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

Requirements for LR model - please explain.

  1. Jun 26, 2010 #1
    Requirements for LR model -- please explain.

    In a recent, and ongoing, thread it was asked what are the requirements that an LR model of entanglement must meet. The following response was given by DrChinese. I still don't understand what he's talking about, so if DrC, or anyone else, would care to clarify his 'clarification' it would be appreciated.

    I don't foresee replying to any responses to my query. Hopefully, some more knowledgeable people than I, including but hopefully not limited to DrC, will reply and I'll then use those replies as a point of departure for future considerations and research. And, of course, if I consider any reply to be obviously incorrect, then I'll post what I currently think about it. Thanks in advance.

  2. jcsd
  3. Jun 26, 2010 #2


    User Avatar
    Science Advisor
    Gold Member

    Re: Requirements for LR model -- please explain.

    Sure, I'll try and I hope others will throw in too. I may not hit it all in one pass.

    First, we would start with EPR before proceeding to Bell. EPR implies that a) there are elements of reality that do not depend on the act of observation; b) you might expect perfect correlations of entangled particles would be observed for some of those elements of reality; and c) generally a local theory might exist in which these elements are consistent with quantum mechanics.

    Second, we add Bell and the explicit realism requirement: there must be simultaneous values for each element of reality, and at least 3 must be demonstrated (since Bell shows this is not possible within the framework of a theory which matches the predictions of QM).

    So the basic idea is: i) demonstrate you have a local realistic theory; ii) demonstrate whether that matches the predictions of QM (or not).

    With the De Raedt model, for example, I can actually simulate - on an event by event basis - data values for 3 different settings of my choosing. So I can actually look at the resultant values - I collect 5000 in my spreadsheet model. Then I can analyze and see how close it comes to QM predictions. Of course it can't match them, it only approaches the Bell inequality limit. But it absolutely passes the test of realism.

    Now, if you put forth a model that does not accomplish this, then you are essentially conceding the main point: you don't have a realistic model. Most LR models being published fail this essential test - which is why no one bothers to refute them. I mean, why bother to refute a model that flunks it main raison d'etre?

    If the author argues they have a realistic model and then don't show how to derive 3 simultaneous values at settings I choose, then they are ignoring Bell entirely. That is equivalent to a magician who claims his magic is real but won't allow you to look up his sleeves. Rabbits don't actually appear ex nihilo.

    (Hey, I used phrases from 2 languages other than English in one post!)

    So the 3 setting requirement is a simple way to find out if Bell was considered or not. It is not enough to say you considered Bell without actually passing the test. Since this test is quite severe, and quite easy to grade as pass/fail. Once you have a successful model, then other requirements can be applied. Such as agreement with experiment, which becomes a big problem.

    Arrivederci. :smile:
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook