How Does Quantum Entanglement Not Violate Causality?

[TMO]

Suppose I have an apparatus A that is entangled with apparatus B. In my reference frame, I observe apparatus A, which simultaneously causes apparatus B to do its thing. However, because there exists a reference frame where apparatus B does its thing before apparatus A, it follows that there exists a reference frame in which B does its thing before it was caused by A to do so.

 

[phinds]

Suppose I have an apparatus A that is entangled with apparatus B. In my reference frame, I observe apparatus A, which simultaneously causes apparatus B to do its thing. However, because there exists a reference frame where apparatus B does its thing before apparatus A, it follows that there exists a reference frame in which B does its thing before it was caused by A to do so.

1) You are over complicating the situation. Either A or B is "observed" first and the other is then locked into a value
2) This does not violate causality because no information is exchanged, you have simply done a measurement on the overall system.

 

[atyy]

Quantum entanglement does violate the classical notion of relativistic causality, if the wave function is considered real. This is the content of Bell's theorem.

However, quantum entanglement does not violate signal causality, ie. it does not allow classical information to be sent faster than light.

 

[TMO]

Quantum entanglement does violate the classical notion of relativistic causality, if the wave function is considered real. This is the content of Bell's theorem.

However, quantum entanglement does not violate signal causality, ie. it does not allow classical information to be sent faster than light.

Thanks for informing me that quantum entanglement does violate the classical notion of relativistic causality. It's good that you included the bit about signal causality, but I already knew that, mostly because every layman article on QM includes it.

 

[atyy]

Thanks for informing me that quantum entanglement does violate the classical notion of relativistic causality. It's good that you included the bit about signal causality, but I already knew that, mostly because every layman article on QM includes it.

The alternative, if the wave function is not considered real and we explicitly exclude hidden variables, then classical relativistic causality may be considered "empty" or "meaningless" so that there is no formal violation.

 

[bhobba]

I observe apparatus A, which simultaneously causes apparatus B to do its thing..

There is your error right there. Observing one part of an entangled system does not cause the other part to do anything any-more than with Berlemann's socks seeing one sock causes the other sock to do anything:
https://cds.cern.ch/record/142461/files/198009299.pdf

From Bells theorem only if you want realism does observing one thing cause the other.

Thanks
Bill

 

[DrChinese]

Suppose I have an apparatus A that is entangled with apparatus B. In my reference frame, I observe apparatus A, which simultaneously causes apparatus B to do its thing. However, because there exists a reference frame where apparatus B does its thing before apparatus A, it follows that there exists a reference frame in which B does its thing before it was caused by A to do so.

There are experiments in which - conforming with your example - particles are entangled AFTER they cease to exist in all reference frames. So OK, there is no local causality. Once you accept that, how is the world any different? You still use the same ol' QM. :)

 

[Derek Potter]

The alternative, if the wave function is not considered real and we explicitly exclude hidden variables, then classical relativistic causality may be considered "empty" or "meaningless" so that there is no formal violation.

Why are physicists so willing to abandon the idea that the things they talk about are real? But, like hippies on acid, they do.

"Nothing is real, Man!"

Unfortunately (?) Bell Inequality violation is observed in real experiments. This drags the issue out of the rarified atmosphere of theory and into the solid world of Alain Aspect's laboratory in Orsay. Causality is violated in the real world unless we abandon direct realism. It's not about whether the wavefunction is real. Our naive belief that what we see is what is there must be modified - the world is incorrigibly wierd, that's for sure. Fortunately, QM provides both the justification and the toolkit to deal with the issue. Naive realism assumes that there are definite outcomes, QM assures us there are superpositions. The data records cannot be real, definite. All possible outcomes coexist. Superposition merely describes how this works. MWI, of course, says "What else is news?"

"Everything is real, Man!"

 

[bhobba]

Why are physicists so willing to abandon the idea that the things they talk about are real?

They aren't unless forced to.

Before going any further though you need to define precisely, and have everyone agree with your definition, what is real. Unless you can do that its a pretty meaningless comment.

My view of real is its what our physical theories describe. But getting everyone to agree on that is a lost cause.

Thanks
Bill

 

[Derek Potter]

of

There are experiments in which - conforming with your example - particles are entangled AFTER they cease to exist in all reference frames.

I'm not sure what that means. How can particles be entangled when there are no particles?
Have you got a simple example of what you mean?

 

[Eye_in_the_Sky]

... I observe apparatus A, which simultaneously causes apparatus B to do its thing.

TMO, I wonder if you would rather say it like this?

According to Quantum Mechanics, there are entanglement scenarios for which the following statement is false:

The 'physical state of affairs' relevant to outcomes at B is independent of the setting at A, and vice versa.

 

[Derek Potter]

They aren't unless forced to.

Nobody's forcing them. Yet they do it.

But getting everyone to agree on that is a lost cause.

Only because they've been panicked into playing at philosophy. It ends up as sophistry that obscures the fact that physics has discovered some interesting, weird even, facts about nature. It needn't be so. Rather than getting stirred up about what it means to be real, it is far more useful to ask what the real world is like.

My view: reality is a primitive, like truth or existence. It makes no sense at all to demand a definition of reality. What possible primitives could such a definition depend on? .

To return to the point, I was reminding atyy that BI violation is observed, it's not merely something in the maths of QM. Thus local causality is violated in the real world and the get-out clause saying that the wavefunction is not necessarily real does not apply. In a stretch QM might be totally wrong, something could turn up tomorrow forcing us to chuck the whole idea of vector states in the bin. BI violation would still occur and the implications for local, real, causality would still be valid.

 

[bhobba]

Nobody's forcing them. Yet they do it.

Its not people that forces them - its the evidence.

My view: reality is a primitive,

Get back to us when everyone agrees with you. That's when the discussion will be meaningful.

Thus local causality is violated in the real world

Well first you need to show locality applies to correlations which the cluster decomposition property excludes. The truth of it is its all semantic verbiage depending on freely chosen definitions.

Thanks
Bill

 

[phinds]

Rather than getting stirred up about what it means to be real, it is far more useful to ask what the real world is like.

So you are going to ask what the "real" world is like without worrying about what "real" means? How do you do that?

 

[Derek Potter]

So you are going to ask what the "real" world is like without worrying about what "real" means? How do you do that?

By experiment and models, same as in any other science.

 

[phinds]

By experiment and models, same as in any other science.

No, that is not how it is done in ANY science. You cannot do experiments or make models for something that is undefined.

 

[bhobba]

By experiment and models, same as in any other science.

That's hardly precise - but that aside again you have the issue of getting everyone's agreement for questions about it to be meaningful.

Really this is an issue in philosophy, not science, and we don't discuss philosophy here. If you want to discuss it there are other forums.

Thanks
Bill

 

[Derek Potter]

No, that is not how it is done in ANY science. You cannot do experiments or make models for something that is undefined.

And neither would I attempt to do so. Nobody investigates whether the real world is real.

 

[bhobba]

Nobody investigates whether the real world is real.

You do know what a tautology is don't you?

Thanks
Bill

 

[Derek Potter]

You do know what a tautology is don't you?

Yes, it's a tautology.

 

[phinds]

Yes, it's a tautology.

Then why are you wasting our time with it?

 

[Derek Potter]

Really this is an issue in philosophy, not science, and we don't discuss philosophy here.

Oh yes you do. I'm trying to wean you off it by saying that "real" has the same meaning in physics as it has in everyday life. Everybody knows what it means but nobody can define it. And there's a good reason for that - it isn't a property. A goblin that doesn't exist is not a particular type of goblin. If that is philosophy, so be it, we certainly have to use a homeopathic amount of philosophy to cure a pathological infection.

 

[Derek Potter]

Then why are you wasting our time with it?

Why are you wasting my time talking about scientific investigation of the meaning of "real"?

 

[aleazk]

"Philosophy: Beacon or Trap?" An extract from the book "Philosophy of Physics" by M.Bunge, my favorite epistemologist (also a physicist and hardcore realist). I strongly recommend this book to anyone interested in physics (yes, I say physics, rather than "anyone just interested in the philosophy of physics").

 

[atyy]

Well first you need to show locality applies to correlations which the cluster decomposition property excludes. The truth of it is its all semantic verbiage depending on freely chosen definitions.

Cluster decomposition does apply to correlations. Cluster decomposition is a property of the Hamiltonian, which applies to all states.

The notion of locality in cluster decomposition is that classical information cannot be transmitted faster than light, which quantum mechanics does respect.

 

[bhobba]

Cluster decomposition does apply to correlations.

Sorry - can't agree:
https://www.physicsforums.com/threads/cluster-decomposition-and-epr-correlations.409861/
'Clearly, these statements formulated as such are incompatible with EPR correlations, and are therefore wrong.'

Correlations must be precluded for it to make sense.

Thanks
Bill

 

[atyy]

Sorry - can't agree:
https://www.physicsforums.com/threads/cluster-decomposition-and-epr-correlations.409861/

Yes, I know that discussion. The pertinent part there is that Weinberg's informal statement of cluster decomposition is too sloppy (clearly violating quantum mechanics if correlations are included - if correlations are not included, then it is a tautology - there are no correlations when there are no correlations). A better informal definition of cluster decomposition would have been "no faster than light signalling". This applies to all states.

 

[bhobba]

Yes, I know that discussion. The pertinent part there is that Weinberg's informal statement of cluster decomposition is too sloppy (clearly violating quantum mechanics if correlations are included - if correlations are not included, then it is a tautology - there are no correlations when there are no correlations). A better informal definition of cluster decomposition would have been "no faster than light signalling". This applies to all states.

It is sloppy - but even if its stated clearly it doesn't apply to correlations. It is mentioned it may be possible to formulate it so it can. But that's not the issue. Precluding correlations leads to no issues with locality in QFT which is embodied in the cluster decomposition property.

Thanks
Bill

 

[Nugatory]

This thread has gone fairly far off the rails, time to close.