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

Bell's mathematical error?

  1. Jun 28, 2010 #1
    I would hope this thread could be limited to Bell-mathematical questions and answers concerning just one given Bell paper -- though it may involve many mathematical questions as we follow it to resolution.

    This thread concerns mathematical questions in one Bell paper and so it should NOT need a lot of words and diversion to resolve it. It is about quantum physics so I think it belongs here --- with help from mathematicians.

    It might help us more if those helping us could say what level of mathematics and quantum physics they have reached. But I respect privacy.

    I believe that logic is at its highest development in mathematics and probability theory -- so I should be not too slow in those areas.

    Let there be no question here about Bell's assumptions BUT

    Question 1. Is there a mathematical error in Bell's theory, as will follow?

    After starting and following the threads "Understanding Bell's logic" and "Understanding Bell's mathematics", I would like to discuss and resolve a clash of the two in Bell's Bertlmann's socks paper -- which is available to all from CERN on-line -- [PLAIN]http://cdsweb.cern.ch/record/142461/files/198009299.pdfpapers [Broken] [/URL] -- the discussion in those threads not appearing to resolve it for me.

    One of Bell's latest papers on EPR, without excuse it should be one of his clearest?

    Look at Bell's equations (11) and (12), and combine them to form

    (Bell 12) = [tex] P(A, B|a, b) = \int d\lambda \rho (\lambda) P_ 1(A|a, \lambda) P_ 2(B|b, \lambda). [/tex]

    It is Bell's supposition (above his (11)) that the variables [tex] \lambda [/tex] allow this decoupling.

    Question 2. Are [tex] P_ 1 [/tex] and [tex]P_ 2 [/tex] Probability Functions?

    Question 3. Is it not the case that Probability Functions map a subset of the sample space to the real interval [0, 1]?

    Question 4. If [tex] P_ 1 [/tex] and [tex]P_ 2 [/tex] are Probability Functions, how do we apply Bell's [tex]\rho (\lambda)[/tex] to such functions?

    Question 5. If they are NOT Probability Functions, what are they, please?

    Question 6. Could you provide an example of the Function that you believe them to be, please?

    Question 7. Bell has [tex] P_ 1 [/tex] and [tex]P_ 2 [/tex]. Why did he not have also [tex] \lambda_ 1 [/tex] and [tex]\lambda_ 2 [/tex] ?

    Question 8. If he did the lambda-separation in Question 7 -- which is allowable under his theory -- how would he have written Bell (12) above?

    Thank you very much.
    Last edited by a moderator: May 4, 2017
  2. jcsd
  3. Jun 28, 2010 #2


    User Avatar
    Science Advisor

    What do you mean by "apply"? [tex]\rho (\lambda)[/tex] is itself just a probability density function, the continuous analogue of a probability function on a discrete sample space. If λ only took a discrete set of values, labeled λi with i ranging from 1-N, then the integral you wrote above could be replaced by the sum [tex]P(A, B|a, b) = \sum_{i=1}^N P_3(\lambda_i) P_ 1(A|a, \lambda_i) P_ 2(B|b, \lambda_i).[/tex] The integral you wrote is just the analogue of this for a λ that can take a continuous range of values.
    What do you mean by λ1 and λ2? Are these just specific values of the random variable λ? If so, [tex]\rho (\lambda)[/tex] already assigns probability densities to each possible specific value of λ.
    Unclear what you mean by "lambda-separation". [tex]P_1[/tex] and [tex]P_2[/tex] aren't based on "separating" a single random variable, rather they are based on separating a joint probability P(AB|a,b,λ) into individual probabilities for A and B. I don't see how you could do anything analogous for [tex]\rho (\lambda)[/tex].

    edit: I suppose that depending on how λ is defined, and what type of local hidden variables theory we're imagining, it might be reasonable to split it into two parts λ1 and λ2, the first describing local hidden variables that influence the outcome of measurement A and the second describing local hidden variables that influence the outcome of measurement B. Is this the sort of thing you were thinking of? For example, if we define λ in terms of all local variables in cross-sections of the two past light cones of the measurements, with the cross-sections taken at some time after the last moment the two past light cones overlap, then λ1 could just refer to all local variables in the cross-section of the past light cone of measurement A, and λ2 could be all local variables in the cross-section of the past light cone of measurement B. In this case your integral could be replaced by the following double integral:

    P(A, B|a, b) = \int \int d\lambda_1 d\lambda_2 \rho (\lambda_1, \lambda_2) P_ 1(A|a, \lambda_1) P_ 2(B|b, \lambda_2).
    Last edited: Jun 28, 2010
  4. Jun 28, 2010 #3


    User Avatar
    Science Advisor
    Gold Member

    I have told you before, and I guess you will discover for yourself: Bell wrote a lot of things and he wrote in a lot of different styles. So no, you won't find much that is clear - or clearer - than the original paper. And for the original paper, like his later ones, you need to consider the audience in order to understand the argument. If you read it like an ordinary text or paper, it will not make sense.

    Now, this is not an "excuse" for Bell. His paper has been cited thousands of times so I really don't think it needs much in the way of excuses. My point is that you will find it easier to follow if you consider the audience. Then the formalities won't matter so much. Also, don't forget that audience would know EPR too.
  5. Jun 28, 2010 #4
    Thank you very much DrC. I do understand your view that Bell was preaching to the converted in so far as he spoke in a language clear to them -- like Danish. But my girlish intuition senses something not quite right in the state of Denmark. The mathematical point I am coming to is in all his EPR papers -- with the beauty of being the same in all languages, as in my next post here responding to JesseM's nicely detailed mathematical reply.
  6. Jun 28, 2010 #5


    User Avatar
    Science Advisor
    Gold Member

    And nice it is!

    The reason I mention this, when you clearly want to speak in the mathematical lingo (and excellent is that), is that you should definitely switch to definitions and representations that match how you think. Use the strongest arguments, not the weakest (as many do on this board). That way you don't go off topic because you are wandering around in the details and imagining Bell is wrong. (If you think Bell is wrong, it is simply one of your definitions that need to change.)

    If you asked 10 different top scientists to derive Bell, you would probably see 10 different representations. And yet all 10, if shown the others, would agree all 10 are actually identical. They won't see the differences you might. If you go to my web site, I give 2 completely different approaches and neither match Bell directly. One follows Mermin (JesseM's scratch card example is identical), the other demonstrates negative probabilities:

    Bell's Theorem and Negative Probabilities

    I guess I should add another that closely follows Bell too, think? :smile:
    Last edited: Jun 28, 2010
  7. Jun 30, 2010 #6
    Thank you JesseM very much for your time and accuracy. I welcome and appreciate your rigor.

    It makes more sense to me to separate Bell's single lambda into two, so your "edit" is the way I see it.

    Question 9. To build a common understanding between physicists and mathematicians, would it be acceptable to call the lambdas "spin-vectors" --- vectors related to angular momentum but unconstrained as to length or orientation?

    Question 10. The lambdas would then constitute an infinite set of random variables in correlated pairs -- with a uniform distribution over 3-space -- which would not take Bell's mathematics in Bell (12) very far -- the result being one-quarter?

    Question 11. Has Bell overlooked the fact that "measurement" is a 2-stage process? For photons, the randomized (but pairwise correlated) lambdas first impact the polarizers. The polarizers reduce an infinite input set to a countable output set which then becomes the local input to each associated analyzer -- Alice's analyzer (orientation a) being the device which delivers outcome A; Bob's analyzer (orientation b) delivers B?

    Question 12. So, mathematically and physically, the density function for each analyzer input is the one Bell must require? NOT the density function for each polarizer?

    In my view the exact QM expectations for any EPRB-experiment then follow readily -- consistent with local-realism?

    Perhaps showing that neither Bell's assumptions nor his logic was at fault?

    I look forward to your critical analysis,

    With my thanks again,

    Last edited: Jun 30, 2010
  8. Jun 30, 2010 #7


    User Avatar
    Science Advisor

    Not sure why you say spin vectors--is the direction of the vector supposed to determine how the particle will respond to each detector setting? If so, how?

    Like I said, I think the most rigorous way to approach Bell's proof is to imagine that each value of λ specifies the value of all local physical facts ('beables') in the cross-section of the past light cone of the measurement at some time t after the last moment that the past light cones of the two measurements overlapped, but before the experimenters have made a choice of what detector settings to use (this is essentially what Bell does on p. 242 of Speakable and Unspeakable in Quantum mechanics, although he actually has λ + c stand for all local variable in a series of cross-sections, the last of which are from a time after the last moment the two light cones overlap). If λ1 gives the complete state of some such cross-section of the past light cone of the measurement which might yield result A, it's then not too hard to see why locality implies that P(A|a,λ1) = P(A|B,a,b,λ1,λ2) [which together with the assumption that P(B|b,λ2)=P(B|a,b,λ1,λ2) implies that P(AB|a,b,λ1,λ2)=P(A|a,λ1)*P(B|b,λ2), so we can arrive at the equation you included in your original post], since outcome B has a spacelike separation from A (as does the choice of measurement setting b, and as does every event whose details are given by λ2), and any past events which could influence both A and B (creating a statistical dependence between them) would be in the overlap of the past light cones of A and B, and thus these events would be at a time earlier than λ1 (and in a local realist universe, the only way these earlier events can have an influence on A is by influencing λ1 which in turn influences A).
    I don't follow, how do you get any definite result from equation (12) if you don't know the value of P(A|a,λ1) and P(B|b,λ2)? Or are you assuming something specific about how the direction of your "spin vectors" interacts with the angle of the polarizer to determine outcomes A or B?

    Also, why do you say the lambdas are in "correlated pairs"? Remember, at this stage in the proof Bell is no longer assuming that if both experimenters choose the same detector setting they are guaranteed to get identical (or opposite) results. On p. 12 he says:
    Under the type of past light cone definition of λ I am using, and which Bell used in that section of "Speakable and Unspeakable in Quantum Mechanics", lambda does not refer to hidden variables at the time of measurement itself, but rather to all local variables in some cross-section of the past light cone of the measurement which might have a causal influence on the values of variables at the actual time of measurement. So, it makes no difference if the measurement is an extended process in which values of hidden variables associated with the particle can change based on interactions with the measurement-apparatus (or for any other reason compatible with locality).
    No, as I've said the different values of lambda which appear in the density function need not specify anything about what is happening at the actual time of measurement.
    Last edited: Jun 30, 2010
  9. Jun 30, 2010 #8
    Thank you JesseM

    1. Spin relating to angular momentum. So they need be vectors. What else could they be? Especially when nothing more is required to reproduce the QM dynamics?

    2. Yes. "The direction of the vector determines how the particle will respond to each detector setting." For a realistic and relevant analogy, consider the dynamics involved with macroscopic wire-grid polarizers. We require nothing more.

    3. You can see that my lambda specification is very broad, and very realistic. So I am not seeking to sneak something under the carpet. The lambdas are unconstrained as to length and orientation; and correlated by whatever angular momentum conservation setting was in place as the particles were co-produced, pairwise correlated.

    4. Question 13. Many physicists resist agreeing with my lambda specifications and provide not one clue as to their reasons. I say that my very broad lambda specifications are NOT inconsistent with QM? Yet even you baulk at the, what is to me, obvious? Why?

    5. Do you require my specifications strengthened or weakened? Why when they appear to be spot-on, realistic, adequate for the problem at hand?

    There are a lot of words and inter-weavings here. Does what follows cut though:-

    How can anyone object to my formulation when I insist that the lambdas are local realistic particle parameters -- they reside with each particle, one lambda-one-N accompanying the N-th particle that Alice will test, one correlated lambda-two-N accompanying the N-th particle that Bob will test?

    Of course. Interested in generality, I write a general formula for a general analyzer -- with a particle's intrinsic spin s in the argument. For spin-half "polarizers" [Stern-Gerlach magnets] s = one-half; for photon-polarizers s = 1.

    The particles ARE correlated. So I recognize it in the formalism.

    Do you accept that this reflects a valid approach?

    Would Occam?

    For me, Bell wanders too far afield -- lost limping in a huge forest when a mere sapling will provide a crutch and escape.

    But is it not the case that CRUCIAL measurement interactions and dynamics are involved "at the time of measurement"?

    Crucial determined quantum dynamics as against quantum mechanics?

    HOW does Bell allow for these dynamics?

    Is he bound to thinking that each "measured" output identifies one-to-one with each input -- he being much opposed to "measurement" for this exact reason?

    Especially considering ---

    Huge numbers of incoming projectiles -- in correlated pairs -- their properties drawn from an infinite set -- reduced, by "measurement", to a countable set -- cardinality 4?

    Do you see that my determined local realistic simplicity cuts through many words, assumptions and much irrelevant discussion?

    Can you point me to something that I do that conflicts with QM or Bell's assumptions?

    Many thanks again, and best regards,

  10. Jun 30, 2010 #9


    User Avatar
    Gold Member

    JenniT, I was wondering what level of mathematics and quantum physics you have reached.
  11. Jun 30, 2010 #10


    User Avatar
    Science Advisor

    But lambda here includes all vectors and all and any information you want, except the local detector settings. That's all. In my https://www.physicsforums.com/showpost.php?p=2743072&postcount=4" to you a month ago I said this, I don't know why you're back at this. Lambda can contain vectors and scalars and complex numbers and whatever you like except the detector settings, which are 'local'. The values of lambda can be as correlated as you like. P1 and P2 can depend on different subsets of lambda if you like. The point of Bell's (12) is that you can separate into P1 and P2 with respect to the local detector settings (or the local temperature in Lyons/Lille).

    I don't see that anyone did object to that. It's not at odds with what Bell asserts in (12). The lambda-dependence of P1 and P2 isn't given or even assumed to be known, so naturally they can depend on subsets of lambda. So introducing [tex]\lambda_1[/tex] and [tex]\lambda_2[/tex] or any other number is fine (but rho will of course then be a function of all lambdas, and you will have to integrate over all of them). What you're describing is a subset of what's allowed by (12).

    I have no idea what you're talking about here. If you're referring to something specific, be specific.

    What do you mean, 'input'? The hidden variables? There's no assumption of a 1-to-1 correspondence at all.

    So? And there are 6 billion people in the world and only one expectation value for their age. What would set theory have to do with it, anyway?

    Don't be silly, not even the Danes http://www.youtube.com/watch?v=s-mOy8VUEBk"
    Last edited by a moderator: Apr 25, 2017
  12. Jul 1, 2010 #11
    Dear morrobay,

    I am as a beginner in these things.

    If there are things that I write that you do not understand, please let me know and I will do my best to correct, improve or explain.

    And do not be misled -- I belong to a scattered minority group of convinced local realists -- more in Einstein's camp than Bell's.

    Best regards,

  13. Jul 1, 2010 #12
    Dear alxm,

    Perhaps I am over sensitive but your reply here looks like you are building a straw woman and putting it down?

    I will attempt to answer with respect and courtesy.

    You start with a BUT. BUT What -- for me -- if I need less than Bell allows?

    It seems that much [NOT all] of what you say here is irrelevant to anything that I have said or believe. Or reinforces agreed points as if I have missed them?

    I suspect you might owe me an apology?

    Your supposed reply to me is in agreement with what I said? So how is it a reply? And how am I "back at it"?

    OK, so [see above]: "lambda here [my emphasis] includes all vectors and all and any information you want, except the local detector settings. ... includes all vectors and all and any information you want, except the local detector settings."

    But the lambdas (plural) here are my "reduced" lambdas because I need nothing more that their representation as pairwise correlated spin vectors -- unconstrained in length and orientation.

    Is this an acceptable general specification as far as a QM specialist is concerned?

    As Science Advisor, please, tell me what else I will need --

    1. To answer my question.

    2. To tackle and understand Bell.

    3. To be consistent with QM.

    If Occam is more to my liking than Bell --- does it matter --- if my proof goes through?

    And please consider this: The detectors are separated and local BUT the detector settings are simply orientations in 3-space -- a very important difference to me.

    And I agree with the P1 and P2 separation -- we agree --- yet you make a point of it.

    Thank you, Yes.

    I am referring to the reduction of an infinite set to a countable set via the "measurement" interaction (dynamics).

    We agree. Thank you.

    Well 6 billion may be a big, but age is hardly an infinite set of lengths and orientations; and what "measurement" would you use to produce an equivalent countable set of cardinality 4?

    Please answer this because I may be missing something -- because I do not see that we are yet at the stage of expectation values? For me, they come later.

    Set theory? Perhaps relating to subsets of the sample space mapped by Probability Functions to the real interval [0, 1]?

    Taler du dansk? :!!)

    I tink so ... and maybe I have just come to like strong robust answers and discussion :!!)

    Thanks for them,

    Last edited by a moderator: Apr 25, 2017
  14. Jul 1, 2010 #13


    User Avatar
    Science Advisor

    You can't reproduce QM dynamics by assuming each particle has a local hidden variable corresponding to a spin vector with a well-defined direction, and that its deflection in a Stern-Gerlach device (up or down) is simply a function of its hidden spin vector and the orientation of the SG device. That would be an example of the type of local hidden variable theory that's ruled out by Bell's theorem. If you think it would be possible to come up with such a local model and reproduce all the QM dynamics, just provide the details of how the spin vector and the SG orientation are supposed to determine the result (or the probabilities of different results), so if I give you a particular angle for the spin vector and a particular angle for the SG orientation, you have a formula for calculating whether the result will be spin-up or spin-down (or what the probabilities of each will be).

    Anyway, the point of Bell's theorem is to be as general as possible, and not make any specific assumptions about how the measurement outcomes are determined. It could be, for example, that each particle would have a large collection of hidden variables associated with it, and would "decide" whether to be deflected upward or downward by a given SG device by making use of some very complicated algorithm. This may not seem very physically plausible, but as long as it's a local hidden variables theory, Bell wants to rule it out. Also, keep in mind that if lambda represents the full state of all local variables in a cross-section from time t of the past light cone of the measurement (with t being after the last moment the past light cones of the two measurements overlap), then that will necessarily include any local hidden variables associated with the particle itself, so if the particle has something like a "spin vector" associated with it, the orientation of that vector at time t will already be included as part of lambda.

    I gave more detail on the rationale for using cross-sections of the past light cones in post #61 here, starting with the paragraph that begins "Let me try a different tack." (the discussion then continued in post #62) I called the past light cone cross-sections PLCCS' there, and offered an analogy of two computers which can exchange date for a while, but after some time t their communication is cut off and afterwards they each simulate a measurement:
    After a discussion of this analogy in terms of how it can be used to understand the "no-conspiracy assumption" in Bell's proof, I continued in post #62:
    This last question was directed at the person I was talking to on that older thread, but I'd like to ask it of you too; if you think the analogy breaks down somewhere, can you point to where? Just as the two computers can no longer communicate after some time t, so there will be some time t that is after the last moment the past light cones of the two measurements overlap, but before the measurements are made (or the experimenter's random choice of which detector setting to use), and nothing in the past light cone of one measurement result at time t or after can have a causal influence on anything in the past light cone of the other measurement result after time t.
    Well, again, can you please provide the formulas to show what the particle will do for any arbitrary combination of vector direction and detector direction? I'm not sure what "macroscopic wire-grid polarizers" are, but note that if you want a macroscopic analogue of particle spin, you can't just imagine something like a macroscopic charged spinning ball. As explained in the first section of this article, a macroscopic charged spinning ball traveling through an external magnetic field could be deflected at a range of angles depending on the angle between its spin axis and the direction of the external field; in contrast, quantum particles with spin like electrons traveling through the same magnetic field will all be deflected in one of two possible directions, corresponding to "spin-up" and "spin-down", regardless of the direction of the field.
    But not as broad as mine, which will already specify all local variables associated with the particle at the time t which is before the measurements are done but after the last moment the past light cones of the measurement overlap (so there can be no causal influence of anything in the past light cone of measurement A at time t or after on anything in the past light cone of measurement B at time t or after, including a causal influence on local variables associated with the particle such as yours). The advantage of my definition is that it allows us to show rigorously why P(AB|a,b,λ1,λ2)=P(A|a,λ1)*P(B|b,λ2), whereas with your definition the rationale for this step isn't so clear.
    Again, because the possibility of a spin vector associated with each particle is already included in my definition, and because my definition allows for a rigorous derivation of the fact that P(AB|a,b,λ1,λ2)=P(A|a,λ1)*P(B|b,λ2).
    If by "strengthened" you mean "made more specific" and "weakened" you mean "made more general", then weakened. Reasons for this above.
    I don't object to this as a hidden-variables theory, I just say that for the purposes of making the derivation rigorous we should define λ in terms of the complete set of local facts in cross-sections of the past light cones of the measurements at some time t, which will naturally include an exact specification of your "local realistic particle parameters" at time t.
    So what is that "general formula"?
    Sure, it's a valid approach to assume the particles will always give perfectly correlated results when measured with the same detector angle, since this is predicted by QM. Bell just wanted to generalize his proof a little to cover the possibility that QM's predictions about this might turn out to be imperfect--he showed that even if there wasn't a perfect correlation with identical detector settings, you'd still be able to use the assumption of local realism to derive some inequalities which are significantly violated by QM. But if you want to stick with the assumption of perfect correlation that's fine.
    All that's important to the proof is that if you already know the complete physical state λ1 of all local variables (including hidden ones) in a cross-section of the past light cone at time t, then the only events which could alter your estimate of different measurement outcomes A (given a known detector setting a) would be other events in the past light cone at times after t. That's enough to guarantee that P(A|B,a,b,λ1,λ2)=P(A|a,λ1), since B,b,λ2 all describe facts about events which lie outside this region of the past light cone of the outcome A. So certainly physical aspects of the measurement device during the measurement might have an effect, but these would be inside the past light cone of the final result after time t.

    Again, just think of the computer analogy. If you already know the complete internal state λ1 of computer A at a time t after it can no longer communicate with the other computers, then whatever happens in the other computers should have no effect on your estimate of the probability of different final outputs for computer A. The computer may be performing all sorts of complicated calculations and simulations (like a simulated particle interacting with a simulated detector) between the time t and the time it gives its final output, and if its processes include a random element (like if it has an internal source of true random noise such as a geiger counter measuring some radioactive decay events) then learning information about its internal state between t and the final output might change your estimate of different possible final outputs, but what happens in the other computers besides A should be irrelevant as far as you're concerned.
    If it is true that identical measurement settings always give identical (or always give opposite) results, then after we have defined λ1 and λ2 in terms of cross-sections of the past light cones of measurement outputs, and shown that local realism implies P(AB|a,b,λ1,λ2)=P(A|a,λ1)*P(B|b,λ2), then from this (and the 'no-conspiracy condition') you can derive the fact that the only way to explain the perfect correlation is to conclude that λ1 predetermined the result A for each of the three possible detector settings, and likewise λ2 predetermined the result B for each of the three possible detector settings. But this isn't an independent assumption, it's derived from the other definitions and local realism + no-conspiracy. And anyway, as I said, Bell also derived some inequalities that would hold even if we dropped the assumption of perfect correlations, in which case the value of λ1 need not give a predetermined measurement result for each of the three detector settings.
    You are just defining lambda too narrowly, so that there is no rigorous way to see that the step P(AB|a,b,λ1,λ2)=P(A|a,λ1)*P(B|b,λ2) is actually justified. Again, Bell's assumptions already cover any type of local realist theory you can imagine, including one like yours where the local hidden variables are vectors associated with each particle. In the type of theory you describe, do you deny it would be possible the define a variable that gives the complete set of local variables associated with any specific point or set of points in spacetime, including the set of all points at time t in the past light cone of the final measurement result? You might say that this variable would include plenty of superfluous information that you think isn't relevant to determining what the final result will be, and that may well be true, but the definition is a coherent one and it allows us to see why the step above would necessarily be true under your (or any) local realist theory.
  15. Jul 1, 2010 #14


    User Avatar
    Science Advisor
    Gold Member

    I ask you: what are you really asserting if you say you are a local realist?

    a) Do unentangled quantum particles have well defined attributes independent of observation?

    b) Are entangled particles unaffected by observations on their distant partner?

    If your answer to either of the above is YES, then how do YOU explain experimental results? You see, Bell is a tool to help us understand these questions. And to assist in framing new experiments around them.

    Personally, I believe the answer to both of the above questions is NO. Experiments, while not altogether conclusive on each individually, strongly support that position. There is essentially no experimental evidence for a YES to either. (However, it is possible that one is a YES and the other is a NO.)
  16. Jul 1, 2010 #15

    Thank you DrC,

    I assert that the real world is local and realistic.

    There are many definitions of these terms.

    In this arena, Einstein has said it all already for me.

    That is: I am not aware of any statement by him, in this arena, that I would deny.

    If there are such statements [by Einstein] which you, with your beliefs, would deny, please let me know.

    YES, of course.

    YES, of course.

    Very simply -- truly -- when constructing a physical theory, I let my mathematics follow the physics.

    I see the interpretative (Feynman, understanding) issues which surround QM as arising from its brilliant origins in mathematics.

    We have much more experimental data now -- so my theory builds only on the physics.

    I have no quibble with any experiment in this area.

    I question the conclusions drawn from them.

    For me, NONLOCALITY is nonsense.

    DrC, your belief is beyond me.

    Maybe you follow that Einstein error when he said -- "It is theory that decides what can be observed."

    Does the inadvertent observation of the cosmic microwave background radiation (or the pigeons) rebut his statement here?

    Not to mention the ether?

    So maybe you are too influenced by Bell's theorem and his massive support base? As to what we observe?

    PS: I very much respect your contributions here.

    Would a paper supporting my claims likely be acceptable in Independent Research?

    Would you help draft its terms in this thread so that we could all be sure of the papers to be cited and the terminology that would be used at IR?

    If you would care to help with that draft we might find that I have missed some subtlety -- or that I have not?

    If I cited just EPR and Bell and Aspect, would that be enough?

    That is all I need.

    In broad terms, would this be what you would like to see?


    Draft Abstract

    Bell's theorem is widely regarded as a serious obstacle to a local-realistic theory of quantum mechanics. Critically examining Bell's formalism and its relevance to quantum dynamics, we refute Bell's theorem and show that local beables co-encode realism (and the binding transformation symmetries of some common properties) with locality (and the causal independence of space-like separated events). Accepting Bell's assumptions and logic (but not his mathematics), we show that stochastically independent micro-probabilities lead to stochastically dependent (entangled) macro-probabilities in full accord with quantum theory and practice. Our findings scotch all Bell-based claims that local-realists must abandon local-action or physical-realism or both.

    Proposed papers to be cited

    EPR, Bell, Aspect.


    DrC, I would sincerely welcome your involvement and assistance.

    I believe the benefit for you might be "Yes, Yes" answers to your above leading questions.

    For me, I might see something with new eyes and have to admit to having over-reached.

    With best regards,

  17. Jul 1, 2010 #16
    Dear JesseM,

    Thank you very much again for this detail.

    I will attempt to answer all your questions in follow-up posts.

    I am not so good with so many words.

    I will be trying to reduce your questions to mathematical ones.

    Might it be possible to reduce your questions if I make some mathematical assertions which cut-through and satisfy you that I am a serious and valid local-realist in the full sense of all Bell's assumptions -- BUT NOT his mathematics?

    I wonder if this next statement both puts your mind at ease and is an acceptable mathematical definition of local-realism to all in this thread:

    (A) Local-realism defined: P(AB|a,b,λ1,λ2)=P(A|a,λ1)*P(B|b,λ2).

    Or, in terms of Mermin's black-boxes, which I prefer -- where G = Green, R = Red. A = {G, R}. B = {G', R'}:

    (A) Local-realism defined: P(GG'|a,b',λ1,λ2)=P(G|a,λ1)*P(G'|b',λ2).

    In studying Bell's work, I start with this definition and then let my mathematics follow the physics of Aspect's experiments.

    To help others and me who are more comfortable in mathematics with not so many words:

    Is there a next definition or formalism that you would like to see asserted by me?

    Mathematics and probability theory being our best developed languages?

    Thank you,

    Last edited: Jul 1, 2010
  18. Jul 1, 2010 #17


    User Avatar
    Science Advisor
    Gold Member

    1. Sure, Einstein believed the moon is there when we're not looking. And he did not believe in spooky action at a distance. I disagree with both of these statements.

    I believe in what I refer to as quantum nonlocality. I believe the Heisenberg Uncertainty Principle (HUP) is fundamental QM and is an expression of something important about the nature of reality. An ordinary photon is not well localized, and could be said to exist in some respects at different points in spacetime; certainly that is true of its probability wave. Ergo, it is quantum nonlocal.

    2. Math, experiments, these are good!

    3. I don't know how much I can help, since my views are opposite. I would certainly encourage you, but honestly IR might be a tough sell since your view goes against established science. You might ask what new perspective you bring in drafting something. Here are some links to Local Realist papers which you may find of interest:

    Where Bell went wrong

    Quantum entanglement, fair sampling, and reality: Is the moon there when nobody looks?

    Realistic interpretation of quantum mechancis

    I have more too, but this would probably be a good start for background. I hope this helps.
  19. Jul 1, 2010 #18
    DrC -- seriously -- when ThomasT and many others meet with you for a celebratory drink -- and pleasantly drink you round-for-round into unconsciousness -- and we all look away -- are you yet with us?

    Seriously, what happens to the moon when no one observes it?

    More later,

  20. Jul 1, 2010 #19


    User Avatar
    Science Advisor
    Gold Member

    When I am uncounscious on the floor, I won't care. :biggrin:

    As often as you look, you collapse the moon into an eigenstate. When no one is looking, nothing changes. Not really too tricky.
  21. Jul 1, 2010 #20
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook