EPR type thought experiment with black hole

[kmdavisjr]

Assuming that I have an atom that produces two photons having opposite polarization, and I send one toward the event horizon of a black hole and the other is sent to a second observer who does not measure his photon's polarization until enough time has passed to insure that the other photon has passed through the event horizon, what happens to the photon that was sent into the black hole? Once the polarization is meausred, the superposition wave collapses, and the other photon, (inside the event horizon), should have the opposite polarization. This seems odd to me because it implies that information can reach us from a black hole because we know what it's polariziation should be by making the measurement of the photon outside the event horizon.

 

[humanino]

Welcome in PF kmdavisjr !

First, there might be reasons rendering impossible the coherence to remain before the photon even cross the horizon. The Hawking radiation can very well cause the ingoing photon to be "measured" before it can escape.

Then, if the coherence can be saved (which is unlikely IMHO), why would that be odd ? There is no actual measurement inside the horizon. The measurement takes place outside the horizon. You can never ensure, by any mean, that the coherence was still here, precisely because of the horizon.

 

[humanino]

It is a very nice question and I do not doubt other people will post their opinion soon.

 

[kmdavisjr]

Welcome in PF kmdavisjr !

First, there might be reasons rendering impossible the coherence to remain before the photon even cross the horizon. The Hawking radiation can very well cause the ingoing photon to be "measured" before it can escape.

Then, if the coherence can be saved (which is unlikely IMHO), why would that be odd ? There is no actual measurement inside the horizon. The measurement takes place outside the horizon. You can never ensure, by any mean, that the coherence was still here, precisely because of the horizon.

Thanks for the kind welcome! I did not have Hawking radiation in mind, but that would do just as well. My issue is looking at the non local property of QM at the most outlandish case to determine how far the theory goes. While I understand that we could never really "check" the properties of the particle inside the black hole, we can infer something about it by measuring the polarization on our particle and infering what happened inside the hole, itself by QM.

 

[humanino]

Well, if you say entanglement still apllies, how is that not metaphysics ? Nobody can tell it applies. If I tell you that the entanglement disappears while crossing the horizon with no other reason than disappointement, how could you prove me wrong ? If it is impossible to check whether QM still aplies or not, would you say this is still a scientific statement ?

Again : the question is really interesting, but I have to play the conter-argument to make a discussion.

 

[vanesch]

we can infer something about it by measuring the polarization on our particle and infering what happened inside the hole, itself by QM.

But this is exactly the way in which the non-locality of QM respects the locality of physical observations as required by SR. Locally, with your one single photon, there's nothing that you can observe that is different from the situation in which it didn't have an entangled brother in the hole (or whereever).
The non-locality in QM is a bit like in the following picture: imagine that 400 billion lightyears away (so way outside the visible universe) there is an exact copy of our current visible universe. There's a copy of you, sitting at your desk, drinking your cup of coffee, exactly as you are doing right now (assuming you are doing that :-). This doesn't affect you in the slightest way. You cannot find out. You cannot know. It is only if you are videotaped right now (and so your twin is videotaped over there) and our remote descentants, 400 billion years from now, discover that tape on a remote, cold planet around a brown dwarf, very very far away from here, that they are left with an open mouth: it is EXACTLY the same tape as the one that was found on what remained of the old Earth !

(ok I admit there are some astrophysical problems with my story, which I'm making up here).

So QM respects locality in that there is no action at a distance. There just is a magical correlation between what are locally random quantum effects.
It is as if the lottery over here and the lottery over there, each one in itself completely unpredictable, always give the same winning number.

In fact, the idea of an action at a distance comes from something that is often mis interpreted: statistical correlation does not point to a causal link (although usually it is always used in that way).
In QM, these correlations are exactly of that non-causal kind: locally you expect something random, and it IS random in exactly the way you expect. It is just that QM "used the same random number generator" to decide this randomness, and that's what is giving you these magical correlations.
I agree that it is very strange. In a way this strangeness has a beauty to it, doesn't it ?

cheers,
Patrick.

 

[kmdavisjr]

vanesch & Humino-

Thanks for the responses. Vanesch statement "In fact, the idea of an action at a distance comes from something that is often mis interpreted: statistical correlation does not point to a causal link (although usually it is always used in that way)." brought the point into focus. The fridge light came on.

I think that this forum is great, btw! Great, tight discussions!

 

[vanesch]

I wrote:

It is just that QM "used the same random number generator" to decide this randomness, and that's what is giving you these magical correlations. I agree that it is very strange.

and I realize that this can give rise to a misinterpretation, if taken too literally. If one thinks that I meant that due to the initial entanglement, both particles "carry physically with them the same random number" then I was misunderstood. Indeed, this is nothing else but a local hidden variable theory, which Bell's theorem shows cannot give rise to the correlations predicted by quantum mechanics (in certain cases).

It is just that if one considers that each quantum experiment, locally, is decided by the Great Blue Quantum Experiment Outcome Decider, then this GBQEOD, who is always everywhere where quantum experiments are performed, magically decided to correlate both (potentially very distant) outcomes. It are the GBQEOD's decisions which are clearly holistic and non-local, unless one considers that these results remain in a quantum superposition until they are brought together by the one and unique observer who can actually observe the correlations, because he sees the measurement results from both experiments.

This is one of the reasons I said in another thread that this non-locality problem disappears when one says that the only true measurement (collapse of the wavefunction) happens in the observer's consciousness.

cheers,
Patrick.

 

[nrqed]

Oops: I meant to reply to the original poster and I replied to Humanino by mistake. Sorry. Here's my reply:



Good question. Believe me or not, I have often wondered about that very issue. In my thought experiment, I was thinking of a large number of entangled pairs of photons. One observer would fall into the black hole with half of the photons, and one would remain outside with the other half. Then they would make their measurements (it does not really matter who measures "first" as seen form any frame.). The question is : if we would compare the results, would the photons still be entangled?

Of course, people will say "this is a meaningless question since the two lists of measurement cannot be compared". (and yet...what if no information is lost in black holes and it is eventually radiated away? Wouldn't the information be recuperable at some point?)


In any case, Istill can't help feeling that something is missing in QM (or in GR or in both ) . Well, people *do* talk about the fact that they are incompatible, but it's usually in the context of Planck scale physics. But no large energies (or ultra short distances or times) are involved in the above thought experiment.

I guess that if we accept the non locality of EPR type experiments, then there is no reason to believe that entangled photons would respect the causality structure of a black hole! So Maybe the correlation would remain. But if the question is truly untestable, it is not a valid question. And yet...

Regards

Pat

 

[Loren Booda]

From my http://www.quantumdream.net :

"Imagine an EPR experiment with initial singlet state located just outside the event horizon (r=R_S) of a Schwarzschild black hole singularity. A detector (at r>R_S), radially perpendicular to the horizon through this photon parent singlet, measures correlation between the infalling photonic events approaching the black hole singularity and their external directly observable correlates. The correlation observed indicates that the captured photon endows the "singularity" (in contrast to just being a mathematical point having only "No Hair" properties mass, charge and angular momentum) with the qualities of any variable compatible to our chosen measurement."

 

[NateTG]

I'm really a layperson regarding this topic. Please ignore the yellow stuff to the left. ;)

Of course, people will say "this is a meaningless question since the two lists of measurement cannot be compared". (and yet...what if no information is lost in black holes and it is eventually radiated away? Wouldn't the information be recuperable at some point?)

According to current theories, spin (or some other analog of angular momentum) is preserved in black holes, so the net spin state of the photons could, theoretically, be observed from outside the black hole by measuring the spin of the black hole.
Alternatively, with colider-generated micro-holes, it might be possible to 'inject' an entangled particle into a black hole, and then measure the net spin of the disintegration products, since small black holes have really short lives.
I'm not sure if a correlation with multiple particles could be measured.

In any case, Istill can't help feeling that something is missing in QM (or in GR or in both ) . Well, people *do* talk about the fact that they are incompatible, but it's usually in the context of Planck scale physics. But no large energies (or ultra short distances or times) are involved in the above thought experiment.

Really, it seems like the Quantum v. Relativity problems - at least as far as EPR is concerned - seem to have more to do with interpretation than with anything else. I haven't seen anything that indicates that an EPR-type arragement allows for FTL communication. Really you could look at it as a refinement of the notion of locality as much as anything else.

 

[humanino]

I just had a funny thought, I think it's worth posting it : imagine a device than can measure only spin up, and let spin down untouched. Imagine there is an ingoing flow of entangled photons in the BH, and many such detectors in the outgoing direction. Could I control the BH angular momentum with this apparatus !?

 

[NateTG]

I just had a funny thought, I think it's worth posting it : imagine a device than can measure only spin up, and let spin down untouched. Imagine there is an ingoing flow of entangled photons in the BH, and many such detectors in the outgoing direction. Could I control the BH angular momentum with this apparatus !?

I don't think I quite understand what you're saying. A spin measurement would indicate either spin up, or spin down (or spin right/spin left, or whatever). I also don't undestand what you mean by 'outgoing direction'.

It should be possible to only send 'spin up' particles into the black hole, but controlling a black hole's spin probably doesn't require anything so sophisticated, you could just use tangential lasers or particle beams.

 

[NateTG]

From my http://www.quantumdream.net :

"Imagine an EPR experiment with initial singlet state located just outside the event horizon (r=R_S) of a Schwarzschild black hole singularity. A detector (at r>R_S), radially perpendicular to the horizon through this photon parent singlet, measures correlation between the infalling photonic events approaching the black hole singularity and their external directly observable correlates. The correlation observed indicates that the captured photon endows the "singularity" (in contrast to just being a mathematical point having only "No Hair" properties mass, charge and angular momentum) with the qualities of any variable compatible to our chosen measurement."

I don't quite follow, I thought that EPR entanglement only pertained to a 'no hair' property - i.e. polarization or spin which correspond to angular momentum. I was also under the impression that black hole no-hair properties were precisely the same properties that a 'particle' could have, so that any particle (entangled or not) could pass all of its properties to the black hole.

 

[humanino]

either spin up, or spin down

Yes of course the "measure only spin up" does not make much sens :uhh:
It was pretty useless

 

[Loren Booda]

NateTG,

I don't quite follow, I thought that EPR entanglement only pertained to a 'no hair' property - i.e. polarization or spin which correspond to angular momentum. I was also under the impression that black hole no-hair properties were precisely the same properties that a 'particle' could have, so that any particle (entangled or not) could pass all of its properties to the black hole.

This is the first time I have read that particles through EPR at most transfer "No Hair" properties. Your criticism noted and enacted. Thanks!

 

[NateTG]

As I wrote, this isn't a strong topic for me, so you should double check that.

 

[vanesch]

Of course, people will say "this is a meaningless question since the two lists of measurement cannot be compared". (and yet...what if no information is lost in black holes and it is eventually radiated away? Wouldn't the information be recuperable at some point?)

This brings in the famous Hawking talk in Dublin. I have to warn that I'm very ignorant of all these issues ; I can only vaguely imagine what Hawking was referring at (a kind of PI over different topologies).
Nevertheless, I find it strange that one can consider *unitary evolution* over billions of years and be disappointed when black holes seem to spoil the party. Measurements (with the projection postulate) *also* give rise to information loss, no ? Aren't these unitary evolution propositions intimately related with a MWI interpretation ?

cheers,
Patrick.

 

[nrqed]

I'm really a layperson regarding this topic. Please ignore the yellow stuff to the left. ;)


According to current theories, spin (or some other analog of angular momentum) is preserved in black holes, so the net spin state of the photons could, theoretically, be observed from outside the black hole by measuring the spin of the black hole.

Very interesting point. Thanks for bringing it up.

So let's say we measure the spin of the BH after each photon is injected. Would we observe a correlation? I guess that most people would say "yes but that does not contradict anything because no "information" is conveyed beyond the event horizon".

I still feel unsatisfied with this but I seem to be the exception.

Really, it seems like the Quantum v. Relativity problems - at least as far as EPR is concerned - seem to have more to do with interpretation than with anything else. I haven't seen anything that indicates that an EPR-type arragement allows for FTL communication. Really you could look at it as a refinement of the notion of locality as much as anything else.

I understand. But the fact that a measurement "here" changes instantaneously the wavefunction over "there" is hard to accept for me. Especially because the notion of "instantaneously" is frame dependent. So in one frame it culd be the measurement of the spin of photon A which causes the collapse of the wavefunction whereas in another frame it's the measurement of the spin of photon B!

I understand that people say "well, we don't see the wavefunction itself so worrying about all this is a moot point". Still...

If we truly believe the standard point of view, we should always point it out whenwe introduce SR: "no timelike events can be correlated in any way *except* when measuring properties of entangled particles in QM in which case you are free to forget everything you learn about SR because an EPR type experiment can't be use by someone to tell someone else who won a baseball game..."

That's what bugs me, but that's of course just a personal opinion.

Pat

 

[humanino]

Signaling, Entanglement, and Quantum Evolution Beyond Cauchy Horizons

Consider a bipartite entangled system half of which falls through the event horizon of an evaporating black hole, while the other half remains coherently accessible to experiments in the exterior region. Beyond complete evaporation, the evolution of the quantum state past the Cauchy horizon cannot remain unitary, raising the questions: How can this evolution be described as a quantum map, and how is causality preserved? What are the possible effects of such nonstandard quantum evolution maps on the behavior of the entangled laboratory partner? More generally, the laws of quantum evolution under extreme conditions in remote regions (not just in evaporating black-hole interiors, but possibly near other naked singularities and regions of extreme spacetime structure) remain untested by observation, and might conceivably be non-unitary or even nonlinear, raising the same questions about the evolution of entangled states. The answers to these questions are subtle, and are linked in unexpected ways to the fundamental laws of quantum mechanics. We show that terrestrial experiments can be designed to probe and constrain exactly how the laws of quantum evolution might be altered, either by black-hole evaporation, or by other extreme processes in remote regions possibly governed by unknown physics.

Ulvi Yurtsever, George Hockney