The notion of locality in (Quantum) Physics should be clearly defined

[Simple question]

QM is described by usual partial differential equations like the Schrödinger equation for the wave function. It's not a stochastic differential equation

Nice ! You finally walked-back your claim that "nature is fundamentally random". Because if the theory is not, you then have no mathematical way to prove it.

The meaning of the state is probabilistic.

So is game theory or statistical mechanics. Still, it does not make those "fundamentally random" nor "mysterious".

That's true. There are attempts to extend the quantum formalism with some stochastic collapse mechanism, but that's not QM anymore but a new theory. There's, however, not the slightest hint that such an alteration is needed anywhere.

You yourself provided those hints. You are just stuck in the past and you cannot accept that Nature behaves as she does and not as you want.

Don't interpret something into what I'm saying, which I never said. QFT as any QT is not realistic, i.e., within this theory not all observables always take determined values.

Everyone knows that QFT is unrealistic it has been mathematically proven. You just don't understand what that means, which is: you cannot make claim about NATURE using it. You can barely describe the most basic setup with it (see **)

Also it's weird to claim that standard local relativsitic QFT were wrong

Another falsehood. You would be hard pressed to quote any such "claim" from anyone on this thread.
You are wrong, not the theory. You seem to identify yourself with it. That is weird.

while in fact it's the most successful class of theories ever discovered!

"class ?" "successful ?". Unsubstantiated opinion holds no water on scientific forum.

Experiment shows that also local relativistic QFT is correct

Experiment within its domain only.

and this excludes spooky actions at a distance by construction.

No it does not. Experiment show that spooky correlation at a distance exist. So the theory does not "exclude" that, because it would mean the theory is wrong or incomplete.
Science prefer experiment over theory, by construction

That's a mathematical property of the theory and cannot be argued away by some "interpretation" gibberish.

So stop that gibberish. Mathematical property are just that, not fact about Nature.

It's among THE key features, and it predicts from the start very well established facts about nature like the CPT symmetry and the relation between spin and statistics.

Cool. So micro-causality has its purpose. I am really not surprised. So how to use it in the simplest setup (see **) ? or in entanglement cases, swapping maybe ?

I've no clue what you mean in #71. Whether I use one equipment to run an experiment 10000 times or whether I build 10000 different equipments doesn't make any difference. It's just preparing large enough ensembles to have a high significance in my statistical tests of the probabilistic predictions of Q(F)T.

** It doesn't surprise me. You are not interested by experiments nor how those can be described by a theory. Not even in principle. You wrote:

The detectors don't negotiate anything. It's just the interaction of the photon with the material around it. Where it will be detected is random,

It's just a contradiction. Those 10000 labs are "prepared" in different light cone. Your own idiosyncratic miss-use of QFT explicitly forfeit its predictive power because you assume micro-causality and local interaction apply. Always everywhere.
All of these events are space like, so what say you ?
Hint: You should have stuck with the minimal (not gibberish) interpretation which is:

Natural science is not for explaining the world, and especially not describing at best as possible with mathematical models. You don't aim to describe nature because it is un-real. And just shut-up and calculate probabilities, even though nature deal in events, not probability of platonic ensemble.

And if one want to shoot only one photon some place, or entangle of few QBit in a QComputer, it is not worthy of "science". The least is the best it can do.

 

[Simple question]

You cannot assume any such thing. All you can assume is what I explicitly said: that for the case of spacelike separated measurements, the predictions of QFT are the same as those of non-relativistic QM.

But that is not the point ! I agree with that perfectly: QFT have no additional claim to make about entanglement of spacelike measurement. @vanhees71 is not agreeing with this. So why do you say *I* made too strong claim ?

The ambiguity is probably on my side because if anything I said that QFT as no claim (note: not contradictory claim) to make about Bod and Alice's local mater and field. They are space-like, so my error is probably to think entanglement swapping involve non-commuting observable.

Anyway the point is that only @vanhees71 bring up micro causality as the only way to (and with godly precision) to describe this, while this has no bearing on Bell's locality tests. How could this be an "interpretation issue" instead of "a plain mistake ?"

There is no such thing except in "objective collapse" interpretations, which, as I think has already been noted in this thread, actually become different theories (i.e., different math from the standard math of QM, making different experimental predictions) when developed fully.

I though so. I doubt I would ever be able to understand how this would explain entanglement "speed". But that's for another thread.

 

[gentzen]

and this excludes spooky actions at a distance by construction.

No it does not. Experiment show that spooky correlation at a distance exist. So the theory does not "exclude" that, because it would mean the theory is wrong or incomplete.
Science prefer experiment over theory, by construction

Because "spooky action at a distance" (or the German equivalent) was coined by Einstein, its meaning is much more technical and fixed, than you seem to assume. This meaning simply does not include nonlocal correlations. (If you want, include the Bohmian type of nonlocality in it, and all the theories/interpretations with an explicit collapse of the wavefunction, but mere "correlations" are too "passive" as to be termed "action" by Einstein.)

"class ?" "successful ?". Unsubstantiated opinion holds no water on scientific forum.

So stop that gibberish. Mathematical property are just that, not fact about Nature.

I get the impression that you are trolling. But maybe my impression is wrong, and this is just your way to have a lively discussion.

 

[PeterDonis]

Thread closed for moderation.

 

[PeterDonis]

Natural science is not for explaining the world

An off topic subthread that was more or less spawned by this comment has been deleted. Please keep discussion in this thread focused on the specific topic of how "locality" is defined in QM. More general discussion of what science is "for" and how it should be done belongs in a separate thread in General Discussion if anyone wants to pursue it.

 

[PeterDonis]

Thread reopened.

 

[PeterDonis]

Brukner and Zeilinger just set the record straight by using the minimal statistical information. That solves all pseudo-problems.

This claim, of course, requires that you adopt the particular interpretation you describe. It should be obvious to you by now that not everyone accepts that interpretation. And the guidelines for this forum make clear that no particular interpretation can be asserted to be "correct". Everyone in the discussion must accept that there are different interpretations of QM that say different, sometimes incompatible things, and that that fact is not going to change as a result of any discussion here.

The only problem that remains is that some philosophers cannot accept that Nature behaves as she does and not as they want. They are still confined in their "classical worldview". That's all that's left.

This is the kind of claim that the guidelines for this forum do not permit. The fact that you prefer a particular interpretation does not allow you to claim that anyone who doesn't accept it is not accepting how Nature behaves. The only things we know about how Nature behaves are the things we see in experiments, and all QM interpretations agree on all experimental predictions.

 

[Demystifier]

A comment on the title of this thread. It's not sufficient that locality (or anything else for that matter) is defined clearly. It also must be defined adequately. The standard QFT definition is clear, but in the context of quantum foundations it is not adequate.

The standard QFT definition of locality talks about observables, not about events (see my first post in this thread). For practical purposes that's perfectly OK, but in quantum foundations one goes beyond pure practicality and tries to understand how exactly the events happen. For that purpose one needs a definition of locality that more directly deals with events, which is why the standard QFT definition is not adequate.

 

[Demystifier]

realistic (which means that all observables always take determined values).

This is an utter nonsense, at least 3 levels.

1. Observable is a self-adjoint operator, an operator cannot take a value (if by "value" one means a number). Operator is a map from a Hilbert space to itself. It can be represented by a matrix. But it cannot be represented by a number, it cannot "take a value".

2. It is not clear what "determined" means. Deterministic, as opposed to stochastic/probabilistic? If so, then it's not what realistic means. Or maybe definite, meaning defined (even when it's not measured)? Yes, that would be a better explanation of realistic, but then it should be said so.

3. The Bell theorem assumes that some variables, which he calls beables, take definite values. These variables may (or may not) have a superficial similarity with some observables. For example, in Bohmian mechanics, the beables are particle positions, which have a superficial similarity with position observables. But the point is that it only refers to some observables, not all observables. For most observables, there are no corresponding beables.

 

[weirdoguy]

Observable is a self-adjoint operator

Some people (e.g. Ballentine, if I recall correctly) differentiate between observables as quantities that can be measured in an experiment, and operators that are associated with them.

 

[Demystifier]

Some people (e.g. Ballentine, if I recall correctly) differentiate between observables as quantities that can be measured in an experiment, and operators that are associated with them.

Yes, but then the post by @vanhees71 would make even less sense, because then "observables" would always commute, not only at spacelike separations.

 

[martinbn]

Some people (e.g. Ballentine, if I recall correctly) differentiate between observables as quantities that can be measured in an experiment, and operators that are associated with them.

I think everyone does. Demystifier just continues with his sophistry.

 

[weirdoguy]

because then "observables" would always commute, not only at spacelike separations.

How do you define commutator of two quantities?

 

[vanhees71]

Yes, but then the post by @vanhees71 would make even less sense, because then "observables" would always commute, not only at spacelike separations.

Why that? Of course there are no self-adjoint operators in the lab nor Hilbert spaces and all that. That's the mathematical description. In the lab you have accelerators, detectors, lasers, and all that theoreticians don't want to get their hands dirty with ;-).

 

[Demystifier]

How do you define commutator of two quantities?

As the commutator under multiplication, $AB-BA$. I guess I don't need to explain what is multiplication of operators, and what is multiplications of real and complex numbers.

 

[Demystifier]

Why that? Of course there are no self-adjoint operators in the lab nor Hilbert spaces and all that. That's the mathematical description. In the lab you have accelerators, detectors, lasers, and all that theoreticians don't want to get their hands dirty with ;-).

Exactly. So if we reserve the name "observable" for operators, then the quantities in the laboratory should not be called "observables". You certainly agree that in science we need precise language, so we should not use the word "observable" for two different things. It is exactly for this purpose that Bell introduced the word "beable", to distinguish it from the "observable".

 

[vanhees71]

Newton's theory of gravitation cannot be local in the here discussed sense, because there's no relativstic spacetime. The Wheeler-Feynman theory hasn't come to anything useful. It was a dead end. In condensed matter physics your field theories are usually not relativistic either and thus also there the microcausality principle cannot be formulated nor does it hold in any sense. Of course, Newtonian approximations are valid in their domain of applicability.

 

[WernerQH]

So, does this mean that your notion of locality must be a fundamental truth about nature?
Or even a fundamental feature of QFT? (There appears to be agreement that QFT is not (yet) in form that would satisfy mathematicians.)

 

[vanhees71]

The funddamental feature is causality in Minkowski space. The assumption of the microcausality condition for local self-adjoint operators that represent observables is a sufficient condition for the relativistic causality condition, and this can be realized in terms of local relativistic QFTs. For the details see Weinberg, QT of Fields, vol.1.

Whether there are other possibilities to construct relativistic QTs that obey the causality constraints of Minkowski spacetime, I don't know. At least I've never found any attempts in this direction in the literature.

 

[Demystifier]

Whether there are other possibilities to construct relativistic QTs that obey the causality constraints of Minkowski spacetime, I don't know. At least I've never found any attempts in this direction in the literature.

There is another possibility, it's string theory. Interestingly, string theory also violates a certain kind of "locality", which is different from both QFT definition of locality and Bell locality. In one paper, I argued that this intrinsic stringy-nonlocality can be avoided, at the expense of making Bell-nonlocality more explicit. https://arxiv.org/abs/hep-th/0605250

 

[vanhees71]

I don't know anything about string theory. What do you mean when you say "Bell-nonlocality"?

 

[Demystifier]

I don't know anything about string theory. What do you mean when you say "Bell-nonlocality"?

Bell nonlocality in string theory means the same as in all other quantum theories. I hope you are not asking me what Bell nonlocality means in quantum theory.

 

[vanhees71]

I ask you what you mean by Bell nonlocality, since locality or nonlocality seems to have completely different meanings when ever these words are used. For me the usual inconsistent lingo of the quantum-foundations community it simply means the violation of Bell's inequalities. We are again back at my plead to make clear definitions of what's meant when the word "locality" or "non-locality" are used!

 

[PeterDonis]

Thread closed for moderation.

 

[PeterDonis]

Some more off topic posts have been deleted. Please keep discussion here on the specific topic of locality in QM. Discussion of fhe philosophy of science is off limits in this thread, and I have issued several zero point warnings for posts that violate that constraint. If you see someone else posting about philosophy of science, or anything else off the thread topic, please do not respond. Use the Report button if you think someone else's post is off limits.

 

[PeterDonis]

Thread reopened.

 

[Demystifier]

I ask you what you mean by Bell nonlocality

I mean the thing you call nonseparability.

 

[vanhees71]

So it simply means entanglement. Why then not saying entanglement instead of using non-locality with an altered meaning. I think the entire "foundational issues" are simply plagued by inprecise language, and that's why it never comes to any conclusion but discusses the same pseudo-problems over and over again. Once even the most stuborn philosophers should realize that on the scientific level the case is closed: It's QT that describes Nature correctly and not "local realistic hidden-variable theories". Science has moved on in the meantime: Entanglement is used for engineering purposes nowadays (quantum cryptography, quantum computing, and all that). Thus QT now becomes part of the engineers' curriculum at the universities of applied sciences!

The true open question in foundational physics is the understanding of the quantum theory of spacetime and/or the gravitational interaction! It'll of course not solved by philosophy but through better and better (astronomical) observations and a new idea from it by theorists.

 

[Demystifier]

I think the entire "foundational issues" are simply plagued by inprecise language, and that's why it never comes to any conclusion but discusses the same pseudo-problems over and over again.

Yes, that's a part of the problem.

Once even the most stuborn philosophers should realize that on the scientific level the case is closed: It's QT that describes Nature correctly and not "local realistic hidden-variable theories".

But a part of the problem is that even scientists are not always sufficiently precise. For example, by "observable" sometimes they mean the operator, and sometimes a thing in the laboratory. Philosophers are motivated to make such things more precise, but in this attempt they produce new imprecisions. In my opinion, a better precision can be achieved by a cooperation between scientists and philosophers.

 

[WernerQH]

Entanglement is used for engineering purposes nowadays (quantum cryptography, quantum computing, and all that).

Why not use the less mysterious term "correlations"? If you would, as I do, see QFT as a statistical theory describing the correlations between isolated events distributed in spacetime, you would find "locality" a very strange starting assumption.

 

[vanhees71]

Entanglement is a very specific kind of correlations. That's why we have all these debates about them!

I don't mean what you mean by "isolated events distributed in spacetime". QFT is just a theory predicting the probabilities for the outcome experiments.

 

[WernerQH]

QFT is just a theory predicting the probabilities for the outcome experiments.

Surely it's more than that. When it is used in cosmology, does it mean that the Universe is just an experiment?
Some people are inclined to think like that, but I did not expect you among them.

 

[Lord Jestocost]

Entanglement is a very specific kind of correlations. That's why we have all these debates about them!

To my mind, the debates are merely about the question:
Why do we experience these correlations in our experiential reality?

 

[vanhees71]

We experience these correlations in our experiential reality (what other reality should be?), because obviously QT is a correct description of Nature and not something invented by EPR what they think should be the right description. That's, how the natural sciences work under the best of all circumstances: You have two well-defined models about how Nature is described (this was of course not given by EPR but by Bell about 30 years later for the model "local, realistic HV theory", while it was established for modern QT already in 1926 ;-)), and you can thus objectively decide which of the models describe the observations better, and that's clearly QT. It's even better: There's not the slightest hint that QT delivers any wrong predictions for the outcome of experiments yet!

 

[AndreasC]

That's, how the natural sciences work under the best of all circumstances: You have two well-defined models about how Nature is described (this was of course not given by EPR but by Bell about 30 years later for the model "local, realistic HV theory", while it was established for modern QT already in 1926 ;-)), and you can thus objectively decide which of the models describe the observations better, and that's clearly QT

So what is your objection to observationally equivalent theories/interpretations?