. Your original statement was wrong because it didn't contain the word signal. Standard QFT forbids FTL propagation of signals, but it doesn't forbid FTL propagation of influences. I know that you don't distinguish signal from influence, but professional scientific literature does.

 

[vanhees71]

For me signal and influence is a synonym. The point is to distinguish causal signals/influences and long-ranged correlations. A correlation does not necessarily imply a causal connection/signal/influence between two events, and space-like separated events can never be causally connected by construction in microcausal relativistic QFT. This is at least standard in all treatments of relativistic QFT in the HEP community. Obviously it's not so often explicitly stated in the quantum-optics literature although of course both communities use the same QED.