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

Experiment from Einstein Bohr debate

  1. Nov 26, 2014 #1
    I was wondering if anyone could comment on this experiment: https://www.atom.uni-frankfurt.de/publications/files/Schmidt2013PRL.pdf, which supposedly experimentally realized Einstein's thought-experimental objection to Bohr at the Fifth Solvay International Conference on Electrons and Photons in 1927. The article I read suggested that this experiment seemed to vindicate the De Broglie-Bohm theory.
     
  2. jcsd
  3. Nov 26, 2014 #2

    atyy

    User Avatar
    Science Advisor

    No, it does not "vindicate" Bohmian mechanics. Everything in that experiment vindicates Bohr.

    You misunderstand what Bohmian mechanics is about. No experiment consistent with quantum mechanics can "vindicate" Bohmian mechanics. But that is not the point. Bohmian Mechanics is a conceptual achievement, just like Wilson's renormalization group. One could erroneously pit Bohmian Mechanics against the Copenhagen interpretation, just like one could pit the Wilsonian viewpoint against "subtracting infinities". Copenhagen and "subtracting infinities" have never been found to be inconsistent with experiment to date. So if it is the Bohmian viewpoint versus Copenhagen, or the Wilsonian viewpoint versus "subtracting infinities", the Bohmian and Wilsonian viewpoints lose. Rather, one should view the Bohmian viewpoint as showing why Copenhagen makes sense, just as the Wilsonian viewpoint shows why "subtracting infinities" makes sense.
     
  4. Nov 26, 2014 #3

    bhobba

    Staff: Mentor

    De-Broglie Bohm is deliberately cooked up to be experimentally indistinguishable from standard QM - that is why its called an interpretation.

    There is no way any experiment can vindicate it without equally vindicating standard QM, Copenhagen, or any other interpretation for that matter.

    Thanks
    Bill
     
  5. Nov 26, 2014 #4
    Theoretically, if someone were to design an experiment that could test both the position and velocity of a particle, wouldn't that have bearing on the interpretation? My understanding of Einstein's thought experiment upon which that experiment was based was that it was designed with precisely that in mind. Am I wrong about that? Or am I wrong in my understanding that this experiment is (at least alleged) to have been based on that thought experiment?
     
  6. Nov 27, 2014 #5

    bhobba

    Staff: Mentor

    But theory says you cant do that - hence its impossible - it would disprove all versions of QM.

    Einstein came up with a number of thought experiments to show it was wrong - but Bohr was up to the challenge. The last one was quite ingenious and Bohr laboured through the night to defeat it - which he did. When Bohr explained its flaw, and interestingly it involved Einstein's own equivalence principle, Einstein tipped his hat to Bohr and accepted QM as correct. From that point on he considered it incomplete not incorrect. In fact he carried a copy of Dirac's Principles of QM around with him greatly admiring the beauty and elegance of Dirac's approach called the transformation theory - which basically goes by the name of QM today.

    His later objections was via the EPR paper where he tried to show it was incomplete.

    Thanks
    Bill
     
  7. Nov 27, 2014 #6

    Demystifier

    User Avatar
    Science Advisor

    What article do you have in mind? The article you gave link to does not even mention De Broglie-Bohm theory.
     
  8. Nov 27, 2014 #7

    Nugatory

    User Avatar

    Staff: Mentor

    If the article you read said that, then it's wrong. It will be hard to say anything more without seeing the article.
     
  9. Nov 27, 2014 #8

    atyy

    User Avatar
    Science Advisor

    You have to understand that in quantum mehanics, when it is said that a particle does not and cannot simultaneously have well-defined position and momentum, the quantities being referred to are the quantum canonically conjugate position and momentum. Since it is impossible in all interpretations of quantum mechanics for a particle to have simultaneously well-defined position and momentum, it is also true in Bohmian Mechanics.
     
  10. Nov 27, 2014 #9

    Demystifier

    User Avatar
    Science Advisor

    Atyy, whenever you say the above (and you say it frequently), I think you should make a link to the discussion between you and me where we explained what that really means, because otherwise people may object that it is not true that in BM particles do not simultaneously have well-defined position and momentum.
     
  11. Nov 27, 2014 #10

    atyy

    User Avatar
    Science Advisor

    Yes, perhaps that can help with the various definitions of "position and momentum". Here is the link to our discussion for the OP: https://www.physicsforums.com/threads/how-to-talk-about-interpretations.775885/.
     
  12. Nov 27, 2014 #11

    vanhees71

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    I've one objection against this comparison. The reason is that Wilson's viewpoint on the renormalization group is a groundbreaking achievement, showing that renormalization is necessary even if no infinities occur at all. It explains, why effective QFTs work, and as far as we know today, all QFTs, also Dyson-renormalizable ones are effective theories, which have their validity domain with respect to the energy-momentum scale involved. In other words it makes the mathematical techniques, developed to tame the infinities occuring in relativistic QFT (like QED or the Standard Model), physical in the sense that this taming is very natural from the point of view of the Wilsonian interpretation of the renormalization-group equations. Nowadays RG techniques are an entire industry used in very many areas of theoretical physics: Particle physics, nuclear physics, condensed-matter physics, statistical mechanics,...

    Compared to this Bohmian mechanics is just a dead end, because it does not provide any new insight into quantum theory. The predictions of Bohmian mechanics are the same as those of non-relativistic quantum theory, and the extension to relativistic quantum field theory is, to my knowledge, not yet achieved at all. Also Bohm's "orbits" are not observed in nature. There is even an experiment by Scully et al which disproves orbits, predicted by Bohmian ideas, but that's for photons, and indeed for massless spin-1 particles the idea of orbits in position space do indeed make the least sense of all examples. So perhaps, this "disproof" is a bit unjust towards Bohmian mechanics, but I've never understood what the advantage of BM might be, except to provide some puzzling exercises for higher mathematics to find Bohm's orbits on top of the solutions of Schrödinger's wave equation.
     
  13. Nov 27, 2014 #12

    DrChinese

    User Avatar
    Science Advisor
    Gold Member

    1. Your question is usually re-phrased (to avoid ambiguity in what a simultaneous measurement of p and q would be) to: can a particle be prepared in a state of known p and q? Of course, all interpretations of QM deny this is possible. But yes, if one day someone did that, it would have a bearing.

    2. The EPR paper was similar to this. However, even in that p and q could not be predicted with certainty. The idea was that you came to that conclusion by assumption (which they felt was reasonable).
     
  14. Nov 27, 2014 #13

    atyy

    User Avatar
    Science Advisor

    I agree with you on the importance of Wilson of course! Now, about Bohmian mechanics - is it as big an achievement as Wilson's? With respect to computational technique, Wilson's contribution was more far reaching. But I think it would be wrong to say that computational technique was Wilson's big contribution. His big contribution was conceptual. My judgement is that Bohm's conceptual contribution to physics was as big as Wilson's. Bohmian Mechanics is the first known solution to the measurement problem for any area of quantum mechanics. Before Bohm, there were some who said that quantum mechanics is fundamentally at odds with a naive conception of reality, that there is no answer to the question of whether the moon is there when you are not looking. The idea behind such a claim would have been von Neumann's erroneous proof that hidden variables cannot exist. By providing a counterexample in the area of non-relativistic quantum mechanics, Bohm showed that von Neumann's proof was not only flawed, but uncompleteable. One cannot say that the measurement problem was not considered a big conceptual problem - the founders of quantum mechanics were obviously perturbed by it, which is why concepts like the Heisenberg cut and the von Neumann chain still survive in some useful form. Even Dirac mentions the measurement problem as a big problem in his 1963 Scientific American essay, and thinks it is too big to solve except that it will probably go away because quantum mechanics is probably not the final theory.

    Now, how about the problem of relativistic quantum mechanics? My feeling here is that using it as a counterexample to the Bohmian viewpoint is like saying Wilson cannot be right that our theories are only effective theories, because in the case of gravity no one has found an example of UV complete gravity, and therefore perturbative quantum general relativity is not an effective theory. Incidentally, because of Wilson, there is (probably) a Bohmian solution for some relativistic quantum field theories such as QED. QED has a lattice formulation, which should reproduce in principle all the predictions of the enormously successful perturbative QED. Lattice QED is also non-relativistic at any finite lattice spacing, and so (probably) has a Bohmian instantiation. The obvious problem for Bohmian theory is that there is not yet a lattice formulation for chiral fermions interacting with non-abelian gauge fields.

    I would like to stress that the Bohmian viewpoint is not a particular theory - but like Wilson's, it is the possibility of many completions of an effective theory. Of course one can still take an "unrealistic" view of quantum mechanics, but the important point is that it is no more necessary in quantum mechanics than in classical mechanics. If one would like an "unrealistic" view, one can do so both in quantum and in classical physics.
     
    Last edited: Nov 27, 2014
  15. Nov 27, 2014 #14

    kith

    User Avatar
    Science Advisor

    In a recent blog post, Lubos Motl linked to a paper by Jeffrey Bub who cites Bohm giving the following example: A particle in a box can be in an eigenstate with arbitrarily high energy. But although we know with certainty that we will obtain a high value in an energy measurement, the particle actually is at rest and has zero momentum. We can't measure this "true" momentum directly, it isn't even correlated with the momentum measurement outcome.

    This makes dBB very different from classical mechanics. The latter itself simply doesn't yield a motivation to abandon realism because physical quantities and measurements have a one-to-one correspondence. In QM we cannot establish such a relationship. So although dBB tells us that we don't have to, abandoning realism is well-motivated (and actually was the crucial step which led Heisenberg to the discovery of QM).

    Bueb also gives an analogy with Lorentz' ether theory. There, in order to save Euclidean geometry, length contraction is explained by hidden forces which act on objects moving through the ether. The SRT avoids these unobservable elements by embracing a different geometry.
     
    Last edited: Nov 27, 2014
  16. Nov 27, 2014 #15
    My understanding was that the experiment was attempting to measure the velocity of a particle by measuring the recoil against a screen in addition to measuring the position, as per Einstein's original thought experiment.

    This footnote to the Wikipedia article on the Bohr Einstein debates is the one that made reference to the experiment seeming to favor the Debroglie-Bohm theory: http://en.wikipedia.org/wiki/Bohr–Einstein_debates#cite_note-7
     
  17. Nov 27, 2014 #16

    atyy

    User Avatar
    Science Advisor

    I think one should distinguish between "effective theory" and "realism". What Bohmian Mechanics argues is that QM is an effective theory, just like statistical mechanics, the standard model, and gravity. Bohmian Mechanics is an argument for the notion of effective theory, and the instrumental/operational approach of Copenhagen being consistent with realism. Realism is a fundamental notion in Copenhagen, because reality lies on one side of the cut, and it is on the side of the cut which is privileged. It is because real experimental results are privileged that the wave function is considered just a tool to calculate the probabilities of real events. In Copenhagen, the classical/quantum cut can be shifted, so any thing can lie on the classical or the quantum side. It is because Copenhagen believes in real experimental outcomes that Copenhagen is agnostic about the notion of the wave function of the universe, which seems not to produce any real events (unless Many-Worlds works). In other words, agnosticism about everything being quantum is due to the belief that reality exists. So Copenhagen is an argument for effective theory, instrumental/operational theory, but not an argument for abandoning realism.

    Also, there are examples in theoretical physics where there is good motivation for inventing degrees of freedom not observable by current technology. These occur when there is a cut. The Wilsonian viewpoint tells us that quantum Einstein gravity has a cut near the Planck scale, and the theory makes sense as a low energy effective theory. It is because we have strong indications that this cut exists that the question of a UV completion of quantum gravity and string theory are well motivated, even though quantum Einstein gravity does not conflict with any known observation. The Bohmian viewpoint is that the classical/quantum cut is an opportunity, as is the cut that motivates string theory.

    To push the analogy further, in quantum gravity it is worth investigating the possibility that the cut can be removed without adding degrees of freedom - this is asymptotic safety. In quantum mechanics the corresponding investigation is the Many-Worlds approach.

    What would be nice of course is if it could be shown that quantum mechanics were a "fixed point" theory. The Valentini proposal for de Broglie-Bohm theory is, I think, in this spirit, but it is technically far from showing it. I do feel that the derivations of quantum mechanics from new axioms like Hardy's are in this spirit. In the RG approach, one writes down all possible theories consistent with the known symmetry constraints. In Hardy's approach, one writes down all possible theories consistent with known (or postulated) operational constraints. There also seem to be some types of emergent theory in which there isn't really an emergence by RG flow down a physical scale, for example the emergence of AdS Einstein gravity from a boundary CFT. But these points refer more to open questions rather than to arguments against realism.
     
    Last edited: Nov 28, 2014
  18. Nov 28, 2014 #17

    Nugatory

    User Avatar

    Staff: Mentor

    The full text of that footnote reads:
    It seems fairly bogus to me.
     
  19. Nov 28, 2014 #18

    Demystifier

    User Avatar
    Science Advisor

  20. Nov 28, 2014 #19

    Demystifier

    User Avatar
    Science Advisor

    That's not really true. All what the mentioned experiment shows is that, in some special cases, the macroscopically visible trajectory may be very different from the microscopic Bohmian trajectory. This can easily be explained by Bohmian nonlocality.

    See also
    Bohm trajectories and protective measurements?
     
  21. Nov 28, 2014 #20

    vanhees71

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    For me there's a big qualitative difference between achievements like Wilson's conceptual understanding of the RG equations, because it leads to computational techniques leading to directly measurable predictions about Nature, and that's what physics is all about. Physics does not aim at keeping our philosophical ideas (or prejudices) about the "workings" of Nature at ease but it's just describing what's objectively quantifiable and measurable in real-world experiments, and there the Wilsonian work of RG is ground breaking in many fields of physics, as written in my previous postings.

    Bohm's non-local mechanics is only an attempt to solve a metaphysical problem concerning "measurements" in quantum theory (QT for me is the mathematical scheme + the minimal statistical interpretation and no more) by introducing unobservable entities into the theory which are tailored such as to lead to the same predictions as QT. This can directly eliminated again via Occam's razor without any change in the physical description and thus is superfluous and thus also misleading as a piece of physics. Now, when I reminded you about the contradiction between trajectories predicted by the Bohm theory,

    M. Scully, Do Bohm trajectories always provide a trustworthy physical picture of particle motion, Physica Scripta T76, 41 (1998),

    and experimental facts, you exactly play the card that the Bohm trajectories are unobservable. So what purpose is it to introduce them in the first place, making QT even more complicated than it already is and giving no merits in terms of physics?
     
  22. Nov 28, 2014 #21

    atyy

    User Avatar
    Science Advisor

    But one of the things you said was that Wilson's achievement was conceptual, not just computational. The conceptual advance was that all our theories, even the standard model which to date has never been falsified, are effective theories. An effective theory means that there are unknown degrees of freedom. If physics is just being able to predict the results of experiments, and our computational methods don't require these unknown degrees of freedom postulated by Wilson, why don't you use Occam's razor and eliminate them and Wilson's concept of effective theory?

    Do you think that quantum mechanics does not have a measurement problem?

    It is the same as Wilson introducing unknown degrees of freedom by his concept of effective theory. In Wilson's effective theory and in the Bohmian viewpoint, the hidden degrees of freedom are hidden only for a given degree of experimental control. In principle, a sufficiently refined experiment will detect deviations from quantum mechanics if the Bohmian viewpoint is correct. One way to see this is that for Bohmian Mechanics to match quantum mechanics, it has to postulate an equivariant distribution of initial conditions. In other words, Bohmian Mechanics solves the measurement problem by making quantum mechanics look like statistical mechanics. However, just as we expect in statistical mechanics that equilibrium is reached from non-equilibrium, in Bohmian Mechanics equivariance is an equilibrium condition that is reached from non-equilibrium. If we can set up experiments before a quantum system reaches equivariance, then Bohmian Mechanics predicts deviations from quantum mechanics.

    Again, keep in mind that the Bohmian viewpoint is like Wilson's , and it is not the claim that there is a unique theory beyond quantum mechanics. Rather it is the idea that quantum mechanics is an effective theory, and experiments are needed for us to determine what the hidden degrees of freedom are - whether it be Bohmian trajectories in the case of the measurement problem or string theory in the Wilsonian view of gravity.
     
  23. Nov 28, 2014 #22

    Demystifier

    User Avatar
    Science Advisor

    For me, the main merit to introduce Bohm trajectories in the first place is to reformulate QM in a form that looks more intuitive. You may say that such a merit is "not physical", but I can reply that it is at least psychological (and add that ultimately any merit is psychological). But good intuition in physics helps also to make actual measurable predictions. So if the Bohmian picture in my mind helps me to make the actual measurable predictions (and it really does, even if it does not produce such an effect to you), then why should I not use this picture, at least as a thinking tool?

    Eventually, it all boils down to what one considers to be "simpler". The standard formulation of QM is simpler at the technical level (it contains a smaller number of equations), but many people find Bohmian formulation simpler at the conceptual level. It can be compared with a choice of programming computers using machine language (which is logically simpler) or a more intuitive language such as Fortran, C++, etc.

    Moreover, to expand the analogy, a hypothetical programmer who does not know how the computer really works might legitimately wonder which language is the "true" computer language, just like an actual physicist legitimately wonders which formulation of QM is the "true" formulation.
     
    Last edited: Nov 28, 2014
  24. Nov 28, 2014 #23

    kith

    User Avatar
    Science Advisor

    If we can propose such experiments, dBB is a different theory than QM and everything is fine.

    The problem is that the very concept of science -namely that we use experimental results to test hypotheses- seems to enforce the equilibrium of dBB. So while in statistical mechanics, the transition to thermodynamic equilibrium as well as sustained non-equilibrium processes can be investigated, the quantum equilibrium of dBB seems to be not accessible by the methods of science.

    So contrary to Wilson's viewpoint, which according to you points to new observable degrees of freedom, dBB is a different way to think about the known ones (and don't get me wrong, I do think it is a quite valuable one for a number of reasons).
     
    Last edited: Nov 28, 2014
  25. Nov 28, 2014 #24

    kith

    User Avatar
    Science Advisor

    Personally, I have a hard time with the consequence that particles with real wavefunctions are at rest. That an electron which is near a proton may be at rest is arguably an even bigger departure from the ideas of classical physics than giving up on realism in QM.

    Also, what is your opinion on predictions unique to dBB? Do you think that what I wrote in my previous post is accurate or do you think such predictions are likely to be made and investigated?
     
  26. Nov 28, 2014 #25

    Demystifier

    User Avatar
    Science Advisor

    Feynman said: "Physics is like sex: sure, it may give some practical results, but that's not why we do it."

    To answer your question, I would paraphrase Feynman by saying: dBB is like sex: perhaps it may give some new measurable results, but that's not why we do it.
     
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook