Does entanglement violate special relativity?

[ZapperZ]

Anyone participating in this thread who does not get back to the main topic and stop the personal attack won't be here for very long.

Zz.

 

[ueit]

I don't think that a reinterpretation of QM can make it local:
http://xxx.lanl.gov/abs/quant-ph/0703071
Essentially, since no interpretation can remove nonlocal elements of the theory (such as wave functions in the configuration space or something equivalent), the theory should be considered nonlocal, irrespective of its interpretation.

This is not true. The wave-function could simply describe correlations that are caused locally by other means, in the same way you can synchronize two or more distant clocks, by light-speed communication with a master clock.

 

[Demystifier]

This is not true. The wave-function could simply describe correlations that are caused locally by other means, in the same way you can synchronize two or more distant clocks, by light-speed communication with a master clock.

In principle, you could be right. But the fact is that nobody yet has been successfull to mathematically reformulate QM in a completely local form. The (unproved) conjecture in the paper above states that such a local reformulation is indeed impossible.

 

[fermi]

I don't think that a reinterpretation of QM can make it local:
http://xxx.lanl.gov/abs/quant-ph/0703071
Essentially, since no interpretation can remove nonlocal elements of the theory (such as wave functions in the configuration space or something equivalent), the theory should be considered nonlocal, irrespective of its interpretation.

I fully agree, but I have a follow-up question: Unlike QM, the quantum field theory (QFT) bends over backwards to be a "local" theory. In fact, it is unclear how to write down a self-consistent non-local field theory. I have always (implicitly) assumed that QFT was a generalization of QM, and it included QM as a special case for slowly moving particles. Does this really mean then, that QFT does not completely include QM? If so, this would mean that QFT is incomplete since QM's entanglement has been shown to be in agreement with all experiments. Perhaps QFT is also non-local in some subtle way?

 

[ueit]

In principle, you could be right. But the fact is that nobody yet has been successfull to mathematically reformulate QM in a completely local form. The (unproved) conjecture in the paper above states that such a local reformulation is indeed impossible.

What if you choose to reformulate BM so that the trajectory of each particle is a function of only the initial conditions at Big-Bang (wave function + particle distribution)? As these initial conditions can be treated as a constant, you have a formally local theory.

 

[Demystifier]

What if you choose to reformulate BM so that the trajectory of each particle is a function of only the initial conditions at Big-Bang (wave function + particle distribution)? As these initial conditions can be treated as a constant, you have a formally local theory.

Such a local reinterpretation does not make the theory formally local in the sense of formal locality as defined in the paper mentioned above. You still must work with nonlocal wave functions, your (otherwise reasonable) idea does not avoid this.

 

[Demystifier]

... I have a follow-up question: Unlike QM, the quantum field theory (QFT) bends over backwards to be a "local" theory. In fact, it is unclear how to write down a self-consistent non-local field theory. I have always (implicitly) assumed that QFT was a generalization of QM, and it included QM as a special case for slowly moving particles. Does this really mean then, that QFT does not completely include QM? If so, this would mean that QFT is incomplete since QM's entanglement has been shown to be in agreement with all experiments. Perhaps QFT is also non-local in some subtle way?

You misunderstood the origin of nonlocality of quantum theory. QFT is not more local than ordinary QM. They are both local, in the sense that the classical Lagrangians/Hamiltonians (which these theories quantize) describe classically local interactions. But nonlocality emerges from entanglement (which does not have a classical analog), which is equally present in both QM and QFT.

Still, it IS questionable whether QFT includes QM as a special case, but for other reasons.

You can see a more detailed (and pedagogic) discussion of all these aspects of quantum theory in
http://xxx.lanl.gov/abs/quant-ph/0609163 [Found.Phys.37:1563-1611,2007]

 

[Isaac_Newton]

Are we off topic (BM etc)? Entanglement correlation and SR:
a) Its not actually information so it may travel FTL and not violate SR.
b) It is information and does travel FTL but does not violate SR. How?
c) There is nothing that travels from one entangled particle to the other.
d) Its all done in information where there is no space anyway.
e) Anything else?

I am on Twitter (Isaac_Newton) because 2 lines of QM is a good byte sized quantity. Join me.

 

[Demystifier]

Are we off topic (BM etc)? Entanglement correlation and SR:
a) Its not actually information so it may travel FTL and not violate SR.
b) It is information and does travel FTL but does not violate SR. How?
c) There is nothing that travels from one entangled particle to the other.
d) Its all done in information where there is no space anyway.
e) Anything else?

I vote for b). How exactly it happens? We don't know. But I have provided one example how it MIGHT happen. This example demonstrates that it is at least possible. Yes, it is based on pilot-wave theory, and yes you are right that pilot waves are not observed. Yet, it does not imply that pilot waves do not exist. You will say that it is not reasonable to believe that unobserved objects exist, and you may be right. Still, the point is that it IS possible to have FTL and SR at the same time. Many people think that it is simply impossible, so finding a counterexample, even if a very unnatural one, is a valuable result.

Of course, I also feel that this counterexample is also very natural. However, naturalness is not something that can be objectively defined, so let us not discuss the issue of naturalness.

And let us avoid personal attacks, OK?

So, what do YOU vote for?

 

[ueit]

Such a local reinterpretation does not make the theory formally local in the sense of formal locality as defined in the paper mentioned above. You still must work with nonlocal wave functions, your (otherwise reasonable) idea does not avoid this.

The only thing you need is the wave function at Big Bang and particle configuration at Big-Bang. Because BM is a deterministic theory you can calculate anything you want from those two parameters (which can be included as constants into the law of motion). You can describe the trajectory of any particle only as a function of these constants and time. The theory is local.

I would say that no deterministic theory can be proven to be non-local because there exists at least a formulation (only in terms of some initial conditions) that does not require any type of interaction, much less a non-local one. Newtonian gravity is another example.

 

[yossell]

They are allowed to affect each other instantly - there is no rule against that.
Information has a very specific meaning in relativity - sending a random value that you can't influence isn't information.

The first sentence surprises me. It is commonly thought that an absolute notion of simultaneity goes against Special Relativity - but 'instantly' seems to introduce exactly such a notion. Is this wrong?

I can see that it's possible to save the predictions of SR if it turns out that the 'privileged' frame of reference, the one where simultaneity in this frame is identical with this absolute notion of simultaneity, is undetectable - we had better not be able to synchronise distant clocks using these affects that are transmitted 'instantly'. And, indeed, the no signalling theorem shows that we can't. But the collapse postulate, understood as a genuine dynamical law, would not be true in different Lorentz frames.

 

[Dmitry67]

The only thing you need is the wave function at Big Bang and particle configuration at Big-Bang. Because BM is a deterministic theory you can calculate anything you want from those two parameters (which can be included as constants into the law of motion). You can describe the trajectory of any particle only as a function of these constants and time. The theory is local.

The return of Laplace determinism :)
particle configuration at Big-Bang defines what I am going to type right now.
Imagine that it is right. Doesnt it mean that the entropy at Big bang was VERY HIGH, contrary to what we know about it?

 

[ueit]

The return of Laplace determinism :)
particle configuration at Big-Bang defines what I am going to type right now.
Imagine that it is right. Doesnt it mean that the entropy at Big bang was VERY HIGH, contrary to what we know about it?

We don't have a theory capable of describing the Big-Bang, therefore I have no idea what the entropy was back then.

Also, BM in its current form does not account for particle creation, so we cannot go too far with this idea. I wanted only to point out that for a generic deterministic theory it is impossible to prove its non-locality in a mathematically rigorous manner.

P.S.

Do you have a problem with the fact that your actions were predetermined at Big-Bang? Do you prefer them to be randomly chosen?

 

[Dmitry67]

P.S.

Do you have a problem with the fact that your actions were predetermined at Big-Bang? Do you prefer them to be randomly chosen?

As you remember, I am MWI fan, so neither is applicaple to me :)

P.S.
And yes, even CI randomness is better then the Laplace determinism. If everything is predetermined then the TOE looks like
* several TOE equations - 1 page
* Appendix 1: Initial conditions, 10^1000000000 pages :)

 

[ueit]

As you remember, I am MWI fan, so neither is applicaple to me :)

P.S.
And yes, even CI randomness is better then the Laplace determinism. If everything is predetermined then the TOE looks like
* several TOE equations - 1 page
* Appendix 1: Initial conditions, 10^1000000000 pages :)

On what exactly do you base your assumptions that the initial conditions must be so complex?

Isn't MWI in the same situation (but messed up because there is no way to know in which branch you are going to be)?

 

[ueit]

As you remember, I am MWI fan, so neither is applicaple to me :)

P.S.
And yes, even CI randomness is better then the Laplace determinism. If everything is predetermined then the TOE looks like
* several TOE equations - 1 page
* Appendix 1: Initial conditions, 10^1000000000 pages :)

On what exactly do you base your assumption that the initial conditions must be so complex?

Isn't MWI in the same situation (but messed up because there is no way to know in which branch you are going to be)?

 

[Ilja]

I would say that no deterministic theory can be proven to be non-local because there exists at least a formulation (only in terms of some initial conditions) that does not require any type of interaction, much less a non-local one. Newtonian gravity is another example.

The notion of causality makes only sense if you assume independent decisions of experimenters.

Superdeterministic theories (everything has to be computed from initial values) simply have no notion of causality, thus, it makes no sense to talk about locality.

But in deterministic theories you can make sense of causality. It is a property of the evolution equations of the theory.

 

[ueit]

The notion of causality makes only sense if you assume independent decisions of experimenters.

Superdeterministic theories (everything has to be computed from initial values) simply have no notion of causality, thus, it makes no sense to talk about locality.

But in deterministic theories you can make sense of causality. It is a property of the evolution equations of the theory.

A fundamental theory that is deterministic is necessary superdeterministic. A deterministic theory at a fundamental level allowing "independent decisions of experimenters" is logically contradictory, therefore it makes no sense to seriously speak about it.

 

[Dmitry67]

A fundamental theory that is deterministic is necessary superdeterministic.

Wrong
It is true only if you assume (like most people do) "single history" which is not a case in MWI

 

[Dmitry67]

On what exactly do you base your assumption that the initial conditions must be so complex?

Isn't MWI in the same situation (but messed up because there is no way to know in which branch you are going to be)?

No, MWI everything can start from an empty space, for example 0
Then there are fluctuations, you get 01 and 10 branches
In each area of space fluctuations ddmore and more entropy: 0001, 0010, 0101, ... etc.
System becomes more and more complex

 

[DrChinese]

A fundamental theory that is deterministic is necessary superdeterministic. A deterministic theory at a fundamental level allowing "independent decisions of experimenters" is logically contradictory, therefore it makes no sense to seriously speak about it.

This is inaccurate, determinism does NOT imply superdeterminism. Even if my decisions are predetermined, that does not mean that the Bell Inequality will be violated by my choices of measurement settings - which is the premise of superdeterminism.

Your superdeterminism is supposed to explain something, and it doesn't. Every particle would need to have all the details of all other particles contained locally to work. We have previously discussed this point in other threads, and concluded that superdeterminism has baggage. Further debate of superdeterminism here would be off-topic.

 

[Isaac_Newton]

We are drifting off into superdeterminism. Read the thread title.

One example of 'instant something' is the collapse of the wave function in every path of a split beam or all over a spherical surface created by a photon spreading out. When a photon is observed then 'it knows' its been observed at every possible place the wave function could be - instantly.

Although its a negative its as strange as correlation 'exchanges' & probably for the same reason. - For me its about information in information space, so I see no problem, but for the rest of you... boy, do you have problems!

 

[Dmitry67]

I don't have problems, I believe in MWI.
The hardest thing of all is to find a black Shroedinger's cat in a dark room, especially if there is no wavefunction collapse.

 

[Demystifier]

The only thing you need is the wave function at Big Bang and particle configuration at Big-Bang.

Yes, but the many-particle wave function at the initial time is still a non-local object, the time evolution has nothing to do with it. Since it is a many-particle wave function, you cannot specify it by specifying psi at each point of space. Instead, for 2 particles you need to specify psi for each PAIR of points on space, and similarly for n particles. This is why it is a nonlocal object even without the time evolution. You probably want to say that there is no nonlocal force, but my point is that there is nonlocal - something.

 

[ueit]

Wrong
It is true only if you assume (like most people do) "single history" which is not a case in MWI

No, MWI is superdeterministic. Once universal wavefunction is specified everything folows. A single branch is not superdeterministic but MWI is not about a single branch.

 

[Demystifier]

This is inaccurate, determinism does NOT imply superdeterminism. Even if my decisions are predetermined, that does not mean that the Bell Inequality will be violated by my choices of measurement settings - which is the premise of superdeterminism.

The point is that predetermined choices are not choices at all. Instead, your arms MUST make the measurement settings that will provide violation of Bell inequalities.

 

[Dmitry67]

No, MWI is superdeterministic. Once universal wavefunction is specified everything folows. A single branch is not superdeterministic but MWI is not about a single branch.

In "bird's view", yes
But for an observer reality looks random
So we agreed on it...

 

[ueit]

No, MWI everything can start from an empty space, for example 0
Then there are fluctuations, you get 01 and 10 branches
In each area of space fluctuations ddmore and more entropy: 0001, 0010, 0101, ... etc.
System becomes more and more complex

As far as I know MWI starts with a wavefunction. That wavefunction evolves deterministicaly in accordance with Schroedinger's equation. You should explain what do you mean by "empty space" and "fluctuations" in this context.

 

[Demystifier]

Also, BM in its current form does not account for particle creation, so we cannot go too far with this idea.

Well, there are some variants of BM that do account for particle creation. For example, the string-theory variant of BM does this in a natural way:
http://xxx.lanl.gov/abs/hep-th/0702060
http://xxx.lanl.gov/abs/0705.3542 [Europhys. Lett.85:20003, 2009]

 

[Demystifier]

The return of Laplace determinism :)
particle configuration at Big-Bang defines what I am going to type right now.
Imagine that it is right. Doesnt it mean that the entropy at Big bang was VERY HIGH, contrary to what we know about it?

No, it doesn't. Nature may choose initial conditions that have a small entropy.