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

How widely accepted is Bell?

  1. Jun 30, 2010 #1
    I apologize for the vagueness of my question, but a more specific version would not fit in the thread title box:

    How widely accepted is the rejection of local realism due to violations of Bell's theorems?

    I also have two follow-up questions:

    How does a rejection of local realism translate to a statement about the ontological stochasticity of quantum mechanics?

    What needs to be rejected to make quantum mechanics ontologically stochastic, locality or realism?
     
  2. jcsd
  3. Jun 30, 2010 #2

    DrChinese

    User Avatar
    Science Advisor
    Gold Member

    Welcome to PhysicsForums, mjpam!

    It is essentially universally accepted by physicists in the field. There are a handful who either reject it or the experiments that test it. Hess, Philipp, Santos are some of the more well known. As far as I know, there are no groups doing any experimental work in which Bell is questioned. Mainstream scientific publications have stopped publishing disproofs or the like. Two excellent recent roundups of the subject of EPR, Bell and related experiments include:

    Genovese, 2007, 106 pages, over 500 references
    http://arxiv.org/abs/quant-ph/0701071

    Cavalcanti et al, 2008, 25 pages, "many" references
    http://arxiv.org/abs/0806.0270
    Generally, all local realistic stochastic interpretations have been either rejected or shown to be non-local. So I would say reject realism if you want a stochastic model. I guess that the interpretations around Bohmian Mechanics might fit that bill.
     
    Last edited: Jun 30, 2010
  4. Jun 30, 2010 #3
    Is it possible to come up with a deterministic model of QM?
     
  5. Jun 30, 2010 #4

    DrChinese

    User Avatar
    Science Advisor
    Gold Member

    http://aps.arxiv.org/abs/quant-ph/0611032

    And we are fortunate to have Demystifier and several others here than can help you to understand non-local deterministic theories such as Bohmian Mechanics (also called Pilot Wave and/or dBB)
     
  6. Jun 30, 2010 #5
    I note that the authors only describe a non-local QFT in that article. My understanding is that non-local models also conflict with general relativity.

    I guess what I'm wondering is: Is the choice among deterministic and stochastic models of QM really just a matter of taste dependent upon the particular physicist's underlying philosophical predilections?
     
  7. Jun 30, 2010 #6

    DrChinese

    User Avatar
    Science Advisor
    Gold Member

    Generally, yes. There are a range of interpretations that keep or throw out various tenets. At this time, they don't really make any distinguishing experimentally verifiable predictions.
     
  8. Jun 30, 2010 #7
    So "local" and "realistic" do not necessarily entail "deterministic"?
     
  9. Jun 30, 2010 #8

    DrChinese

    User Avatar
    Science Advisor
    Gold Member

    Well, that is meat for several other threads here. :smile:

    But the usual use of these words is that Bell excludes local hidden variable theories. These are sometimes called local realistic theories, local deterministic theories, locally causal theories, local non-contextual theories... you get the idea.
     
  10. Jun 30, 2010 #9
    Dear mjpam,

    I like the drift of your questions so I encourage you to continue to probe and question.

    I am not sure of the above "necessity" BUT ---

    I am certain that QM reflects an underlying local, realistic, deterministic reality

    --- with NO spooky actions at a distance.

    JenniT
     
  11. Jun 30, 2010 #10

    alxm

    User Avatar
    Science Advisor

    Bell is generally accepted as said. But I think it's also the case that most working physicists simply don't spend a lot of time bothering to think about it. Relatively few work in the field of the foundations of QM. It's not that people aren't not interested in the questions as much as the fact that we simply don't have many ideas on how to get answers.

    I think the whole issue of determinism is really over-hyped as well. People forget to distinguish between "determinism" and "predictability". The fact that a theory is deterministic doesn't mean that it's predictable - even in principle, because the theory can limit (even in principle) what you can find out about the initial conditions of the system. deBB being an example of this.

    I don't care if God plays dice, it's next week's Lotto numbers I want to know.
     
  12. Jun 30, 2010 #11
    I would count myself as one who accepts Bell's theorem and the constraints it imposes, and would answer so if polled. What I apparently disagree with many on is the generality of those constraints, and the generality of the limits it imposes.
     
  13. Jun 30, 2010 #12

    I totally agree with this statement.
     
  14. Jun 30, 2010 #13
    The "converse" goes for stochastic models: just because a model is stochastic doesn't mean that it unpredictable.

    One of my biggest pet peeves is when someone uses the word "random" to mean "unpredictable" and then applies it to a predictable scientific theory to claim that the theory is not random.
     
  15. Jul 1, 2010 #14
    Good point(s). But Bell is 'said' in different, and conflicting, ways. Hence the controversy.

    I agree, and find your contributions to ascertaining the true meaning of Bell's work quite interesting. And, I have the feeling that somewhere in your many posts on this you've hit on the (physical) crux of the matter. So, please, keep thinking about this and refining your ideas and posting.

    Not sure what you mean. Example?
     
  16. Jul 1, 2010 #15
    I am not sure what you mean either. Do you have an example?

    I would hold that a random variable means that one can predict within a deviation -- but that the deviation of any individual experiment can be so large as to have a real result which does not agree with what is "likely". I wouldn't (for example) say that something which occasionally tunnels 16 feet to the left is actually "predictable" when it tunnels so rarely as to be ignored in practice ... nor would I call the lottery "predictable" when predictions generally are only applicable to how likely one is to fail at guessing the next outcome of the lottery.

    I'll quote Alxm's line since I picked on smart Alx in another thread and don't want to totally bash the guy.
    "I don't care if God plays dice, It's next week's lotto numbers I want to know."

    At very least (in the case of tunneling), one must admit that stochastic rejection of some non-locality is implicitly rejected by the majority... that is the very meaning of "improbable". If one allows it "in principle", but never believes it will happen "observably" ... I don't think that really is different from what I am saying.

    Perhaps the OP's question has some subtlety being glossed over... ?
     
  17. Jul 1, 2010 #16

    DrChinese

    User Avatar
    Science Advisor
    Gold Member

    For those asking about the above: I think the idea expressed is that random and stochastic are not the same things, even though they are sometimes interchanged. As well that stochastic does not always mean indeterministic.
     
  18. Jul 1, 2010 #17
    In what instances is this true?
     
  19. Jul 1, 2010 #18

    DrChinese

    User Avatar
    Science Advisor
    Gold Member

    I thought that was your point, that random, indeterministic and stochastic are interchanged when they should not be.
     
  20. Jul 1, 2010 #19
    Then I misunderstood what you were saying.:redface:
     
  21. Jul 1, 2010 #20

    DrChinese

    User Avatar
    Science Advisor
    Gold Member

    Well, just forget what I said. Not that important. :smile:

    In the meantime, tell us what you see in stochastic theories that is of interest to you.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: How widely accepted is Bell?
  1. How important is Bell? (Replies: 36)

Loading...