Can absence of hidden variables save locality?

[Fra]

To the thread description I see no reason to abandon locality. Everything I've seen where people that claim that QM is non-local are confusing correlations with causality.

In my book, correlations between two remote things have nothing to do with non-locality.

Non-locality is if you have remote information influence a local decision (beyond coincidences that is!) or where you have FTL communication.

What one can do is discuss how causality is inferred, from correlations. Then one will see that it's not possible to infer a confident causation, ALL there is are correlations, that leads to various degrees of confidence in EXPECTED causation. This EXPECTED causation is all there is IMO. There need not be a "realist style" causation.

So I think all the way through this analysis do various levels of realism sneak in, that sometimes distort the reasoning.

/Fredrik

 

[Varon]

To the thread description I see no reason to abandon locality. Everything I've seen where people that claim that QM is non-local are confusing correlations with causality.

In my book, correlations between two remote things have nothing to do with non-locality.

Non-locality is if you have remote information influence a local decision (beyond coincidences that is!) or where you have FTL communication.

What one can do is discuss how causality is inferred, from correlations. Then one will see that it's not possible to infer a confident causation, ALL there is are correlations, that leads to various degrees of confidence in EXPECTED causation. This EXPECTED causation is all there is IMO. There need not be a "realist style" causation.

So I think all the way through this analysis do various levels of realism sneak in, that sometimes distort the reasoning.

/Fredrik

You have a book? What's the title?

Do you think a new field (higgs like) can function like a preferred frame? What are possible candidate of bonafide preferred frame? If an instantaneous signal can use it to commuinicate between say between 100 billion light years distance. Causality is not affected, isn't it.

 

[Fra]

No I don't haven't published any books, I meant "in my book" as in the idiom.
http://www.usingenglish.com/reference/idioms/in+my+book.html

in my book = in my opinion

/Fredrik

 

[Varon]

No I don't haven't published any books, I meant "in my book" as in the idiom.
http://www.usingenglish.com/reference/idioms/in+my+book.html

in my book = in my opinion

/Fredrik

Ok. English is not my native language so I don't get certain colloquial or slang.

 

[Fra]

English isn't my native language either. I think your english appears very good to me though.

I see no need for a preferred frame of any kind. All there is are IMO interacting observerframes. The conventional view is that in such cases there are objective transformation rules that defines the observer invariants. But I object to that since this inference is done by an external observer.

/Fredrik

 

[zonde]

Well, some physicists say that quantum field theory is "local" in the specific sense that the field operators at each point (which I think give the probabilities or expectation values for the outcome of some measurement at that point) evolve in a way that's described by local Lorentz-invariant equations--see the top of p. 3 here for example. And there is also the fact that it's been proven impossible to use entanglement in quantum field theory to communicate FTL.

If field operators evolve in Lorentz-invariant fashion then I think it's the same locality and I don't see where it makes the difference.
After all it's simultaneous clicks in two detectors and as long as coincidence tend toward 100% matching as detectors tend toward 100% efficiency it does not affect the reasoning whether reason for clicks are beables or fairly localizable configuration of field.

Why "retro causality"? The "toy model" I suggested in [post=1557143]this post[/post] could be simulated on a regular pair of separated computers with no need for any signals from the future...

Well, with your model it's not "retro causality" but FTL interaction nonethless.
When you pair up different worlds you determine the past that should be already determined. From perspective of single world it is no different then FTL interaction.

 

[Demystifier]

I would be interested in how you got L -> H out of their reasoning. I think you will see H is an assumption, not a deduction.

Well, to explain it, the crucial thing is to give a more careful DEFINITION of H appropriate for the context of the EPR paper. In that context H (hidden variable) is any additional quantity NOT given by the wave function.

The rest is easy. Let us assume QM and L. Consider the state
|psi> = |up>|down> + |up>|down>
before the measurement. Let the measurement reveal that the left particle is in the state |up>. Now from QM we know with certainty that the right particle is in the state |down>.

Now consider an additional assumption that there is no H. This means that the new knowledge must also be described by a wave function, because there is nothing else at disposal (due to the no-H assumption). That means that now the total state must be
|psi'> = |up>|down>
However, the transition from |psi> to |psi'> is not local. On the other hand, L was an assumption. In other words, the additional no-H assumption led to a contradiction. Therefore, as QM and L are taken as assumptions, the no-H cannot be true. It must be false. Therefore, H must be true. Q.E.D.

Note that this is not exactly the same reasoning as in the EPR paper, but is somewhat simplified in order to extract what I need.

If your objection is: "Yes, but such a definition of H is not the one appropriate for the Bell theorem", then you are absolutely right. And that's exactly why the reasoning in the first post is not correct. You cannot consistently combine two different definitions of H and pretend that they are the same.

 

[Demystifier]

Ah, (2) and (4) can both be true if L is false.

Of course .. duh.

Yes.