Experiment from Einstein Bohr debate

[jbmolineux]

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.

 

[atyy]

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.

 

[bhobba]

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

 

[jbmolineux]

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?

 

[bhobba]

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?

But theory says you can't 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

 

[Demystifier]

The article I read suggested that this experiment seemed to vindicate the De Broglie-Bohm theory.

What article do you have in mind? The article you gave link to does not even mention De Broglie-Bohm theory.

 

[Nugatory]

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.

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

 

[atyy]

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?

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.

 

[Demystifier]

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.

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.

 

[atyy]

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.

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/.

 

[vanhees71]

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.

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 occurring 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.

 

[DrChinese]

1. 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?

2. My understanding of Einstein's thought experiment upon which that experiment was based was that it was designed with precisely that in mind.

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).

 

[atyy]

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 occurring 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.

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.

 

[kith]

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.

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.

 

[jbmolineux]

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

 

[atyy]

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.

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.

 

[Nugatory]

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

The full text of that footnote reads:

<If more experiments begin to confirm this view &ndash; that a particle is in fact equally present (so, in a way, spread) in all the points of an orbital and that all such points indeed interact with the environment &ndash; it could possibly undermine the [[Copenhagen interpretation]] and favour the [[De Broglie-Bohm theory]], which has assumed such spread of properties (e.g. mass, charge). A deterministic theory (through a kind of guiding towards the probability centre) could possibly explain better and more universally than a fully indeterministic one how can a particle (i.e. a common wavefunction for a continuum of points) remain coherent in time.

It seems fairly bogus to me.

 

[Demystifier]

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/.

And the crucial posts are #26, #31, #32.

 

[Demystifier]

There is even an experiment by Scully et al which disproves orbits, predicted by Bohmian ideas

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?

 

[vanhees71]

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?

 

[atyy]

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.

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?

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.

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

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?

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.

 

[Demystifier]

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?

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.

 

[kith]

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.

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).

 

[kith]

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).

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?

 

[Demystifier]

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?

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.

 

[vanhees71]

Well, sex is not always necessarily good, and Bohmian mechanics is bad sex ;-).

 

[Demystifier]

http://en.wikipedia.org/wiki/De_gustibus_non_est_disputandum :)

 

[jbmolineux]

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/.

I read that post, and I find a bit of a paradox in what seems to be the consensus of the community.

(1) On the one hand, it is being said that interpretations are a "psychological preference, and not science"
(2) On the other hand, it is being said that newcomers/layman don't have enough knowledge to choose the interpretation

It seems to me that there are only two possibilities. Either:

• the interpretations stem from the theories and experiment, and are matters of science (and thus should be at least in theory falsifiable), and are in this case not within the purview of non-scientists to comment on...or….
• The interpretations do NOT stem from the theories / experiments, are not falsifiable by any possible future experiments, and are therefore not matters of "science" but are instead matters for personal psychological / philosophical preference (I believe "taste" is the term Heisenberg used), but which therefore the scientist has no more "expertise" than the reasonably-informed layperson (assuming such a layperson is willing to trust the scientists so far as the actual science is concerned)

Which is it?

I was also wondering what the general view is on Karl Popper's postscript on QM. Is there consensus on the reasons it is considered invalid? (Which I assume is the case based on my understanding of the current situation.)

Also, would someone be able to give me a succinct layman's version (if such is even possible) of why dBB is considered to fail?

Thanks!

 

[Demystifier]

I read that post, and I find a bit of a paradox in what seems to be the consensus of the community.

(1) On the one hand, it is being said that interpretations are a "psychological preference, and not science"
(2) On the other hand, it is being said that newcomers/layman don't have enough knowledge to choose the interpretation

It seems to me that there are only two possibilities. Either:

• the interpretations stem from the theories and experiment, and are matters of science (and thus should be at least in theory falsifiable), and are in this case not within the purview of non-scientists to comment on...or….
• The interpretations do NOT stem from the theories / experiments, are not falsifiable by any possible future experiments, and are therefore not matters of "science" but are instead matters for personal psychological / philosophical preference (I believe "taste" is the term Heisenberg used), but which therefore the scientist has no more "expertise" than the reasonably-informed layperson (assuming such a layperson is willing to trust the scientists so far as the actual science is concerned)

Which is it?

Unfortunately, there is no simple answer. Interpretations both "are" and "aren't" a part of science. To fully understand an interpretation, you certainly need to have some knowledge of science. On the other hand, interpretations also have something additional (let us call it "philosophy") which science in a narrow sense does not possess. The simplest (but not most accurate) way to explain it is through the formula

interpretation = science + philosophy

But one should also add that there is no strict boundary between science and philosophy, nor there is a strict definition of either science or philosophy. Every scientist uses some philosophy to make some interpretations, but different scientists differ in how much they do it.

 

[atyy]

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).

Let's take the specific case of quantum gravity. The Wilsonian viewpoint is that new degrees of freedom enter near the Planck scale, yet they have not been observed, and may never be observed. If string theory is scientific, then so is dBB. But in both cases, apart from the specific theory (string theory, dBB), the lesson is the same even if no deviations from present theory are observed: quantum general relativity and quantum mechanics make sense as effective theories, even though they have cuts.

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

Just like the large extra dimensions of string theory, there may be manifestations of dBB at low energies: http://arxiv.org/abs/1306.1579. However, I think both the low energy stringy scenarios and low energy dBB scenarios are unlikely.

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.

The general prediction of the Wilsonian and the dBB viewpoints is that the standard model and quantum mechanics are incomplete. It doesn't mean that the next theory will be a complete theory or a realistic theory, since the general viewpoint is that we have sequence of effective theories. For example, the theory that replaces the standard model might be the MSSM (not sure that's still in play, but you get the idea), and the next theory that replaces QM might be another operational theory like http://arxiv.org/abs/1105.4464.

In a sense both the Wilsonian and the dBB viewpoints have already been vindicated. In the Wilsonian case, there are massive neutrinos which will probably make the new standard model non-renormalizable. In the case of dBB, there is lots of evidence that reality exists - Copenhagen depends on it, and privileges it.

 

[atyy]

(1) On the one hand, it is being said that interpretations are a "psychological preference, and not science"
(2) On the other hand, it is being said that newcomers/layman don't have enough knowledge to choose the interpretation!

There are two sorts of interpretations: pure and impure. A pure interpretation produces predictions identical to those of quantum mechanics. An impure interpretation is consistent with quantum mechanics in a large regime, but predicts deviations from quantum mechanics beyond that regime.

Given any two pure interpretations, it is a matter of aesthetics, and laymen (like me - I'm a biologist) can indeed pick what they like, and there is no problem with changing one's favourite interpretation every second. However, the big caveat is that there is only one known pure interpretation, which is Copenhagen. Given that only Copenhagen has been completely successful in matching all known observations, laymen should not pick dBB or MWI as alternatives. I don't believe anyone has shown a dBB standard model of particle physics, and there is still no consensus even among proponents as to how probabilities enter MWI. So you can pick any pure interpretation you like, but there is only one available.

Now I am going to get myself into trouble, but I am going to try this argument. In this thread, I have been arguing for dBB as an impure interpretation, and not as any specific theory, but as a viewpoint. Given that I have argued for Copenhagen first and foremost, what is the role of the dBB viewpoint? Here by Copenhagen I mean an interpretation with a classical/quantum cut, and agnosticism about the reality of the wave function., ie. Copenhagen is the operational view that we can divide the world into measuring apparatus/quantum system, and that the wave function is a tool to calculate the probabilities of measurement outcomes. dBB-Copenhagen is the flavour of Copenhagen which says that there is nothing mysterious with Copenhagen, and Copenhagen makes no challenge to naive realism. dBB-Copenhagen says that Copenhagen makes complete sense as an effective theory, and is consistent with the moon existing when we are not looking at it. Some people may prefer unreal Copenhagen, or consistent histories Copenhagen, or quantum Bayesianism Copenhagen.

Unlike a pure interpretation, an impure interpretation can pass from aesthetic preference to being preferred by experiments, because an impure interpretation predicts deviations from quantum mechanics. However, there are at present no deviations from quantum mechanics, so we have to go with some flavour of Copenhagen.

 

[vanhees71]

An interpretation of a physical theory in the scientific sense is the association of elements of the mathematical formalism (experience shows that there are no valuable non-mathematical theories) with observables in real nature. The theory predicts relations between observables that can be checked by experiment. Only that parts of the theory are science which lead to predictions for observables in Nature that can be falsified by experiment, if the theory turns out not to be an accurate description of Nature.

Another meaning of interpretation of a physical theory is metaphysical or philosophical and not part of science. It's the association of physical theories with a certain world view and just subject to private believes of individuals. They cannot be disproven by experiment, even in principle.

 

[jbmolineux]

Atyy, if CI is the only interpretation consistent with known observations is Copenhagen, why do only 42% of physicists in this poll (http://arxiv.org/pdf/1301.1069v1.pdf) support it? That question is not meant to be rhetorical, but actually seeking information. Is this poll not a good representation of the spectrum of opinion?

Do others agree that the CI is the only interpretation consistent with known experiments?

You seem to be using the words "quantum mechanics" in a way that is at time synonymous with "known experimental results," and at another time in a way that is almost synonymous with "Copenhagen Interpretation." Thus you say that "a pure interpretation produces predictions consistent with quantum mechanics" (by which I assume you mean "known experimental results in quantum mechanics"). But later you say "since there are no deviations from quantum mechanics we have to go with some flavor of Copenhagen." If you meaning is that same as above, this statement would translate into "among the known experimental results of QM, there are no known deviations from the experimental results of QM, so we have to go with some flavor of Copenhagen." That would be quite circular, so I'm assuming you mean something different here.

Vanhees, isn't evolution a non-mathematical theory? Can it be falsified by experiment?

Demystifier, from your formula, by the subtractive property of equality…

science = interpretation - philosophy

That can't be right! =)

No takers on the Popper question or the one about why dBB is considered to fail?

Thanks everyone!

 

[atyy]

Atyy, if CI is the only interpretation consistent with known observations is Copenhagen, why do only 42% of physicists in this poll (http://arxiv.org/pdf/1301.1069v1.pdf) support it? That question is not meant to be rhetorical, but actually seeking information. Is this poll not a good representation of the spectrum of opinion?

Well, many physicists are misinformed - everyone should support Copenhagen. :) Actually, if you add all the Copenhagen-style interpretations in the poll (Copenhagen + Information-based + Quantum Bayesianism), the percentage rises to 72% (edit: oops, the maximum possible is not 100%). The major alternative to Copenhagen in that poll is MWI with 18%. However, even proponents of MWI will agree that there is no consensus that the technical problems are all solved, as you can read for example in http://www.preposterousuniverse.com...ion-of-quantum-mechanics-is-probably-correct/ and http://arxiv.org/abs/0712.0149. Copenhagen itself does have the measurement problem, so it really is not an interpretation, especially since by quantum mechanics I do mean the Copenhagen interpretation. Anyway, these polls are just physicists having fun and should not be taken too seriously, as Leifer points out http://mattleifer.info/2013/09/17/q...surveys-on-the-foundations-of-quantum-theory/.

You seem to be using the words "quantum mechanics" in a way that is at time synonymous with "known experimental results," and at another time in a way that is almost synonymous with "Copenhagen Interpretation." Thus you say that "a pure interpretation produces predictions consistent with quantum mechanics" (by which I assume you mean "known experimental results in quantum mechanics"). But later you say "since there are no deviations from quantum mechanics we have to go with some flavor of Copenhagen." If you meaning is that same as above, this statement would translate into "among the known experimental results of QM, there are no known deviations from the experimental results of QM, so we have to go with some flavor of Copenhagen." That would be quite circular, so I'm assuming you mean something different here.

Yes, it is circular. I do take Copenhagen to be the default interpretation of quantum mechanics.

No takers on the Popper question or the one about why dBB is considered to fail?

So far there is no demonstration that dBB is able to reproduce all of the standard model of particle physics. In particular, it is unknown if there is a Bohmian explanation of chiral fermions interacting with non-abelian gauge fields.

But in non-relativistic physics dBB does work. And I don't think it is unreasonable to hope for a dBB theory that can include chiral fermions and non-abelian gauge fields, since there are lattice theorists working on the problem. So for the sake of discussion, let's assume dBB works. I would argue that it is against the spirit of dBB to believe in it without direct evidence for the hidden variables. The reason is that there are demonstrably many dBB-like theories consistent with non-relativistic quantum mechanics, and to believe in dBB without any violation of quantum mechanics would make dBB unreal, since it is an arbitrary picking of one of many possibilities without experimental evidence. The whole spirit of dBB is that reality exists outside of your head, whereas picking one arbitrarily without experimental support for it against other hidden variable theories would make dBB exist only in your head.

However, I do think one can believe in a Bohmian viewpoint that hidden variables exist, without committing to any particular hidden variable theory until it is directly detected. Till then I believe the Bohmian viewpoint would be to view Copenhagen with wave functions that are not necessarily real as an excellent effective theory consistent with naive realism. Good Bohmians should support Copenhagen. :)

 

[jbmolineux]

Atyy, I take CI to involve essentially the idea that particles do not exist in the classical sense, and therefore do not have a position and velocity simultaneously, and that therefore reality is ultimately not deterministic but probabilistic.

My understanding is that it is also possible to interpret the equations and experiments in other ways that would preserve (non-local) determinism --either in the sense that there are "hidden variables" or that the particles have a realistic "spread."

But it is (as I understand it) either actually impossible, or at least beyond our current ability to design an experiment that can test between the two interpretations (which I believe is largely what makes them "interpretations" and not science). From what I understand, the Bell test experiments ruled out "local deterministic" interpretations, but either the CI, or a non-local determinism would both be possible interpretations that match the experimental data. Is that correct?

I actually agree with what I understand to be the conclusions of what you are saying. If you take "Copenhagen Interpretation" to mean "that which matches known experimental results," then the CI is the only reasonable interpretation. And if you then allow for the CI to be "agnostic about the reality of the wave function," and even agnostic about the possibility of hidden variables --then that seems very close to the view that makes the most sense to me in my current limited state of knowledge.

But in a previous thread you said, "It is impossible to know both the position and momentum of a quantum particle, because it does not and cannot have both position and momentum simultaneously." My understanding was that some non-CI interpretations would not necessarily agree with that--or at least that it would be possible to believe that the particles DID have a momentum and position without contradicting any known experimental results. Einstein was at least reasonable familiar with most of the relevant theoretical results, and my understanding was that he did not agree with that. Moreover, it seems to be actually impossible to design an experiment that PROVES that particles don't have a momentum or velocity, even if it IS possible to prove that it's impossible to MEASURE both a particle's momentum and velocity.

But perhaps I am missing something. My understand was that the Heisenberg accomplished the latter, and not the former. Am I wrong? I know I have been here directed to "nut it out myself" on this point, and try to understand how uncertainty follows from the postulate of QM, which I have every intention of doing. I have already ordered the books I was directed to order by Bill, but I live in Nepal and there is no consistent reliable way to ship things here! In the meantime, could someone clarify for me which of the following is the case?

(a) from the postulates of QM it is possible to prove that it is impossible to measure both a particle's momentum and velocity, or…
(b) from the postulates of QM is it possible to prove that a particle does not have both momentum and velocity (and if it is this former, could someone attempt to give me a layman's understanding of how you could even theoretically prove the non-existence of something)

In my current state of ignorance I can't help but feel that there is a double-meaning in the way you're using the term "CI." When it comes to arguing between interpretations, the CI is defined as "that which matches all know experimental data." But when it comes to applying that interpretation to make statements about reality, you're bringing back in the historical assumptions of the Copenhagen Interpretation to mean something that one of the most brilliant physicists of the century never accepted (leading me to have some confidence that it is possible to understand the math and experiments and still not believe in those assumptions).

Surely I am missing something! What is it?

 

[atyy]

I actually agree with what I understand to be the conclusions of what you are saying. If you take "Copenhagen Interpretation" to mean "that which matches known experimental results," then the CI is the only reasonable interpretation. And if you then allow for the CI to be "agnostic about the reality of the wave function," and even agnostic about the possibility of hidden variables --then that seems very close to the view that makes the most sense to me in my current limited state of knowledge

Yes that is close to what I mean by CI. It is true that CI is not a well-defined term, and different people use it differently. I mean CI as a sort of minimal interpretation that should be consistent with most or all other possible interpretations. To be specific, I am thinking of something like the interpretation given at the start of Landau and Lifshitz's textbook on quantum mechanics.

But in a previous thread you said, "It is impossible to know both the position and momentum of a quantum particle, because it does not and cannot have both position and momentum simultaneously." My understanding was that some non-CI interpretations would not necessarily agree with that--or at least that it would be possible to believe that the particles DID have a momentum and position without contradicting any known experimental results. Einstein was at least reasonable familiar with most of the relevant theoretical results, and my understanding was that he did not agree with that. Moreover, it seems to be actually impossible to design an experiment that PROVES that particles don't have a momentum or velocity, even if it IS possible to prove that it's impossible to MEASURE both a particle's momentum and velocity.

Yes, this is technically tricky, but it is consistent with the meaning of CI I have just outlined. The main problem causing the confusion is that there are several different definitions of position and momentum. A particle may not have a certain type of position and momentum, but it can of course have another type of position and momentum. In quantum mechanics, the most usual definition of position and momentum are the non-commuting canonically conjugate position and momentum. It can be shown that in general and for all hidden variable interpretations, including any of the many possible variants of the Bohmian interpretation, that a particle cannot have simultaneously well-defined non-commuting canonically conjugate position and momentum. In Bohmian mechanics, it is possible for a particle to have a different sort of position and momentum, defined using the Hamilton-Jacobi formalism.

 

[jbmolineux]

It can be shown that in general and for all hidden variable interpretations, including any of the many possible variants of the Bohmian interpretation, that a particle cannot have simultaneously well-defined non-commuting canonically conjugate position and momentum. In Bohmian mechanics, it is possible for a particle to have a different sort of position and momentum, defined using the Hamilton-Jacobi formalism.

This is over my head at the moment, but let me try to understand. You're saying that which of my two options (a or b) is the case depends on how you define position and momentum, right?

But in the case of "well-defined non-commuting canonically conjugate position and momentum," it literally does not have both, correct? Does the concept of "well-defined non-commuting canonically conjugate position and momentum" involve defining position and momentum by measurability?

 

[atyy]

This is over my head at the moment, but let me try to understand. You're saying that which of my two options (a or b) is the case depends on how you define position and momentum, right?

Yes.

But in the case of "well-defined non-commuting canonically conjugate position and momentum," it literally does not have both, correct?

Yes, in Copenhagen, and in any hidden variable interpretation.

Does the concept of "well-defined non-commuting canonically conjugate position and momentum" involve defining position and momentum by measurability?

Something like that. Here by position and momentum I mean non-commuting canonically conjugate position and momentum. Let X stand for a quantum observable, say position or momentum. In quantum mechanics, an accurate measurement of X is a procedure that yields a certain distribution of outcomes for a given wave function. An inaccurate measurement of X is a procedure that yields a slightly different distribution of outcomes on the same given wave function. Position and momentum cannot be simultaneously accurately measured, because their respective measurement procedures require different setups which cannot be put in the same location. In addition it can be shown that position and momentum cannot be simultaneously well-defined properties of a particle in any hidden variable interpretation (there are a couple of additional assumptions needed, which are usually fulfilled). Quantum position and momentum are defined in this way because in the classical limit of quantum mechanics these become the position and momentum of the classical Hamiltonian formalism.

 

[vanhees71]

Yes that is close to what I mean by CI. It is true that CI is not a well-defined term, and different people use it differently. I mean CI as a sort of minimal interpretation that should be consistent with most or all other possible interpretations. To be specific, I am thinking of something like the interpretation given at the start of Landau and Lifshitz's textbook on quantum mechanics.

If you understand this as "Copenhagen Interpretation", I'm also a follower of CI, but as you said, CI is not a clearly defined interpretation, because there are already profound differences between Heisenberg's and Bohr's point of view which where closely collaborating in the early years of QT. I prefer to call, what I think is the only scientifically founded interpretation, the "Minimal Statistical Interpretation", which can also be subsumed under the various flavors of CI.

The poll by Schlosshauer et al is for sure not representative, because it was taken among "33 paticipants of a conference on the foundations of quantum mechanics". I think this is a very small subject of specialist thinking (not to say speculating ;-)) about the "foundations of quantum mechanics". In my experience these specialists tend to be quite far from the main-stream use of QT in the larger scientific community which is more interested in the physics than the metaphysics of the subject (particle and nuclear physicists like myself or the condensed matter physicists). I'd guess, without having made a poll among my colleagues, most of them follow the Minimal Interpretation or more something like the "shut up and calculate/measure" interpretation. They just use quantum mechanics as a theory to predict probabilities like cross sections of reactions occurring in particle/nuclear collisions or the properties of quantum many-body systems (usually in or close to thermal equilibrium) and then check these predictions by experiment. I'd count this vast majority of quantum practitioners as Copenhagenians in the broad sense as atyy put it above.

 

[Ilja]

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.

Regarding relativistic QFT, you are wrong, the first example of a relativistic field theory in the Bohmian approach can be found already at the original paper (for the EM field). Essentially one can consider Bohmian versions in a straightforward way whenever the Hamiltonian is quadratic in the momentum variables. Which is not a problem in QFT where the Hamiltonian of a field is usually a variant of $\pi^2+\partial_i\phi^2+V(\phi)$.

"Disproofs" of de Broglie-Bohm theory are usually the consequences of incorrect understanding of what the theory tells. The proponents of this theory are sometimes not innocent in this connection. The greatest error is IMHO the focus on many particle theory (one should start with the general variant of a general configuration space, and not a special choice of the configuration space).

The advantages of BM are quite obvious. It is realistic (in the precise meaning used by Bell in his theorem), deterministic and causal. Different from many words and inconsistent histories (SCNR) it makes sense. Different from Copenhagen, it has no distinction between classical and quantum domain, but covers above domains. It does not have to focus on the uncertain notion of measurement, because measurements are described by the same math as usual evolution. And it has no measurement problem, because the collapse appears as a natural consequence of the general formula for definition of an effective wave function for a subsystem.

Then, it allows to derive (via Valentinis subquantum H-theorem) the quantum equilibrium, and, thus, the Born rule.

And, no, you cannot eliminate the Bohmian trajectories - because everything you see are trajectories. And you, yourself, are also described by some trajectory. All you can, is to eliminate it quite inconsistently from some artificially subdivided part of the universe named "quantum part". This would introduce an additional artificial subdivision (quantum vs. classical parts of the world) with unclear connections and different equations, and, even worse, unclear boundaries between them. So, removing the trajectories from dBB gives horrible results - the only point in favour of this horror is that we are already used to this horror, have even a name for it (Copenhagen interpretation) and are used to "shut up and calculate".

 

[Ilja]

Well, sex is not always necessarily good, and Bohmian mechanics is bad sex ;-).

Couldn't resist to consider the question which interpretation can be associated with which type of sexual behaviour:

Many worlds: Total promiscuity, you cannot resist even sex with the ugliest person you can imagine, every imaginable perversion really happens.
Consistent histories: Many different sex with different persons at the same time, telling everybody a consistent history about fidelity.
Copenhagen: Man and women live separately, sometimes they meet, have short sex, nobody really knows what the other thinks about this.
Bohm: Back to classical laws - sex as part of the marriage, with love, fidelity and all this. But, let's not forget, in some quantum corners some sex happens which is not classical at all, and seems to be against the law of the majority.
Transactional interpretation: The name tells it - prostitution.
Bayesian interpretation: All what counts are the subjective expectations - no real sex, only nice dreams alone at home.
Ithaca: All what counts is information - no real sex, but a lot of pornography.

So, your choice ;-)

 

[Dale]

I think this thread has run its course.