An argument against Bohmian mechanics?

[stevendaryl]

Fine, thus my initial question, why do everyone acts in these debates as if we were in 1928 and NRQM was the last word about quantum things. I suspect one reason is that not very many people of the already not so big set that is comfortable with NRQM knows enough about QFT.

Because the interpretation problems with quantum mechanics are really not changed by going to QFT. QFT is an extra complication that doesn't seem to make any difference.

Well, what I'm saying is that such an extension would have to renounce to the basic tenet of Bohm's original theory and therefore it would be a different interpretation altogether. If this is not a definite argument against Bohmian mechanics I don't know what it is.

What is it that you think of as the basic tenet of Bohm's theory? In NRQM, every observation boils down to (according to some people, anyway) an observation of position. So the Bohm theory makes position into a privileged variable, and describes all dynamics in terms of position.

Bell suggests that for QFT, since there is not a fixed number of particles, (and particle identity is lost completely), the more basic observation is not "the particle is at this position", but "there is some particle of type X at this position". I don't think that's a huge change.

 

[RockyMarciano]

Because the interpretation problems with quantum mechanics are really not changed by going to QFT. QFT is an extra complication that doesn't seem to make any difference.

Well, I would say that the upgrade from NRQM to QFT is quite a big change, but let's suppose the only change (not exactly trivial) was that QFT is explicitly excluding influences exceeding c speed. That change is enough to make the choice we talked about above and that you agreed with. And therefore to explicitly claim that classical determinism understood as "existence of predetermined values at all possible measurements" is false.

What is it that you think of as the basic tenet of Bohm's theory?

Classical determinism as the "existence of predetermined values at all possible measurements", which is false according to QFT as per above.
I really don't know what in this logical chain of arguments are you doubting, can you pinpoint it?

 

[stevendaryl]

Well, I would say that the upgrade from NRQM to QFT is quite a big change, but let's suppose the only change (not exactly trivial) was that QFT is explicitly excluding influences exceeding c speed.

That's barking up the wrong tree. It's true that NRQM allows for instantaneous forces, and so it allows instantaneous interactions. But in the EPR experiment, there are no interaction terms involving the two particles. So whatever mediation of the correlations might be involved, it is not particle-particle interactions.

Going to QFT doesn't make any difference, because the interactions that are excluded (FTL particle-particle interactions) weren't present in the nonrelativistic case, either.

 

[stevendaryl]

Classical determinism as the "existence of predetermined values at all possible measurements", which is false according to QFT as per above.

To exclude a Bohmian type model, you have to point to an experimental result that is inconsistent with any deterministic realistic model. That hasn't been done. I'm pretty sure it's impossible for an experiment to rule out a deterministic model. You can rule out a local deterministic model, but you certainly cannot rule out all deterministic models.

There's probably a theorem to that effect. You have a theory, QFT, that predicts a set of possible histories of observations, with certain probabilities. You can always turn such a theory into a deterministic model by introducing a hidden variable that specifies which history we're in. Then everything just plays out as specified in that history.

The complicating factor is to avoid superdeterminism. You want a theory where the choices made by experimenters are "free" while the results of their measurements are determined. I don't think that that's a lot harder. Instead of having a "history" variable, you could have a "history function", which would take as input the choices made by experimenters, and would return a history.

 

[RockyMarciano]

That's barking up the wrong tree. It's true that NRQM allows for instantaneous forces, and so it allows instantaneous interactions. But in the EPR experiment, there are no interaction terms involving the two particles. So whatever mediation of the correlations might be involved, it is not particle-particle interactions.

Going to QFT doesn't make any difference, because the interactions that are excluded (FTL particle-particle interactions) weren't present in the nonrelativistic case, either.

There is no need to invoke particles either in Bell's theorem or in QFT so I don't know how bringing in particles changes anything here. But anyway if you agree with the choice in Bell's theorem and accept experiments violating Bell's inequalities then your claiming QFT doesn't make any difference is just saying that just with Bell one can discard Bohmian mechanics.

 

[stevendaryl]

There is no need to invoke particles either in Bell's theorem or in QFT so I don't know how bringing in particles changes anything here. But anyway if you agree with the choice in Bell's theorem and accept experiments violating Bell's inequalities then your claiming QFT doesn't make any difference is just saying that just with Bell one can discard Bohmian mechanics.

No, what I agreed to was that "no FTL" plus "classical determinism" is ruled out. Bohmian models give up "no FTL", but keep "classical determinism".

The only way to rule out classical determinism is to come up with an experiment that is inconsistent with classical determinism. There is no such experiment (as far as I know). I don't think it's possible to have such an experiment. This conclusion is really independent of the details of QFT.

 

[RockyMarciano]

No, what I agreed to was that "no FTL" plus "classical determinism" is ruled out. Bohmian models give up "no FTL", but keep "classical determinism".

Ok, and QFT keeps "no FTL" and give up "classical determinism", how can they be not in conflict?

The only way to rule out classical determinism is to come up with an experiment that is inconsistent with classical determinism. There is no such experiment (as far as I know). I don't think it's possible to have such an experiment.

No, but there are experiments ruling out "no FTL" plus "classical determinism", and this in combination with QFT experiments that confirm QFT is right in its choice of "no FTL" and therefore impossible to have also classical determinism which you just agreed that having both is ruled out, clearly leave no room for Bohmian models. So I don't agree the only way to rule out Bohmian models is by a single experiment that is inconsistent with classical determinism, you can use a combination of inputs if they are logically connected.
This has no relation IMO with the possibility of superdeterminism.

 

[stevendaryl]

Ok, and QFT keeps "no FTL" and give up "classical determinism", how can they be not in conflict?

Because "no FTL" is an unobservable difference. If the Bohmian model predicted that FTL could be used for communication, then it would be in conflict with QFT, but it doesn't. There is FTL influences in a Bohmian model, but they cannot be used for communication. Those influences are unobservable.

 

[RockyMarciano]

Because "no FTL" is an unobservable difference. If the Bohmian model predicted that FTL could be used for communication, then it would be in conflict with QFT, but it doesn't. There is FTL influences in a Bohmian model, but they cannot be used for communication. Those influences are unobservable.

I'm referring to the conflict about one keeping and the other ruling out the Bohmian fundamental claim of the "existence of predetermined values at all possible measurements".

 

[stevendaryl]

No, but there are experiments ruling out "no FTL" plus "classical determinism", and this in combination with QFT experiments that confirm QFT...

The logic of this is wrong. You are reasoning that:

• Experiments confirm $A$.
• $A$ contradicts $B$.
• Therefore, experiments rule out $B$.

This is invalid. Confirmation of $A$ doesn't rule out $B$ unless it is a disconfirmation of $B$.
For experiments to rule out $B$ and not $A$, you need some observation that is predicted by $B$ but not $A$. What observation do you think rules out Bohmian-style models?

 

[stevendaryl]

I'm referring to the conflict about one keeping and the other ruling out the Bohmian fundamental claim of the "existence of predetermined values at all possible measurements".

Yes, that's a conflict between the two theories, but it isn't a testable conflict. You can never prove determinism or nondeterminism experimentally. You can only disprove particular deterministic predictions.

 

[RockyMarciano]

Yes, that's a conflict between the two theories, but it isn't a testable conflict. You can never prove determinism or nondeterminism experimentally. You can only disprove particular deterministic predictions.

I think I might understand what you mean by this. But I'm not saying that determinism is testable, philosophical positions as broad as that are not empirically testable and that is not what I'm proposing.
But the conflict between QFT and Bohmian mechanics is enough for me(and I think it should be for most) as a definite argument against Bohmian mechanics.

 

[RockyMarciano]

The logic of this is wrong. You are reasoning that:

• Experiments confirm $A$.
• $A$ contradicts $B$.
• Therefore, experiments rule out $B$.

This is invalid. Confirmation of $A$ doesn't rule out $B$ unless it is a disconfirmation of $B$.
For experiments to rule out $B$ and not $A$, you need some observation that is predicted by $B$ but not $A$.

Hmm... You might want to go through this logic analysis again.

What observation do you think rules out Bohmian-style models?

I'm not saying that a single observation rules out Bohmian models, but the combination of experiments violating Bell's inequalities plus those confirming QFT plus the cluster decomposition theorem of QFT are a convincing argument against Bohmian interpretations of QM. Since Bohmian models are at least nominally following the math of QM it is obvious that it doesn't rule it out to the extent that this combination doesn't rule out quantum theory but reinforces it empirically and theoretically.

 

[stevendaryl]

Hmm... You might want to go through this logic analysis again.

Okay, I'll go through it again. To say that an experiment confirms a theory doesn't mean that it proves the theory is true. It simply means that one prediction of the theory has been proved true.

So for example, suppose we have two theories:

1. Theory A predicts that in years divisible by 32, there is a comet above Las Vegas, but in odd-numbered years, there is no comet.
   
2. Theory B predicts that in years that are composite numbers, there is a comet above Las Vegas, but in prime-numbered years, there is no comet.

These theories contradict each other, but seeing a comet above Las Vegas in 2016 confirms both.

I'm not saying that a single observation rules out Bohmian models, but the combination of experiments violating Bell's inequalities plus those confirming QFT plus the cluster decomposition theorem of QFT are a convincing argument against Bohmian interpretations of QM.

Yes, but I'm saying that's wrong.

 

[atyy]

Where is the minimal interpretation disproven, i.e., why is it wrong to say that there is no cut? Where is it proven that the classical behavior of macroscopic objects are due to dynamics that contradicts the standard quantum dynamics? Where is the measurement problem, i.e., is there an real-world experiment that cannot be described by minimally interpreted QT?

The minimal interpretation contains a Heisenberg cut. Without the cut, the Born rule cannot be applied. Without the Born rule, the minimal interpretation is inconsistent with all available evidence.

 

[PeterDonis]

I'll just ignore the trolling fragrance of your last remarks.

Just for the record, martinbn's response was perfectly appropriate. "A" level threads assume graduate level knowledge of the subject matter.

 

[Demystifier]

You are free to make your own distinction of realism and deterrminism but in the context of Bell's theorem realism is usually referring to classical determinism.

It's true, but only for people who misunderstood Bell. Bell himself considered realism and determinism as very different categories. In fact, instead of using the word "realism" he coined his own word beable.

 

[vanhees71]

I would say that you are making it into a philosophical discussion by assuming that there is no technical answer. The question of whether something does not follow from certain assumptions is a technical question, not a philosophical question.

For me the technical answer is that the Born postulate is an independent postulate of QT that cannot be derived from the other postulates. I just accept the probabilistic nature of nature as claimed by QT. As long as there's not a deterministic theory with at least the same success as QT in describing the observable phenomena, I don't know, why I should give up the clear minimal interpretation of QT, which works for everything observed so far.

Also now I cannot see any clear distinction between the clearly defined notion of determinism and the unsharp philosophical term "realism", also not in Bell's book; he likes to talk about beables instead of observables, but that also obcures the subject. For me observable is what's measurable in the way it's done every day in labs all over the world. In the same way there's a operational definition of states: It's a preparation procedure or an equivalence class of preparation procedures which admits the definition of a statistical operator to describe the situation defined by this preparation(s), and it's content is probabilistic, and there cannot be known more than these probabilities. The probabilistic prediction can be verified on an ensemble of such preparations. That's it, and everything else beyond that is philosophical speculation rather than "hard science".

 

[vanhees71]

The minimal interpretation contains a Heisenberg cut. Without the cut, the Born rule cannot be applied. Without the Born rule, the minimal interpretation is inconsistent with all available evidence.

Again, I don't understand this. The Born rule is applied everywhere, where QT is applied to describe observed phenomena. Nobody talks about a cut. You have some preparation procedure (e.g., the bunches of protons running in the LHC) and measurement devices (e.g., the big detectors ATLAS, CMS, LHCb, and ALICE). Then as much "statistics" is taken as possible (i.e., you make a lot of pp collisions at the LHC energy and observe a lot of things with the detectors to measure spectra, cross sections, etc. etc.). I don't need new artificial words like beable or the like. Physics is just what's done in the lab, the evaluation of the data by experimentalists and the description (aka modeling/simulating) in terms of QT by theoreticians.

 

[RockyMarciano]

Okay, I'll go through it again. To say that an experiment confirms a theory doesn't mean that it proves the theory is true. It simply means that one prediction of the theory has been proved true.

So for example, suppose we have two theories:

1. Theory A predicts that in years divisible by 32, there is a comet above Las Vegas, but in odd-numbered years, there is no comet.
   
2. Theory B predicts that in years that are composite numbers, there is a comet above Las Vegas, but in prime-numbered years, there is no comet.

These theories contradict each other, but seeing a comet above Las Vegas in 2016 confirms both.
Yes, but I'm saying that's wrong.

Ok, I must not be conveying what I mean right because what you write has nothing to do with my point. I think I am making it unnnecessarily complicated, because my argument is really simple.
First it is not about different theories since Bohmian mechanics claims to be an interpretation, not a different theory from QM, so my argument is about clarifying how the Bohman interpretation of what is going on in quantum mechanics is wrong in the light of Bell experiments and the assumption that I'm justifying with QFT but could be justified in many ways, so you might think QFT is not needed to assume it and that is fine with me, of "no ftl communication".

So if one assumes "no ftl communication" and Bells theorem in the terms that I'm going to specify now with a reference that we could all agree about its use of terminology just for the sake of the argument here, it follows that the Bohmian interpretation that holds that predetermined values for mesurements results and predetermined initial positions for particles and well defined trajectories is false for QM.

My reference is the paper by Hensen recently published in Nature "Loophole-free Bell inequality violation using electron spins separated by 1.3 kilometres" .

In the second paragraph(arxiv version 1508.05949v1) of the paper one can read: "
The remarkable discovery made by Bell is that in any theory of physics that is both local (physical infuences do not propagate faster than light) and realistic (physical properties are defined prior to and independent of observation) these correlations are bounded more strongly than in quantum theory."
So here it can be seen the same split of "local realism" into "local" and "realistic" parts and the same definitions for each separate term that I've been using all along and that has been critiziced and considered as not meeting the requirements of a graduate level is also used.

Morover, given this separation of the concept of "local realistic theory" allowing the requirement of requiring separately both "physical infuences do not propagate faster than light" and "physical properties are defined prior to and independent of observation", one can conclude that IF we assume "physical infuences do not propagate faster than light" ( a reasonable assumption for serious scientists whether or not one assumes it due to QFT, but not an assumption of the theorem of Bell), then violation of the Bell inequalities reject all theories whose "physical properties are defined prior to and independent of observation".
The Bohmian interpretation claims that QM is such a theory in which "physical properties are defined prior to and independent of observation".

Hanson, Ronald. "Loophole-free Bell inequality violation using electron spins separated by 1.3 kilometres". Nature. 526: 682–686.

 

[stevendaryl]

Ok, I must not be conveying what I mean right because what you write has nothing to do with my point. I think I am making it unnnecessarily complicated, because my argument is really simple.
First it is not about different theories since Bohmian mechanics claims to be an interpretation, not a different theory from QM, so my argument is about clarifying how the Bohman interpretation of what is going on in quantum mechanics is wrong in the light of Bell experiments and the assumption that I'm justifying with QFT but could be justified in many ways, so you might think QFT is not needed to assume it and that is fine with me, of "no ftl communication".

Definitely, Bohmian mechanics has FTL influences. But my point is that rejecting something for that reason only is a philosophical preference, not empirical. I actually dislike it for that reason, as well.

 

[RockyMarciano]

Definitely, Bohmian mechanics has FTL influences. But my point is that rejecting something for that reason only is a philosophical preference, not empirical. I actually dislike it for that reason, as well.

Ok, but note that Bohmian mechanics is an interpretation, it has no empirical content, and therefore it can only be refuted in logical or philosophical grounds, made reasonable by empirical reasons perhaps but ultimately by reasoning.

 

[atyy]

Bohmian mechanics has empirical content. Bohmian mechanics disallows FTL under the assumption of a kind of equilibrium. However, for nonequilibrium, FTL and deviations from quantum mechanics are predicted.

 

[vanhees71]

Now, I'm confused. I thought that the physically testable predictions of BM are identical with non-relativistic QM. Then it's no surprise that FTL propagation of observable signals is possible since non-relativistic physics doesn't know the "relativistic speed limit" of $c$. Now, if applied to situations, where a non-relativistic QM treatment is appropriate (which holds for a large part of atomic, molecular and condensed-matter physics!) BM shouldn't be empirically testable against minimally interpreted QM, because all observable consequences are the same, and as physical theories BM an non-relativistic QM are thus indistinguishable. So BM cannot have more empirical content than non-relativistic QM. So my question is, what do you think is an observable phenomenon that is predicted to be different than what you get from standard QM.

I've no clue, how to repair BM for the relativistic case. Sometimes I hear the claim, there is something similar as for non-relatvistic ("first quantization") QM for relativistic QFT, but I've not seen this worked out in a detaile mathematical way yet.

 

[Demystifier]

Sometimes I hear the claim, there is something similar as for non-relatvistic ("first quantization") QM for relativistic QFT, but I've not seen this worked out in a detaile mathematical way yet.

You haven't seen that only because you didn't read papers that I linked several times for you.

 

[Demystifier]

I thought that the physically testable predictions of BM are identical with non-relativistic QM.

They are identical in the so-called quantum equilibrium. The standard version of BM assumes equilibrium. But Valentini speculated that Universe at the Big Bang might not have been in the equilibrium, which could have observable cosmological consequences different from the standard (inflation) theory.

 

[Demystifier]

Again, I don't understand this.

Why do I have a feeling that you think of your lack of understanding as a virtue rather than a drawback?

 

[vanhees71]

That's indeed interesting. Is there any chance to observe these consequences? It's of course very difficult to observe anything from that early epoch. The only recent hope was the BICEP2 result on the CMBR polarization but that unfortunately literally decayed to dust :-(.

 

[vanhees71]

Why do I have a feeling that you think of your lack of understanding as a virtue rather than a drawback?

I didn't think it to be a virtue. That's why I ask!

 

[vanhees71]

You haven't seen that only because you didn't read papers that I linked several times for you.

Well, I obviously didn't see the merit of these papers (as I still lack to understand what's the merit of BM in the non-relativistic context either), but if there are really observable differences between BM and QT, I have to revise my point of view and take it as another theory to be checked against experiments. That's how science works: Make an empirically testable prediction and adopt your view, as soon as the prediction is verified or falsified.

 

[Demystifier]

That's indeed interesting. Is there any chance to observe these consequences?

I don't know. Personally I don't find it very interesting, because the assumed initial violation of equilibrium is too ad hoc. The initial equilibrium can be violated in an infinite number of different ways, and Valentini assumed one particular form without any good reason.

 

[Demystifier]

(as I still lack to understand what's the merit of BM in the non-relativistic context either)

Only because you refuse to explain to yourself why do you attempt to explain the magic tricks even when you don't check your ideas experimentally.

 

[vanhees71]

Hm, out of curiosity of course, but when I can't check my ideas experimentally, I'm not doing of much value for science. Of course I mean that I make an assumption or claim which doesn't lead to observable consequences even in principle. Of course, it's a great thing that Einstein predicted gravitational waves 100 years before physicists were able to really detect them, but that it took so long was due to the technical difficulty to get the result, and gravitational waves are observable predictions of GR (of course, Einstein was himself not sure about this for a while; of course, he got the waves in the weak-field limit of GR, but then he couldn't find wave-like solutions for the full non-linear theory, and that stayed for quite a while this way, but that's another story).

 

[vanhees71]

I don't know. Personally I don't find it very interesting, because the assumed initial violation of equilibrium is too ad hoc. The initial equilibrium can be violated in an infinite number of different ways, and Valentini assumed one particular form without any good reason.

Then the question is, whether there's a reason for the universe to start in "quantum equilibrium" and not in another state. Of course, a lot about cosmology is pure speculation, like the puzzle concerning the matter-antimatter asymmetry. It's always assumed that the universe started with a symmetric state (i.e., equal amount of matter and antimatter), but that's an assumption (although a pretty "natural" one). Also inflation is an ad-hoc assumption to solve some puzzles like the flatness and horizon problems, but it's in no way verified today by observations.

 

[atyy]

I've no clue, how to repair BM for the relativistic case. Sometimes I hear the claim, there is something similar as for non-relatvistic ("first quantization") QM for relativistic QFT, but I've not seen this worked out in a detaile mathematical way yet.

One way to repair BM for the relativistic case is to use lattice gauge theory - which is non-relativistic, but for small enough lattice spacing will be consistent with experiment. Also, given the Wilsonian viewpoint, we can use lattice gauge theory as the basis of QED.

This is only a partial solution, as there is no lattice standard model due to the chiral fermion problem.