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

Back up a step, before Bell's theorem

  1. Feb 15, 2008 #1
    Hi, think of me as an armchair philosopher I suppose, who is very interested in the science of quantum right now.

    Bell's theorem led to a finding of nonlocality. I need to back up a step and ask, why. Why is it apparently believed that particles don't have a definite state until they are measured? I can easily accept that we don't know any of their state until we measure them, but why don't they have that same state regardless? I am sort of lucky right now, in that it makes perfect sense to me that entangled particles, when measured, are found to have the same spin, etc. I suppose the right answer to my question will disabuse me of my comfort level :)

    It's something to do with those interference patterns I suppose; maybe I just didn't get something. But isn't it possible for a particle to have state, and "simply" not be revealing it? Wouldn't that be a nice finding? (i.e., for those of you who don't really want to believe in nonlocality, nor in many worlds.) How is it disproven?

    Thank you in advance for any responses, that can better my understanding.
     
  2. jcsd
  3. Feb 15, 2008 #2

    Fra

    User Avatar

    philosophy

    Hello Ted and Welcome!

    Here's my personal comment.
    Others may disagree.

    IMO a key question to ask is:

    How can you possibly distinguish the "particles definite state" from your information about this state? :)

    The way I see it, the distinction your are somehow trying to make from first impressions, seems to lack realistic basis.

    All you can do without actually performing a measurement, is to ponder the set of "possible answers" you expect to get to a specific question. And this is exactly what QM is about.

    This is I think, chould be very plausible even to a philosopher.

    This may seem superficial, but if you think again. On what information are all your desicions and actions based? I presume on the information you have at hand, right? If you are badly informed, you make bad decisions. The conclusion itself might be dead one, but starting with a flawed premise, the inductions will be similarly affected.

    Perhaps the basis of physical interactions at subatomic level can be understood in terms of responses to local information? All subsystmes in nature, respond to the environment based on their expectations, not on what is correct. There isn't even a sensible notion of objective measure of correct!

    The closest you get is the "collective opinion".

    The obsession to find out what things "really are" may be totally uncalled for! Simply because MAYBE all interaction in nature, are ruled by relative information anway!! This suggest means that information is more fundamental than ontological backgrounds.

    Objectivity may still emerge as a result of everybody and everything evolving together. Clearly agreement is more constructive, disagreement is destructive and inconsistecies simply aren't preserved in evolution.

    /Fredrik
     
  4. Feb 15, 2008 #3

    Fra

    User Avatar

    An good example of the principle is game theory and economy.

    Everybody know how strange the stock market is. The exchange rate of a company stock is not determined by the "real value" of the company, or what is going on at the company floor. It is largely determined by the market actors, expectations on this company! And what they THINK the case is, and is going to be, with this company.

    So understanding modern economy, is as much to understand the expectations of the actors on each other! Its' bascailyl about the dynamics of mutual expectations.

    This insight, is IMO, no too unlike the one in modern physics, especially if you see it from the philosophical perspective.

    /Fredrik
     
    Last edited: Feb 15, 2008
  5. Feb 15, 2008 #4
    Bell's two famous papers on this should answer your question.

    Quantum nonlocality can be inferred without Bell's theorem through interpretation of what the physical meaning of the quantum theory might be.

    Quantum nonlocality might or might not correspond to some sort of superluminal motion in the physical world. No one knows, and, as far as anyone does know, this is an unanswerable question.

    Anyway, strictly speaking, quantum nonlocality refers to instantaneous-action-at-a-distance which is physically meaningless.
    None of the models which involve definite pre-measurement states can account for all of the experimental data. That is, afaik anyway, all existing classical models are inadequate to a certain extent.

    Does this mean that there are no definite pre-measurement states? As far as viable extant models are concerned ... yes. As far as nature is concerned ... no one knows -- and quantum theory doesn't provide any precise qualitative way of talking about what happens between emission and detection.


    There's just no way to know what is actually, qualitatively occuring in the deep reality beneath the level of the instruments.

    As far as anyone can be concerned there's no way to objectively talk about definite physical particle states aside from the qualitative experimental results.

    Quantum states are not (at least not necessarily) real physical states. They're part of the mathematical apparatus that's been developed to account for and relate the results of quantum experiments. They weren't intended to be a physical description of deep quantum reality -- and how or where quantum theory might actually correspond to reality beneath the instrumental level, hidden from our senses, is unknown (and if the principles of orthodox quantum theory are correct then it's unknowable).

    Quantum entanglement is (apparently) a product of physical traits shared by spatially separated disturbances. These shared traits are produced via common origin, past interaction, or the imparting of a common torque (etc.) to spatially separated disturbances.

    So, it should make sense to you that, for example, the angular momenta of opposite moving disturbances produced by the same atomic transition should be related so as to produce predictable, coincidental spin or polarization results.

    Keep in mind that certainly much of and maybe most of the quantum world has not been revealed to us.

    There are some qualitatively definite physical things happening in that deep reality. It would be absurd to say otherwise. But so far physics has a very incomplete comprehension of just what those things might be.
     
    Last edited: Feb 15, 2008
  6. Feb 15, 2008 #5

    JesseM

    User Avatar
    Science Advisor

    You could assume that, but what Bell's theorem proved was that you can't explain the statistics of different particles by assuming they have preexisting states which are influenced only by events in their past light cone--the only way to explain the statistics using hidden states is to imagine that a measurement of one particle can change the other particle's hidden state in an FTL manner (as I understand it this is essentially what's going on in the Bohmian interpretation of QM). For more on how Bell's theorem proves this, you might want to read this thread and this one.
     
  7. Feb 15, 2008 #6
    Wow, everyone, what awesome replies! It's great to find such an active site, that shares a key interest. I will enjoy reading each one, carefully, and readng those links also. It'll take me a little while.
     
  8. Feb 16, 2008 #7

    DrChinese

    User Avatar
    Science Advisor
    Gold Member

    Welcome to PhysicsForums!

    Ted, what you are describing is called "Realism". This was a position held by many scientists, including Einstein. Many of those same scientists, also including Einstein, believed that "action-at-a-distance" was not position. They therefore believed in "Locality". This was pre-Bell, of course, as your thread title mentions.

    But it turns out that the "realistic" assumption has ramifications that were overlooked prior to Bell's 1964 paper. Consider the idea that photon spin is definite, but not revealed, as you describe. Then you would probably agree to something like the following:

    Where Match(angle, angle) represents the likelihood of getting the same answer - either two "yes" or two "no". MisMatch(angle, angle) represents the likelihood of getting the different answers - one "yes" and one "no".

    [1] Match(0 degrees,67.5 degrees) >= 0%

    That's an easy one to agree to, all we are saying is that sometimes they will match and sometimes they won't.

    If you can accept that, and you are a "realist", then you also believe that there is a definite, but unrevealed value, for 45 degrees as well. You believe in the existence of definite values for 3 angles (or more) simultaneously. Occasionally, the value at that angle (45 degrees) would either match the other two, or it would be different. Even if it was rarely different than the other two, its likelihood would always be in the range of 0% to 100%. Again, seems very simple and innocuous.

    [2] Match(0 degrees,67.5 degrees) where also MisMatch(0 degrees, 45 degrees)
    >= 0%


    So all that is saying is that the various combinations of selecting 3 angles to consider are greater than or equal to 0%. That should be true of any combination of possible outcomes. That is, if you are a realist.

    I won't repeat Bell's proof here - you can read it elsewhere or see the exact proof at my website - but it turns out that:

    If Quantum Mechanics is correct, then the value of [2] above is actually LESS THAN ZERO, a seemingly impossible result. In fact the value turns out to be -10.36%. Experiments show that QM is correct. Therefore, you must reject the idea that there are definite values at 3 angles simultaneously.

    Bell's Theorem and Negative Probabilities shows the proof, which is fully equivalent to the standard Bell proof.
     
    Last edited: Feb 16, 2008
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Back up a step, before Bell's theorem
  1. Bell's theorem (Replies: 264)

Loading...