Relativity & Quantum Theory: Is Locality Violated?

[DrChinese]

Ok, after reading some of DocC's articles, I feel a bit better...but can anyone explain to me why hidden variables COULDN'T back up a local theory because of Bell's Experiment? Wasn't that the whole point of hidden variables? And it is impossible to test locality? The idea that 'standard QM' must imply non-locality...how is that?

PLEASE dumb it down a lot for me...as my brain refuses to work with me to jump over this hurdle. If you wish, feel free to message me privately.

Assuming you are familiar with the Heisenberg Uncertainty Principle (HUP): A first step is to ask the question: Does the HUP describe reality? Or is there a deeper level of reality and we just can't see it?

If you believe that the HUP is incomplete, and there is a deeper level of reality in which non-commuting variables are well-defined in violation of the HUP: You are a "realist". Einstein was a realist. If you are a realist, then Bell's Theorem has important implications for you.

 

[ttn]

If you believe that the HUP is incomplete, and there is a deeper level of reality in which non-commuting variables are well-defined in violation of the HUP: You are a "realist". Einstein was a realist. If you are a realist, then Bell's Theorem has important implications for you.

Hogwash. Bohm's theory is "realist" in this sense, yet is entirely consistent with all the Bell-related experimental facts.

The correct statement is: if you're a *local* "realist", then Bell's Theorem has important implications for you -- namely, your views are inconsistent with experimental data. But this is also true for *local* "non-realists" -- such a view is *also* inconsistent with experimental data. Bottom line: Bell's Theorem says nothing about "realism" one way or the other. What it says is: you can't have a local theory.

 

[ttn]

Sorry I didn't see that one in the headlines, Why is the evening news so slow on this.
What "hidden variable theory" does this.

Bohmian Mechanics.

And please don't say BM that would be stupid.

Oops. You'll have to explain why it's stupid.

BM is a non-local theroy that just uses a non-local guide wave as a proxy for a hidden variable - that's not the same as a hidden variable.

BM is non-local, that's true. But what is your other point here? You're denying that BM is a hidden variable theory? This makes me think you don't know what the term "hidden variable theory" means. Generally people use this phrase to mean a theory that attempts to solve the measurement problem by adding some kind of structure to orthodox QM (i.e., by denying the orthodox completeness doctrine). As I said before, what BM adds to the wave-function description of orthodox QM is definite particle positions (which then follow definite, deterministic trajectories at all times, even, e.g., when the wave-function isn't a position eigenstate).

Edit: Oh, by the way, the "non-local guide wave" that you mention is not a hidden variable nor a proxy for one. It's the wave function -- the same dynamical object that already appears in orthodox QM. It's the particle positions that are the "hidden variable" (though, as has been pointed out many times, this is a silly misnomer since, if anything is "hidden", it's the wave function, not the particle positions!).

 

[LnGrrrR]

DrChinese,

I would say that, based on instinct alone, I am a 'realist' (there should be some 'magical' way to determine both velocity and position that we have not discovered yet).

Assuming that I were to be a realist, how does Bell's Theorem show that this can not exist?

 

[Hurkyl]

I would say that, based on instinct alone, I am a 'realist' (there should be some 'magical' way to determine both velocity and position that we have not discovered yet).

Assuming that I were to be a realist, how does Bell's Theorem show that this can not exist?

It doesn't. But if you further assume some notion of statistical noninteraction of separated events (e.g. I don't need to know what's going on at Alpha Centauri to figure out the odds of a result here on Earth), then the math will yield that certain possibilities have negative probabilities.

 

[RandallB]

LnGrrR
Since this thread is well off topic from the OP
I’ll continue comments on Entanglement in a new thread And try to “Dumb it down" there

Ok, after reading some of DocC's articles, I feel a bit better...but ….
PLEASE dumb it down a lot for me...as my brain refuses to work with me to jump over this hurdle. If you wish, feel free to message me privately.

GO TO:
https://www.physicsforums.com/showthread.php?t=116036"

 

[UglyDuckling]

You have misunderstood Bell's Theorem. The point is not to "distinguish between hidden variables and quantum mechanics." The point is to distinguish between local theories and non-local theories.

The argument is really quite simple. Orthodox QM is a non-local theory. Einstein pointed this out long ago, and noted that (perhaps) a local theory could be constructed by supplementing OQM with "hidden variables." Bell's Theorem then showed that *even* by adding hidden variables, you can't have a local theory that agrees with experiment. So locality cannot be saved, period. The question of the existence of hidden variables simply isn't touched here. Today, there exist both hidden variable theories and non-hidden-variable theories that are empirically viable. What's important and interesting, though, is that both kinds of theories are non-local. That's the lesson of Bell: in order to agree with experiment, a theory has to be non-local, period.

(Assuming we don't go off into parallel-universes MWI la-la-land... )




If you think about it, this claim (which is a disturbingly widespread misconception) is really quite stupid and obviously wrong. The whole idea of "hidden variables" is to *supplement* the wave-function-description with some other variables, some additional structure. Bohmian Mechanics is of course the nicest example here, where the wave function is supplemented by actual particle positions.

Anyway, here's my point: you have some theory (orthodox QM) which people erroneously think is local; then you're going to *add* some structure to the theory; and suddenly (people erroneously think) the theory has to be made nonlocal in order to make the same predictions the earlier theory was able to make locally without the hidden variables? That's preposterous on its face. Just as an obvious counterexample, you could add deliberately pointless hidden variables -- variables that *did nothing* dynamically and didn't affect in any way the outcomes predicted already by orthodox QM. And yet people think that adding such variables will render a still-local theory in disagreement with experiment all of the sudden? *How* could it possibly do that?

Think of it this way: the basic issue here in all these Bell inequality tests is how strongly correlated outcomes in two places can be. Orthodox QM says they should be correlated a certain way. Now isn't it just obvious that if you *add* some additional variables to orthodox QM (i.e., write down a hidden variable theory) you'll maybe be able to get *stronger* correlations? How could adding structure to the theory somehow force the correlations to be weaker? It just makes no sense. Yet this is implicit in the common (mistaken) view that Bell's theorem is all about "QM vs hidden variables." It isn't, and what these people are missing is the crucial fact that orthodox QM is already non-local. So when you compare the correlations predicted by orthodox QM to those of a local hidden variable theory, you are comparing a non-local theory to a local theory. And then it's no surprise that the non-local theory can predict stronger correlations between distant measurements! But then it's outrageously stupid to conclude that hidden variables are thus refuted.

You are of course correct when you say “the point is to distinguish between local theories and non-local theories.” However for a local theory to be valid quantum mechanics necessarily needed to be supplemented with hidden variables! These hidden variables in turn needed to be local. The local nature of these variables is implicit in Bell’s theorem and the experimental design. Bell’s theory coupled with Aspect experiment showed that local hidden variables cannot explain the correlations between spatial remote sets of observations.

This result does not eliminate hidden variables it merely says local hidden variables alone cannot explain quantum behaviour; there is an apparent non-local element influencing of the quantum world.

The lack of ontology associated with quantum mechanics leaves plenty of room for conjecture about the nature of hidden variables be they local or non-local.

 

[ttn]

You are of course correct when you say “the point is to distinguish between local theories and non-local theories.” However for a local theory to be valid quantum mechanics necessarily needed to be supplemented with hidden variables! These hidden variables in turn needed to be local.

Maybe we agree, then, but I can't quite follow exactly what you're saying here. Are you referring basically to the EPR argument here? I think it's correct that what EPR showed was that the only way to "save" locality for quantum mechanics, was to introduce local hidden variables. (That is, they showed that QM without hidden variables -- regular orthodox QM -- is already a nonlocal theory, but that one could in principle get rid of the nonlocality by rejecting the completeness doctrine and adding hidden variables.) Is that also what you're saying?

The local nature of these variables is implicit in Bell’s theorem and the experimental design. Bell’s theory coupled with Aspect experiment showed that local hidden variables cannot explain the correlations between spatial remote sets of observations.

Agreed. Bell's theorem (+ experiment) shows that the kind of theory argued for by EPR (argued for, that is, as the only way to locally account for certain correlations) doesn't work. Such a theory will have to make some other predictions that don't agree with experiment.

This result does not eliminate hidden variables it merely says local hidden variables alone cannot explain quantum behaviour; there is an apparent non-local element influencing of the quantum world.

I agree with the last bit, but I don't know why you put the first part that way. It doesn't "merely" say that "local hidden variables alone cannot explain quantum behaviour." The whole crucial point here that so many people miss is that it says *more* than that -- it says that no local theory *at all* (whether it has hidden variables or not) can explain quantum behavior, i.e., the empirical facts.

The lack of ontology associated with quantum mechanics leaves plenty of room for conjecture about the nature of hidden variables be they local or non-local.

Conjecture is cheap. Here's what's actually known to be true: local hidden variable theories are ruled out; indeed, local theories of any kind are ruled out. So we need a nonlocal theory. And then we have several extant options: orthodox QM (which is nonlocal and has no hidden variables -- and btw suffers from the measurement problem) and Bohm's theory (which is nonlocal and does have hidden variables -- which means no measurement problem). And there are some others, too, such as the GRW type theories. And of course there is always MWI, which is a kind of counterexample to the claim that we need a nonlocal theory... but not a counter example to the more precise claim that we need a nonlocal theory *in order to explain the observed experimental results*. MWI doesn't explain those observations; it dismisses them as delusional.

 

[RandallB]

Bohmian Mechanics.

Oops. You'll have to explain why it's stupid.

Bohm's theory is non-local and has hidden variables – so big deal.
What’s stupid is to argue that hidden variables in a non-local theory has anything to do with the local hidden variables in Einstein’s view, EPR, or a classical view.

It’s grossly miss-leading to those trying to understand hidden variables.
Do everyone a favor, when describing BM hidden variables at least call them “non-local hidden variables” at least until you can describe them in something other than a non-local theory.

Since this thread has long ago addressed the issue of these things not affecting the “Validity of Relativity” from the OP.
I’d recommend that detailed debate on the various issues this discussion has split into move into individual threads that can focus on one topic/issue.

See you in another thread.

 

[ttn]

Bohm's theory is non-local and has hidden variables – so big deal.
What’s stupid is to argue that hidden variables in a non-local theory has anything to do with the local hidden variables in Einstein’s view, EPR, or a classical view.

It’s grossly miss-leading to those trying to understand hidden variables.
Do everyone a favor, when describing BM hidden variables at least call them “non-local hidden variables” at least until you can describe them in something other than a non-local theory.

I don't get your point. There are (in principle) local hidden variable theories, and there are also nonlocal hidden variable theories. You suggest these have nothing to do with each other, but obviously they do: they're both hidden variable theories.

Since this thread has long ago addressed the issue of these things not affecting the “Validity of Relativity” from the OP.

Do you mean that we've reached some consensus about the "validity of relativity"? It seems to me there are still lots of people around who want to say that there is no problem here for relativity, because the problematic non-locality only afflicts hidden variable theories. I have been at pains to argue that this conclusion is too narrow, that the real conclusion from EPR/Bell is that no local theory *at all* is empirically viable -- and that therefore relativity is, so to speak, broken.

But there seems to be nobody left who's interested in actually debating this point. So I guess either they've all come around to the side of the truth, or they're too scared to discuss it. Oh well.

 

[LnGrrrR]

TTN, get in on the 'dumb'd down bell experiment' and help me out with understanding this stuff there. :)

 

[UglyDuckling]

I certainly agree about that. I just think it's important to stress that, even though we don't yet know exactly "what it's all about" -- we *do* know (already) that whatever that ultimate physical theory looks like, it has to be nonlocal. We don't know everything about nature, but we do know at least one thing: it contains causal influences that are superluminal. And I think this is a serious problem for relativity. (Bell thought so too.)

Before we can conclude that relativity is in trouble we need to eliminate all possible explanations of how the correlations found Aspect’s result can be obtained without information getting around instantaneously. Although the accepted wisdom is that if Bell’s inequality is exceeded then we have a non-local quantum world I believe there is a large loop-hole in this thinking. What the combined outcome of Bell’s theorem and Aspect’s experiment may do; is restrict our models of locality (and the form of the associated space-time arenas) to ones that are free of contradictions between quantum mechanics and relativity.

The loophole in Bell’s theorem, as I see it, lies in the relationship between the source of the “photons” and the detectors. If we are to avoid the instantaneous transmission of information (between quantum objects not human observers) then what happens at the source cannot be independent of what is happening at the detectors! In Aspect’s experiment both sets of detectors are looking at “photons” originating from the same event; the cascade of the calcium atom. The states of both detectors at the moment of observation affects what happens at the source when the “particles” become entangled Since the detectors have common interdependency with the source they cannot be independent of each other. But the routes of interdependency all have non space-like intervals therefore are compliant with relativity.

I don’t propose to elaborate on this deduction in this post but I would like your comments and any proposals you have to eliminate this loophole?



Otherwise the Bell Aspect result still leaves us with two possibilities: -

a non-local quantum world

or a quantum world with a modified form of locality where the contradictions between relativity and quantum mechanics are eliminated.

I look forward to your response.

 

[ttn]

Before we can conclude that relativity is in trouble we need to eliminate all possible explanations of how the correlations found Aspect’s result can be obtained without information getting around instantaneously. Although the accepted wisdom is that if Bell’s inequality is exceeded then we have a non-local quantum world I believe there is a large loop-hole in this thinking. What the combined outcome of Bell’s theorem and Aspect’s experiment may do; is restrict our models of locality (and the form of the associated space-time arenas) to ones that are free of contradictions between quantum mechanics and relativity.

The loophole in Bell’s theorem, as I see it, lies in the relationship between the source of the “photons” and the detectors. If we are to avoid the instantaneous transmission of information (between quantum objects not human observers) then what happens at the source cannot be independent of what is happening at the detectors! In Aspect’s experiment both sets of detectors are looking at “photons” originating from the same event; the cascade of the calcium atom. The states of both detectors at the moment of observation affects what happens at the source when the “particles” become entangled Since the detectors have common interdependency with the source they cannot be independent of each other. But the routes of interdependency all have non space-like intervals therefore are compliant with relativity.

Let me make sure I understand. Bell's Theorem proves that the outcomes in one wing of the apparatus must depend on the setting of the polarizer in the other/distant wing of the apparatus. Your point is that this dependence could be "mediated" by the particle source at the center -- i.e., the joint state of the particle pair could be affected in some way by the settings of (say) both polarizers, so that the particle on one side "knows about" the distant setting. Is that the idea? Presumably the information about the settings would travel ("backwards", from detectors to source) at the speed of light or slower. So then this would be a local (i.e., relativistically causal) mechanism by which the correlations could be explained.

Have I got that basically right?

The problem is this: In this mechanism, the distant setting that a given particle "knows about" isn't (necessarily) the *current* setting of that distant polarizer; it's the setting of that distant polarizer a time 2L/c ago (where L is the source-polarizer distance on each side). So, if the orientations of the polarizers were to be randomly set while the particles are in flight (i.e., after the particle pairs have been emitted) then the particles on each side will have "bad info" some of the time, and the QM correlations won't be able to be reproduced. This "loophole" (called sometimes the "locality loophole" in the literature) is well-known. It is because of it that the "delayed choice" experiments (in which the orientations are randomly flipped around while the particles are in flight) were crucial.

I don’t propose to elaborate on this deduction in this post but I would like your comments and any proposals you have to eliminate this loophole?

It's already been eliminated.

Otherwise the Bell Aspect result still leaves us with two possibilities: -

a non-local quantum world

or a quantum world with a modified form of locality where the contradictions between relativity and quantum mechanics are eliminated.

I look forward to your response.

One could of course still "modify" the definition of locality to still be able to say that nature is local. But that's just playing with words. What's important is what's actually established, not what you call it. And what's established is that there exist superluminal causal influences. And this makes relativity unhappy.

 

[vanesch]

My point is: even to discuss anything, you've got to take some things as given. We normally take as given (at least) stuff like the real existence of the macroscopic physical objects around us (e.g., as you did when you accepted the real existence of certain letters spelled out on your computer screen just now). ...which is *all* I'm doing when I neglect to mention your beloved MWI "counterexample" to the claim that nature isn't local.

Uh. What you seem to take as a given, is not only that stuff like the real existence of macroscopic physical objects, really exists, but that what you see of them is ALL that exists. THIS is the assumption that kills off MWI. MWI allows your screen to exist, but tells you at the same time, that there are other versions of your screen, and explains you also why you don't see these other versions of your screen: in that what you see of your screen is only an aspect of what really exists of your screen.

I would also be shocked if a theory told me that my screen doesn't exist - in that there is NOT ONE SINGLE ASPECT OF REALITY that remotely corresponds to something as a screen. But that's not what MWI tells you. It tells you that what you see of your screen exists all right, but that OTHER things exist too (other screens which you don't see). Is that so terribly shocking ? It can be conceived as *useless* (see my dragon in the other thread). But if that "useless" idea SAVES RELATIVITY, I find it worth to consider.

Imagine classical mechanics, and the phase space of the classical universe. Now, imagine that there is another phase space of something else, and that there is no interaction between the degrees of freedom of that second phase space and of the first. I think you'll agree with me that there will be no observational evidence, by creatures "living" in the first phase space, to find out anything about the second. But DOES that mean that it is nonsense to consider that second phase space, if it allows for a symmetry principle to be introduced which explains also some behaviour in the first phase space ?

This is in fact exactly what happens thanks to decoherence in MWI: the wavefunction is a collection of individual essentially classical systems which do not interact. Each term, individually, lives the life of an essentially classical universe for which you could set up a phase space, and decoherence forbids any interaction between these systems. Unless some specific quantum experiment is executed, in which case there IS something that shines through of these "other" universes: the separation is then not perfect.

I agree with you that one needs some a priori things to start with. But the more sophisticated the science you want to do, the more careful one should pick one's initial viewpoints - and I find, honestly, some common sense idea that the things that we perceive are the ONLY things that exist, not up to the standard of what we should take as starting point. My starting point is that nature must be described by a mathematical object (which puts me definitely in the reductionist camp, and in the "realist" camp), and that this mathematical object must have properties we discover one by one, and which we call "fundamental principles". And from this object, using a general rule, must be deducible what is observed. In the lab, in everyday life and so on.

Sure, something like that, though I don't like the precise way you phrased it. How about "under the assumption that our normal everyday perception of the familiar macroscopic external world (of such things as tables, books, and instrument-pointers) isn't delusional."

No, I don't say that your everyday perception is delusional. I simply say that your everyday perception may simply not see ALL what is. It is not because you see a closed book on the table, that you can conclude that there may not exist also a universe in which the same book is open, on the table. That doesn't mean that there is not an aspect of reality which corresponds to "closed book on the table" of course. It would be shocking if it weren't (and then you'd be right about "delusional"). But how can you possibly KNOW FOR SURE that there is not, at the same time, a parallel world where similar but different things happen ?
I agree that this sounds like Star Trek or worse, and I agree that there would be no reason to do so... if it didn't accomplish anything. But IF it saves relativity, hey, that's good enough for me ! Also, you could say that if you NEVER EVER observed anything about these "parallel worlds" then they have no reason to exist. True. The point is, *sometimes* we can take it that what we observe ARE things that are related to these parallel worlds. They are the typical quantum experiments. Bell type situations, two-slit experiments...

The point I will keep coming back to forever is this: *without* that assumption, there can be no such thing as science, period. Science can't exist without (among other things) the idea of *evidence* -- if there can be no evidence for a proposition, then there's no way to distinguish the true from the false, and no way to do science (or think generally).

I agree with that. But things may be a bit more subtle than "what I see is what exists, and ONLY what exists". It is a good starting hypothesis. But if you then find out that things are nicer formulated if you say that there exist things that you do not see directly, most of the time.

Imagine a different history. Imagine that Bohm or Bell never came along, that Einstein never made his objections the way he did etc... and that nobody ever realized that there were these "non-local" effects. On the other hand, imagine that the battle went on over the Copenhagen interpretation, and that, free of "non-local" ideas, a guy like Everett came along and proposed something like MWI. Most people would probably react the way you do: COME ON! Too crazy! There are no parallel worlds, the superposition principle doesn't apply to macroscopic objects ! Show me an experiment where the indication of such a thing may be seen, or use Occam's razor !

Well, Everett might then have set up an EPR style experiment. He would say: How am I going to put a macroscopic object into superposition, and not allow directly decoherence to kill off my superposition ? Answer: by using locality! I take an entangled photon pair, send them off to distant observers, which obviously cannot interact within a spacelike interval, and as such, I have, for a while, a superposition of one of the observers wrt the other, which cannot decohere immediately (as long as the light cones do not intersect).
I start with the famous |up>|down> - |down>|up>,
the first photon goes to the left to Alice, the second photon goes to see Bob, and this now means that, for a short while, as there is no interaction possible (thanks to locality) between Alice and Bob, they cannot decohere together. If that time is sufficient for Alice to pick her direction, and to see her outcomes, and Bob does the same, then it is sufficient for me TO SHOW INTERFERENCE between the different branches of Bob wrt to Alice to prove that they were in a superposition. And what's this interference ? IT ARE THE FAMOUS EPR CORRELATIONS.
In other words, EPR correlations are nothing else but a quantum interference experiment of macroscopic systems (even observers), where the phases have been preserved by the two branches thanks to spacelike separation (to avoid common decoherence).

As the the Broglie wavelength of macroscopic things is too small to hope for a two-slit experiment, this trick of using locality as a way of preserving for a while the phase relation between two macroscopic quantum systems (here called Alice and Bob) to show an interference pattern when they come together, is, IMO, the clearest way to establish the quantum nature of big objects.

And if literally *seeing* something in front of your face doesn't count as evidence, nothing ever will. We *have* to accept the veracity of direct perception, or else (leaving aside crazy mystics) we have no access to reality at all, and there's no more point trying to do physics (or anything else).

Sure. This means that there must be some aspect of reality which must correspond to what we see in the front of your face. But we should also be wary of the "filter of our perception". Otherwise we arrive at questions of whether the objects in a mirror are real, and how deep is a mirror :-) It makes for great fairy tales too. There's a balance to be found between the necessity to accept certain of our perceptions as being relevant, and the trap of naive realism.

So IF you have some indications of the existence of things that you do not see in front of your face, that may be sufficient to consider them, no ?

 

[DrChinese]

Bell's Theorem proves that the outcomes in one wing of the apparatus must depend on the setting of the polarizer in the other/distant wing of the apparatus.

This comment is for everyone EXCEPT ttn (in order to prevent a repeat of earlier discussion):

This is NOT a generally accepted conclusion from Bell's Theorem. It is well known that there is no dependency between the outcome at Alice based on a setting at Bob. QM predicts a statistical relationship between the outcomes and the settings at both Alice and Bob (4 observables). This is seen in actual experiments.

ttn pushes the above quoted viewpoint as if it is accepted. He has published to the preprint archives on his hypothesis. I do not have the link handy at this time, but it is certainly interesting even if not in keeping with generally accepted physics.

 

[ttn]

This is NOT a generally accepted conclusion from Bell's Theorem.

Very true. My intention isn't to fool people into thinking my view is generally accepted. My intention is only to clarify the issues, the truth. If we cannot resist appealing to authorities, though, I would urge people not to concern themselves too much with the muddled masses -- look at what Bell himself
said, though, wrote, etc. He understood his own theorem better than the vast majority of later commentators.

It is well known that there is no dependency between the outcome at Alice based on a setting at Bob.

That may be widely believed, but it can't be "well known" since it isn't true. Or perhaps someone would like to put forward a theory which can explain the observed correlations (of actual, real, measurement outcomes) locally (i.e., a theory that's consistent with Bell's "local causality" condition)?

QM predicts a statistical relationship between the outcomes and the settings at both Alice and Bob (4 observables). This is seen in actual experiments.

Irrelevant. Lots of theories "predict a statistical relationship between the outcomes" that is "seen in actual experiments." My claim is simply that *all* of those theories violate Bell's local causality condition. Anyone who thinks my claim is false is free -- nay, encouraged -- to post a counterexample.

But don't get me wrong. My claim (that no local theory can account for the observed correlations) isn't based on the lack of a counterexample -- it's based on the positive proof offered by the EPR argument and Bell Theorem together. This is why I'm so confident that nobody will be able to post a counterexample!

 

[ttn]

Uh. What you seem to take as a given, is not only that stuff like the real existence of macroscopic physical objects, really exists, but that what you see of them is ALL that exists. THIS is the assumption that kills off MWI. MWI allows your screen to exist, but tells you at the same time, that there are other versions of your screen, and explains you also why you don't see these other versions of your screen: in that what you see of your screen is only an aspect of what really exists of your screen.

Well, we've been over this before, so I'll keep this short. But I think the apparent reasonableness of the case you're trying to make here depends, in a crucial way, on what turns out to be an equivocation about "ALL". My point is that you have to be able to accept the direct testimony of your eyes, because (as good scientific empiricists) there is no other source at all for information about reality. In response you say: sure, but we shouldn't assume that reality is *limited* to what we directly perceive.

I agree 100% with your point, formulated that way. Indeed, part of being a realist (as opposed to an idealist) is to accept that reality is out there, independent of us, and that there's more of it out there than we're aware of at anyone moment or point in history. (To deny this is to accept "esse est percipi"!)

Let me frame my response around the example of Schroedinger's cat. My point is basically this: when you open the box and look inside and (say) see that the cat is alive, you must accept that the cat is *really* alive. This is *the real* state of the cat. Of course, this only goes so far -- there are a *lot* of facts about the universe (including many about the cat) which are left out. For example, the price of tea in china, the position of a certain leaf on a certain tree in tokyo, the precise contents of the cat's stomach, etc. So I am not at all arguing that there exists nothing beyond the simple fact (the living cat) which you perceive.

What I am saying, though, is that these "other facts" which aren't arrived at by the perceptual act in question, better not *contradict* what you got from direct perception. If the cat is alive -- but, also, unperceived by you, the price of tea in china went up $.05 today, that is fine. But when you say: "I perceive that the cat is alive -- ah, but also, the cat is dead" you lose me. That's not an additional fact out there in nature, a fact about some other entity (or about some detailed aspect of the same entity). It's an alleged fact about the same aspect of the same thing, and it contradicts what you perceive. That, I cannot accept.

This is really just the law of non-contradiction, right? When I say "the cat is alive" I mean the cat is *really* alive. So if the cat really is alive, then that statement is true (and I would call my perception veridical). But if the cat *isn't* really alive, yet I believe it to be, then that belief is false -- I'm deluded. I just don't see any way around this. If, *really*, the cat is part of a massively entangled state involving superpositions of alive and dead, then I'm just plain deluded if I think "it's alive". You try to get around this by talking about "parallel universes" and such, but that's not really what your theory (MWI) says. It's not in another universe that the cat is not just alive, but also, kinda, dead. There's only one universe (even according to the mis-named MWI). It's just that the correct description of the state of the contents of that universe is a wave function with lots of branches, some of which "contain" a dead cat, and some a living cat. But those living and dead cats aren't different objects in different places. They all refer to the same one collection of goop. The MWI claim is that the actual state of the cat is neither alive nor dead. Hence, any belief that "the cat is really alive" is just false. It's a delusion. And so, according to MWI, everything we believe (even the testimony of our eyes about things right in front of us) is delusional.

And I just find that all way too crazy to take seriously -- even though, I confess, it is a logically-self-consistent way to explain at least one individual's (mine, I guess!) conscious experience.

I agree that this sounds like Star Trek or worse, and I agree that there would be no reason to do so... if it didn't accomplish anything. But IF it saves relativity, hey, that's good enough for me !

Well then we disagree about that. The more interesting question is this: what content remains in relativity after you reject all that MWI requires you to reject?

Here's a nice quote from Tim Maudlin on this point: "Physicists have been tremendously resistant to any claims of non-locality, mostly on the assumption (which is not a theorem) that non-locality is inconsistent with Relativity. The calculus seems to be that one ought to be willing to pay *any* price -- even the renunciation of pretensions to accurately describe the world -- to preserve the theory of Relativity. But the only possible view that would make sense of this obsessive attachment to Relativity is a thoroughly realistic one! These physicists seem to be so certain that Relativity is the last word in space-time structure that they are willing even to forego any coherent account of the entities that inhabit space-time."

This is more targeted toward "anti-realists" than MWI people, but I think the point is still relevant. The lengths you go to to "save relativity" better not be so extreme as to leave relativity without any real content or meaning. Otherwise your attempt to "save" it just backfires.

 

[LnGrrrR]

TTN,

I happen to agree with your above post. (Of course, I'm not very 'learned' in this area, but for philosophical reasons, I tend to lean towards your PoV.)

The MWI arguments to me seems far too much like solipsism.

 

[DrChinese]

That may be widely believed, but it can't be "well known" since it isn't true.

There is no dependency on the outcome at Alice (+ or -) based on a setting at Bob (measured in degrees). I.e. if you vary Bob's setting, you do not see any change in the +/- pattern at Alice. We have been over this a hundred times.

The only time a pattern emerges is when you correlate Alice's setting, Bob's setting, Alice's outcome, and Bob's outcome (4 observables). And the resulting pattern follows the predictions of standard QM, which respects special relativity. And my description IS a generally accepted description of the physics.

That said, none of this comprises an absolute disproof of non-local effects. Bell's Theorem allows a degree of latitude.

 

[DrChinese]

The MWI arguments to me seems far too much like solipsism.

The interesting thing is that MWI and Bohmian Mechanics BOTH hypothesize the existence of forces/wave/worlds which cannot, in principle, be observed directly. So why would one be "more plausible" than the other? Or more palatable? I think it simply comes back to personal preference, not logic.

 

[ttn]

The interesting thing is that MWI and Bohmian Mechanics BOTH hypothesize the existence of forces/wave/worlds which cannot, in principle, be observed directly. So why would one be "more plausible" than the other? Or more palatable? I think it simply comes back to personal preference, not logic.

There is a huge difference. In Bohm's theory, there is a very physical/tangible difference between branches of the wave function -- namely, most of them are "empty" while one is "full". That is to say, there's one branch that *in fact has the particles in it*. There's an *actual* configuration for all the particles. So there is no weird mystery about (say) Schroedinger's cat -- assuming that what you refer to with the word "cat" is all of the particles the cat is built from, then the cat really is definitely alive or dead, period. No ambiguity, no parallel universes, etc.

 

[ttn]

There is no dependency on the outcome at Alice (+ or -) based on a setting at Bob (measured in degrees). I.e. if you vary Bob's setting, you do not see any change in the +/- pattern at Alice. We have been over this a hundred times.

Yet you still remain confused about what's at issue. When you say "there is no dependency", what do you mean *exactly*? According to some particular theory, Alice's outcome "doesn't depend" on Bob's setting? If so, what theory? Or do you mean that the observed relative frequency of +/- is (empirically) independent of Bob's setting? These are very different claims, and it's not even clear how *either* is relevant to what we're talking about here.

Here's what is actually true: no theory respecting Bell's "local causality" condition (for which I can give the equations if anyone cares) can be in agreement with the observed correlations. So the results on one side *do* depend on the results on the other -- even if the dependence is "washed out" in the noise in a way that prevents *signalling*.

The only time a pattern emerges is when you correlate Alice's setting, Bob's setting, Alice's outcome, and Bob's outcome (4 observables). And the resulting pattern follows the predictions of standard QM, which respects special relativity.

So the collapse postulate is consistent with special relativity? What *exactly* do you mean when you say that "standard QM" "respects special relativity." Do you mean that standard QM is "signal local"? That it is "Bell Local"? Something else? It's vagueness over questions like this that have led to such decades-long confusion about what Bell proved. So do your part to achieve clarity by being precise!

And my description IS a generally accepted description of the physics.

So much the worse for "the physicists" then, since your description (so far) is very vague... But I hope you will fix this!

That said, none of this comprises an absolute disproof of non-local effects. Bell's Theorem allows a degree of latitude.

What does that mean? Maybe both sides are right? I don't think there's any "latitude". There's controversy, yes, but that just means somebody's confused. The claim I'm making (no bell local theory can agree with experiment, hence nature violates bell locality) is perfectly definite and is either true or false. I'm either right or wrong. No latitude.

Unless MWI is true, then maybe in one branch of the wf I'm right, and in another I'm wrong...

 

[LnGrrrR]

DrC,

Perhaps it is a preference thing. I personally find the idea of universes we can never know about distasteful. :)

 

[ttn]

The interesting thing is that MWI and Bohmian Mechanics BOTH hypothesize the existence of forces/wave/worlds which cannot, in principle, be observed directly. So why would one be "more plausible" than the other? Or more palatable? I think it simply comes back to personal preference, not logic.

After responding to this above, I realized I might have misunderstood you. I thought you were referring to the fact that, in Bohm's theory, the same "total wave function" exists that MWI says exists. So I pointed out that there's a huge difference, that the *actual particle configuration* in Bohm's theory picks out one branch of the wave function as dynamically special (in that *only* it affects the subsequent time evolution of the configuration).

But maybe you meant something more general -- namely, that Bohm's theory has "hidden variables" which you seem to think "cannot, in principle, be observed directly." If you remember that the "hidden variables" in Bohm's theory are *particle positions*, you will realize this is a silly claim. The weird, spooky, mysterious, dubiously-tangible, not-directly-observable entity in Bohmian Mechanics isn't the particle positions -- it's the wave function. But surely nobody can criticize Bohmian Mechanics on *that* point, unless they're able to put forward some version of quantum theory which doesn't have wave functions in it! My point: you're treading on very thin ice if, as an advocate of "conventional QM", you criticize Bohm's theory for containing "unobservables." The only unobservables in Bohm's theory are shared with conventional QM. And the *observables* in Bohm's theory (the particle positions) are precisely what allows that theory to solve the various ambiguities that plague the conventional theory (specifically, the measurement problem).

 

[UglyDuckling]

DrC,

Perhaps it is a preference thing. I personally find the idea of universes we can never know about distasteful. :)

I agree! The universe ought to be simple and elegant not distasteful.

Personally I believe the ideas of non-locality, MWI, Bohmian mechanics, entangles particles are all spawned from a near universal misunderstanding of the special theory of relativity.

The problems of quantum weirdness, particle entanglement, HUV and the apparent instantaneous collapse of the wave function should be exhaustively addressed within the framework of relativity before we allow ourselves the indulgency of these exotic fantasies.

 

[UglyDuckling]

Let me make sure I understand. Bell's Theorem proves that the outcomes in one wing of the apparatus must depend on the setting of the polarizer in the other/distant wing of the apparatus. Your point is that this dependence could be "mediated" by the particle source at the center -- i.e., the joint state of the particle pair could be affected in some way by the settings of (say) both polarizers, so that the particle on one side "knows about" the distant setting. Is that the idea? Presumably the information about the settings would travel ("backwards", from detectors to source) at the speed of light or slower. So then this would be a local (i.e., relativistically causal) mechanism by which the correlations could be explained.

Have I got that basically right?

You start off OK but then lose the plot about half way through when you start talking about information traveling backwards in time. Special relativity does not require a signal to go backwards in time. This is merely an illusion caused by the way we have to measure and represent distance and time.

The problem is this: In this mechanism, the distant setting that a given particle "knows about" isn't (necessarily) the *current* setting of that distant polarizer; it's the setting of that distant polarizer a time 2L/c ago (where L is the source-polarizer distance on each side). So, if the orientations of the polarizers were to be randomly set while the particles are in flight (i.e., after the particle pairs have been emitted) then the particles on each side will have "bad info" some of the time, and the QM correlations won't be able to be reproduced. This "loophole" (called sometimes the "locality loophole" in the literature) is well-known. It is because of it that the "delayed choice" experiments (in which the orientations are randomly flipped around while the particles are in flight) were crucial.

Again you misinterpreting what is happening in space-time. Events that are fixed in space and time have to be reinterpreted for space-time.

For Aspect's Experiment; in space-time the cascade event and detection events are contiguous and any communication between the events must be instantaneous. Therefore, again your argument about "bad info" because of the delay in the settings is incorrect. The apparent 2L/c delay is due to the way we measure and represent events in space and time.

Since when has quantum mechanics said anything about particles in flight?

Talk of altering the settings while the "particles are in flight" has no validity.

It's already been eliminated.

As your argument was so full of errors I think the validity of this statement is very doubtful.

The possibility of relativity acting as the agent of entanglement cannot be so easily dismissed.

One could of course still "modify" the definition of locality to still be able to say that nature is local. But that's just playing with words. What's important is what's actually established, not what you call it. And what's established is that there exist superluminal causal influences. And this makes relativity unhappy.

You cannot say that “superluminal causal influences exist”; especially on the strength the argument above!

Relativity is quite capable of encompassing the violation of Bell’s inequality.

In fact the violation of Bell’s inequality doesn’t make the universe non-local but makes locality more interesting and will eventually elevate the status special relativity when its full implications in explaining what’s going on in the quantum world are fully understood

 

[ttn]

You start off OK but then lose the plot about half way through when you start talking about information traveling backwards in time. Special relativity does not require a signal to go backwards in time. This is merely an illusion caused by the way we have to measure and represent distance and time.

I said "backwards" -- not "backwards in time." What I meant is "backwards" relative to the direction the particles go; the information in question flies from the polarizers to the particle source, not (like the particles) from the source to the polarizers, right? That was your idea.

Again you misinterpreting what is happening in space-time. Events that are fixed in space and time have to be reinterpreted for space-time.

For Aspect's Experiment; in space-time the cascade event and detection events are contiguous and any communication between the events must be instantaneous. Therefore, again your argument about "bad info" because of the delay in the settings is incorrect. The apparent 2L/c delay is due to the way we measure and represent events in space and time.

Since when has quantum mechanics said anything about particles in flight?

Talk of altering the settings while the "particles are in flight" has no validity.

Huh? Are you denying that it takes some time for the particles to get from the particle source to the polarizers (which are, at Innsbruck say, several kilometers away)?

In fact the violation of Bell’s inequality doesn’t make the universe non-local but makes locality more interesting and will eventually elevate the status special relativity when its full implications in explaining what’s going on in the quantum world are fully understood

Well, I don't agree. But I anxiously await your producing an example of a Bell Local theory (i.e., a relativistically-locally causal theory) that agrees with experiment.

 

[UglyDuckling]

I said "backwards" -- not "backwards in time." What I meant is "backwards" relative to the direction the particles go; the information in question flies from the polarizers to the particle source, not (like the particles) from the source to the polarizers, right? That was your idea..

The information dosen't have to fly anywhere. Three events are contiguous and therefore can/will share information about their states. Of course nature does not permit us to see what goes on at quantum level. We must rely on our ability to measure distance and time with rules and clocks and then crudely represent the locations of the events on inertially referenced space-time diagrams. Which gives us a very different view of the world to what quantum objects experience. Fortuneately special relativity enables us to piece together the relationship between the events and helps us understand why sets of spatially separated results can be correlated.

Huh? Are you denying that it takes some time for the particles to get from the particle source to the polarizers (which are, at Innsbruck say, several kilometers away)?

Yes I am denying, it takes some time for the particles to get from the source to the polarisers. Special relativity shows the source and the polarisers at three events to be superpositioned and therefore can freely interact. The perceived timing of the events is a function of the way we measure and represent their locations.

Well, I don't agree. But I anxiously await your producing an example of a Bell Local theory (i.e., a relativistically-locally causal theory) that agrees with experiment.

I don't have to. It already exists. Its called special relativity.

You just have to make the effort to understand it properly

 

[vanesch]

I personally find the idea of universes we can never know about distasteful. :)

I always had difficulties with derivatives... I really don't like them

I know. It is the eternal, and in fact the ONLY objection to the "straightforward" interpretation of QM: "Naaah, too crazy !"

History is full of "naah, too crazy" statements: men evolved from apes ? Naah, too crazy ! The Earth is round ? Naah, too crazy ! The Earth turns around the sun ? Naah, too crazy ! Spacetime is curved ? Naah, too crazy ! The universe once was a hot place that is expanding ? Naaah, too crazy ! Long ago there were big beasts running over the Earth's surface, which looked like dragons ? Naah, too crazy ! ...

My PoV is that we should learn some modesty concerning our intuitive concepts, which induced us all too often in saying "naah, too crazy", and instead learn to take at face value what the formalism of science tells us. Mind you, this is no excuse to expressly postulate crazy things, because crazy things must be true !
But, if you find a formalism which makes correct predictions (quantum formalism) ; principles which seem to hold (like Lorentz invariance) and you can build a system in which all these requirements are met, but the only objection you have is "naah, too crazy", well,... you should think twice.

 

[ttn]

The information dosen't have to fly anywhere. Three events are contiguous and therefore can/will share information about their states.

I don't think you know what "contiguous" means!

Yes I am denying, it takes some time for the particles to get from the source to the polarisers. Special relativity shows the source and the polarisers at three events to be superpositioned and therefore can freely interact. The perceived timing of the events is a function of the way we measure and represent their locations.

All right, now it's just painfully clear you don't know what you're talking about.

Thanks for the discussion.

 

[LnGrrrR]

I always had difficulties with derivatives... I really don't like them

I know. It is the eternal, and in fact the ONLY objection to the "straightforward" interpretation of QM: "Naaah, too crazy !"

History is full of "naah, too crazy" statements: men evolved from apes ? Naah, too crazy ! The Earth is round ? Naah, too crazy ! The Earth turns around the sun ? Naah, too crazy ! Spacetime is curved ? Naah, too crazy ! The universe once was a hot place that is expanding ? Naaah, too crazy ! Long ago there were big beasts running over the Earth's surface, which looked like dragons ? Naah, too crazy ! ...

My PoV is that we should learn some modesty concerning our intuitive concepts, which induced us all too often in saying "naah, too crazy", and instead learn to take at face value what the formalism of science tells us. Mind you, this is no excuse to expressly postulate crazy things, because crazy things must be true !
But, if you find a formalism which makes correct predictions (quantum formalism) ; principles which seem to hold (like Lorentz invariance) and you can build a system in which all these requirements are met, but the only objection you have is "naah, too crazy", well,... you should think twice.

Oh yeah, I'm not saying MWI can't be correct. I just find it distasteful.

The thing is, to me, that MWI is not very useful unless you can provide evidence for it. Now, I know that MWI does fix some 'problems' in QM, but I think we should give the other methods some more time to try to work on it before we have to relegate ourselves to believing in another realities that can not seen to be (as of now) empirically proven.

I can see the appeal of MWI (after reading up on it), but for me, it doesn't seem to be the definition that seems 'right'.

 

[UglyDuckling]

I don't think you know what "contiguous" means!

It means touching or next to. As far relativity is concerned it means the proper interval between events has zero magnitude.





All right, now it's just painfully clear you don't know what you're talking about.

Thanks for the discussion.

OK if I don't Know what I'm talking about! Explain to me why spatially separated quantum objects should not become super-positioned at locations on their world lines where their "proper" separation has zero magnitude.

 

[DrChinese]

Yet you still remain confused about what's at issue. When you say "there is no dependency", what do you mean *exactly*? According to some particular theory, Alice's outcome "doesn't depend" on Bob's setting? If so, what theory? Or do you mean that the observed relative frequency of +/- is (empirically) independent of Bob's setting? These are very different claims, and it's not even clear how *either* is relevant to what we're talking about here.

I'm confused? What IS the dependency between Alice's outcome and Bob's setting if I'm wrong? I have never heard any postulated. That is 2 specific observables, not 3, not 4, just 2.

My point is simple. This thread should be about the relativity and QM, not about your personal pet hypothesis - which we have already discussed in other threads.

 

[vanesch]

The thing is, to me, that MWI is not very useful unless you can provide evidence for it. Now, I know that MWI does fix some 'problems' in QM, but I think we should give the other methods some more time to try to work on it before we have to relegate ourselves to believing in another realities that can not seen to be (as of now) empirically proven.

I can see the appeal of MWI (after reading up on it), but for me, it doesn't seem to be the definition that seems 'right'.

Let me tell you a secret: to me neither ! Where it helps me, is to put aside, for the moment, considerations of which I *really* think they are totally misguided, and create a lot of problems and questions, where they shouldn't.

The first of these considerations is some positivist talk about "undescribable reality" and "what only counts is outcomes of measurement, not to say what "really" happens". This, to me, is a far greater offense to science - to give up on talking about what really happens - than any counter-intuitive ontology. Even though Popper is there to remind us that we should concentrate on the empirically verifiable, that doesn't mean that all there is to science, is just relationships between observations. If that were true, I really don't see the point in doing anything else but purely applied technological scientific research, where we can put these relationships to good use to improve our lives. What's the point in "knowing the relationship between observations with a telescope and a radiotelescope" ? We want, somehow, to get indications of what "really" goes on, no ?

The second of these considerations is the putting down of certain principles, which have nevertheless, served us extremely well. I'm thinking of Lorentz invariance. Never we've been able to find explicit violations of Lorentz invariance, and it is a MAJOR guiding tool in, say, considering different possible interactions in elementary particle theory (you start by writing a Lorentz-invariant Lagrangian). But for the sake of the co-existence of an intuitive ontology, and quantum mechanics, we'd have to kiss goodbye to this otherwise so powerful principle ?
I'm also thinking of the superposition principle (which is the founding principle of quantum theory). Never, when empirically challenged, the superposition principle has failed (in that one expected an interference pattern based upon the superposition principle, and one didn't observe one). Rather on the contrary: in EPR experiments, interferences have been observed which are a direct consequence of the superposition principle.

The rigorous application of these two principles leads, with our current tools, to something like MWI. All the rest falls or in one of the above considerations, or into speculation of what a future theory might look like. I can have some sympathy for the last viewpoint, which is: quantum theory makes correct predictions, but probably one day we will find what's REALLY going on, and this will then probably be totally different from what you think is going on. We will then also understand all our current confusions. My answer is: sure! One day, we will maybe know all that. When ? 100 years from now ? 500 years from now ? What does that help us NOW ? Would it have been helpful to tell people, 400 years ago, that all that talk about "the Earth is going around the sun" or vice versa is useless, counter-intuitive talk, and that what only counts is the GOOD AGREEMENT between observations theory ; that things like "the force of gravity" are hopelessly misguided concepts, that they would find out 300 years later that things are different than they seem to be and that, well, 400 years later, we'd have a "much clearer view on all this physics" ?

So once the homework is: accept the superposition principle, and accept Lorentz invariance as two fundamental guiding principles, and tell me now what fundamental ontology we can think of, I do not see many alternatives.

 

[ttn]

I'm confused? What IS the dependency between Alice's outcome and Bob's setting if I'm wrong? I have never heard any postulated. That is 2 specific observables, not 3, not 4, just 2.

Bell wrote down a precise mathematical condition which, he argued, should be satisfied by any "relativistically local" theory. It amounts to this: the probability that a theory assigns to a given event should depend *only* on facts in the past light cone of that event (in the sense that the probabilities shouldn't change if one, in addition, specifies something outside the past light cone). This is called Bell Locality. Any theory that violates this criterion obviously contains some kind of physical mechanism by which space-like separated events can causally affect one another (though what the mechanism *is* will of course vary from theory to theory).

*That* is the "dependency" that exists between Alice's and Bob's experiments.

You say you've never heard this postulated before, but that's definitely not true. Conventional QM contains just this kind of dependency, in the collapse postulate: which angle Alice sets her device to instantaneously influences the state of Bob's particle (it collapses to one or the other eigenstates of the relevant spin/polarization component). Bohm's theory also includes such a nonlocal mechanism (by which what Alice does at her end affects the state of Bob's particle), though what the mechanism is is different because of the different ontology of the theory.

Can you provide an example of a theory which does *not* contain such a nonlocal mechanism (i.e., which is Bell Local) yet which agrees with experiment? No. Because it's a *theorem* that all Bell Local theories which are capable of explaining the observed correlations for one class of possible experiments, are inconsistent with another class. That is: no Bell Local theories are empirically viable.

So what exactly are you disagreeing with?

My point is simple. This thread should be about the relativity and QM, not about your personal pet hypothesis - which we have already discussed in other threads.

"Relativity and QM" *is* what I'm talking about.

And why do you constantly focus on personalities rather than facts/proofs? First off, the argument I'm sketching above isn't "my personal pet hypothesis." This is precisely the view taken by Bell (just read any of his papers, e.g., Bertlmann's Sox, or La Nouvelle Cuisine), David Albert (read his beautiful book "Quantum Mechanics and Experience"), Tim Maudlin (read his beautiful book "Quantum Nonlocality and Relativity"), Shelly Goldstein (read his insightful encyclopedia article on Bohmian Mechanics), and a number of other well-known physicists and philosophers of science. No, the view is not universally accepted, but then neither are all sorts of eminently true things. And second, even if I were the only person in the world taking this view, that wouldn't make one bit of difference as to its truth. Scrutinize my argument and tell me what's wrong with it if you're going to so vehemently disagree. Unless the point of this discussion board isn't to clarify issues and help us all achieve truth, but simply to list, as dogma, widely accepted viewpoints whether they are logical and true or not. In which case, I'm outta here.

 

[vanesch]

Bell wrote down a precise mathematical condition which, he argued, should be satisfied by any "relativistically local" theory. It amounts to this: the probability that a theory assigns to a given event should depend *only* on facts in the past light cone of that event (in the sense that the probabilities shouldn't change if one, in addition, specifies something outside the past light cone). This is called Bell Locality. Any theory that violates this criterion obviously contains some kind of physical mechanism by which space-like separated events can causally affect one another (though what the mechanism *is* will of course vary from theory to theory).

Well, you can take that in different ways. If you only look upon the OUTCOMES, then clearly, the probability of what happened at Alice only depends of what is in this event's past lightcone: the probability to have a click or not, only depends on the source, and on her local setting. In most experiments, this probability is simply 50%.
IT DOESN'T MAKE SENSE TO TALK ABOUT A CORRELATION at this point !
Now, the same can be said for Bob. So Bob has only events to his disposal, of which the probability to click or not, is only dependent on the source and on HIS setting. Alice's settings do not influence Bob's probability of seeing a click or not.

The only way to find a "strange" result is by comparing Alice's and Bob's observations. But this must happen at a LATER event, where Bob and Alice's measurement actions are in the past lightcone of this "comparing" event. It is only NOW, at THIS EVENT that it makes sense to talk about the probabilities of hits which were seen simultaneously, or in opposition (correlation or anti correlation). So, at the event where it makes sense to talk about the "probability of a correlated hit", both Alice's and Bob's measurement are in the past lightcone. And at Alice's place, it doesn't make sense to talk about Bob's probabilities and vice versa, because these results are not available to her.

In fact, to arrive at Bell's statement, you have to make AN EXTRA hypothesis: that is that Bob's and Alice's POTENTIAL outcomes are part of an overall common probability measure, no matter their actual settings.

This extra hypothesis would then allow you to create a (hidden or not) variable, distributed according to this overall common probability measure, and reproduce the correlations ; and that's what Bell proved, cannot happen and agree with QM predictions.

Can you provide an example of a theory which does *not* contain such a nonlocal mechanism (i.e., which is Bell Local) yet which agrees with experiment? No. Because it's a *theorem* that all Bell Local theories which are capable of explaining the observed correlations for one class of possible experiments, are inconsistent with another class. That is: no Bell Local theories are empirically viable.

Yes, for the nth time: MWI ! What MWI "violates" is the extra hypothesis of the existence of an overall common probability measure of POTENTIAL outcomes, because the only terms appearing in the overall wavefunction are those that got correlated with the ACTUAL measurement apparatus (and whose Hilbert norms hence give you the probabilities of observation at Alice, at Bob, and later, when they come together, for their correlations). But it respects the probabilities of all observable things at a certain event, to depend only on what happens in the past light cone. As such, the CORRELATIONS don't "exist" until Alice and Bob come together, and the "potential measurement outcomes" are not considered. In other words, there is no overall probability measure. And hence no Bell theorem.

 

[LnGrrrR]

So once the homework is: accept the superposition principle, and accept Lorentz invariance as two fundamental guiding principles, and tell me now what fundamental ontology we can think of, I do not see many alternatives.

Yes, I can see your point from here. I also detest the idea of 'just worrying about the probabilities' for the same reason you do...it just seems like 'quitting'.

Now, you all certainly know much more than I (though I am slowly...ever so painfully learning), but I think I can see where you're coming from. Right now, MWI is certainly one of the, if not the most, 'valid' ideas, given what we know.

However, it almost seems to be a 'god of the gaps' idea (if you're not sure what this means, I'll let you know). Without having read up a great deal on Lorentz transformations or anything, it seems to me (again, more on a gut feeling than anything else) that QM right now seems to be where classic Newtonian physics was.

Newton did an amazing job describing how certain things function in our world, and for most things, it works great...however, Einstein of course improved upon Newton's mechanics and came up with a theory that not only covered what his could cover, but even more so.

I think that I'm 'holding out' for something like the Theory of Relativity to happen to QM...mayhaps you could say I'm looking for a 'hidden variable' of some sort. But I'm still holding out that we will discover a means that will rule out MWI.

It bothers me enough to think that there are things where we just CAN NOT know (HUP)...but to think that there are other REALITIES we can't get to? Far too 'unfair'...it's like reading one book by an author and finding it immensely satisfying. Then having joy upon hearing he's written THOUSANDS of books...and despair to realize you can't read any other one but the one in your hands. :)

I don't think I've done enough research to form an honest opinion about which 'interpretation' is best though.

 

[selfAdjoint]

Yes, I can see your point from here. I also detest the idea of 'just worrying about the probabilities' for the same reason you do...it just seems like 'quitting'

With the deepest respect I have to say this: when I was in grad school back in the sixties it was the era of the radical student movement. And the the leaders used to say "I know this{whatever they were urging } is not fair or prudent, but WE HAVE TO START SOMEWHERE. Otherwise we're just giving up!" In that case such reasoning led to bombing a campus building at UW Madison, and killing a mathematical researcher.

"Don't be a quitter, no matter what" is a little demon that whispers in peoples' ears to lead them into perdition.

 

[DrChinese]

*That* is the "dependency" that exists between Alice's and Bob's experiments.

You say you've never heard this postulated before, but that's definitely not true. Conventional QM contains just this kind of dependency, in the collapse postulate: which angle Alice sets her device to instantaneously influences the state of Bob's particle (it collapses to one or the other eigenstates of the relevant spin/polarization component).

As you vary Alice's setting, there is no change in Bob's outcome and vice versa. So where is the dependency? There is no prediction by oQM that there is any demonstrable or (non-local) physical connection between these. And there is no way to test Bob's particle alone and determine if it is a collapsed eigenstate or not. Only when you add more observables do you have anything to talk about.

 

[LnGrrrR]

With the deepest respect I have to say this: when I was in grad school back in the sixties it was the era of the radical student movement. And the the leaders used to say "I know this{whatever they were urging } is not fair or prudent, but WE HAVE TO START SOMEWHERE. Otherwise we're just giving up!" In that case such reasoning led to bombing a campus building at UW Madison, and killing a mathematical researcher.

"Don't be a quitter, no matter what" is a little demon that whispers in peoples' ears to lead them into perdition.

I fail to see how your first comment leads to anything but a non-sequitur...

Tell me, how do you know when the right to just 'start' somewhere is? What is wrong with trying to determine the methods behind QM? I personally see nothing wrong with trying to determine it.

Sure, there are some fundamental things we may never know (for instance, /why/ gravity occurs) but certainly we should try to investigate and determine answers about our world.

I don't see why scientists should 'quit' trying to determine answers to any subject. As far as demons...I don't believe in that stuff ;)

 

[vanesch]

Yes, I can see your point from here. I also detest the idea of 'just worrying about the probabilities' for the same reason you do...it just seems like 'quitting'.

Now, you all certainly know much more than I (though I am slowly...ever so painfully learning), but I think I can see where you're coming from. Right now, MWI is certainly one of the, if not the most, 'valid' ideas, given what we know.

However, it almost seems to be a 'god of the gaps' idea (if you're not sure what this means, I'll let you know). Without having read up a great deal on Lorentz transformations or anything, it seems to me (again, more on a gut feeling than anything else) that QM right now seems to be where classic Newtonian physics was.

I agree with all that. In fact, the major challenge is not the interpretation of quantum theory, but the unification, the clash, or no matter how you call it, between quantum theory and general relativity. As long as the jury is out on that one, I'd say: hold your bets.
But does that mean that in the mean time, we shouldn't have a picture to work with ?

Newton did an amazing job describing how certain things function in our world, and for most things, it works great...however, Einstein of course improved upon Newton's mechanics and came up with a theory that not only covered what his could cover, but even more so.

I think that I'm 'holding out' for something like the Theory of Relativity to happen to QM...mayhaps you could say I'm looking for a 'hidden variable' of some sort. But I'm still holding out that we will discover a means that will rule out MWI.

Sure. And usually, the surprise comes from an unexpected side. It's not said that things become simpler. Maybe. Maybe not.

It bothers me enough to think that there are things where we just CAN NOT know (HUP)...but to think that there are other REALITIES we can't get to? Far too 'unfair'...it's like reading one book by an author and finding it immensely satisfying. Then having joy upon hearing he's written THOUSANDS of books...and despair to realize you can't read any other one but the one in your hands. :)

I don't think I've done enough research to form an honest opinion about which 'interpretation' is best though.

It's like getting married, you know...

 

[ttn]

Well, you can take that in different ways. If you only look upon the OUTCOMES, then clearly, the probability of what happened at Alice only depends of what is in this event's past lightcone: the probability to have a click or not, only depends on the source, and on her local setting. In most experiments, this probability is simply 50%.

This involves an equivocation. There are two different things that "probability" could mean here. The way you're using it above, you mean the relative frequency of a certain occurrence. The way I was using it before was to refer to the probabilities that some *particular theory* assigns to a given happening. You're right that there's no way to just look at Alice's data and determine whether Bell Locality is violated. This is because Bell Locality is primarily a locality requirement *for theories*. Any claims made about nature (based on experiment) have to be filtered through (so to speak) an argument about theories. This is why I'm always careful to say things like: "no Bell Local theory can be in agreement with the observations -- thus nature violates Bell Locality."

IT DOESN'T MAKE SENSE TO TALK ABOUT A CORRELATION at this point ! Now, the same can be said for Bob. So Bob has only events to his disposal, of which the probability to click or not, is only dependent on the source and on HIS setting. Alice's settings do not influence Bob's probability of seeing a click or not.

Meaning: Alice's settings do not influence the relative frequency of Bob seeing a click. That's true. It's basically equivalent to what's usually called "signal locality." The deeper question (that Bell asked) is: can a locally causal theory explain the outcomes? The answer turns out to be no (modulo your usual objection about MWI... )

The only way to find a "strange" result is by comparing Alice's and Bob's observations. But this must happen at a LATER event, where Bob and Alice's measurement actions are in the past lightcone of this "comparing" event. It is only NOW, at THIS EVENT that it makes sense to talk about the probabilities of hits which were seen simultaneously, or in opposition (correlation or anti correlation). So, at the event where it makes sense to talk about the "probability of a correlated hit", both Alice's and Bob's measurement are in the past lightcone. And at Alice's place, it doesn't make sense to talk about Bob's probabilities and vice versa, because these results are not available to her.

But (like a good MWI-er ) you're then denying that the measurements really have outcomes prior to Alice and Bob meeting up later! This is like my saying: there's no fact of the matter about whether some distant star has supernova'ed or not (in say the rest frame of the earth) because facts about that distant star only become "real" when the "information" about them gets to my eyes.

What happened to the *realism* I thought we agreed about??

In fact, to arrive at Bell's statement, you have to make AN EXTRA hypothesis: that is that Bob's and Alice's POTENTIAL outcomes are part of an overall common probability measure, no matter their actual settings.

This extra hypothesis would then allow you to create a (hidden or not) variable, distributed according to this overall common probability measure, and reproduce the correlations ; and that's what Bell proved, cannot happen and agree with QM predictions.

You can look at it this way, but I think it's clearer to say that the "extra hypothesis" you need is simply that Alice's and Bob's individual experiments really do have definite outcomes, independent of whether the two of them get together later for coffee and to compare notes. With this assumption alone, the EPR argument gets you (from the postulate of Bell Locality) the existence of local hidden variables which determine (in advance, so to speak) the outcomes on each side. And then Bell's Theorem demonstrates that this kind of model is at odds with experiment. So from two assumptions (that the experiments on each side actually have definite outcomes, and Bell Locality) you get a contradiction with experiment. So either (what I always say, because I'm unwilling to deny the actual outcomes) nature violates Bell Locality... or (what you always say, because you're in love with locality) we have to give up the idea that Alice's and Bob's individual experiments had definite outcomes.

Yes, for the nth time: MWI ! What MWI "violates" is the extra hypothesis of the existence of an overall common probability measure of POTENTIAL outcomes, because the only terms appearing in the overall wavefunction are those that got correlated with the ACTUAL measurement apparatus (and whose Hilbert norms hence give you the probabilities of observation at Alice, at Bob, and later, when they come together, for their correlations). But it respects the probabilities of all observable things at a certain event, to depend only on what happens in the past light cone. As such, the CORRELATIONS don't "exist" until Alice and Bob come together, and the "potential measurement outcomes" are not considered. In other words, there is no overall probability measure. And hence no Bell theorem.

Yes, I think we agree about all this. We just disagree about which of the two premises (locality, or actual outcomes) is more reasonable to give up.

 

[ttn]

I don't think I've done enough research to form an honest opinion about which 'interpretation' is best though.

I thought I'd second my own earlier recommendation of David Albert's book, "Quantum Mechanics and Experience." It is far and away the best book out there if you want to really understand what the issues are and what the possible interpretations are all about. There are a few weird things in the book (speculations about how this or that interpretation might deal with a situation in which some crazy device is hooked up directly to some poor guy's brain, etc.) but for the most part the book is a 100% honest, and 100% clear presentation of the actual problems with the conventional view, and the various proposed ways of dealing with the problems (which means, basically: GRW type theories, MWI type theories, and Bohm type theories).

 

[ttn]

As you vary Alice's setting, there is no change in Bob's outcome and vice versa. So where is the dependency? There is no prediction by oQM that there is any demonstrable or (non-local) physical connection between these. And there is no way to test Bob's particle alone and determine if it is a collapsed eigenstate or not. Only when you add more observables do you have anything to talk about.

Sigh. I've tried so many times to explain this, but we just never seem to get anywhere. Should I try again? Maybe one last time...

You have to allow that there's a difference between two things:

1. Alice can affect the relative frequency of some occurence near Bob by turning a knob (or something else she has direct control over). For example, the frequency of a certain outcome of a certain experiment near Bob has one value if Alice's knob is turned to the left, and has a different value if Alice's knob is turned to the right. In the simplest case, the frequencies are 1 or 0, so, say, when the knob is turned left, a light bulb near Bob lights up; when it's turned right, the bulb goes dim.

2. Some particular theory says that the fundamental dynamical probability of a certain event depends on happenings outside the backwards light cone of the event.

The first involves (evidently) a *controllable* causal mechanism. It allows Alice to send a *signal* to Bob. And note that we don't have to know anything about what theory is or isn't true in order to demonstrate the existence of something like 1. We just discover (say) that turning a certain knob over here, causes some regular correlated happening over there. It's a purely *empirical* statement.

The second, on the other hand, is a statement not about data from an experiment, but about a *theory*. If some theory works this way, we would say that this theory is nonlocal, right? We'd say that this theory violated relativity's prohibition on superluminal causation, right?

Are you with me so far? If so, I'll continue later... If not, no point going further.

 

[UglyDuckling]

Help! Help! Help! Help! Help! Help! Help! Help! Help! Help! Help! Help! Help!

Will someone please put me out of my misery regarding “the locality loophole”? ttn says “I don’t know what I’m talking about”, I have tried to find a rational valid reason why he should make this allegation but so far it escapes me. So would someone please! please! tell me what I’m missing!

I will restate my reasoning so you can identify more easily where my logic is going wrong.

The assertion is; given experimental results that violate Bell’s inequality then “locality” can be recovered if we assume that certain events in the experimental procedure are, at quantum level, super-positioned.

For Aspect’s experiment these events are: -
1. The moment of the calcium atom decays.
2. The moments at the polarisers when by the conventional theory the “photons” would be passing through them.
3. Finally the moments at the detectors when they pick up a count.

The argument is the quantum systems at these events become super-positioned and act as a single entity. Communication between the components of the system becomes instantaneous and any changes in the quantum states of the component system can take place in accordance with the laws of conservation without special relativity being violated.

I’ve tried thinking of why ttn should so vehemently oppose this viewpoint and I’ve come up with the following possibilities.

1. The events in question are separated in space and time and therefore could not possibly be super-positioned in the manner proposed? The answer here is the events can all be connected by zero proper interval paths so mathematically there is no reason they should not be super-positioned.
2. Since the component quantum systems are super-positioned there is no requirement for a photon to mediate the electromagnetic force? As quantum mechanics does not provide an ontology this cannot be his grounds for objection. Super-positioning, for instance, as the mediator of electromagnetic force would have a neutral on the dynamics of QED.
3. The only other reason I’ve thought of is that the proposal is incompatible with his belief that super-luminal influences are an established fact and he may think it is inconsistent with Bohm’s interpretation of quantum mechanics. But I’m sure ttn would not let a fondness for a pet theory come in the way of rational debate. So that can’t be the answer!

So just what is it I’m missing about “the locality loophole” problem.

Again.help!

 

[LnGrrrR]

It's like getting married, you know...

Which is another issue I'm very undecided on...

I certainly see your point of "go with MWI now until we understand better otherwise". However, I think a part of me refuses to believe/acknowledge that idea as correct. Whether it's due to lack of information or just something about the way I'm configured, I have no clue. :)

 

[LnGrrrR]

I thought I'd second my own earlier recommendation of David Albert's book, "Quantum Mechanics and Experience." It is far and away the best book out there if you want to really understand what the issues are and what the possible interpretations are all about. There are a few weird things in the book (speculations about how this or that interpretation might deal with a situation in which some crazy device is hooked up directly to some poor guy's brain, etc.) but for the most part the book is a 100% honest, and 100% clear presentation of the actual problems with the conventional view, and the various proposed ways of dealing with the problems (which means, basically: GRW type theories, MWI type theories, and Bohm type theories).

Thanks for the recommendation! The bookstores near me...well...there aren't any anymore. (Thanks Katrina!) Maybe next time I go to Mobile, AL, I'll take a look...or just rely on Amazon.com :)

 

[DrChinese]

The first involves (evidently) a *controllable* causal mechanism. It allows Alice to send a *signal* to Bob. And note that we don't have to know anything about what theory is or isn't true in order to demonstrate the existence of something like 1. We just discover (say) that turning a certain knob over here, causes some regular correlated happening over there. It's a purely *empirical* statement.

The second, on the other hand, is a statement not about data from an experiment, but about a *theory*. If some theory works this way, we would say that this theory is nonlocal, right? We'd say that this theory violated relativity's prohibition on superluminal causation, right?

Are you with me so far? If so, I'll continue later... If not, no point going further.

I am with you so far...

 

[ttn]

I am with you so far...

Spectacular. OK. So, a violation of "1" from the previous post is a violation of what's usually called "signal locality." A violation of "2" from the previous post is a violation of "Bell Locality." This is all just defining of terms, so nothing really to worry about.

By way of continuing, here are some important facts, in no particular order:

* All experiments to date suggest that Signal Locality is true. There is every reason to believe that "the one true theory" (whatever that is exactly) should be Signal Local.

* It is possible for a theory to violate Bell Locality (and, I think we agreed, therefore be at odds with relativity's prohibition on superluminal causation) but nevertheless *respect* Signal Locality. Here's a silly example: suppose there are these two boxes and whenever you and a friend look into the two boxes, you see balls that are the same color. We just take that as an empirical fact. Now suppose somebody proposes the following theory: when nobody's looking, the balls are grey; but then as soon as the first person (you or your friend) look into one of the boxes, the ball in the looked-in box flips a coin and turns itself either red or blue (at random) -- *and*, simultaneously, the distant ball turns itself the same color. Let's leave aside the question of whether there could possiblly be any good reason to believe such a theory (probably not). My point here is simply that this theory violates Bell Locality (since the color of the one ball is determined, according to this theory, by the space-like separated outcome of that random coin flip) -- but that, despite violating Bell Locality, the theory is consistent with Signal Locality (basically, because what one person sees when he opens his box is either a red/blue ball, seemingly at random, and there's no way to infer anything about what happened or didn't happen at the distant box from this). If this theory is right, there exist relativity-violating interactions, but it is impossible to *use* these interactions to communicate with your distant friend (i.e., you can't send a signal).


So... still with me? Do we agree that, in principle, a theory can violate Bell Locality (and hence not be consistent with relativity) yet still not support the sending of signals faster than light?

 

[vanesch]

So... still with me? Do we agree that, in principle, a theory can violate Bell Locality (and hence not be consistent with relativity) yet still not support the sending of signals faster than light?

I know that you want to talk to Dr C. and I don't want to interrupt your discussion. But I wanted to make something clear. We know that signal locality is a less severe requirement than Bell locality. However, we should not forget where this "locality" requirement comes from: it comes from the idea that spacetime has a geometrical Minkowski structure ; from this, it follows that the laws of nature should not depend on how we LABEL the events in Minkowski space - which results that ALL physically relevant expressions should be writable in Lorentz-invariant form. In fact, we need a second assumption: that in all these different ways of labeling the events in Minkowski space, the "t" coordinate represents somehow time for A thinkable observer, which should not be placed before the fact that something that can be set up to be influenced at t1 can determine the probabilities of something at t2, with t2 < t1, for this particular observer ; this, simply because of the obvious paradox that would result: the observer could set up a device that determines the relevant probabilities at t2, and then wait until t1 to change things such that what he observed at t2 would not be true.
This last idea comes from the fact - the observed fact - that we can obtain results at a certain ta and THEN take decisions based upon that result, at a tb > ta, but not vice versa. In other words, we seem to be able to influence the future, but not the past.
This, plus the geometrical structure of Minkowski space, makes us "require locality". But BEFORE it even makes sense to talk about such a locality, one must adhere to the *Minkowski* GEOMETRY of spacetime ; and this, in turn, means, that all physically relevant quantities that could ever appear in a theory, must be representations of the Lorentz group.

As such, it doesn't make, IMO, much sense to *even talk* about signal locality if the Minkowski geometry of spacetime is not respected. There is not the remotest motivation to even think about signal locality if the Minkowski geometry is not there. Of course, there can be a kind of "conspiration" for a theory, which doesn't respect the Minkowski geometry, to nevertheless manifest requirements that ONLY MADE SENSE for that geometry, but there's absolutely no fundamental reason for this to happen. It is a bit as if one discovered that, say, all international banc transfer amounts in the last 20 years, expressed in dollars, are prime numbers.
Of course this is possible. But there's no reason to it. In the same way, a theory that doesn't respect the Minkowski geometry of spacetime (is not expressible in a Lorentz-invariant way) has absolutely no "reason" to respect information locality. It's of course possible. But it would be a very weird thing, as with the banc tranfers being prime numbers.