Delayed choice, pseudo-retrocausality, massive particles

[Quantum of Solace]

Double-slit experiments have been done to death, Obtaining and preserving which-path information (or otherwise) for a particle or photon causes its entangled partner to show the probability distribution with or without the interference pattern, violating Bell's inequality and thus proving nonlocality from our understanding of time and space, albeit without violating causality.

Kim-et-al's delayed choice experiment brought something new to the table, whereby preserving or obscuring which-path-information for photon B caused its entangled photon A to fall in a probability distribution which showed correlation in a similar fashion, even though photon A had *already had its position recorded*.

From our notion of time and space this seems to suggest not only nonlocality, but retrocausality. But an explanation that seems to make sense is that since the photon is traveling at c , thus exists as a line in spacetime instantaneously joining its creation and destruction. i.e. in the photon's frame of reference there is no time. (infinite time dilation).

There are a number of quantum eraser experiments for massive particles which show post-slit collapse of the wavefunction, but I'm curious to know if there have been any experiments similar to Kim-et-al, whereby one partner is recorded before the quantum erasure is decided, using entangled massive particles traveling at a mere fraction of c? If this too displayed seeming retrocausality it would be much more spooky as it then couldn't be explained by the infinite time dilation as in the photon experiments.

 

[bhobba]

From our notion of time and space this seems to suggest not only nonlocality, but retrocausality. But

It may sugrest that to you, but Bells theorem says otherwise:
http://www.drchinese.com/Bells_Theorem.htm

an explanation that seems to make sense is that since the photon is traveling at c , thus exists as a line in spacetime instantaneously joining its creation and destruction. i.e. in the photon's frame of reference there is no time. (infinite time dilation).

That's a common misconception, but false.

SR is a theory about transformations between inertial frames. There is no way to attach an inertial frame to an object traveling a c, since its never at rest, the statement it has no time is meaningless.

Thanks
Bill

 

[Quantum of Solace]

t may sugrest that to you, but Bells theorem says otherwise:
http://www.drchinese.com/Bells_Theorem.htm

Didn't I mention that in my first paragraph? As for retrocausality of course I'm referring to the illusion of such. Just as there is an illusion of causality with any entangled-pair measurement.

That's a common misconception, but false.

SR is a theory about transformations between inertial frames. There is no way to attach an inertial frame to an object traveling a c, since its never at rest, the statement it has no time is meaningless.

Thank you very much for your answer. If that's false then I'm suddenly wanting for any kind of rational description of how the interference pattern can be defined after one particle has recorded the result in our timeframe. If the answer is purely mathematical I don't suppose I'll ever be comfortable with that.

As for SR... It's very tempting to take your quote that time is meaningless and use that to say I'm essentially saying the same thing.

I appreciate very much that you took the time to answer my very naive questions. As I like to say "there are no stupid questions" ... but people don't always see it that way.

 

[bhobba]

If that's false then I'm suddenly wanting for any kind of rational description of how the interference pattern can be defined after one particle has recorded the result in our timeframe. If the answer is purely mathematical I don't suppose I'll ever be comfortable with that.

There are a number of misconceptions - but they are VERY common.

Mathematics simply describes what's happening - no answer is purely mathematical any more than an answer in English is purely English.

Secondly what is said about the double slit in beginning texts and popularization's using the wave particle duality is, charitably, too simplistic, and uncharitably wrong. Here is a much better explanation:
https://arxiv.org/ftp/quant-ph/papers/0703/0703126.pdf

Its simply this. Since just behind the slit we know it position then its momentum is unknown and can go in any direction. When two slits are open superposition comes into play and you get interference.

If you would like to discuss the double slit that paper is a good reference point.

As for SR... It's very tempting to take your quote that time is meaningless and use that to say I'm essentially saying the same thing.

I am not sure its the same as what I am saying. In physics time is what a clock attached to an object measures. You can never catch up to an object traveling at c so time is simply not a meaningful concept. Its not as you said 'in the photon's frame of reference there is no time' - its that since it has no frame of reference time is not a valid concept.

Thanks
Bill

 

[DrChinese]

There are a number of quantum eraser experiments for massive particles which show post-slit collapse of the wavefunction, but I'm curious to know if there have been any experiments similar to Kim-et-al, whereby one partner is recorded before the quantum erasure is decided, using entangled massive particles traveling at a mere fraction of c? If this too displayed seeming retrocausality it would be much more spooky as it then couldn't be explained by the infinite time dilation as in the photon experiments.

Here are a few experimental references, they don't involve massive particles though.

While bhobba (and most) reject retrocausality, there are certainly elements of QM that might cause you to scratch your head on that question. The space-time diagram of these experiments respects c but only in the sense that the diagram goes forward AND backward in time. You might be interested to learn: There is no quantum non-locality that does not follow this pattern (of c respected going both forward and backward in time). At least none I am aware of*. And as bhobba points out, this is not proof of anything.

http://arxiv.org/abs/quant-ph/0201134
You can entangle particles AFTER they are detected, see page 5 for a discussion.

http://arxiv.org/abs/1209.4191
You can entangle photons that never co-existed, and therefore never interacted.
"The role of the timing and order of quantum measurements is not just a fundamental question of quantum mechanics, but also a puzzling one. Any part of a quantum system that has finished evolving, can be measured immediately or saved for later, without affecting the final results, regardless of the continued evolution of the rest of the system. In addition, the non-locality of quantum mechanics, as manifested by entanglement, does not apply only to particles with spatial separation, but also with temporal separation. Here we demonstrate these principles by generating and fully characterizing an entangled pair of photons that never coexisted. Using entanglement swapping between two temporally separated photon pairs we entangle one photon from the first pair with another photon from the second pair. The first photon was detected even before the other was created. The observed quantum correlations manifest the non-locality of quantum mechanics in spacetime."


*A possible exception: It is possible to entangle particles from phase-locked separated sources, of course synchronization of those requires conventional signalling.

 

[bhobba]

While bhobba (and most) reject retrocausality,

That's true.

But it must be said its an opinion, an interpretation I don't hold to. We have interpretations where such is the case (look up transactional interpretation) so is a perfectly valid view.

No interpretation is better than any other, or worse for that matter. Choice of interpretation reveals more about the prejudices, psychology etc of the person concerned than actual science.

Thanks
Bill

 

[Strilanc]

]Kim-et-al's delayed choice experiment brought something new to the table, whereby preserving or obscuring which-path-information for photon B caused its entangled photon A to fall in a probability distribution which showed correlation in a similar fashion, even though photon A had *already had its position recorded*.

First thing's first, delayed erasure doesn't work the way you think it does. The interference pattern doesn't "come back", you measure the "choice" photon in a particular way then use it to group outcomes into buckets. Within each bucket there will be an interference pattern.

Because recovering the interference patterns requires the measurement outcome of the choice photon, there's no retro-causality required. The situation is more reasonably described as the earlier signal photon measurement outcome telling you how the choice photon's measurement result will be biased if you go with erasure.

Personally I find it all very obvious when simplified into a circuit diagram, though that's probably more to do with my familiarity with them than anything intrinsic:

curious to know if there have been any experiments similar to Kim-et-al, whereby one partner is recorded before the quantum erasure is decided, using entangled massive particles traveling at a mere fraction of c? If this too displayed seeming retrocausality it would be much more spooky as it then couldn't be explained by the infinite time dilation as in the photon experiments.

Such experiments are called Bell tests, and yes they have been done with the measurements being space-like separated.

But, as I mentioned above, this doesn't demonstrate retrocausality.

 

[DrChinese]

In case it wasn't clear from my post #5 above:

All entangled particle pairs (observed by Alice and Bob, who freely choose how to measure their respective particles) are connected as part of a larger experimental context, in which the the component elements have interacted in the past with one or more of the other components. The relationship (which is indirect) between Alice and Bob, however, need not be local (bounded by c). The resulting relationship (correlation) is an example of quantum nonlocality.

If you attempt to trace a "causal" connection for explaining how Alice's choice of setting apparently affects Bob's outcome, and/or vice versa: then that connection will trace a zigzag path forward and backward in time in a fashion that respects c. So that is how the "retrocausal" part shows up, if you want to call it that. That should not be taken as proof of anything, as the Bohmian interpretation - featuring explicit nonlocal action - accomplishes something similar with exclusively forward-in-time causality. And other interpretations have other mechanisms.

The Zigzag Limit of Entanglement
But the interesting thing is that the observable entangled effect, regardless of the interpretation, never extends farther in space and time than can be diagrammed by the zigzag between the context's components. So that's the head scratcher. Why? Is there a "retrocausal" mechanism at work? If there isn't, why is that zigzag a limitation on the extent of entanglement?

 

[jimgraber]

I would appreciate references for that Zigzag limit.
TIA.
Jim Graber

 

[DrChinese]

I would appreciate references for that Zigzag limit.
TIA.
Jim Graber

I don't have anything specific. Just something I noticed from looking at a number of configurations of entangled systems - usually photons so c is more evident. The references I provided are good examples - teleportation. But even regular parametric down conversion of photons has the same characteristic.

I do plan to put together a diagram as a visual though. I think that would help. On the other hand, if anyone has a counterexample... I would love to see that.

 

[Schneibster]

What we really want to know about is in the Delayed Choice Quantum Eraser of Kim and Scully et alii. Since I have the links I'll post them:
Kim et al. (2000): https://arxiv.org/abs/quant-ph/9903047
The original Scully/Druhl (1982) paper (unfortunately not open access): http://journals.aps.org/pra/abstract/10.1103/PhysRevA.25.2208

Now, I suggest some simple modifications to this experiment. That this happens is nice. It's time now to quantify it.

What we want to know is exactly how quickly modification of the interference pattern happens when the idler path in the experiment is shifted from all erasure to all detection and back. What we expect is that it will be instantaneous; but to be sure, we should measure it. @Strilanc's analysis above (very nice, thanks for that, it captures all the salient features of the DCQE) shows us that this is not simple; we won't be able do it in real time because the interference pattern is not obvious on inspection, because we must separate the detections at the idler eraser detectors A and B in order to see it (or the processing at the idler welcher weg detectors C and D in order to not see it) and correlate them in order to detect interference. However, this is simple enough to do, and the delay between the detection of signal and idler photons can be adjusted to any convenient value to allow enough time to do the pre-processing of the idler detections simply by sending the signal photons down a fiber optic cable to introduce a time delay.

Then we substitute mirrors with an actuator for the beam splitters to choose whether the idlers go to the erasure path or the welcher weg path, and retrodict the timing of appearance and disappearance of the interference pattern and compare it with the timing of the mirror flips. We can also shorten and lengthen both the signal and idler paths with fiber optic cables of the desired length. This will let us determine whether the speed of the "transmission of entanglement" really is infinite (which is what we expect) or whether there is a time delay (unexpected but interesting) or a reverse time delay (totally unexpected and very weird).

I know that Jack Cramer up at UW (the originator of the Transactional Interpretation) was agitating to do something like this and string the fiber optic cables down the LIGO tunnels, but I never heard anything more about it. I'm not sure he was thinking about the flip mirrors either. I'm surprised no one seems to have done it and further surprised Cramer wanted to use the LIGO tunnels; maybe he was thinking that having the fiber optic cables straight would reduce the loss.

On edit: also note that if the idler path is longer than the signal path, this gives the appearance that the interference in the signal could be detected before the determination of whether the corresponding idler photons went to the erasure or welcher weg path; however, @Strilanc's analysis shows us this is wrong, because we can never determine whether there is interference without correlating the arrivals of idler and signal photons, after correction for path length. This is explicit entanglement. It's why people think of retrocausality when they first see this experiment.

 

[jimgraber]

I did trace through the entangled but never simultaneously existing example. It does seem more or less obvious to me, but perhaps hard to prove. As far as I can remember, I have never seen anything published on this.
Thanks for the reply.,
Jim Graber

 

[Quantum of Solace]

Some great info here. Some I was aware of, but I had not seen the experiment at http://arxiv.org/abs/1209.4191 showing entanglement of non-coexisting particles. It does seem 'obvious' if you accept kim-et-al's result, but it's an even more striking example of quantum spookiness.

Its not as you said 'in the photon's frame of reference there is no time' - its that since it has no frame of reference time is not a valid concept.

Well to me this leads, intuitively (huge mistake, I know) to the concept that any imposition of correlation on the detection preceding the quantum erasure is perfectly valid, as time is meaningless. Sorry if I'm being stubborn here, but what you said doesn't obviously contradict what I was trying to say.

Anyway despite the wealth of information you have kindly provided, it would seem there have not yet been experiments whereby massive particles have shown similar entanglement over time-like separation (into the past... so to speak). I look forward to seeing them when they do become feasible.

 

[Strilanc]

Some great info here. Some I was aware of, but I had not seen the experiment at http://arxiv.org/abs/1209.4191 showing entanglement of non-coexisting particles.

That is such a strange experiment to me. Not because the result is strange, but because the framing acts like it's somehow surprising.

They entangle two qubits A and B, measure A in various ways, create another entangled pair of qubits C and D, teleport B into D via its entanglement with C, then... act surprised when D contains quantum information entangled with the measurement of A.

But of course D contains quantum information entangled with A, they literally moved that information there in step 4!

Quantum information is fungible. It doesn't go away just because you moved it into a different physical qubit. Even if the moving is done with teleportation. Even if the qubit is 'newer' than the quantum information. To expect otherwise is like being surprised that a new hard drive can store old photos.

 

[DrChinese]

That is such a strange experiment to me. Not because the result is strange, but because the framing acts like it's somehow surprising.

They entangle two qubits A and B, measure A in various ways, create another entangled pair of qubits C and D, teleport B into D via its entanglement with C, then... act surprised when D contains quantum information entangled with the measurement of A.

But of course D contains quantum information entangled with A, they literally moved that information there in step 4!.

Well I guess different people are surprised at different things. A and D no longer exist when the decision to entangle them is made. That's reasonably strange in my book.

Of course, nothing about that is surprising per the QM formalism itself.

 

[bhobba]

Well to me this leads, intuitively (huge mistake, I know) to the concept that any imposition of correlation on the detection preceding the quantum erasure is perfectly valid, as time is meaningless.

You are tying yourself into conceptual knots.

Whats going on in delayed choice is well known:
http://quantum.phys.cmu.edu/CQT/index.html

In particular see:
http://quantum.phys.cmu.edu/CQT/chaps/cqt20.pdf

It has nothing to do with time as a concept for photons. The very concept of time doesn't apply - which is NOT what you are saying. What you are saying is time doesn't exist for photons or something similar. There is a BIG logical difference. Since the concept doesn't even apply you can't even mention it in an answer - yet you want to do it. Its like if you want to discuss the smell of mathematical expressions - the very concept concept doesn't even apply so you can't talk about it.

Why hasn't the experiment been done with massive particles? It may have been - I don't really know. But it not a particularly enlightening thing to do because what's going on is well known and independent of mass.

Thanks
Bill

 

[Quantum of Solace]

Got it. (I think...) I was confused about the speed of quantum propagation being relevant to time. Also about quantum mechanics having consequences that were constrained by time. The wealth of studies here and the associated maths shows clearly that quantum mechanics transcends time. That's pretty damned weird, but I start to get used to the concept. Thanks again for the explanations and references.

edit: and please don't misunderstand me about causality... of course the correlation cannot be seen until information has been communicated. I did try to maintain that any suggestion of nonlocality or retrocausality was an illusion that forbade information transfer.

 

[bhobba]

The wealth of studies here and the associated maths shows clearly that quantum mechanics transcends time.

I don't even know what you mean by that.

But basically QM is simply a reasonable extension of probability theory:
http://www.scottaaronson.com/democritus/lec9.html

Thanks
Bill

 

[rootone]

Retrocausality?
If somebody unexpectedly contacts me to say that they intend to be arriving at my local bus station at 2pm tomorrow, so I change my plans so I can meet them.
Would that be it?

 

[harrylin]

[..]
They entangle two qubits A and B, measure A in various ways, create another entangled pair of qubits C and D, teleport B into D via its entanglement with C, then... act surprised when D contains quantum information entangled with the measurement of A.
But of course D contains quantum information entangled with A, they literally moved that information there in step 4!
[..] like being surprised that a new hard drive can store old photos.

But... doesn't Bell's theorem effectively claim on statistical grounds that there can be no such quantum information contained in entanglement (and thus also not be next transported) before the measurement?
What am I missing here?

 

[georgir]

But... doesn't Bell's theorem effectively claim on statistical grounds that there can be no such quantum information contained in entanglement (and thus also not be next transported) before the measurement?
What am I missing here?

Eh, I feel like I'm the one missing something about your post now, because I've seen you around long enough to know you know more than me about it :p But anyway...
Bell's claim is for classical, local variables. "Quantum information", if you define it as some variables common (and change-able) for the whole system obviously are non-local and are not covered. "Teleport B into D" just can't happen classically and locally.

 

[Quantum of Solace]

They entangle two qubits A and B, measure A in various ways, create another entangled pair of qubits C and D, teleport B into D via its entanglement with C, then... act surprised when D contains quantum information entangled with the measurement of A.

I thought it was at least as interesting as Kim et al's delayed choice experiment, as the measurement of qubit D is proven through violation of Bell's inequality to have acted nonlocally on qubit A despite that qubit A was born and died before qubit D existed. I suppose in QM it's a given but I though it an even more striking demonstration.

Why hasn't the experiment been done with massive particles? It may have been - I don't really know. But it not a particularly enlightening thing to do because what's going on is well known and independent of mass.

Again, it might be well known theoretically but there aren't many experiments that demonstrate it even with photons, so it would help verify the correctness of the theory if this nonlocality could act backwards in time for massive particles as well.

 

[bhobba]

Again, it might be well known theoretically but there aren't many experiments that demonstrate it even with photons, so it would help verify the correctness of the theory if this nonlocality could act backwards in time for massive particles as well.

I think you misunderstand why experiments are done. If we have an explanation that explains an experiment we do not change things in it for fun. There must be a reason. There is no reason to suppose massive objects make any difference.

Your musings about 'time' are simply incorrect, non locality acting backwards - well what you even mean by that I have no idea, are not valid reasons to do such experiments. That's not to say it may not be done, it even may have been done I simply do not know, but a valid reason why it would be interesting needs to be offered.

Thanks
Bill

 

[PeterDonis]

Thread closed for moderation.

Edit: the thread will remain closed