Local realism ruled out? (was: Photon entanglement and )

[DrChinese]

Seems similar to double slit experiment in a sense that photon's context wave splits over two paths with exception that paths do not end at the same place.
At polarization independent beam splitter photon goes down one path but photon's context wave (pilot wave if you like) goes down by both paths. At PBS photon's context wave interacts with other photon's (empty) context wave and creates entanglement. Only requirement is that both photons come from common source and polarization of one of photons is rotated by 90deg (polarizer + HWP at 45deg relative to polarization axis of polarizer).

That seems to be common recipe for creating polarization entangled photons.

I don't think it is actually a requirement that they come from a common source (there does need to be phase matching). I think that it the easier way by far to create the pairs. They are no longer polarization entangled when they start their independent processes.

 

[yoda jedi]

You don't accept that QM can be considered as a valid theory

and/or maybe a transitory one.

quantum mechanics is just plain wrong.

or inconsistent or incomplete.

many interpretations suggests that there is no one satisfactory interpretation.

and are "INTERPRETATIONS"....

That was the thinking of the EPR paper, and it is very reasonable.

thanks !
but i think for example, the electron spin have just 2 possible values, spin-up and spin-down,
right ? then emitting a pair of electron at once, you measure first one value, right ? and then is spin down, then the other electron "have to be" spin up, they can have their values from the beginning.

excuse my english please.

a spin can be up and the other down, or both up ? or both down ?
but that it can be argued in any case, that results are predetermined (from the beginning).







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

Proceedings Vol. 7421
Andrei Khrennikov

The main aim of this review is to show that the common conclusion that Bell's argument implies that any attempt to proceed beyond quantum mechanics induces a nonlocal model was not totally justified. Our analysis of Bell's argument demonstrates that violation of Bell's inequality implies neither "death of realism" nor nonlocality. This violation is just a sign of non-Kolmogorovness of statistical data - impossibility to put statistical data collected in a few different experiments (corresponding to incompatible settings of polarization beam splitters) in one probability space.



.

 

[DrChinese]

1. and maybe a transitory one.

2. or just incomplete.

...

3. not 100 %, respect to ?
to non-contextuality ? contextuality ?
or independence ?

a spin can be up and the other down, or both up ? or both down ?
but that it can be argued in any case, that results are predetermined (from the beginning).


4. Proceedings Vol. 7421
Andrei Khrennikov

The main aim of this review is to show that the common conclusion that Bell's argument implies that any attempt to proceed beyond quantum mechanics induces a nonlocal model was not totally justified. Our analysis of Bell's argument demonstrates that violation of Bell's inequality implies neither "death of realism" nor nonlocality. This violation is just a sign of non-Kolmogorovness of statistical data - impossibility to put statistical data collected in a few different experiments (corresponding to incompatible settings of polarization beam splitters) in one probability space.

1. Sure, a better (more useful) theory could come along any day. In fact, I hope one does.

2. The incompleteness argument was made in EPR and is now soundly rejected.

3. The correlation is 85% when Alice measures at 0 degrees and Bob measures at 22.5 degrees. So not 100%.

4. A fair example of his thinking is from his paper: "Complete account of randomness in the
EPR-Bohm-Bell experiment" which is at: http://arxiv.org/PS_cache/arxiv/pdf/0806/0806.0445v1.pdf"

He derives the usually CHSH inequality but ends up with result CHSH<=8 when the accepted result is CHSH<=2. See his (2) on page 6. So he concludes that experiment does not rule out local reality, since a typical experimental result is aboout 2.4. The point is that makes no sense. If you know how CHSH is derived, that is like proving that 1=3. It really isn't worth time to dispute the logic here, as no one really accepts it as useful in the first place.

In the paper you reference, he attacks the Kolmogorov axioms - this has been raised previously as an objection and it basically defeats the local realistic agenda if accepted. So not much there.

Suffice it to say, publication is not equivalent to general acceptance per se, especially with pure theoretical work. Just ask anyone who publishes if they expect their work to end up in a textbook. In sum: Khrennikov's work is not generally accepted, and I don't see any merit to his argument as it stands.

He does have some interesting papers, though:

"A Conclusive Experimentation Evidences that Mental States Follow Quantum Mechanics." (I actually like this, strange as it seems.)

"Quantum-like Representation of Extensive Form Games: Wine Testing Game" (I am a fan of reds.)

He also has some interesting papers about random fields, including: "Demystification of quantum entanglement".

 

[Dmitry67]

Can I ask my question again.
Why people are attacking that tiny partucular area?
There are no people who are trying to bring flogiston or either. Whats so special with LR?

 

[DrChinese]

Can I ask my question again.
Why people are attacking that tiny partucular area?
There are no people who are trying to bring flogiston or either. Whats so special with LR?

If someone asserts something on this board that is not generally accepted as if it is traditional science (i.e. they label it as fact, not an opinion)... that is the invitation.

On the other hand, generally accepted science is not "proven" so much as "supported". It usually makes useful predictions which can be used to rule out that theory. When folks make predictions which are contradicted by experiment, then it is time to change their theories. QM does not make any predictions at this time which are inconsistent with entanglement experiments. LR does. QM is supported and LR is not.

That is a generally accepted statement within the physics community, like it or not.

 

[yoda jedi]

2. The incompleteness argument was made in EPR and is now soundly rejected.

i refer to the "information loss problem" not to epr.



(and other details).

 

[DrChinese]

i refer to the "information loss problem" not to epr.
(and other details).

Not sure I follow. Can you be more specific? Are you talking about with black holes? If so I don't see the issue as being relevant.

 

[ThomasT]

A local realist believes that the results at Alice are not dependent on the results at Bob.

If a local realist is one who believes that LHV theories of entangled states haven't been definitively ruled out, then I'm a local realist. If a local realist is one who believes that such a theory is likely to be forthcoming, then I'm not a local realist.

Either way, I believe that in order to produce entangled states experimentally, then it's necessary that the results at Alice and Bob be statistically interdependent (which is accomplished entirely via local interactions/transmissions during the pairing process), which will also ensure that any associated Bell-type inequality will be violated.

Therefore, any correlations between Alice and Bob must be SPURIOUS because they are both causally connected to some other (prior) event.

Why would a common causal connection to some other (prior) event make them spurious? In fact we only see predictable correlations between results at Alice and Bob for two settings.

 

[DrChinese]

Why would a common causal connection to some other (prior) event make them spurious? In fact we only see predictable correlations between results at Alice and Bob for two settings.

Because the results at Alice are not causally related to the results at Bob. They are instead both causally related to some other prior event.

On the other hand, entanglement acts "as if" there is a causal connection.

 

[ThomasT]

Because the results at Alice are not causally related to the results at Bob. They are instead both causally related to some other prior event.

On the other hand, entanglement acts "as if" there is a causal connection.

Thanks for the replies. My latest was hurried and I had to cut it short. I had a couple more comments/questions which it looks like will have to wait until later, or tomorrow.

 

[yoda jedi]

why the glue ?

LOCAL REALISM ruled out?

"which concept, locality or realism, is the problem?"

i understand, is a type of realism (are jointly false).
a realism that is local.

cos the real, observed or not, exist.









-------------------
Bell inequalities are based on poincare relativity, have to be seen what happen in a sitter relativity.

 

[Demystifier]

Generally, Bell says:
No physical theory of local Hidden Variables can ever reproduce all of the predictions of Quantum Mechanics.

It is important to understand the assumptions under which the Bell theorem is obtained. As stressed, e.g., in
http://xxx.lanl.gov/pdf/quant-ph/0702225 [Rev. Mod. Phys. Vol. 81, No. 2, pp. 865-942 (2009)]
page 3, the assumptions are:
(i) realism
(ii) locality
(iii) free will
The theorem says that QM violates at least one of these 3 assumptions.

In particular, it is possible to retain both realism and locality if you give up free will. That's the idea of superdeterminism. The problem with that option is that it is very difficult to construct an explicit local-realistic model that has the same predictions as QM. ('t Hooft has attempts in this direction, but I don't think that these attempts are very successfull.)

On the other hand, if you give up locality, then it is easy to construct a nonlocal-realistic model consistent with QM. The simplest known model of that sort is Bohmian mechanics, which also turns out to be a superdeterministic model (no free will).

 

[Demystifier]

The http://arxiv.org/abs/0908.3408" won't be ruled out anytime soon.

Indeed. In fact, probably it will never be ruled out.

 

[Demystifier]

I agree with Akhmeteli that there exist experimental loopholes which do not allow us to say WITH ABSOLUTE CERTAINTY that nature is nonlocal.
However, what I don't understand is - why does it matter?

I mean, the above is true for ANY property of nature, we are not absolutely certain about anything. So why it is nonlocality, and not some other property of nature, that is questioned so frequently by serious physicists? Why nonlocality seems so difficult or weird to them?

 

[DrChinese]

Indeed. In fact, probably it will never be ruled out.

And indeed, it is a loophole for all physical theories, not just quantum mechanics. Relativity, evolution, big bang... all can be equally well explained by superdeterminism. With a mere wave of the hand, at that!

 

[ThomasT]

It is important to understand the assumptions under which the Bell theorem is obtained. As stressed, e.g., in
http://xxx.lanl.gov/pdf/quant-ph/0702225 [Rev. Mod. Phys. Vol. 81, No. 2, pp. 865-942 (2009)]
page 3, the assumptions are:
(i) realism
(ii) locality
(iii) free will
The theorem says that QM violates at least one of these 3 assumptions.

I don't think this is the clearest (read: correct) way to talk about what Bell's theorem means.

In particular, it is possible to retain both realism and locality if you give up free will. That's the idea of superdeterminism.

Superdeterminism and free will have nothing to do with it. It just has to do with the formal expression of locality.

The problem with that option is that it is very difficult to construct an explicit local-realistic model that has the same predictions as QM. ('t Hooft has attempts in this direction, but I don't think that these attempts are very successfull.)

Bell advanced a certain generic formulation for LHV models of quantum entangled states whose salient formal characteristic was assumed to be necessary for any LHV model of a quantum entangled state.

But that has by no means been proven to be the case.

The current state of affairs is that there's no formal expression of locality that is compatible with entanglement experimental designs and the salient feature of SQM formalization of entangled states (nonseparability or nonfactorability). Does 't Hooft's match all the precictions of SQM? Is it explicitly local?

A successful LHV model of entangled states can't be rendered in the straightforward factorable form proffered by Bell, because this alone contradicts a necessary condition of entanglement experiments which is the statistical interdependency of Alice's and Bob's results (outcome dependence) -- ie., Bell's locality condition is ambiguous.

On the other hand, if you give up locality, then it is easy to construct a nonlocal-realistic model consistent with QM. The simplest known model of that sort is Bohmian mechanics, which also turns out to be a superdeterministic model (no free will).

What's so realistic about the quantum potential and instantaneous action-at-a-distance?

Besides, there's no reason to give up locality.

So why it is nonlocality, and not some other property of nature, that is questioned so frequently by serious physicists?

Because of Bell?

Why nonlocality seems so difficult or weird to them?

It has no empirical foundation. Just an easy explanation for entanglement corrolations.

 

[MaxwellsDemon]

I agree with Akhmeteli that there exist experimental loopholes which do not allow us to say WITH ABSOLUTE CERTAINTY that nature is nonlocal.
However, what I don't understand is - why does it matter?

I mean, the above is true for ANY property of nature, we are not absolutely certain about anything. So why it is nonlocality, and not some other property of nature, that is questioned so frequently by serious physicists? Why nonlocality seems so difficult or weird to them?

I think locality is valued for a couple different reasons:

1. Simplicity, the assumption of locality makes calculations easier. For example, you can describe the physics of a game of billiards without having to worry about whether the planet Jupiter is moving through the constellation Gemini. You don't have to consider the effect of some ancient dinosaur's sneeze in your calculation when figuring out what force you should hit the 8 ball with.

2. Reductionism. Reductionism is the idea that you can completely describe the whole in terms of its parts and their interactions. Non-locality would be bad news for reductionism I think. (I can elaborate on why I suspect that if you would like...) Since reductionism seems to be the paradigm favored in modern science, an abandonment of locality would be distasteful to most.

 

[akhmeteli]

I agree with Akhmeteli that there exist experimental loopholes which do not allow us to say WITH ABSOLUTE CERTAINTY that nature is nonlocal.
However, what I don't understand is - why does it matter?

I mean, the above is true for ANY property of nature, we are not absolutely certain about anything. So why it is nonlocality, and not some other property of nature, that is questioned so frequently by serious physicists? Why nonlocality seems so difficult or weird to them?

I have been trying to avoid answering this question, but I guess your post was the last straw:-)

First, why was I reluctant to answer (of course, I cannot answer for other people)? Because, as long as there are no clear-cut theoretical or experimental reasons to reject locality (and I don't believe there are such reasons), this is a matter of personal preferences, maybe philosophical views, which a) take a long time to outline and b) are not very appropriate for this forum. So I'll try to stick to physical reasoning. I believe relativity strongly favors locality, and partial differential equations, which are all-pervading in physics, also favor locality. The absense of faster-than-light signaling also tends to suggest locality.

However, this is still a matter of preferences, so let me ask you, all other things equal, would you prefer a local theory, or a nonlocal one?

 

[DrChinese]

I think locality is valued for a couple different reasons:

1. Simplicity, the assumption of locality makes calculations easier. For example, you can describe the physics of a game of billiards without having to worry about whether the planet Jupiter is moving through the constellation Gemini. You don't have to consider the effect of some ancient dinosaur's sneeze in your calculation when figuring out what force you should hit the 8 ball with.

2. Reductionism. Reductionism is the idea that you can completely describe the whole in terms of its parts and their interactions. Non-locality would be bad news for reductionism I think. (I can elaborate on why I suspect that if you would like...) Since reductionism seems to be the paradigm favored in modern science, an abandonment of locality would be distasteful to most.

You don't need to change your calculations as is, Demystifier has tried to show this on many occasions. So that doesn't seem a fair critique. And if you did need to change them, that would actually be a near-proof of non-locality.

As to reductionism, the alternative is to abandon realism (which I am personally OK with). So that is probably equally distasteful if it comes down to taste.

 

[RUTA]

However, this is still a matter of preferences, so let me ask you, all other things equal, would you prefer a local theory, or a nonlocal one?

Causally local, constitutively non-local, i.e., non-separable.

 

[MaxwellsDemon]

Special Relativity is certainly local, but I would argue that General Relativity is not. In GR, the geometry is a global description, not a local one. Locally the geometry is flat, its only on a large scale that spacetime curvature comes into play. I would think that the principle of general covariance (where all “regular” derivatives in local laws are replaced with covariant derivatives when talking about large scale phenomena) is where this difference is most apparent. The covariant derivative still applies locally, but the extra term added in is dependent on the overall geometry. The curvature is something extra that requires a knowledge of the energy-momentum distribution in a region that goes beyond simply knowing the distribution in the here and now. The fact that we have to change our calculations in GR depending on the global geometric features of a region suggests to me that it is locality that needs to be abandoned. To me, abandoning realism is far more distasteful anyway. I prefer to think that concepts like position and momentum aren’t just ideas I have about nature, or biases from my human way of thinking, but that they have some objective foundation in reality. Even if they don’t exist exactly as I conceive of them, I’d like to think that a concrete objective phenomenon can be related to my ideas in some way. Color and temperature don’t exist as I perceive them, but there are still well defined objective things like wavelengths of light and atomic vibrations that can be related to my sensory experiences.

 

[Demystifier]

and partial differential equations, which are all-pervading in physics, also favor locality.

Schrodinger equation for two particles is a partial differential equation, so it is local but in the (6+1)-dimensional CONFIGURATION space, not in the ordinary (3+1)-dimensional space. This is exactly why QM is nonlocal (or nonseparable to be more precise) in the ordinary space, even though it is described by a partial differential equation. You may say that the world is still local, but then you must also say that the world contains a huge number of dimensions (3 new space dimensions for each particle). Does such a multi-dimensional local world makes you more happy?

 

[Demystifier]

And indeed, it is a loophole for all physical theories, not just quantum mechanics. Relativity, evolution, big bang... all can be equally well explained by superdeterminism. With a mere wave of the hand, at that!

Exactly.

 

[akhmeteli]

Schrodinger equation for two particles is a partial differential equation, so it is local but in the (6+1)-dimensional CONFIGURATION space, not in the ordinary (3+1)-dimensional space. This is exactly why QM is nonlocal (or nonseparable to be more precise) in the ordinary space, even though it is described by a partial differential equation. You may say that the world is still local, but then you must also say that the world contains a huge number of dimensions (3 new space dimensions for each particle). Does such a multi-dimensional local world makes you more happy?

First, I said PDE favor locality, not dictate it.

Second, we don't know what the final theory will look like: according to your papers and our previous discussions, you don't even believe all predictions of quantum theory will be confirmed experimentally.

Third, for pretty much any system A of (nonlinear) PDE in 3+1 dimensions one can construct a system of linear differential equations in the Fock space, which is equivalent to A on the set of solutions of A (see the outline of this result by Kowalski/Steeb in my post https://www.physicsforums.com/showpost.php?p=1825523&postcount=90 - some time ago I read about this result in nightlight's posts). That means that if quantum unitary evolution is successfully described by a linear system of equations in the Fock space (which is broader than any configuration space), you cannot be sure that system cannot be successfully replaced by a system of nonlinear equations in 3+1 dimensions. Therefore, you cannot be sure the system in the Fock space describes nonlocal reality.

And I would very much appreciate your answer to my question: everything els being equal, would you prefer a local theory, or a nonlocal one?

 

[yoda jedi]

Can you be more specific?

have not a unified dynamics for microscopic and macroscopic systems.
its physically incomplete.

 

[RUTA]

Special Relativity is certainly local, but I would argue that General Relativity is not. In GR, the geometry is a global description, not a local one. Locally the geometry is flat, its only on a large scale that spacetime curvature comes into play. I would think that the principle of general covariance (where all “regular” derivatives in local laws are replaced with covariant derivatives when talking about large scale phenomena) is where this difference is most apparent. The covariant derivative still applies locally, but the extra term added in is dependent on the overall geometry. The curvature is something extra that requires a knowledge of the energy-momentum distribution in a region that goes beyond simply knowing the distribution in the here and now. The fact that we have to change our calculations in GR depending on the global geometric features of a region suggests to me that it is locality that needs to be abandoned. To me, abandoning realism is far more distasteful anyway. I prefer to think that concepts like position and momentum aren’t just ideas I have about nature, or biases from my human way of thinking, but that they have some objective foundation in reality. Even if they don’t exist exactly as I conceive of them, I’d like to think that a concrete objective phenomenon can be related to my ideas in some way. Color and temperature don’t exist as I perceive them, but there are still well defined objective things like wavelengths of light and atomic vibrations that can be related to my sensory experiences.

First, it looks like you’ve conflated causal and constitutive locality. Your argument for the “nonlocality” of the covariant derivative is of the constitutive variety. See Howard, D., “Spacetime and Separability: Problems of Identity and Individuation in Fundamental Physics” in Potentiality, Entanglement and Passion-at-a-Distance, edited by R.S. Cohen et al., Kluwer Academic, Great Britain, 1997, pp. 113-141. Then you argue to keep “realism,” but realism in this sense is associated with constitutive locality, i.e., that entanglement violates causal locality and/or realism per EPR --> causal and/or constitutive nonlocality per Healey and Howard, for example. See also Healey, R.: Holism and Nonseparability in Physics: In: Zalta, E.N. (ed.) The Stanford Encyclopedia of Philosophy (Spring 2009 Edition), http://plato.stanford.edu/archives/spr2009/entries/physics-holism. For the term “constitutive locality” see Healey, R.: Gauging What’s Real: The Conceptual Foundations of Gauge Theories. Oxford University Press, Oxford (2007).

Essentially, EPR said there are quantum “objects” which possesses definite properties in and of themselves (realism) that are revealed by measurements independent of what’s being done to entangled partners at space-like separated events (causal locality). If you keep the causality requirement, you can explain the entangled outcomes by saying the quantum objects’ properties are not possessed in and of themselves, but they are “co-possessed” by entangled partners. That’s constitutive nonlocality/nonseparability.

Second, I don't agree that your argument establishes the constitutive nonlocality of the covariant derivative. As a differential geometry prof once emphasized, despite being definable via parallel transport, the covariant derivative is a local object independent of the choice of curve along which you parallel transport at a point on the manifold. You do need to input a vector in the tangent space of said point if by "covariant derivative" you mean the exterior derivative so restricted, but it's still local. See Misner, C.W., Thorne, K.S., Wheeler, J.A.: Gravitation. W.H. Freeman, San Francisco (1973).

So, while the measurement devices and outcomes are separated (constitutively local), the properties of the objects being measured are not per constitutive nonlocality. It’s hard to imagine (for most people, anyway) how nonseparability would be modeled, as the rest of your post indicates. If you’d like to see how we model constitutive nonlocality via discrete path integrals over graphs, see arXiv 0908.4348. It’s in the “revise and resubmit” mode at Foundations of Physics, but substantively it’s sound (at least the referees and editors had no complaints about its substance—if you find a mistake, please let us know).

 

[Demystifier]

And I would very much appreciate your answer to my question: everything els being equal, would you prefer a local theory, or a nonlocal one?

I don't understand what do you mean by "everything else".

 

[akhmeteli]

I don't understand what do you mean by "everything else".

That means just that - "everything else" :-), but since you insist:-), let me define it as follows. Let us imagine for a moment that in ten or twenty years from now, as a result of progress both in theory and experiment or just because some god told us the ultimate truth:-), we have a final quantum theory, which is fully self-consistent and perfectly agrees with all experiments. Would you prefer this theory to be local or nonlocal, provided that your own well-being and success of your own research does not depend on whether it is local or nonlocal?

You may wonder why I am persistently asking this question - because I would like to save some time:-) If you answer "yes", it'll be easier for me to answer or for you to understand why some people prefer locality, if you answer "no", maybe you'll be able to explain to me why you personally prefer nonlocality. So I am just trying to understand if your current preference for nonlocality can be explained by your personal preferences or you just believe that the current case for nonlocality is too strong to even think about a possibility of a local theory.

 

[Demystifier]

Akhmeteli, I would prefer nonlocality. The reasons are simple. First, because the Bell theorem strongly suggests (I will not say proves) that the quantum world is nonlocal, while experiments confirm the predictions of quantum mechanics. Second, because the wave function is a single mathematical object describing all particles at once, and nobody knows a reformulation of quantum mechanics in which this fact can be avoided. See also
http://xxx.lanl.gov/abs/quant-ph/0703071

 

[jambaugh]

Akhmeteli, I would prefer nonlocality. The reasons are simple. First, because the Bell theorem strongly suggests (I will not say proves) that the quantum world is nonlocal, while experiments confirm the predictions of quantum mechanics.

(If I may hum yet another chorus of the CI song...)

It is hard to say an RAA argument "suggests" which of the prior assumptions should be considered false. It rather points that the assumptions as a whole are mutually inconsistent. However to my mind we can't have "reality" if we sacrifice local causality if one is to accept special relativity.

SR + violation of local causality implies future actions can affect past states of reality. What then is the meaning of "reality" if it is not objectively defined and immutable once in the past?

I think it no more strange to reject absolute reality in QM than to reject absolute time in SR. It just takes some getting used to. The alternative is not nihilism or illusory mind created phenomena. It is a relative actuality of observed phenomena without the underlying assumption of a clockwork objective mechanism. We don't assume, we don't deny we simply pay attention only to the scientifically meaningful observations and observables without painting our own prejudices about what must lie beneath.

Alternative "interpretations" always remind me of the epicycles invented to hold onto the Platonic perfection of circular motion in spite of the evidence to the contrary in observed planetary behavior. I think the "absolute objective reality" hypothesis is similar to the Platonism of old in this sense.

 

[Demystifier]

SR + violation of local causality implies future actions can affect past states of reality. What then is the meaning of "reality" if it is not objectively defined and immutable once in the past?

There is a way out of this problem. Due to violation of local causality, some properties of the system in the past are determined by some properties of the system in the future. However, it does not mean that the past can be changed. Since there is only one future (the one that will actually happen), there is only one past as well (the one that has actually happened). Once the past is known, the future cannot be changed in a way that would contradict the known past.

See also the attachment in
https://www.physicsforums.com/showpost.php?p=2455753&postcount=109

 

[jambaugh]

There is a way out of this problem. Due to violation of local causality, some properties of the system in the past are determined by some properties of the system in the future. However, it does not mean that the past can be changed. Since there is only one future (the one that will actually happen), there is only one past as well (the one that has actually happened). Once the past is known, the future cannot be changed in a way that would contradict the known past.

See also the attachment in
https://www.physicsforums.com/showpost.php?p=2455753&postcount=109

I understand that. But that just boils it all down to "known" past i.e. observables instead of states. As a philosophical foundation, the "reality" of the "unobserved past" is meaningless in this context so why continue to work with it? The reason for invoking the "reality hypothesis" is no less invalid given this "way out".

If you are going to work with "tentative reality" then call it what it is, classes of possible observations. My point is that one can still reject the absolute reality of what is not observed (past, present, and future) while retaining the desired local causality. The reverse just isn't possible, your comments not withstanding.

Either you have violation of local causality with its implied invalidation of (unobserved) reality or you have local causality with QM+Bell invalidation of unobserved reality.

So reality being lost, we can still retain local causality if it, by itself, is consistent with observation. We know it to be consistent with predicted observations in QM, via the "no Bell telephones" theorem.

 

[Demystifier]

I understand that. But that just boils it all down to "known" past i.e. observables instead of states. As a philosophical foundation, the "reality" of the "unobserved past" is meaningless in this context so why continue to work with it? The reason for invoking the "reality hypothesis" is no less invalid given this "way out".

Well, this way out works even if you replace the word "known" by the word "real". I am not saying here that reality is necessary or needed or desirable (nor I'm saying that it is not), but I AM saying that reality may be compatible with SR and nonlocality. Maybe there is no reality, but SR+nonlocality are not a valid argument against reality.

 

[yoda jedi]

Well, this way out works even if you replace the word "known" by the word "real". I am not saying here that reality is necessary or needed or desirable (nor I'm saying that it is not), but I AM saying that reality may be compatible with SR and nonlocality. Maybe there is no reality, but SR+nonlocality are not a valid argument against reality.

and in any case, the fact, that if there is no CFD, does not mean that there is no reality,
cos we can ask, what do you observe ? (or measure) its something, and then, something its reality.

 

[Peter Morgan]

Well, this way out works even if you replace the word "known" by the word "real". I am not saying here that reality is necessary or needed or desirable (nor I'm saying that it is not), but I AM saying that reality may be compatible with SR and nonlocality. Maybe there is no reality, but SR+nonlocality are not a valid argument against reality.

All through this thread, I find no distinction made between dynamical nonlocality and nonlocality of initial conditions (not that the distinction is much made in the literature). Dynamical locality is essentially preferred by classical physics. Initial conditions of a classical dynamics, however, are essentially always nonlocal, whether the dynamics are Newtonian or Lorentz invariant, because at a given time we have to specify the position and momentum of all particles, everywhere on a space-like hyperplane. Furthermore, classically, if we observe some phenomenon that requires a weird set of initial conditions, then that just means that the initial conditions in the past were also weird. This is all that superdeterminism is --- if what we observe now is weird, the setup must have been weird too. It makes no difference whether we introduce local or nonlocal dynamics. All of which is to say, Demystifier, that I say with you that "reality may be compatible with SR and nonlocality".

The only superdeterminism that is required to model the Bell-EPR situation, however, is superdeterminism of the evolution of probability densities. That is, if the probability density now is weird, then the probability density in the past must also have been weird. Superdeterminism of the state of a classical deterministic dynamics is not necessary. Amongst other consequences, it's therefore not necessary to impinge much on free will, unless, I suppose, one wants to deny that probability can be applied to model people's microbehaviour.

Another distinction not introduced here, in my look through, is contextuality. It's well-established that noncontextuality alone is enough to derive Bell inequalities. The distinction can also be put in terms of whether we regard settings of an instrument as parameters of a model or as observables in the model. Noncontextuality is rather against the spirit of classical particle modeling, and arguably can be thought of as anti-realist relative to particle properties, but it is not against the spirit of classical field models. Indeed, for field systems at thermal equilibrium the global configuration of an experimental apparatus condition the thermal equilibrium state, just as the Copenhagen interpretation insists it should. Think heat equation in contact with various heat reservoirs. I include a sketch of an experiment that gives more detail (which is of course needed) in a recent preprint, http://arxiv.org/abs/1001.4993" (this says nothing against any of the sophisticated interpretations that are out there, each of which gives its own interesting way of thinking about QM, and each of which a person may reasonably find more-or-less in tune with their own intuitive preferences).

Happy hunting!

 

[DrChinese]

All through this thread, I find no distinction made between dynamical nonlocality and nonlocality of initial conditions (not that the distinction is much made in the literature). Dynamical locality is essentially preferred by classical physics. Initial conditions of a classical dynamics, however, are essentially always nonlocal, whether the dynamics are Newtonian or Lorentz invariant, because at a given time we have to specify the position and momentum of all particles, everywhere on a space-like hyperplane. Furthermore, classically, if we observe some phenomenon that requires a weird set of initial conditions, then that just means that the initial conditions in the past were also weird. This is all that superdeterminism is --- if what we observe now is weird, the setup must have been weird too. It makes no difference whether we introduce local or nonlocal dynamics. All of which is to say, Demystifier, that I say with you that "reality may be compatible with SR and nonlocality".

The only superdeterminism that is required to model the Bell-EPR situation, however, is superdeterminism of the evolution of probability densities. That is, if the probability density now is weird, then the probability density in the past must also have been weird. Superdeterminism of the state of a classical deterministic dynamics is not necessary. Amongst other consequences, it's therefore not necessary to impinge much on free will, unless, I suppose, one wants to deny that probability can be applied to model people's microbehaviour.

Another distinction not introduced here, in my look through, is contextuality. It's well-established that noncontextuality alone is enough to derive Bell inequalities. The distinction can also be put in terms of whether we regard settings of an instrument as parameters of a model or as observables in the model. Noncontextuality is rather against the spirit of classical particle modeling, and arguably can be thought of as anti-realist relative to particle properties, but it is not against the spirit of classical field models. Indeed, for field systems at thermal equilibrium the global configuration of an experimental apparatus condition the thermal equilibrium state, just as the Copenhagen interpretation insists it should. Think heat equation in contact with various heat reservoirs. I include a sketch of an experiment that gives more detail (which is of course needed) in a recent preprint, http://arxiv.org/abs/1001.4993" (this says nothing against any of the sophisticated interpretations that are out there, each of which gives its own interesting way of thinking about QM, and each of which a person may reasonably find more-or-less in tune with their own intuitive preferences).

Happy hunting!

I saw that a few days ago on the arxiv and just starting reading it. For those interested, it comments on a paper by Navascues and Wunderlich regarding classic-quantum correspondence. It also has some good references.

 

[nikman]

FWIW in "Quantum nonlocality vs. Einstein locality" Dieter Zeh makes a distinction between "dynamic" and "kinematic" (quote):

Quantum theory is kinematically nonlocal, while the theory of relativity (including relativistic quantum field theory) requires dynamical locality ("Einstein locality"). How can these two elements of the theory (well based on experimental results) be simultaneously meaningful and compatible? How can dynamical locality even be defined in terms of kinematically nonlocal concepts?

http://www.rzuser.uni-heidelberg.de/~as3/nonlocality.html

 

[Peter Morgan]

I saw that a few days ago on the arxiv and just starting reading it. For those interested, it comments on a paper by Navascues and Wunderlich regarding classic-quantum correspondence. It also has some good references.

Thanks, DrC, and I'd be glad of your comments as always, here or by e-mail. There's an after-thought to this Comment, which is that a friend pointed out that the arXiv version of the paper it comments on does not includes the word "field" at all. To appreciate the details of the argument therefore requires the Proc.Roy.Soc.A paper. I'm somewhat curious whether the published version only introduces the classical field concept because a referee introduced the question (which might slightly improve the chances of the Comment being accepted, because the Proc.Roy.Soc.A editorial procedure for Comments includes the original paper's referee if the editors decided to send it to referees).

Fortunately, I believe the published version is freely available at http://rspa.royalsocietypublishing.org/content/466/2115/881" because of the Proc.Roy.Soc.A anniversary celebrations.

The chance of this Comment being accepted by Proc.Roy.Soc.A is small. The editorial board will presumably understand that discontent would be expressed in some quarters if they were to accept it, so I presume they will only accept it if it touches something of their own interests in the question.

 

[yoda jedi]

"A glance beyond the quantum model"

Happy hunting!

http://arxiv.org/PS_cache/arxiv/pdf/0907/0907.0372v1.pdf

...Here we propose a fundamental axiom that we believe any reasonable post-quantum theory should satisfy, namely, that such a theory should recover classical physics in the macroscopic limit...

coincidence (a correlation, or better yet a metaphysical corelation), i am reading:


On the Classical Limit of Quantum Mechanics
http://www.springerlink.com/content/p57117239x631547/fulltext.pdf


...In spite of many results of the standard approach, it is not yet clear how to explain within standard quantum mechanics the classical motion of macroscopic bodies.....


but suffers the same problems that standard quantum theory, is a patchwork proto-theory.

 

[Peter Morgan]

FWIW in "Quantum nonlocality vs. Einstein locality" Dieter Zeh makes a distinction between "dynamic" and "kinematic" (quote):

Quantum theory is kinematically nonlocal, while the theory of relativity (including relativistic quantum field theory) requires dynamical locality ("Einstein locality"). How can these two elements of the theory (well based on experimental results) be simultaneously meaningful and compatible? How can dynamical locality even be defined in terms of kinematically nonlocal concepts?

http://www.rzuser.uni-heidelberg.de/~as3/nonlocality.html

Thanks for this. Definitely worthwhile. I'm not as familiar with Zeh's thinking on environmental decoherence as I should be.

As an aside, I went to Foundations of Physics for Zeh's most recently mentioned paper on his web-site, "Quantum discreteness is an illusion", which is not yet published but is available as an "online first" paper. The quality of the (69!) papers in the "online first" queue (that's probably 6 months ahead) shows signs of 't Hooft's tenure as editor starting to make a very big difference. The list of authors who have decided to publish at FoP is close to stellar.

 

[ThomasT]

Akhmeteli, I would prefer nonlocality. The reasons are simple. First, because the Bell theorem strongly suggests (I will not say proves) that the quantum world is nonlocal ...

Is Bell's theorem about the way the quantum world is, or is it about limitations on the formalization of entangled states?

 

[nikman]

As an aside, I went to Foundations of Physics for Zeh's most recently mentioned paper on his web-site, "Quantum discreteness is an illusion", which is not yet published but is available as an "online first" paper. The quality of the (69!) papers in the "online first" queue (that's probably 6 months ahead) shows signs of 't Hooft's tenure as editor starting to make a very big difference. The list of authors who have decided to publish at FoP is close to stellar.

One was gratified that 't Hooft published Suarez's "Nonlocal 'Realistic' Leggett Models" paper. It's hard to imagine two scientists with more starkly contrasting world-views than that pair.

http://www.springerlink.com/content/v5652005u01628h2/

or, for you members of the vast unfunded public, grab it gratis:

http://www.quantumphil.org/SuarezFOOP201R2.pdf

 

[yoda jedi]

Well, this way out works even if you replace the word "known" by the word "real". I am not saying here that reality is necessary or needed or desirable (nor I'm saying that it is not), but I AM saying that reality may be compatible with SR and nonlocality. Maybe there is no reality, but SR+nonlocality are not a valid argument against reality.

and in any case, the fact, that if there is no CFD, does not mean that there is no reality,
cos we can ask, what do you observe ? (or measure) its something, and then, something its reality.

same thing for suarez:
http://arxiv.org/PS_cache/arxiv/pdf/0705/0705.3974v1.pdf

.....Additionally, Bohm’s objective description can no longer be considered completely “realistic” since in experiments involving entangled polarized photon pairs neither of the two photons carries a definite polarization when it leaves the source....

 

[akhmeteli]

Akhmeteli, I would prefer nonlocality. The reasons are simple. First, because the Bell theorem strongly suggests (I will not say proves) that the quantum world is nonlocal, while experiments confirm the predictions of quantum mechanics. Second, because the wave function is a single mathematical object describing all particles at once, and nobody knows a reformulation of quantum mechanics in which this fact can be avoided. See also
http://xxx.lanl.gov/abs/quant-ph/0703071

Thank you for your answer. So it looks like you prefer nonlocality not because you like it more than locality, but because you think theory and experiment favor it. However, as I argued starting this thread, there are no no-go theorems or no-go experiments ruling out locality, so I don't have your reasons to favor nonlocality, and I regard it as a radical notion, and the burden of proof is very high for such radical ideas. As for the absence of a local reformulation of quantum mechanics, let me give you an example. For a quarter of a century after formulation of modern quantum mechanics, the de Broglie - Bohm interpretation, while existed (in the form offered by de Broglie), was dead, for all intents and purposes. Nevertheless, it was resuscitated by Bohm. And even now, as far as I know, there is no generally recognized relativistic form of this interpretation (actually, you told me that some time ago, and I don't think much has changed since then). But I guess you believe there will be such relativistic form in the future, and even have your own suggestions. So we don't know what can happen in the future. And I mentioned one possibility how a nonlocal theory can be a local theory in disguise.

So it started with your question: what's so special about locality. I gave you my reasons. Again, if there were some iron-clad no-go arguments, I would have to accept nonlocality. So far I see no reasons for that.

 

[Demystifier]

while the theory of relativity (including relativistic quantum field theory) requires dynamical locality ("Einstein locality").

You have a too narrow view of the concept of relativity. If by relativity one means only that the laws of physics do not depend on the choice of spacetime coordinates, then relativity does not require locality.

 

[Demystifier]

theory and experiment favor it. However, as I argued starting this thread, there are no no-go theorems or no-go experiments ruling out locality, so I don't have your reasons to favor nonlocality, and I regard it as a radical notion, and the burden of proof is very high for such radical ideas.

I still don't understand your logic. So I'll start with a question. Do you agree that theory and experiment favor nonlocality? (I'm not asking if they definitely prove it, because they don't. I'm only asking if they favor it.)

And I mentioned one possibility how a nonlocal theory can be a local theory in disguise.

If you mean your idea that a single charged particle guided by the wave function can be viewed as being guided by the electromagnetic potential (which is an interesting idea), then it has nothing to do with locality and nonlocality. To say anything about nonlocality, you must consider a system of at least two entangled particles.

 

[Maaneli]

If you mean your idea that a single charged particle guided by the wave function can be viewed as being guided by the electromagnetic potential (which is an interesting idea), then it has nothing to do with locality and nonlocality. To say anything about nonlocality, you must consider a system of at least two entangled particles.

I think Akhmeteli is referring to this claim in one of his prior posts:

... for pretty much any system A of (nonlinear) PDE in 3+1 dimensions one can construct a system of linear differential equations in the Fock space, which is equivalent to A on the set of solutions of A (see the outline of this result by Kowalski/Steeb in my post https://www.physicsforums.com/showpos...3&postcount=90 - some time ago I read about this result in nightlight's posts). That means that if quantum unitary evolution is successfully described by a linear system of equations in the Fock space (which is broader than any configuration space), you cannot be sure that system cannot be successfully replaced by a system of nonlinear equations in 3+1 dimensions. Therefore, you cannot be sure the system in the Fock space describes nonlocal reality.

 

[jambaugh]

You have a too narrow view of the concept of relativity. If by relativity one means only that the laws of physics do not depend on the choice of spacetime coordinates, then relativity does not require locality.

Not as such, of course. (Special) Relativity however sets the stage for the argument of local causality. If SR is valid and the light-cone is not a causal horizon then we have the potential of (not certainty of) constructing causal paradoxes. So absent local causality how are such paradoxes prohibited? If we actually (in our conceptual model of how nature works) allow causal feedback, future to past, it seems to me then we must invoke a "meta-time" over which such phenomena would decay out or reinforce to a caustic threshold or stable oscillation, (the local "reality" oscillating w.r.t. this meta-time). This was an attractive idea to me once, as e.g. a model for superposition and interference phenomena. But eventually I rejected it as fanciful and meaningless.

The problem as I see it is this sort of speculation is not operationally meaningful. It's no different than supposing an invisible aether, or Everette many worlds. Sure you can speculate but you can't test within the bounds of science. Such phenomena are by their nature beyond observation. Again I see the "reality" of it as meaningless within the context of science. That isn't an argument, just the results of my many internal arguments over past years.

What in the end do we mean by "reality"? Generally it is the reality of a universe of objects with always defined (though not always observed) objective properties or states of being.
Classically that is either the particles or the field quantities at each point of space.

Quantum mechanically we work in a language of phenomena, observables and observations and interactions between systems. I think it incorrect to objectify the mathematical constructs (esp. wave function=hilbert space vector). For that matter I think we should abandon the use of the Hilbert spaces all together except in the mathematics of constructing the Lie algebras and groups where the observables and dynamics are represented.

As to causal locality, that is easily enough described within QM and QFT via the structure of the dynamics. And it is easily enough tested both conceptually via though experiments, and in the lab. The only evidence I can conceive of, for true non-local causation is a classical FTL signal, e.g. a "Bell telephone". If you can't send a classical signal then you aren't talking about observable non-locality and thus speculating beyond the scope of science.

Now having said all that, I do think that if we're ever going to succeed at merging GR and QM we'll need to start with a "pre-local" theory. That is to say a theory of interacting quantum systems out of which condenses the macroscopic classical world. In which case I envision the local causality definition to be rather reversed. Nearness is ultimately defined by causal interactions and the causal structure of interacting systems ultimately defines the light-cones, and space-time metric structure. Objects are spatially close because they look close. This means they strongly interact with our eyes and our flashlights, or our radar antennas, or our fingers or sticks in our hands.

So ultimately I think causality will by definition be local because locality is ultimately based on causal interactions. At the microscopic quantum level this may break down, but not in the causal aspect, but rather the loss of meaning to geometry (and possibly even topology) at the small scale.

 

[nikman]

You have a too narrow view of the concept of relativity. If by relativity one means only that the laws of physics do not depend on the choice of spacetime coordinates, then relativity does not require locality.

You should really take that up with Herr Professor-Doktor Heinz-Dieter Zeh. I suspect he's making a deeper distinction relating to fundamental correlation and causality. Kinematics does of course rear its head in Relativity with the Lorentz contraction.

 

[RUTA]

Objects are spatially close because they look close. This means they strongly interact with our eyes and our flashlights, or our radar antennas, or our fingers or sticks in our hands.

It's possible to have a distant object be brighter than a closer object, e.g., the Sun is much brighter than this computer screen. Likewise the angle subtended by an object doesn't discriminate relative spatial distance. How do you envision relating distance and interaction? And, how do you see your approach giving a Lorentz invariant result, since it can't give a definite spatial separation and be Lorentz invariant?