Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

A Another loophole-free test of Bell's theorem

  1. Nov 11, 2015 #1

    DrChinese

    User Avatar
    Science Advisor
    Gold Member

    This just showed up from a team led by Zeilinger, for those interested in loophole-free Bell tests:

    http://arxiv.org/abs/1511.03190

    A significant-loophole-free test of Bell's theorem with entangled photons

    Marissa Giustina, Marijn A. M. Versteegh, Soeren Wengerowsky, Johannes Handsteiner, Armin Hochrainer, Kevin Phelan, Fabian Steinlechner, Johannes Kofler, Jan-Ake Larsson, Carlos Abellan, Waldimar Amaya, Valerio Pruneri, Morgan W. Mitchell, Joern Beyer, Thomas Gerrits, Adriana E. Lita, Lynden K. Shalm, Sae Woo Nam, Thomas Scheidl, Rupert Ursin, Bernhard Wittmann, Anton Zeilinger

    (Submitted on 10 Nov 2015)

    "Local realism is the worldview in which physical properties of objects exist independently of measurement and where physical influences cannot travel faster than the speed of light. Bell's theorem states that this worldview is incompatible with the predictions of quantum mechanics, as is expressed in Bell's inequalities. Previous experiments convincingly supported the quantum predictions. Yet, every experiment performed to date required assumptions that provide loopholes for a local realist explanation. Here we report a Bell test that closes the most significant of these loopholes simultaneously. Using a well-optimized source of entangled photons, rapid setting generation, and highly efficient superconducting detectors, we observe a violation of a Bell inequality with high statistical significance."
     
  2. jcsd
  3. Nov 11, 2015 #2

    DrChinese

    User Avatar
    Science Advisor
    Gold Member

    And apparently today was as well another team's choice to post their results:

    http://arxiv.org/abs/1511.03189

    A strong loophole-free test of local realism

    Lynden K. Shalm, Evan Meyer-Scott, Bradley G. Christensen, Peter Bierhorst, Michael A. Wayne, Martin J. Stevens, Thomas Gerrits, Scott Glancy, Deny R. Hamel, Michael S. Allman, Kevin J. Coakley, Shellee D. Dyer, Carson Hodge, Adriana E. Lita, Varun B. Verma, Camilla Lambrocco, Edward Tortorici, Alan L. Migdall, Yanbao Zhang, Daniel R. Kumor, William H. Farr, Francesco Marsili, Matthew D. Shaw, Jeffrey A. Stern, Carlos Abellán, Waldimar Amaya, Valerio Pruneri, Thomas Jennewein, Morgan W. Mitchell, Paul G. Kwiat, Joshua C. Bienfang, Richard P. Mirin, Emanuel Knill, Sae Woo Nam

    (Submitted on 10 Nov 2015)

    "We present a loophole-free violation of local realism using entangled photon pairs. We ensure that all relevant events in our Bell test are spacelike separated by placing the parties far enough apart and by using fast random number generators and high-speed polarization measurements. A high-quality polarization-entangled source of photons, combined with high-efficiency, low-noise, single-photon detectors, allows us to make measurements without requiring any fair-sampling assumptions. Using a hypothesis test, we compute p-values as small as 5.9×10−9 for our Bell violation while maintaining the spacelike separation of our events. We estimate the degree to which a local realistic system could predict our measurement choices. Accounting for this predictability, our smallest adjusted p-value is 2.3×10−7. We therefore reject the hypothesis that local realism governs our experiment."
     
  4. Nov 11, 2015 #3

    DrChinese

    User Avatar
    Science Advisor
    Gold Member

    There is some overlap in the teams for the above papers. We have already had some threads talking about similar results published by another team, which also features some cross-members.

    http://arxiv.org/abs/1508.05949
     
  5. Nov 11, 2015 #4

    ZapperZ

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2016 Award

    So the caveat here is the definition of "loophole free". It appears that they are using the criteria of minimal set of assumption set out by Larsson (Ref. 16) to define what is meant by "loophole free". Is this stronger than the most recent claim made by Hensen et al.? It doesn't appear to be that way to me.

    Zz.
     
  6. Nov 11, 2015 #5

    DrChinese

    User Avatar
    Science Advisor
    Gold Member

    My post 3 is a reference to the Hensen et al paper, which was the first of the 3. Morgan W. Mitchell was a member of all 3 teams. Larsson (who you mentioned) is a member of the Zeilinger team.

    So I think they are all using similar criteria. The main loopholes addressed are the locality, fair sampling and freedom of choice loopholes. (Freedom of choice meaning that the measurement settings are selected independently at times and in a manner at which they cannot affect the outcomes in a local realistic world.)
     
  7. Nov 11, 2015 #6

    Mentz114

    User Avatar
    Gold Member

    They find the probability of getting their value for the J statistic from local realism is of the order 10-31. Effectively zero.

    Amazing.

    It's good this so well understood otherwise a lot of heads would be getting scratched (irony).
     
  8. Nov 11, 2015 #7
    All of the experiments linked to in this thread simultaneously close the detection loophole, coincidence loophole, and the locality loophole. The remaining "loophole", superdeterminism (or non-freedom of choice), can't be closed by experiment anyway. It's a purely philosophical objection to Bell's theorem.

    All in all, it's quite a year for Bell experiments.
     
  9. Nov 11, 2015 #8
    Looks like all of these three groups explicitly assume that measurement outcomes are fixed at some point. This assumption seems to be in contradiction with unitary evolution of quantum theory. As far as I understand, if unitary evolution is correct, measurement outcome, strictly speaking, can never be final. Does that mean they eliminate local realism assuming quantum theory is wrong?
     
  10. Nov 11, 2015 #9
    Interestingly, these two new claims (apparently choreographed to appear the same day) came out a day before the fatal signalling loophole was identified in the raw data of the most recent "loophole free" test from August 2015. The new claims haven't provided the raw data yet, hence it would be wise to hold off on the champagne.
     
  11. Nov 12, 2015 #10

    DrChinese

    User Avatar
    Science Advisor
    Gold Member

    I would not call this analysis convincing enough to be fatal. I am not sure it means anything - I suspect a firm rebuttal from the authors is likely quite soon. Unfortunately, the Hensen paper has a relatively small data set as it takes a longer time to accumulate each heralded event.

    But true enough, all 3 of these experiments are new and therefore subject to peer scrutiny.
     
  12. Nov 12, 2015 #11
    There is no "signaling loophole" in Bell tests. If someone tries to explain the results by invoking FTL signaling, then obviously the underlying model is not local, so local realistic models have anyway been ruled out. In this particular paper the author notes that the randomized settings of the angles should be independent of whether the entanglement swap was successful or not (aka heralded; these events are recorded with space-like separation), so the number of successful swaps should be approximately equal for all of the four settings combinations. In the Hensen et.al. experiment, they are rather unequal. I did a chi-square test of this with a null hypothesis of a uniform distribution, and got a p-value of 0.046. But no one in their right mind would reject the correctness of special relativity based on a p-value of 4.6 % in a single experiment.
     
    Last edited: Nov 12, 2015
  13. Nov 12, 2015 #12

    Mentz114

    User Avatar
    Gold Member

    Seems logical.

    Surely this must be true unconditionally, as a sign that the experimental setup is behaving as expected.

    That is a big deviation from what one would expect. Either we have a rare-event run ( odds about 20:1 against) or the setup is flawed. It must reduce confidence in the hypothesis that the inequality is violated ( also about 20:1 with the 'uncorrected' data).

    More data is needed. A lot more.
     
  14. Nov 12, 2015 #13
    Yes. But given what we know about the experimental setup, it is hard to come up with a plausible flaw that could generate this kind of dependence. Which makes the "rare-event run" hypothesis that more edible.

    (The author of the paper mentioned earlier in the thread speculates about FTL signaling as an explanation. This is of course even less likely than all other alternatives, since in that case, the anomaly could actually be used to facilitate FTL communication).
     
    Last edited: Nov 12, 2015
  15. Nov 14, 2015 #14
    Don't overlook Kaiser et al's cosmic photons proposal. I believe Zeilinger (among others) has also been thinking along those lines. So anyway are we entering an era when metaphysical issues will be settled by experiment? It's already hard enough for philosophy postdocs to find careers.
     
  16. Nov 14, 2015 #15

    zonde

    User Avatar
    Gold Member

    I like how seriously these experimenters approached the question of RNG. They XORed outputs from two different QRNG and pseudorandom source (!).
    For pseudorandom number source they used different data ... including movie Back to the Future :smile:
     
  17. Nov 16, 2015 #16

    f95toli

    User Avatar
    Science Advisor
    Gold Member

    Hensen has always been very careful about stating exactly what they measure and what they can/can not rule out. He gave a talk at a conference I attended recently (Quantum Information Processing and Communication in Leeds, UK) and he spent about a 1/3 of his talk talking about potential loopholes. However, most of these are philosophical and/or would violate SR; as far as I can remember there were no "serious" loopholes in their experiment.
     
  18. Nov 16, 2015 #17

    gill1109

    User Avatar
    Gold Member

    That's right. They all assume that measurement outcomes are fixed at some quite definite point. They also assume that randomly determined measurement settings were also fixed at a slightly earlier definite point.

    If you believe in the many world theory then (it seems to me) you do not believe in any ordinary reality at all. There are no "measurement outcomes". Not ever. They are somehow illusory. Many world theorists do not have any problem with non-locality because there is no classical world where measurements actually have outcomes.

    In fact the pages of the journals where the results are published do not exist either because they remain forever in quantum superposition with the pages which would have been published if the measurement outcomes had been different.
     
  19. Nov 16, 2015 #18
    As one can see from my posts here or from my articles, I have nothing to do with many worlds. And I believe there is reality out there (at least I don't think there is a valid reason to doubt it:-) ). Ordinary reality? I guess it depends on how you define "ordinary":-)

    Obviously, you are sarcastic here:-) Obviously, Schroedinger was sarcastic about the Shroedinger's cat:-) However, I listened to lectures of Haroche and Wineland once (It was at APS March meeting last year). Both of them mentioned the Schroedinger's cat quite seriously, and one of them said (it was Haroche, if I am not mistaken): "There is no fundamental decoherence." So which of the following do you dispute?

    1). Unitary evolution, strictly speaking, is not compatible with definite outcomes of measurements.
    2). Unitary evolution is always correct (meaning it will never be in disagreement with experimental results).

    If you dispute 1), I'd like to hear your arguments. If you dispute 2), then, strictly speaking, you don't think quantum theory is correct. In this case, the assumptions of the Bell's theorem are not correct, so what's there to discuss?

    So you may ask me, if measurement outcomes are never final, do I discard all experimental physics? No, I don't. Do I contradict myself then? I am not sure. For example, I very much like the approach of http://arxiv.org/abs/1107.2138 (published in Phys. Rep.). I interpret their results as follows (maybe they will disagree though:-)): while measurement outcomes are never final, the time required for the outcomes to change can be comparable to the Poincare's recurrence time. But I am not sure if the recurrence time is that big in the experiments that we discuss. Anyway, I believe the experimentalists' assumption of fixed outcomes is wrong if quantum theory is right. Then what did they demonstrate? That local realism is wrong assuming quantum theory is wrong?
     
  20. Nov 16, 2015 #19
    Mind you, I mean no disrespect. The experimentalists have pushed the envelope, so now we know more than we knew last year. However, we should be careful with the conclusions.
     
  21. Nov 17, 2015 #20
    They do believe in the objective reality of the universal wavefunction, don't they?

    Yes, there are measurement outcomes. Each measurement is supposed to "split" the universe, and the outcome of the measurement correpsonds to one of the succeeding universes that result from the split.

    Can you please elaborate a bit more on this?

    As far as I can see, non-locality does not seem any more natural with repsect to the MWI than it does with respect to the Copenhagen interpretation.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Another loophole-free test of Bell's theorem
Loading...