Quantum nonlocality and "spooky action at a distance"

[gentzen]

No, you missed a fine detail of my point. I think Vanhees did too.

There is no bell style HV, yes. I am with all this.

Which is a very specific HV with additional assumptions on how the average action is formed, which i think is the problem.

I am not saying that it is 100% certain that there are no HV, only that so far all we know points in that direction. What makes you think that there could be HV (non Bell type of course)?

To Fra: Do you have an opinion on A. Neumaier's approach? It comes perhaps a bit more from the mathematical/logical/"statistical" side than your and Lee Smolin's more physical approach, but it also tries to overcome the limitations of a pure instrumental approach, and has a commitment to a single world. I know, it may seem strange that it seems to leave QT nearly unmodified, and still claims to have ("non-separable") HV hidden in plain sight, but ... What is your opinion? At least those are non-Bell style HV!

 

[martinbn]

To Fra: Do you have an opinion on A. Neumaier's approach? It comes perhaps a bit more from the mathematical/logical/"statistical" side than your and Lee Smolin's more physical approach, but it also tries to overcome the limitations of a pure instrumental approach, and has a commitment to a single world. I know, it may seem strange that it seems to leave QT nearly unmodified, and still claims to have ("non-separable") HV hidden in plain sight, but ... What is your opinion? At least those are non-Bell style HV!

I realize that is off topic but what is his approach? What are the hidden variables?

 

[PeterDonis]

But everything we've learned since then points to no hidden variables!

Not necessarily. The Bohmian interpretation has hidden variables and matches the predictions of standard QM. They're just nonlocal hidden variables--the particle positions are affected instantaneously by changes in the wave function anywhere in the universe.

 

[vanhees71]

That's consistent, of course, only in non-relativistic QM and at odds with relativistic QFT, which obeys by construction relativistic causality, i.e., no "action at a distance"!

 

[PeterDonis]

That's consistent, of course, only in non-relativistic QM

The original Bohmian interpretation was, yes. There are more recent relativistic versions (including, IIRC, one published by @Demystifier). Of course there is dispute in the literature about whether these versions actually work--just as there are disputes in the literature about every QM interpretation. I'm not trying to argue that the Bohmian interpretation is "correct", just giving it as an example of a hidden variable interpretation that is not ruled out by Bell's Theorem.

at odds with relativistic QFT, which obeys by construction relativistic causality, i.e., no "action at a distance"!

More precisely, no FTL signaling and spacelike separated measurements must commute. But that does not mean that a particular model cannot have unobservable nonlocal hidden variables that "change faster than light". It just means those variables cannot result in FTL signaling or spacelike separated measurements not commuting.

 

[vanhees71]

What are such unobservable nonlocal HIVs good for, particularly since it's claimed to solve some apparent philosophical quibbles about the physical meaning of the theory? It's just even more puzzling than the theory itself, which is consistent with the (at least special) relativistic spacetime description and causality, i.e., such ficticious hidden non-local variables breaking the intrinsic consistency of standard QFT do not solve apparent metaphysical problems but create even more serious new ones.

 

[PeterDonis]

What are such unobservable nonlocal HIVs good for

For constructing a model that has an actual mechanism for producing counterintuitive results of QM like violations of the Bell inequalities.

You might not think that's a useful goal, but others apparently do.

It's just even more puzzling than the theory itself

Perhaps to you, but not to everyone. It's certainly more concrete than just punting and saying, well, QM just predicts probabilities and doesn't tell you any actual mechanism for how things happen.

 

[Structure seeker]

I usually avoid the word "reality", because its meaning is uncertain ;-))

It's simply the collection of all true statements.

Having prepared the system in a given quantum state, you know the probabilities for the outcome of measurments of observables on the system, no more, no less, and that's what describes all observations with utmost precision.

Is this true or not? If it is, the behaviour of wavefunctions under measurement is part of reality as defined above. Is the statement "the system is in quantum state X at time t" true? The same applies. If it's not true, the state is not part of reality and therefore you can't use the idea that nature doesn't care what we think. All else is bullshit unless you come with an underlying consistent alternative for the belief in reality.

 

[PeterDonis]

It's simply the collection of all true statements.

This seems confused. Statements are statements about reality; a statement is not the same thing as the thing the statement is about.

Rather than try to make up your own definition, I would strongly recommend looking at the literature to see how the term "reality" is defined by the actual workers in the field. And I think you will find that there is not one single accepted definition of that term, nor is there even agreement in the literature that the term should be used at all.

Is this true or not?

Of course it is; it's basic QM. But that doesn't mean the rest of your claims are necessarily correct. See above.

 

[Fra]

What makes you think that there could be HV (non Bell type of course)?

There can be other arguments, but stay on track, what I had in mind here was that it was also Einsteins only (within reason) logical possibility explanation to quantum entanglement. As all other options seems just unacceptable.

The the total interference patterns, does NOT appear as if, the spins are predetermined but random, is not hte same thing. (which I would only guess Einstein would have agree with)

Ie. the problem isn't randomness, the problem is the quantum inseparability. So you must solve the latter, but not the first. So thinking there have to be some sort of "non-bell" HV, doesn't meen you have a theory for it, it seems like just a reasonable logical conclusion,

(Bell's theorem would in theory solve both, as if the HV was known, the outcome would be predetermined. And we know this gives wrong interference paterns!)

/Fredrik

 

[Fra]

To Fra: Do you have an opinion on A. Neumaier's approach? It comes perhaps a bit more from the mathematical/logical/"statistical" side than your and Lee Smolin's more physical approach, but it also tries to overcome the limitations of a pure instrumental approach, and has a commitment to a single world. I know, it may seem strange that it seems to leave QT nearly unmodified, and still claims to have ("non-separable") HV hidden in plain sight, but ... What is your opinion? At least those are non-Bell style HV!

I share alot of Nemaiers critique of current motivation of the statistical methods, so "something" needs to be done.

It's interesting but his solution is quite different to what I have in mind. His solution also, as far as I understand, does not aspire to provide insight into unification of forces without too much fine tuning (which is one of my motivators). I think Neumaiers treatise is also fairly "conservative", I just don't find that it solves all the problems I want to have answered, but it may solve some.

/Fredrik

 

[vanhees71]

It's simply the collection of all true statements.

Is this true or not? If it is, the behaviour of wavefunctions under measurement is part of reality as defined above. Is the statement "the system is in quantum state X at time t" true? The same applies. If it's not true, the state is not part of reality and therefore you can't use the idea that nature doesn't care what we think. All else is bullshit unless you come with an underlying consistent alternative for the belief in reality.

There is no general behavior of quantum states (wave functions are a very special case only) "under measurement". To describe its behavior you have to describe the interaction of the measured system with the measurement apparatus as you describe any interaction with the system with "its environment".

I still don't know, what you mean by "reality". For me "reality" are the objective observations we make in Nature. For physics we define quite refined quantified observations in terms of measurable observables, defined by an equivalence class of measurement processes for such a quantity/observable.

 

[Structure seeker]

I still don't know, what you mean by "reality". For me "reality" are the objective observations we make in Nature.

From propositional logic and the belief that statements are either true, false or some third undetermined value, I simply mean that reality is the statements (of all possible statements) that are true.
For me that is independent of being observed, measured or anything. They're true or they are not true. If that cannot work in our model of nature, quantum physics has left the foundation of logic and is therefore not anymore an exact science.

 

[GarberMoisha]

The problem is less about quantum theory and more with insisting on having objects in time and space, instead of events.
Reality as all the collection of events unfolding, has zero conceptual issues with QT. And if anything can be deduced from QT it is that this universe is about events(measurements and observations) Aka reality. Anything more would require a belief in a sophisticated structure of invisible hidden variables which may also resemble a conspiracy.

 

[Structure seeker]

And of course it is fine when the statement "the state is X at time t" has an unknown or third truth value, but then it is not yet acceptable to state "it is that way regardless of how we think it should be" (nature doesn't care what we think).

 

[Morbert]

From propositional logic and the belief that statements are either true, false or some third undetermined value, I simply mean that reality is the statements (of all possible statements) that are true.
For me that is independent of being observed, measured or anything. They're true or they are not true. If that cannot work in our model of nature, quantum physics has left the foundation of logic and is therefore not anymore an exact science.

This idea (which sounds a bit like Wittgenstein's declaration "The world is all that is the case") is intuitive enough in a classical setting, but issues arise in a quantum setting.

Given some classical system, we can construct some maximally fine-grained structure of propositions about the system, and assign probabilities to the propositions based on what we know about the system.

Given some quantum system, we can construct multiple maximally fine-grained structures of propositions, each internally consistent, but mutually incompatible with one another. If reality is accounted for by some set of propositions about it, then it raises the question: from which of these internally consistent, mutually incompatible, unprivileged structures do we construct our set of propositions?

 

[Structure seeker]

Can you give an example, @Morbert ?

 

[gentzen]

I realize that is off topic but what is his approach? What are the hidden variables?

He uses the names q-expectation and q-correlation for the formal expectations and correlations of the mathematical model, in order to have a clear separation between statements about the mathematical model, and statements about the physical world. For QM, he considers q-expectations as functions of time t, for QFT he considers q-expectation as functions of spacetime (t,x,y,z) and q-correlations as functions of multiple spacetime points (t1, x1, y1, z1), (t2, x2, y2, z2), ... Then he takes for example those q-expectations for which the low frequency components (spatial or temporal, or both, as is most reasonable for the given model) should basically be measurable (in practice), and declares their high frequency components (i.e. those beyond the reach of actually conceivable measurement devices) as hidden variables. And similar for q-correlations where some low order correlation for close spacetime points should basically be measurable, declares their far distant, or higher order correlations, or high frequency parts (i.e. those beyond the reach of actually conceivable measurement devices) as hidden variables.

More abstractly, his HV hidden in plain sight are those "variables" which are obviously part of the mathematical model, and which "the mathematical model itself cannot simply exclude from being measurable", but which are impossible to measure in the physical world for all practical purposes.

 

[vanhees71]

This idea (which sounds a bit like Wittgenstein's declaration "The world is all that is the case") is intuitive enough in a classical setting, but issues arise in a quantum setting.

Given some classical system, we can construct some maximally fine-grained structure of propositions about the system, and assign probabilities to the propositions based on what we know about the system.

Given some quantum system, we can construct multiple maximally fine-grained structures of propositions, each internally consistent, but mutually incompatible with one another. If reality is accounted for by some set of propositions about it, then it raises the question: from which of these internally consistent, mutually incompatible, unprivileged structures do we construct our set of propositions?

I'd say that a system is completely determined if it is prepared in a pure state $\hat{\rho}=|\psi \rangle \langle \psi|$ with $\langle \psi|\psi \rangle=1$. This can be achieved, e.g., by a von Neumann filter measurement of a complete set of compatible observables. Then $|\psi \rangle$ is defined up to a phase (i.e., the state is uniquely defined) as the common eigenvector of the corresponding self-adjoint operators representing these observables.

This is also "complete information" in the sense of the von-Neumann entropy
$$S=-\text{Tr} (\hat{\rho} \ln \hat{\rho}).$$
Which is $S=0$ for a pure state.

 

[Morbert]

Can you give an example, @Morbert ?

Consider a photon. We can construct fine-grained propositions about the linear polarization of the photon, and build a boolean algebra from these propositions. We can also construct fine-grained propositions about the circular polarization of the photon, and build a boolean algebra from these propositions. But we cannot combine these propositions into a single boolean algebra.

One tactic for dealing with this multiplicity of algebras is to abandon realism: We no longer seek accounts for isolated microscopic systems, and we only concern ourselves with classical data produced by measurements on microscopic system. With this approach, the experimenter can simply select whichever algebra is most suitable for their experiment.

 

[PeterDonis]

From propositional logic and the belief that statements are either true, false or some third undetermined value

Propositional logic has no third value. Statements are either true or false. We might not know which value a particular statement has, but that doesn't mean it has some third value; it just means our knowledge of the truth values of statements is a separate thing from the truth values themselves.

Whether such propositional logic is applicable to QM is a different question. There is a whole literature on quantum logic which suggests that it is not.

Again, you should not be making this up on your own. You should be looking at the literature. Personal speculations are off limits here.

 

[PeterDonis]

For me "reality" are the objective observations we make in Nature.

Is this how "realism" is defined in the QM literature?

 

[gentzen]

From propositional logic and the belief that statements are either true, false or some third undetermined value

Propositional logic has no third value. Statements are either true or false.

This is actually an important distinction between a theory and a model (for propositional logic, and also for predicate logic): For a model, statements are either true or false. For a theory, there can also be statements about which the theory stays silent.

 

[vanhees71]

Is this how "realism" is defined in the QM literature?

I don't know, how "realism" is defined in the QM literature, because it has quite varying meaning.

I think in most contexts it means the assumption that all observables always take determined values, no matter whether they are known due to the information about the system or not. That's the case in classical statistical mechanics, i.e., in a many-body system all (point) particles always have determined positions and momenta, but we cannot know them due to the complexity of the system and thus we describe them by (single-particle or in some cases also two- and few-particle) phase-space distribution functions. Here the probabilistic nature of the description is only due to the lack of information due to the complexity of the situation but not inherent in the underlying classical theory ("point-particle mechanics").

In QT, in the standard minimal statistical interpretation, however observables only take determined values, if the system is prepared in a corresponding state. Even being "completely prepared", i.e., being prepared in a pure state, does not imply that all observables take determined values, and this is a inherent property of Nature and not just due to the lack of our knowledge about the minute microscopic details.

That, however, seems not to be the generally accepted meaning, and it's even not so easy to decide, whether this really is the meaning intended by Bell, when he talks about "realistic, local theories", but I understand it as such, at least in his seminal paper.

 

[PeterDonis]

I don't know, how "realism" is defined in the QM literature, because it has quite varying meaning.

Then we shouldn't be using terms like "reality" in this thread at all, since we don't have a good basis for assigning them a useful meaning.

 

[GarberMoisha]

I will give it a try.
Reality is quantum fields - the rock bottom of fundamental "stuff" that everything is made of.
Quantum fields are mathematical.
It is astonishing we are able to discover the code that everything runs on. This code is the laws of physics, which are mathematical.
Math is the language of reality and its essence. The ground of being.

 

[Structure seeker]

We can construct fine-grained propositions about the linear polarization of the photon, and build a boolean algebra from these propositions. We can also construct fine-grained propositions about the circular polarization of the photon, and build a boolean algebra from these propositions. But we cannot combine these propositions into a single boolean algebra.

Circular and linear polarization are both limiting values of elliptical polarization, according to https://en.m.wikipedia.org/wiki/Elliptical_polarization. Does that resolve the issue?

In general, superposition of states doesn't mean state A and B at the same time. That assumes a classical reality in a quantum setting, which doesn't work. A superposed (non-pure) state is a defined quantum state (including the phases of the summand pure states). The statement that it has exactly this superposed state is true, all other values for the state (including A and B) are false.

 

[vanhees71]

Then we shouldn't be using terms like "reality" in this thread at all, since we don't have a good basis for assigning them a useful meaning.

Exactly! As with "locality/non-locality" one has to clearly state what you mean by the term, because it's no longer well-defined as a common scientific term at all. Perferrably you define it in well defined mathmematics.

 

[PeterDonis]

A superposed (non-pure) state

This is not correct. A pure state can be a superposed state. Superposition is basis dependent. For example, the pure state "spin-z up" of a qubit can also be written as a superposition of "spin-x up" and "spin-x down".

 

[PeterDonis]

I will give it a try.

Please don't. As I have already pointed out in response to others, you should not be making up your own definition of "reality". Either find a paper in the literature that gives a definition generally accepted in the appropriate physics community, and use that, or refrain from using the word at all.

 

[Structure seeker]

For example, the pure state "spin-z up" of a qubit can also be written as a superposition of "spin-x up" and "spin-x down".

True, I assumed orthogonal basis states A and B. Superposition is basis-dependent indeed.

 

[kered rettop]

And violating a Bell inequality means local noncontextual theories are ruled out. There is nothing to debate about this point, it’s been standard science for nearly a half century.

Why would we try to convince people to reject the very thing Einstein called “spooky action at a distance”? He knew full well what that meant, and that if the EPR argument was wrong (as Bell later proved) that would be what we would be left with.

So why wouldn’t we simply point folks to the preferred PF term? Rather than persuading them that nothing is going on, when obviously something is. Either: the quantum context is nonlocal; or there is nonlocal action. That’s Bell, plain and simple. And that what Quantum Nonlocality is.

I provided references saying exactly this. No one can provide a single reference to the contrary. Weihs et al 1998, Violation of Bell’s Inequality Under Strict Einsteinian Locality Conditions… and so on. There’s nothing to interpret on this point.

It seems to me that in reality: for many here, their preferred interpretation actually denies the Bell result, I.e. that they actually believe we live in a local realistic world. It’s that point of view that should be called out clearly - instead of trying to talk posters into denial of the extremely sophisticated experiments being performed that show that Einsteinian locality is not respected. Again, hundreds of papers back up that position. And none say otherwise. If we call ourselves scientists, why back away from published conclusions?

About 10 years after I graduated, I wasted a long time trying to devise a local realistic model to replicate the EPR correlations! Alas, I had not realised that achieving full correlation when using parallel detectors - and zero correlation when crossed - was not sufficient. And I have seen lecturers at prestigious universities make the same mistake. I blame all those "left and right glove" analogies, which can only replicate locally real theories, thereby creating the impression that there is nothing odd happening!

 

[Morbert]

I don't know, what Bell means by "locality".

From La nouvelle cuisine: A sketch of spacetime regions 1 and 2, with corresponding past light cones.

Bell says a theory is causally local if there is some sufficient specification of events at 3, such that a measurement at 2 can not give us any new information about events at 1.

For me locality is the assumption of QFT that we describe everything with a Hamilton density consisting of field operators at the same space-time point, have the usual local realization of the Poincare group on the field operators, and that self-adjoint operators representing local observables (like the em. energy-momentum tensor, which contains the description of photon detection, using the usual dipole approximation) obey the microcausality constraints.

Bell's motivation:

Ordinary ‘local’ quantum field theory does have a causal structure. As everyone knows [...] Could the no-superluminal-signalling of ‘local’ quantum field theory be regarded as an adequate formulation of the fundamental causal structure of physical theory? I do not think so. For although ‘local commutativity’ has a nice sharp-looking mathematical appearance, the concepts involved in relating it to causal structure are not very satisfactory

There are possible responses to Bell here, but the important point is the site of disagreement is over the adequacy of the microcausality constraint in accounting for the causal structure of a theory, rather than the mere stipulation of a condition itself.

[edit to add] - I also suspect a statistical interpretation would itself be interpreted as a refusal to grapple with the issue Bell raises, rather than a solution to it. If a quantum theory is a theory of ensembles, then it does not address the causal relations between individual events.

 

[vanhees71]

My problem with these statement by Bell is that it is not clear, what he really means. In standard local QFT the Hamiltonian is given by a Hamiltonian density, obeying the microcausality condition with all operators representing local observables (including itself, representing energy density, which also makes it unique for gauge theories, where you have to use the gauge-invariant/covariant Belinfante energy-momentum tensor). This implies that there are no causal influences on any local observables over space-like separated distances. That's a mathematical fact and establishs a causality structure. I don't know, which problem Bell is addressing. From an earlier discussion I think to have understood that Bell rightfully stresses the importance of gauge-in/covariance and then calls observable beables, as if renaming things with funny sounding new word creations would solve any problems.