'speed' of spooky action vs delayed choice

[Quantum of Solace]

I have seen many times mention of measuring the 'speed' of state-change over quantum entanglement, and potentially measuring it.

re: wave/particle 'duality' this paper explains everything extremely clearly and dispels such popular notions from my mind for a single particle:
https://arxiv.org/ftp/quant-ph/papers/0703/0703126.pdf
That is not my line of enquiry.

Here, an explicit attempt to measure the 'speed' of quantum weirdness in entangled particles:
http://arxiv.org/abs/1303.0614
"The lower bound is at least 10,000 times the speed of light"

Yet delayed choice experiments record the state or position of particle A before a decision is even made about which measurement or state to impose on particle B. Yet the Bell test still shows the inequality as being violated, proving that independent local variables cannot explain the correlation which becomes apparent upon comparison of the data from A and B. Quantum spookiness again, except not just spacelike separation but temporal separation.

http://arxiv.org/abs/quant-ph/0201134
http://arxiv.org/abs/1209.4191

Does this not prove that the 'speed' of quantum entanglement is >infinity i.e. independent of time?

The conclusion paragraphs of these studies seem to suggest so... I know there are dissenting opinions on this but what theories are given most credence?

 

[Paul Colby]

Does this not prove that the 'speed' of quantum entanglement is >infinity i.e. independent of time?

In this case I think this is likely. To have a speed, something must be moving. Can one define what that something is in this case?

 

[Quantum of Solace]

I don't think it can. I think the concept of 'speed' is irrelevant. I think in the quantum world time is irrelevant, and so-called nonlocality is irrelevant to us in our time frame because no information can be gleaned until the datasets are compared and sorted, which can only happen at <lightlike frames. "Causality" implies the transfer of information and this never happens. But something outside our frame of reference is clearly happening otherwise Bell's name would be sullied.

 

[Tollendal]

Dear Quantum of Solace,

Why couldn't we accept the reiteratelly verification that Universe isn't local? That simple act of humility would simplify so much the discussion!
Suppose I take a pair of gloves and put each of them in a box. I keep one with me and give the other to you, who then take a rocket to the Moon and there open it. Instantaneously you will know what glove remained with me - there is no "communication" between us, as the situation was deffined the very moment I closed the boxes. Well, I chose a very simple example, I concur, but that is cartesian: first simple, then the more complex. It seems to me the same procedure can be aplied to more complicated cases.

Einstein didn't understand it. The "ghostly action at distance" makes no sense at all!

 

[stevendaryl]

The speed of nonlocal quantum effects is undefined, as far as pure quantum mechanics is concerned.

The essence of quantum nonlocality is entanglement:

• You have a composite quantum system made up of two subsystems.
• The state of the composite system is entangled, in the sense that it fails to factor into a product of states of the component systems. In the simplest case, the composite state $|\Psi\rangle$ is the superposition of two factorizable states: $|\Psi\rangle = a |\psi_A\rangle |\phi_A\rangle + b |\psi_B\rangle |\phi_B \rangle$, where $|\psi_A\rangle$ and $|\psi_B\rangle$ are two possible states for the first subsystem, and $|\phi_A\rangle$ and $|\phi_B\rangle$ are two possible states for the second subsystem.
• You observe the first system to be in state $|\psi_A\rangle$
• You immediately know that the second system is in state $|\phi_A\rangle$

The question of timing amounts to asking: When did the second system make the transition to being definitely in state $|\phi_A\rangle$? But pure quantum mechanics tells us that there can be no physical content to such an answer. As far as anything you could possibly observe, it's as if the first system was always in state $|\psi_A\rangle$ and the second system was always in state $|\phi_A\rangle$, we just didn't know which until we measured it. But we know that that can't be the answer, because Bell's theorem shows that quantum probabilities can't be interpreted as due to ignorance of the quantum state (without nonlocal interactions).

 

[Khashishi]

It's silly to associate a speed with a correlation. You can correlate something with something in the past. What then? Is the speed negative?

 

[Quantum of Solace]

That's my point. The glove analogy doesn't work as Bell's theorem proves the hidden state (left or right?) doesn't exist until one of the gloves is observed. The most logical conclusion is that observing B affected the state of A, and if A was already recorded then yes, the speed is > infinite (time taken is negative)

What else? Maybe A's state was fixed and whatever happens to B is hitherto determined. This seems possible for any half-mirror situation but what if the experiment allows conscious choice by the scientist? Does the state of A determine that too?

 

[Quantum of Solace]

It's silly to associate a speed with a correlation. You can correlate something with something in the past. What then? Is the speed negative?

Bell's theorem says yes. I'm fine with that, formal QM allows it and causality is not violated. My post was asking... If one accepts the formalism and results of delayed choice experiments, why are people still asking about the "speed" of entangled interaction.

 

[Strilanc]

What they mean by 'the speed of entanglement was at least X' is that, if the correlations were caused by magical classical fairies running back and forth between the two experiments, the fairies would have to run at least that fast to replicate the results of the experiment.

This doesn't mean reality actually uses fairies, or less esoteric forms of signalling. It just means that if it's all being done with hidden signals then they have to go at least such and such speed. And since those speeds exceed light speed, Einstein gets quite cross with you if you start taking the signals idea too seriously.

I don't think it's silly to test these kinds of things. Quantum mechanics says that entanglement shouldn't stop working just because you'd have to ferry signals back and forth really really fast, so we go out and check that entanglement keeps working in the regime where hidden signals would have to go really really fast. You don't really expect to find anything, but you bothered to check and bothering to check is what science is all about.

 

[Paul Colby]

Bell's theorem says yes. I'm fine with that, formal QM allows it and causality is not violated. My post was asking... If one accepts the formalism and results of delayed choice experiments, why are people still asking about the "speed" of entangled interaction.

I think there is considerable terminology drift in such discussions. Calling it "entangled interaction" is misleading and wrong minded because there is no interaction. The dependence on Alice and Bob's measurement device setting is of the very same type of system-observer dependence inherent in every QM measurement local or non-local. Quantum systems don't have preexisting values the way classical systems do. A correlation is not an interaction. An observer dependent correlation is not an interaction.

 

[stevendaryl]

I think there is considerable terminology drift in such discussions. Calling it "entangled interaction" is misleading and wrong minded because there is no interaction. The dependence on Alice and Bob's measurement device setting is of the very same type of system-observer dependence inherent in every QM measurement local or non-local. Quantum systems don't have preexisting values the way classical systems do. A correlation is not an interaction. An observer dependent correlation is not an interaction.

Hmm. Well, classical correlations certainly don't imply interaction. That's because things can be correlated through their common past, and this correlation persists in the future. For example, Carol takes a pair of shoes, and puts one in a box for Alice, and another in a box for Bob, and then Alice and Bob open their respective boxes. Clearly, if Alice finds a left shoe, she knows instantly that Bob will find a right shoe. We know that this correlation doesn't imply any interaction between Alice and Bob, because we can reason that Alice's box must have contained a left shoe, all along, and Bob's box must have contained a right shoe, all along.

In the quantum case, however, the correlations cannot be explained in terms of "Bob's particle must have been spin-up, all along", so it's less clear that there is no interaction. What you definitely can say is that you can prove that there is no interaction that could possibly be used to communicate FTL between Alice and Bob.

 

[stevendaryl]

In the quantum case, however, the correlations cannot be explained in terms of "Bob's particle must have been spin-up, all along"

Well, I guess quantum correlations can be explained this way if we go to retrocausal or superdeterministic interpretations.

 

[Paul Colby]

Hmm. Well, classical correlations certainly don't imply interaction. That's because things can be correlated through their common past, and this correlation persists in the future. For example, Carol takes a pair of shoes, and puts one in a box for Alice, and another in a box for Bob, and then Alice and Bob open their respective boxes. Clearly, if Alice finds a left shoe, she knows instantly that Bob will find a right shoe. We know that this correlation doesn't imply any interaction between Alice and Bob, because we can reason that Alice's box must have contained a left shoe, all along, and Bob's box must have contained a right shoe, all along.

In the quantum case, however, the correlations cannot be explained in terms of "Bob's particle must have been spin-up, all along", so it's less clear that there is no interaction. What you definitely can say is that you can prove that there is no interaction that could possibly be used to communicate FTL between Alice and Bob.

This all follows from quantum systems not having preexisting values, ever. Measurement always "affects" the results. I guess I would require there being an energy associated with such an "interaction" There should be some transfer of energy evident in order to qualify as an interaction. Even the transfer of information is associated with some energy. So, I question the language used in this case.

 

[Paul Colby]

In the quantum case, however, the correlations cannot be explained in terms of "Bob's particle must have been spin-up, all along", so it's less clear that there is no interaction. What you definitely can say is that you can prove that there is no interaction that could possibly be used to communicate FTL between Alice and Bob.

Can't more than this be claimed? When taken in pairs, every particle pair measured by Alice and Bob are independent of every other pair measure by them. This allows one to construct (by editing a larger data set[1]) two measurement histories in which Alice's data set is identical between the two while Bob's is different because he selected different detection settings relative to Alice. One could view Bob's results as changing because his relative setting to Alice's is changing. Alice's two data sets are totally identical and thus unaffected by the relative setting to Bob.

[1] Editing data post mortem is a common technique used in delayed choice experiments so I'm well within my rights to employ it here.

 

[Quantum of Solace]

Bell proves that measurement of the state of B affects the state of A, no matter how separated in time or space. Call it 'interacts' or whatever you will... there is no transfer of energy or information so causality is not violated.

The experiment at
http://arxiv.org/abs/1303.0614
attemps to measure a 'speed' at which this effect propagates. I would agree this is an interesting thing to try to measure, were it not that:

My question was simply, isn't this ALREADY moot, given that measuring particle B has already shown, through violation of Bell's inequality, to have affected the state of particle A, when particle A has already been recorded, and in the case of the second experiment I quoted, even when particle B's existence has no overlap with particle A.

The only conclusion I can draw from that is that so-called 'speed' in the delayed choice experiments has been shown to be >infinite (if one were to try to imagine that time, and thus a speed, were relevant measurements.)

Formal QM has no consideration for time in this wavefunction 'collapse', as expertly explained by stephendaryl. I take from that, that time is not relevant, thus neither is 'speed' a valid parameter.

Knowing the results of the delayed-choice experiments, and knowing that QM has no consideration of time, I then went back and wondered why they did the measuring experiment at all.

Thanks for all your answers and replies.

 

[DrChinese]

Bell proves that measurement of the state of B affects the state of A, no matter how separated in time or space. Call it 'interacts' or whatever you will... there is no transfer of energy or information so causality is not violated.

The experiment at
http://arxiv.org/abs/1303.0614
attemps to measure a 'speed' at which this effect propagates. I would agree this is an interesting thing to try to measure, were it not that:

My question was simply, isn't this ALREADY moot, given that measuring particle B has already shown, through violation of Bell's inequality, to have affected the state of particle A, when particle A has already been recorded, and in the case of the second experiment I quoted, even when particle B's existence has no overlap with particle A.

...

Knowing the results of the delayed-choice experiments, and knowing that QM has no consideration of time, I then went back and wondered why they did the measuring experiment at all.

Doing the experiment - especially for something like this - is important in confirming the predictions of QM. It certainly lays to rest the idea that there is some obvious finite speed.

I often cite the experiment you mention precisely because it drives home the point that the relationship between Alice and Bob can be timelike separated just as it can be spacelike separated. We don't really have good language to express this outside of the math, so it is often lost in descriptions. An entangled quantum system has a spacelike and timelike extent that defies a simple verbal description.

 

[BillTre]

How about this as an alternative explanation that could be shift the problem:
An initial superposition of the two possible states:
1) involves both boxes (or whatever), one is A or B while the other is correspondingly B or A
2) the superposition breaks down when some interaction forces one state or the other
3) the universe bifurcates into a pair of new derivative multiverses, each with a different resultant state
4) coordination or speed of communication between the two separated parts is not the issue, but the universe bifurcation might be. It all should change faster than the speed of light indicating it is something beyond relativity.

 

[bahamagreen]

The language does seem to be missing some analogy. There is a similar question sometimes about the "speed of gravity", and a possible analogy that might help in the entanglement question?

If there were no gravity, but all objects grew in diameter at the same rate, the apparent movements among the objects would look like gravitational attraction. Objects' motions very far apart would still appear as if gravity was instantaneous because it would not be the result of a mediation but a universal attribute (growing size) that underlies the relative accelerations appearing without any mediation.

Somehow, this universal attribute seems unlike the usual "hidden variable"... is there a similar analogy attribute that helps picture the entanglement correlations without mediation?