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

About violation of Bell's inequalities

  1. Sep 16, 2006 #1
    About violation of Bell ineqalities (and the Loch Ness monster)
    Assuming no loophole experiments can be done... even then I am not sure of what this would mean beyond the fact that QM's prediction on correlation work fine (and where is the surprise there?).

    I know that assuming some statistical properties of yet to be proven valid Hidden Variables (HVs) theories, one could replace what cannot be measured by what is measured (no measure of hidden variable can be done, and not much is gained when replacing HVs by realism as Wigner did). But let's be serious: HVs are probably as likely to be relevant as the Loch Ness monster (same thing for naive realism, i.e. realism that does not tae account of the Uncertainty Principle).
    Assume that a theory shows that if you put a goat and a goat that has seen the Loch Ness monster on some small boat, that boat will immediately sink. Prosper Youyou shows experimentally that a goat weights less than half the weight that can be carried by the small boat. Then Prosper claims that he has experimentally disproved the Loch Nees and goat theory. Is Prosper entitled to that claim, or has Propser only proved a well-established fact about goats?
    Bell's historical paper starts with a strong opinion on the EPR paper that contradicts Einstein's own expressed views: that QM is non-local. This paper of Bell paper then shows that any (naive) HV theory where HVs have the same statistics that QM would also need to be non-local. The strong claim is the one about QM since no one has ever validated HVs.
    The proofs about HVs can ONLY be theoretical as long as no one has experimentally shown HV theories to make sense (of course, that experimental proof could be simultaneous with the proof that Bell's Inequalities are violated). But till otherwise proved, Bell's Inequalities are about very hypothetical entities and no experimental proof of their range of validity makes sense.
  2. jcsd
  3. Sep 20, 2006 #2


    User Avatar
    Science Advisor
    Gold Member

    I think I see your point. For a local HV theory to be tenable (i.e. taken seriously), one of the following is necessary:

    a) There should be some evidence that a Hidden Variable is present, such as a violation of the Uncertainty Principle.
    b) The theory should make a claim that the QM correlation statistics (cos^2 relationship) are wrong; this would be very surprising indeed.

    So either way, some fundamental element of QM - which has already been experimentally verified - would need to be contradicted first.
  4. Sep 21, 2006 #3
    A LHV theory is required because QM only gives statistical answers. Such a theory doesn't need to contradict QM in any way, on the contrary, it should give the same statistical results (for example, it should be able to predict, or at least postdict, the timing of a radioactive decay but it doesn't need to predict a different decay probability).

    Bell's Inequalities are about a null set of LHV theories, the set of logically impossible ones. The assumption of free will, required for Bell's proof to work, is incompatible with the assumption that the universe is deterministic. It is a circular argument that can be rejected on logical grounds, without any appeal to QM or Aspect's experiments.
  5. Sep 22, 2006 #4
    Bell's Inequalities are about a null set of LHV theories, the set of logically impossible ones. The assumption of free will, required for Bell's proof to work, is incompatible with the assumption that the universe is deterministic. It is a circular argument that can be rejected on logical grounds, without any appeal to QM or Aspect's experiments. ***

    True, but on the other hand this zero measure class is about the most intuitive ones. But if one looks at the crazy ideas going around in modern theoretical physics, I indeed agree that some liberty SHOULD be granted; actually such extra degrees of freedom are already suggested by what we know to be correct. :smile: Local realism is correct, a simple notion of reality might be misguided (or QM is incorrect which is still a logical possibility).

  6. Sep 22, 2006 #5
    How does this work with the Aspect Experiment of 1982? Smoke, and mirrors? Was non-local causality not verified?
  7. Sep 22, 2006 #6
    Aspect's Experiment just like any other of this type assumes that the detector settings are random. Such an assumption cannot be granted when randomness itself is to be tested, therefore no relevant conclusion can come up from these experiments. A correct test of a LHV theory must include the time evolution of the detector itself (including human experimenters or randomizers).
    In a fully deterministic system it is trivial to reproduce QM's predictions. Here there is such an example:


    Last edited by a moderator: Apr 22, 2017
  8. Sep 22, 2006 #7


    User Avatar
    Science Advisor
    Gold Member

    This is not accurate at a number of levels.

    First, strict Bell tests have been performed in which the detector settings are altered at a point in time too late for an effect travelling at c to have any bearing on the outcome. While there are a few who do not accept any Bell tests as valid (mainly on philosophical grounds), it is a generally accepted result within the physics community.

    Second, the strange reference to cellular automata is not peer reviewed. Quite honestly, I was not impressed with the line of thinking. It completely ignores the basics of the EPR discussion and does not map to the actual problem in any way. But the author is pretty proud of himself - he claims to have an entirely new physics which should replace all existing physics. You be the judge. :rofl:
  9. Sep 22, 2006 #8
    Euh, your webpage isn't peer reviewed either, right ?! Second, what are the basics of the EPR discussions ?? There is no consensus about that at all. I did not finish reading the paper (and will save comments for later, I am a bit worried myself) but the authors clearly state that they go for predeterminism and that is sufficient. Perhaps you could demonstrate that you actually read the paper since somehow I think you did not touch the Selleri book either yet.
  10. Sep 22, 2006 #9
    I'm still not sold. I would buy verifiable replication, by another source, i.e. a peer review. Others must be able to reproduce it repeatedly.
  11. Sep 23, 2006 #10
    One worry which immediately came to my mind after a very brief look is the issue of apparatus complexity. I mean, usually one expects the internal dynamics of the apparati to be so complex that it would be virtually impossible to read its mind. Anyway, more details regarding such issues would be welcome.

    Last edited: Sep 23, 2006
  12. Sep 23, 2006 #11
    Bottom line folks, I work with a top-of-the-line spectrophotometer every day, and while it ain't a supercollider, it does give me real empirical data (in nm) from which I can do research. Deviation in the equipment? .0005% okay, it might be wrong that much of the time, but repeatability reduces the error to a satisfactory deviation. It's not numbers on a page representing suppositions. NO measurement, NO verification.Only an example, not a sample relevant to this discussion.
    Last edited: Sep 23, 2006
  13. Sep 23, 2006 #12
    How the detector settings are changed are simply irrelevant. In a deterministic setup they are correlated with everething else since the big-bang. No signal has to travel between them and the particles. The correlations are a consequence of their common past.

    To be more clear I'll quote several writings.

    In “Speakable and Unspeakable in Quantum Mechanics”, chapter 12, J.S. Bell writes:

    He then tries to argue that this assumptions should be granted and ends with:

    I say that in a deterministic setup free choice does not exist. Assuming it, makes the argument circular, therefore useless.

    Richard Gill, in the article:

    http://www.math.uu.nl/people/gill/Preprints/vaxjo.pdf" [Broken]

    ...in which predeterminism is considered among the possible interpretations of Bell’s experiments, acknowledges this circularity:

    Here there is a pretty recent, peer reviewed article ( American Journal of Physics -- June 2001 -- Volume 69, Issue 6, pp. 655-701):

    http://arxiv.org/PS_cache/quant-ph/pdf/0209/0209123.pdf" [Broken]

    As I've explained above there is no controversy that Bell's proof does not work when a logically consistent determinism is assumed. Petrov's article simply exploits this. It remains to be seen if CA is a fertile idea, I think it is
    Last edited by a moderator: May 2, 2017
  14. Sep 23, 2006 #13
    Repetition builds the concept of constants. Working with variables won't solve anything. We can only rely on that which is constantly available for analytical reference. Repeatable. SHOW ME! again and again, now we're gettting somewhere!
  15. Sep 23, 2006 #14
    We are not debating the raw data but its interpretation. What exactly do you want me to show you?
  16. Sep 23, 2006 #15
    Sorry, You have already shown me a lot. I rely on empirical data proven to me by our current equipment, but theory is provided by you. My facility is one of analysis, but the input by people like you allows me some measure of diagnosis of these results. Everyone doesn't have the analytical mind that you have, thus we must rely on you. It is a mutually beneficial relationship, in that you provide, what I cannot. I have the lab.
  17. Sep 23, 2006 #16


    User Avatar
    Science Advisor
    Gold Member

    Ya, my isn't peer reviewed... but I simply present a digest of mainstream items that are. Not exactly an apples to apples comparison, is it?

    As to Selleri's book... I already told you I have never even seen it for sale. If you can send me yours, I promise I will look at it. :smile:
  18. Sep 23, 2006 #17


    User Avatar
    Science Advisor
    Gold Member

    From a theoretical (philosophy of science) viewpoint, this is a useless statement. You may as well say that Bell test results are due to the active hand of God personally manipulating the results to make it appear that QM is correct. And why do I say from a theoretical viewpoint? Because these statements explain exactly nothing and are totally superfluous. As well as being completely untestable. According to predeterminism, there is a previous "cause" of the outcome of Bell tests. OK, where is it? Oh, I forgot. It is the entire state of the universe. So that explains how local realism works. That's what I call circular logic.
  19. Sep 23, 2006 #18
    You are wrong here, 85 percent of scientists have no particular attitude towards this subject whatsoever and basically have made piece with the fact that they will never have a deep understanding of their calculations, but hey it works. :rolleyes: Then you have the 15 percent fanatics, which admittedly mainly speak in esoteric terms such as nonlocality, consciousness and all that. My point of view is ``less talking, coming up with better alternatives and doing finally a conclusive experiment'' and especially to keep an open mind.

    Hehe, you can buy it from amazon.com; it is a good investment I promise you :wink: (on my atheist soul :rofl: )

  20. Sep 23, 2006 #19


    User Avatar
    Science Advisor
    Gold Member

    Always good advice... :smile:
  21. Sep 23, 2006 #20
    Are you saying you have, or have direct access to the equipment and facilities to support ".....coming up with better alternatives and doing finally a conclusive experiment’’?
    Where if someone where to provide a reasonable and rational alternative to the current explanations -- And more importantly translate that into a performable experiment to test that alternative – You would actually be in a position to have such a test run or considered for testing in your Lab?

    I’m assuming this would be an optical lab with detectors suitable for testing correlations of EPR photon polarity entanglement. (Stern–Gerlach measurements of spin seem so much harder to deal with in every way – testing, test design, explaining, and understanding).

    What an amazing offer on a forum!!
    – of course it’s understood your not just taking in any suggestions
    – you’d need some detail as to how any suggested test could produce something new in its results.
    A very brief description of your lab capabilities would be nice!
  22. Sep 24, 2006 #21
    Useless or not, this follows as a logical necessity from the assumption of determinism employed in Bell's reasoning. If, for whatever reason, you don't like determinism you may say so and move forward, without the need of useless, circular arguments (I equate here determinism with local realism. It is true that stochastic, local-realistic theories have been proposed but I do not maintain such a position)

    I may, but I'm not. This is very unparsimonious.

    They are not superfluous, they show that deterministic local-realism is compatible with QM. That's all I'm saying. Bell's theorem is simply irrelevant . This conclusion is not suppose to explain anything but to open the possibility of deterministic theories which could give an explanation.

    Why do you say that a deterministic theory cannot be testable? The cellular automata model, for example, is testable in principle. I'm not saying CA is the answer, the theory must be developed first, but it is a good example.

    You are attacking a straw-man here. See above.

    Now, can I ask you in turn, what better explanation do you propose? And no, I'll not take "shut up and calculate" or "it's random" or something of this sort as an explanation.
  23. Sep 24, 2006 #22
    I see a significant problem with taking this as adequate ‘proof’ that “Bell's theorem is simply irrelevant”. Presuming you define a deterministic theory that explains correlations; by what test exactly do you propose to verify and prove the claim if not Bell.
    Did you have a new kind of test to offer to DrDave (I hope he will comment) to check such future theories against?

    Bell designed his theorem to give a HVT to opportunity to do just that. Like it or not I’ve seen no other proposed test that might detect the difference between a Non-local theory (QM, BM MWI etc) and a local deterministic theory.
    In fact IMO when a deterministic theory does come along; It will want, It will need, and It must use, BELL to prove its point.
  24. Sep 24, 2006 #23
    The paper cited clearly intends to violate a particular form of the Bell inequalities. Perhaps let us concentrate on the paper so that we can figure out how the program works and how the data could violate the Bell inequalities. I just looked through the program instructions and did not quite get some things straight away (I guess ueit knows this since he praises the paper) :
    (a) there is a left/right assymetry in rule 30, why ??? It does not seem natural to me to break this symmetry.
    (b) the source emits left and right travelling virtual particles first which are *determining* the detector settings ?
    (c) the colour of the detection does *not* depend upon the detector switch state line but only upon the particle spin !
    (d) it is not obvious at first sight that given equal detector settings left and right, the spin measurements are the same, is that proven somewhere?
    (e) In the simulation, I had almost 50 percent of the time single hits (and that is extremely bad)! How are these processed in the statistics (substraction I guess, but then it is well known you can violate the Bell inequalities by ``Bell local'' realist theories)??

    Anyway, some comments would be welcome before I would put some more time into this.

    Last edited: Sep 24, 2006
  25. Sep 24, 2006 #24


    User Avatar
    Science Advisor
    Gold Member

    First: I don't see where determinism - or indeterminism - is a function of Bell's Theorem. Perhaps you can point that particular piece out to us specifically (as in a reference).

    Second: you will not find it so easy to develop a local realistic theory - even one which is deterministic in the manner you have cited - which can reproduce the predictions of QM. Anything less than a full example will satisfy only those who don't know any better. The requirements are many.

    Third: The premise you are starting from makes little sense: you will have to completely tear down the Uncertainty Principle to achieve your goals (since it is the source of the EPR Paradox). And then we would be left wondering why the Uncertainty Principle even appears to exist! Ah, determinism! What an answer! As said previously, this explains absolutely nothing. I'll stick with Bell's Theorem at this time.
  26. Sep 25, 2006 #25
    If I understand correctly, your position is that we should accept a fallacious argument as sound because we don't have a better one. Needless to say, I disagree. If we cannot distinguish between determinism and non-determinism then that's it.
    It is sad to see great physicists claiming that it is foolish to investigate LHV theories because they have been proven wrong by Bell and see all kinds of SF-like ideas proposed as serious explanations.
    And MWI is said to be both local and deterministic although I disagree.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook