GHZ Experiment: Detections vs. Hidden Variables

[trosten]

Have anyone done the GHZ experiment? As I understand it from a book I read QM predicts detections that hidden variables theories can't produce at all! If theese detections has been done then all hidden variables should be out of question even if the Bell experiments haven't been flawless!?

 

[Caroline Thompson]

Have anyone done the GHZ experiment? As I understand it from a book I read QM predicts detections that hidden variables theories can't produce at all! If theese detections has been done then all hidden variables should be out of question even if the Bell experiments haven't been flawless!?

Yes, there had been at least one experimental test:
Jian-Wel Pan, Dik Boumeester, Matthew Daniell, Harald Weinfurter and Anton Zeilinger, "Experimental test of quantum nonlocality in three-photon Greenberger-Horne-Zeilinger entanglement", Nature 403, 515-518 (2000).

This did support QM, but I think there must be something terribly wrong with the supposed "local realist" theory that they assume. I mean to look into it. It would help if I could get hold of something nearer to the raw data and not just the published "normalised" results. Detection rates were very low, so a variation of the standard "detection loophole" would have applied, but the published results show such severe contrast to the supposed local realist predictions that I think there must be more to it than this. Perhaps the LR prediction is one that applies to the geometrically symmetrical case covered in:

Daniel M. Greenberger, Michael A. Horne, Abner Shimony, Anton Zeilinger, Bell's theorem without inequalities, Am. J. Phys. 58 (12), 1131 (1990)​

The actual experiment may have had completely different symmetry properties. As I said, though, I intend to look into it. The theory is horrendous (much worse than that for the Bell inequalities) so I shall have to build up my courage to tackle it!

Caroline
http://freespace.virgin.net/ch.thompson1/

 

[trosten]

keep it up!

 

[Hurkyl]

Y'all make it sound as if it's a foregone conclusion that the experiment will be consistent with LR.

 

[trosten]

I don't believe that, I believe QM will prove to be complete. But I think caroline has got a point that a lot of the experiments to disprove LR with bells theorem have been flaved.

I spoke to a teacher at my school and he said that recently other researches have shown that there haven't been any experiment that have dissproven LR so far since all experiments have contained missed photons etc. All though the research show that its highly unlikely that LR is correct, it still isn't proven it isnt.

 

[Caroline Thompson]

I don't believe that, I believe QM will prove to be complete. But I think caroline has got a point that a lot of the experiments to disprove LR with bells theorem have been flaved.

I spoke to a teacher at my school and he said that recently other researches have shown that there haven't been any experiment that have dissproven LR so far since all experiments have contained missed photons etc. All though the research show that its highly unlikely that LR is correct, it still isn't proven it isnt.

Re ordinary Bell tests, I think that some day a sufficiently careful experiment will show that the QM predictions can give wrong numerical predictions.

Re the GHZ test, though, something seems to have gone wildly wrong. There is a fairly thorough coverage of the history of the test in Amir D Aczel, “Entanglement: The greatest mystery in physics”, Four Walls Eight Windows, New York, 2001. Though Aczel, despite many one-to-one discussions with some of the people involved, seems to have come away with a totally false impression of the validity of the Bell test experiments, he is the best source I know for the history from Bell's paper to about 1998. Apparently the GHZ test was discussed at conferences and privately then given publicity by Mermin before the official paper had been published. Something odd was going on here ...

Caroline
http://freespace.virgin.net/ch.thompson1/

 

[Hans de Vries]

Re ordinary Bell tests, I think that some day a sufficiently careful experiment will show that the QM predictions can give wrong numerical predictions.

Re the GHZ test, though, something seems to have gone wildly wrong. There is a fairly thorough coverage of the history of the test in Amir D Aczel, “Entanglement: The greatest mystery in physics”, Four Walls Eight Windows, New York, 2001. Though Aczel, despite many one-to-one discussions with some of the people involved, seems to have come away with a totally false impression of the validity of the Bell test experiments, he is the best source I know for the history from Bell's paper to about 1998. Apparently the GHZ test was discussed at conferences and privately then given publicity by Mermin before the official paper had been published. Something odd was going on here ...

Caroline
http://freespace.virgin.net/ch.thompson1/

"Violations of Local Realism in the Innsbruck GHZ experiment"
http://arxiv.org/PS_cache/quant-ph/pdf/9811/9811013.pdf

Not really straightforward conclusive... Some of the quotes:

Quote 1)
"Some observed events do not follow
the usual pattern of the GHZ correlations. The aim of
this paper is to show, following the ideas of ref [4], that
those ‘wrong’ events are irrelevant in the derivation of
a GHZ-type contradiction for the quantum predictions"

Quote 2)
"It will be shown that in any local realistic theory those
wrong events must happen (or not) irrespective of what
observables are chosen to be measured by the remote
observers at the three spatially separated stations"

Quote 3)
"Thus, the GHZ argumentation
can be confined to only the “right” events."


Regards, Hans

 

[uknwe]

My understanding of the GHZ experiment is that it does not rule out situations where the underlying hidden variable is different between distinct outcomes.

This seems a rather weak interpretation of how a local realistic hidden variable might work i.e that the hidden variable is tied to some immediate property of the object under measurement as opposed to something more holistic.

Bohm's work on hidden variables although not dealing with GHZ explicitly (as far as i know) points n this direction