Questions about Bell: Answering Philosophical Difficulties

[JesseM]

Right, under the assumptions of (1) statistical independence and (2) the validity of extant attempts at a description of the reality underlying the instrumental preparations, then "it could not possibly explain the correlations seen by QM".

(1) Only if by "statistical independence" you are referring to the idea that at the moment the particles are created, whatever properties they are assigned (the 'common cause' which makes sure they both later give the same answers when measured on the same angle) are statistically independent of the future choices of the experimenters about what angle to set their polarizers--the source does not have a "crystal ball" to see into the future behavior of the experimenters as in superdeterminism. No other assumptions of statistical independence are being made here.

(2) Can you be more specific about what you mean by "extant attempts at a description of the reality underlying the instrumental preparations"? The only reality assumed by Bell was local realism, and the fact that each particle must, when created, have been given predetermined answers to what response they'd give to each possible detector angle, with the predetermined answers being the same for each particle (the common cause). The second follows from the first, as there is no other way to explain how particles could always give the same response to the same detector angle besides predetermined answers, if you rule out FTL conspiracies between the particles.

I don't think that anyalyzing this stuff with analogies like washing socks, or lotto cards, etc., though I appreciate your efforts, will provide any insight into what's happening in optical Bell experiments.

This seems like a knee-jerk reaction you haven't put any thought into. After all, if the light has predetermined answers to what response it will give to each of the three possible polarizer settings (as must be true under local realism--do you disagree?), then this is very much like the case where each card has a predetermined hidden fruit under each square. Where, specifically, do you think the analogy breaks down?

The light in optical Bell experiments is behaving much as it behaves in an ordinary polariscopic setup.

No it isn't. In an ordinary polariscopic setup it is impossible to set things up so that each of two experimenters always gets yes/no answers on each trial, and are picking between three possible detector settings, and when they pick the same detector setting they always get the same answer, but when they pick different detector settings they get the same answer less than 1/3 of the time. And yet this is what QM predicts can happen for entangled photons if the three polarizer angles are 0, 60, and 120.

It isn't known what's happening at the level of the light-polarizer interaction. QM assumes only that polarizers A and B are analyzing the same optical disturbance for a given coincidence interval.

What are you talking about? Bell didn't make any assumptions about "optical disturbances", he just pointed out that under local realism, if experimenters always get the same answer when they choose the same detector setting and both make their measurements at the same time, that must be because the "things" (you don't have to assume they're 'optical disturbances' or anything so specific) that are measured by each detector must have been assigned identical predetermined answers to what result they'd give for each detector setting at some point when they were in the same local region of space. Do you think it is possible for local realism to be true yet this second assumption to be false? If so, then you haven't thought things through carefully enough, but I can explain why this assumption follows necessarily from local realism if you wish.

 

[RandallB]

This is how Malus' Law was discovered a few hundred years ago, and it is, in my view, strikingly similar to what's happening in simple A-B optical Bell tests.

Well yah duh,
But saying it is “strikingly similar … to Bell” - only tells me you do not understand what Bell was looking for, let alone the point the observations apparently make.
Take some time to read Bell and understand what the straight Bell inequity line from 100% to 0% is.
How QM predictions and actual observations produce a “violation” by going above and below that line with the curve of a sine wave across the same range.

And finally how Local theories can only predict the same sine wave shape but limited to a range of 75% to 25% which keeps it inside (on the 50% side) of the Bell inequity line. The point is a true local theory has yet to explain how to match the observations as non-locals can.

You need to bring yourself up to speed on the real experiment and these issues or you will never be able to keep up with the discussions. Certainly before claiming you know what “Bell's theorem doesn't contradict” you have a lot more asking questions and understanding to attain before doing that.
Otherwise you will only generate pointless arguments here.

 

[ThomasT]

(1) Only if by "statistical independence" you are referring to the idea that at the moment the particles are created, whatever properties they are assigned (the 'common cause' which makes sure they both later give the same answers when measured on the same angle) are statistically independent of the future choices of the experimenters about what angle to set their polarizers--the source does not have a "crystal ball" to see into the future behavior of the experimenters as in superdeterminism. No other assumptions of statistical independence are being made here.

The assumption of statistical independence in Bell's formulation has to do with setting the probability of coincidental detection equal to the product of the probability of detection at A and the probability of detection at B: P(A,B) = P(A) P(B).

Factorability of the joint probability has been taken to represent locality. But, it doesn't represent locality. It represents statistical independence between events (probability of detection) at A and events (probability of detection) at B during any given coincidence interval.

That there is no such statistical independence is evident, and is dictated by the experimental design(s) necessary to test Bell's theorem (ie., necessary to produce entangled pairs).

So, as far as I can tell, Bell didn't make a locality assumption per se.

(2) Can you be more specific about what you mean by "extant attempts at a description of the reality underlying the instrumental preparations"? The only reality assumed by Bell was local realism, and the fact that each particle must, when created, have been given predetermined answers to what response they'd give to each possible detector angle, with the predetermined answers being the same for each particle (the common cause). The second follows from the first, as there is no other way to explain how particles could always give the same response to the same detector angle besides predetermined answers, if you rule out FTL conspiracies between the particles.

We know that attempts, to date, to describe the light-polarizer interactions in realistic mathematical expressions have not matched the qm predictions for all settings.

It's interesting that if we just don't say anything realistic about these interactions (except that the polarizers are interacting with the same disturbance during a given coincidence interval), vis quantum theory, then the probability of joint detection can be fairly accurately calculated.

Apparently whatever is being emitted and is incident on the polarizers is not behaving according to classical polarization theories.

This seems like a knee-jerk reaction you haven't put any thought into. After all, if the light has predetermined answers to what response it will give to each of the three possible polarizer settings (as must be true under local realism--do you disagree?), then this is very much like the case where each card has a predetermined hidden fruit under each square. Where, specifically, do you think the analogy breaks down?

It just seems to me like an unproductive way to think about this stuff. After all, people have been mulling over Bell's theorem for half a century with no agreement as to it's meaning. Why not try a different perspective?

Seeing the connection between simple A-B optical Bell tests and the classic polariscope might prove to be quite, er, fruitful.

No it isn't. In an ordinary polariscopic setup it is impossible to set things up so that each of two experimenters always gets yes/no answers on each trial, and are picking between three possible detector settings, and when they pick the same detector setting they always get the same answer, but when they pick different detector settings they get the same answer less than 1/3 of the time. And yet this is what QM predicts can happen for entangled photons if the three polarizer angles are 0, 60, and 120.

The similarity between the two setups that I see involves the same light extending between polarizers A and B (forget about the emitter in the optical Bell tests), and the same detection rate angular dependence.

What are you talking about? Bell didn't make any assumptions about "optical disturbances", he just pointed out that under local realism, if experimenters always get the same answer when they choose the same detector setting and both make their measurements at the same time, that must be because the "things" (you don't have to assume they're 'optical disturbances' or anything so specific) that are measured by each detector must have been assigned identical predetermined answers to what result they'd give for each detector setting at some point when they were in the same local region of space. Do you think it is possible for local realism to be true yet this second assumption to be false? If so, then you haven't thought things through carefully enough, but I can explain why this assumption follows necessarily from local realism if you wish.

The statistical independence representation and the assignment of specific emission property values together constitute what's usually called, misleadingly I think, the assumption of local realism.

The, obviously wrong, statistical independence assumption has nothing to do with locality. We're left with the possibility of realistic representations being contradicted.

If experimenters always get the same answer when they choose the same polarizer setting during a coincidence interval, that might be because the polarizers are analyzing the same thing (which was produced during the same emission interval). This is what quantum theory assumes.

 

[ThomasT]

Certainly before claiming you know what “Bell's theorem doesn't contradict” you have a lot more asking questions and understanding to attain before doing that.

One of my contentions is that Bell's theorem doesn't actually make a locality assumption. If you think it does, then point out where you think it is in his formulation.

If Bell's locality condition isn't, in reality, a locality condition, then Bell's theorem doesn't contradict locality.

 

[JesseM]

The assumption of statistical independence in Bell's formulation has to do with setting the probability of coincidental detection equal to the product of the probability of detection at A and the probability of detection at B: P(A,B) = P(A) P(B).

You'd have to be more specific about what "A" and "B" are supposed to represent here. For example, if A="experimenter 1 measures at angle 120, gets result spin-up", and B="experimenter 2 measures at angle 120, gets result spin-down" then it is certainly not true that Bell assumed that P(A,B) = P(A)*P(B)...if each experimenter has a 1/3 chance of choosing angle 120, then P(A) = P(B) = 1/6 (because on any given angle, there is a 1/2 chance of getting spin-up and a 1/2 chance of getting spin-down), but P(A,B) is not 1/6*1/6 = 1/36, but rather 1/18 (because there's a 1/3*1/3 = 1/9 chance that both experimenters choose angle 120, but if they both do it's guaranteed they'll get opposite spins, so there's a 1/2 chance experimenter 1 will get spin-up and experimenter 2 will get spin-down, and a 1/2 chance experimenter 1 will get spin-down and experimenter 2 will get spin-up).

We know that attempts, to date, to describe the light-polarizer interactions in realistic mathematical expressions have not matched the qm predictions for all settings.

Really? In what experiments have QM predictions not matched the results?

It's interesting that if we just don't say anything realistic about these interactions (except that the polarizers are interacting with the same disturbance during a given coincidence interval), vis quantum theory, then the probability of joint detection can be fairly accurately calculated.

I really have no idea what you're talking about here. Can you actually give the specifics of this calculation that you say gives a "fairly accurate" match for the probability of joint detection?

This seems like a knee-jerk reaction you haven't put any thought into. After all, if the light has predetermined answers to what response it will give to each of the three possible polarizer settings (as must be true under local realism--do you disagree?), then this is very much like the case where each card has a predetermined hidden fruit under each square. Where, specifically, do you think the analogy breaks down?

It just seems to me like an unproductive way to think about this stuff. After all, people have been mulling over Bell's theorem for half a century with no agreement as to it's meaning.

Where did you get that idea? As far as I know, all physicists agree on the meaning: that Bell's theorem shows the predictions of QM for entangled particles are inconsistent with local realism. So what lack of agreement are you referring to?

Also, just waving away my argument with the word "unproductive" again suggests a knee-jerk reaction that you haven't put any thought into. It would be like if I presented a proof that there can be no largest prime number, and instead of addressing flaws in the proof, you just said "it seems like an unproductive way to think about prime numbers...people have been mulling over the prime numbers for years with no agreement as to their meaning." Sorry, but proofs are proofs, you have to address the specifics if you want to dispute their conclusions. And everything about my example involving lotto cards maps directly to the real experimental setup involving particles, if you don't see how this works I can explain, though it should be pretty obvious if you give it any thought (to get you started, the two cards map to the two particles being measured, the three possible boxes either person can scratch map to the three possible detector angles each experimenter can choose from, and the hidden fruits behind each box map to the notion that each particle has been assigned a predetermined answer to the result it will give when measured on any of the three possible angles).

Seeing the connection between simple A-B optical Bell tests and the classic polariscope might prove to be quite, er, fruitful.

What "connection" is that? All your statements are so hopelessly vague. Please give specifics, like numerical predictions that you think are the same.

The similarity between the two setups that I see involves the same light extending between polarizers A and B (forget about the emitter in the optical Bell tests), and the same detection rate angular dependence.

What detection rate angular dependence? You never give any specifics. There is nothing in the classical optics setup that says that if one experimenter has his polarizer set to angle phi and the other has his polarizer set to angle xi, then the probability of them getting the "same result" in some sense will be cos^2(phi - xi); that is a purely quantum rule for the detection rate angular dependence, which has nothing to do with Malus' law (which has to do with the angle between the polarization of the wave and the detector, not to do with the angle between two detectors).

The statistical independence representation and the assignment of specific emission property values together constitute what's usually called, misleadingly I think, the assumption of local realism.

As I said, it seems to me that the kind of "statistical independence" you referred to above is not assumed by Bell or any other physicist--I think you're just confused, or else you're being overly vague about what "A" and "B" are supposed to represent.

If experimenters always get the same answer when they choose the same polarizer setting during a coincidence interval, that might be because the polarizers are analyzing the same thing (which was produced during the same emission interval). This is what quantum theory assumes.

That's what Bell assumed too, that they were both analyzing two things with properties that were sufficiently "the same" to ensure they had the same predetermined answers to any possible measurement (or opposite answers, depending on the specific experiment being discussed). But he showed that even if you assume this, then under local realism this leads to conflicts with QM over the predicted statistics in trials where the experimenters choose different detector settings. I already explained this with the lotto card example, which you apparently refuse to look at--would you be happier if I performed some trivial editing on that example so that it was no longer about lotto cards, but instead about particles whose spin are measured at two different detectors? None of the math would need to be any different, but if you have some kind of psychological block about mapping analogies to the actual physical situation they're supposed to be analogous to, perhaps it would help you to see that the proof really is completely straightforward.

 

[DrChinese]

Factorability of the joint probability has been taken to represent locality. But, it doesn't represent locality. It represents statistical independence between events (probability of detection) at A and events (probability of detection) at B during any given coincidence interval.

...

If Bell's locality condition isn't, in reality, a locality condition, then Bell's theorem doesn't contradict locality.

Actually, I somewhat agree with these statements. I also think that the separability requirement does not strictly represent locality. Bell says that the vital assumption is that the setting of Alice does not affect the outcome at Bob (and vice versa). So I do believe a locality assumption is represented. I usually refer to this as Bell Locality to distinguish it from other possible representations.

But if you are, in fact, a local realist... then it is a little difficult to maintain that Bell's Theorem is not talking to you. The entire idea of Bell was to show that you need to account for Alice and Bob's space-like separated results being correlated in a way that local realism does not allow. Specifically, the results cannot match the predictions of Quantum Mechanics.

 

[JesseM]

Actually, I somewhat agree with these statements. I also think that the separability requirement does not strictly represent locality. Bell says that the vital assumption is that the setting of Alice does not affect the outcome at Bob (and vice versa).

Ah, so when ThomasT wrote P(A,B) = P(A)*P(B), this may have been an equation that was actually presented in a proof of Bell's theorem, but based on what you say here I'd guess it was presented with the understanding that A and B were only supposed to represent the choice of settings made by Alice and Bob, not the results they obtained with these settings. In this case I do agree the equation should hold as long as Alice and Bob are making their choices independently, but I am not sure that ThomasT was clear on the limited scope of the equation. Hopefully you'd agree with my point here:

You'd have to be more specific about what "A" and "B" are supposed to represent here. For example, if A="experimenter 1 measures at angle 120, gets result spin-up", and B="experimenter 2 measures at angle 120, gets result spin-down" then it is certainly not true that Bell assumed that P(A,B) = P(A)*P(B)...if each experimenter has a 1/3 chance of choosing angle 120, then P(A) = P(B) = 1/6 (because on any given angle, there is a 1/2 chance of getting spin-up and a 1/2 chance of getting spin-down), but P(A,B) is not 1/6*1/6 = 1/36, but rather 1/18 (because there's a 1/3*1/3 = 1/9 chance that both experimenters choose angle 120, but if they both do it's guaranteed they'll get opposite spins, so there's a 1/2 chance experimenter 1 will get spin-up and experimenter 2 will get spin-down, and a 1/2 chance experimenter 1 will get spin-down and experimenter 2 will get spin-up).

 

[DrChinese]

Hopefully you'd agree with my point here:

I don't think we have any significant disagreements on this topic... :)

The issue is really with the person who is arguing that Bell's Theorem does not rule out Local Realism. The burden is really on them to provide a qualifying theory that can match QM. If you already are convinced that either realism or locality can be abandoned, there isn't much left to argue about. It just becomes semantics.

But if you are a local realist, there is a big hill to climb, and attacking Bell's assumptions is a waste of time. So what if there is a little rust around some element of Bell's brilliant paper? Just put forward a qualifying local realistic theory!

So my question is: ThomasT, are you a local realist?

 

[vanesch]

One of my contentions is that Bell's theorem doesn't actually make a locality assumption. If you think it does, then point out where you think it is in his formulation.

If Bell's locality condition isn't, in reality, a locality condition, then Bell's theorem doesn't contradict locality.

Bell uses TWO assumptions to be able to write:
P(A,B,lambda) = P(A,lambda) P(B,lambda).

The first assumption is locality. Now, locality has a slightly different definition depending on whether we have to do with a deterministic theory or with a stochastic theory. In a deterministic theory, the definition is simple: the time evolution of an ontological physical quantity at a space(time) point is entirely determined by the values of the ontological physical quantities defined in a close neighbourhood of said space(time) point.

This by itself already assumes that we have postulated ontological physical quantities, and that they are fields over space(time). Indeed, it doesn't make sense to talk about locality about physical quantities who are not attached to a point in space(time). It also assumes that we have given the full list of ontological (observable or non-observable) quantities.

In practice, this comes down to requiring that the time evolution of all ontological physical quantities is given by a set of partial differential equations.

Relativity requires on top of that, "an upper limit of propagation speed", which comes down requiring that the Green's functions of the partial differential equations vanish outside of the light cone.

This is locality for deterministic theories.

Things become a bit more difficult for stochastic theories. In a stochastic theory, physical quantities are not determined uniquely by the "current state", only their *probabilities* are determined by the "current ontological state". The thing is that probabilities are not physical quantities, because they depend on the conditions one imposes. Well, here one requires the following for locality. We still assume that there are ontological physical quantities associated to each point in space(time).

The conditional probability for an ontological physical quantity at point P to evolve into one or another value, given all the values of the ontological physical quantities within a neighbourhood of point P, remains unchanged when one adds extra conditions concerning the physical values of remote, or past, events.

If that's the case, then the stochastic theory is said to be local.

Let us understand this definition. Assume that we are at point P, at instant t0, and we look at a physical ontological quantity X. At t0+dt, X can take on certain values. Now, if we don't know anything about the physical situation, then we can say for instance that these potential values of X are distributed according to a certain distribution (say, uniform). One would think that "the more we know", the more "refined" our probabilities for X at P and at t0 + dt will be. For instance, the probability to have X0, knowing that at t and P, we had another physical quantity Y = Y0, will be different than if we didn't know Y to be equal to Y0. And if we know about Z = Z0 at P and t0, then that changes again our probabilities for X at t0 + dt. And if we know about Z = Z1 at another point, Q, then this still changes our probability of X at t0+dt.
But IF WE TAKE INTO ACCOUNT all the ontological physical quantities in a neighbourhood of P, at time t0, which we call collectively ALL0, then we find a certain probability P(X0|ALL0) to have X = X0 at t0 + dt, and this is "all the useful information we need and that will tell us something about X0". So if now we ADD another condition:
P(X0 | ALL0 AND STUFF) and "STUFF" is a condition on an ontological physical variable somewhere else, or in the past, then:
P(X0 | ALL0) = P(X0 | ALL0 AND STUFF)

In other words, knowing something extra won't change anything to the probability distribution of X anymore. The neighbourhood of P, and all ontological physical variables, specified everything there was to know.

If that's the case, we call our stochastic theory "local". Notice - and that is very important - that if our stochastic theory is actually a deterministic theory, then both definitions of locality coincide. The only difference is that the probability values will be 1 or 0.

Bell needs this definition to be able to write that P(A,lambda) is not dependent on B (the choice at Bob's). But note the "lambda": it stands for "all the ontological physical variables that are present at Alice". Lambda contains actually a bit more (the part sent to Bob), but we know that ONCE we have the "local" part, that normally, P won't change anymore.

So it is in writing P(A,lambda) (choice at Alice: local quantity, and variables dependent on the incoming particle, whatever they are), and not P(A,B,lambda), Bell uses locality of a stochastic theory to find the probability of having "up" with choice A.
We can write a similar thing Q(B,lambda) at Bob: the probability for Bob to find "up".

The second thing he needs, is that the probability to find, say, (up,up) (written: R(A,B,lambda) ) is now given by the product of the probability of "up" at Alice and the probability of "up" at Bob.

R(A,B,lambda) = P(A,lambda) x P(B,lambda).

HERE, we use the assumption of stochastic independence of our FULLY DETERMINED probabilities. Note that we don't write: R(A,B) = P(A) x P(B). No, we use lambda: for a given (unknowable in practice, but assumed to be given in theory) fully determined ontological state. This is the assumption of no superdeterminism.

Point is: with R(A,B,lambda), we can't do anything because we don't know lambda. So we will have to weight over lambda.

We use again locality in assuming that there is a P(lambda), a certain probability distribution of the ontological physical quantities sent out by the source, which doesn't depend on the choices A and B.

And we use again no superdeterminism when we apply:
integral over lambda of R(A,B,lambda) x P(lambda) to obtain the probability to have "up,up" without any lambda condition.

 

[ThomasT]

So my question is: ThomasT, are you a local realist?

Well, I don't have any new local realistic theory to offer.

It is, as you've indicated, a problem of semantics.

Thank you to you, Jesse, Randall, vanesch, etc. for taking the time to provide thoughtful comments and criticisms.

Now I will go over vanesch's latest post in this thread point by point.

 

[ThomasT]

Bell uses TWO assumptions to be able to write:
P(A,B,lambda) = P(A,lambda) P(B,lambda).

The first assumption is locality. Now, locality has a slightly different definition depending on whether we have to do with a deterministic theory or with a stochastic theory. In a deterministic theory, the definition is simple: the time evolution of an ontological physical quantity at a space(time) point is entirely determined by the values of the ontological physical quantities defined in a close neighbourhood of said space(time) point.

This by itself already assumes that we have postulated ontological physical quantities, and that they are fields over space(time). Indeed, it doesn't make sense to talk about locality about physical quantities who are not attached to a point in space(time). It also assumes that we have given the full list of ontological (observable or non-observable) quantities.

In practice, this comes down to requiring that the time evolution of all ontological physical quantities is given by a set of partial differential equations.

Relativity requires on top of that, "an upper limit of propagation speed", which comes down requiring that the Green's functions of the partial differential equations vanish outside of the light cone.

This is locality for deterministic theories.

Things become a bit more difficult for stochastic theories. In a stochastic theory, physical quantities are not determined uniquely by the "current state", only their *probabilities* are determined by the "current ontological state". The thing is that probabilities are not physical quantities, because they depend on the conditions one imposes. Well, here one requires the following for locality. We still assume that there are ontological physical quantities associated to each point in space(time).

The conditional probability for an ontological physical quantity at point P to evolve into one or another value, given all the values of the ontological physical quantities within a neighbourhood of point P, remains unchanged when one adds extra conditions concerning the physical values of remote, or past, events.

If that's the case, then the stochastic theory is said to be local.

Let us understand this definition. Assume that we are at point P, at instant t0, and we look at a physical ontological quantity X. At t0+dt, X can take on certain values. Now, if we don't know anything about the physical situation, then we can say for instance that these potential values of X are distributed according to a certain distribution (say, uniform). One would think that "the more we know", the more "refined" our probabilities for X at P and at t0 + dt will be. For instance, the probability to have X0, knowing that at t and P, we had another physical quantity Y = Y0, will be different than if we didn't know Y to be equal to Y0. And if we know about Z = Z0 at P and t0, then that changes again our probabilities for X at t0 + dt. And if we know about Z = Z1 at another point, Q, then this still changes our probability of X at t0+dt.
But IF WE TAKE INTO ACCOUNT all the ontological physical quantities in a neighbourhood of P, at time t0, which we call collectively ALL0, then we find a certain probability P(X0|ALL0) to have X = X0 at t0 + dt, and this is "all the useful information we need and that will tell us something about X0". So if now we ADD another condition:
P(X0 | ALL0 AND STUFF) and "STUFF" is a condition on an ontological physical variable somewhere else, or in the past, then:
P(X0 | ALL0) = P(X0 | ALL0 AND STUFF)

In other words, knowing something extra won't change anything to the probability distribution of X anymore. The neighbourhood of P, and all ontological physical variables, specified everything there was to know.

If that's the case, we call our stochastic theory "local". Notice - and that is very important - that if our stochastic theory is actually a deterministic theory, then both definitions of locality coincide. The only difference is that the probability values will be 1 or 0.

Bell needs this definition to be able to write that P(A,lambda) is not dependent on B (the choice at Bob's). But note the "lambda": it stands for "all the ontological physical variables that are present at Alice". Lambda contains actually a bit more (the part sent to Bob), but we know that ONCE we have the "local" part, that normally, P won't change anymore.

So it is in writing P(A,lambda) (choice at Alice: local quantity, and variables dependent on the incoming particle, whatever they are), and not P(A,B,lambda), Bell uses locality of a stochastic theory to find the probability of having "up" with choice A.
We can write a similar thing Q(B,lambda) at Bob: the probability for Bob to find "up".

The second thing he needs, is that the probability to find, say, (up,up) (written: R(A,B,lambda) ) is now given by the product of the probability of "up" at Alice and the probability of "up" at Bob.

R(A,B,lambda) = P(A,lambda) x P(B,lambda).

HERE, we use the assumption of stochastic independence of our FULLY DETERMINED probabilities. Note that we don't write: R(A,B) = P(A) x P(B). No, we use lambda: for a given (unknowable in practice, but assumed to be given in theory) fully determined ontological state. This is the assumption of no superdeterminism.

Point is: with R(A,B,lambda), we can't do anything because we don't know lambda. So we will have to weight over lambda.

We use again locality in assuming that there is a P(lambda), a certain probability distribution of the ontological physical quantities sent out by the source, which doesn't depend on the choices A and B.

And we use again no superdeterminism when we apply:
integral over lambda of R(A,B,lambda) x P(lambda) to obtain the probability to have "up,up" without any lambda condition.

Thanks for the lengthy explanation. I don't think that Bell's theorem or experimental violations of Bell inequalities tell us anything about whether nature harbours nonlocal or instantaneous action at a distance forces or connections (or whatever).

Some conventions:

A = rate of detection at A per unit of time
B = rate of detection at B per unit of time
(A,B) = rate of coincidental detection per unit of time
a = setting of polarizer at A
b = setting of polarizer at B
(a,b) = |a-b| = joint setting of polarizers = angular difference between settings
P = probability (ie., rate of detection per unit of time normalized to 1)

In a standard EPR-Bell test:

the individual detection rate
without polarizers
P(A) = A
P(B) = B
P(A) = P(B)

the individual detection rate
with polarizers (averaged over all possible settings)
P(A) = P(A|a) = P(A|b) = P(A|a,b) = .5A = .5
P(B) = P(B|b) = P(B|a) = P(B|a,b) = .5B = .5

the coincidental detection rate
without reference to polarizer settings
P(A,B) = P(A) P(B) = .25

the coincidental detection rate
with reference to polarizer settings (any given (a,b))
P(A,B|a,b) = cos^2(a,b) (the joint probability approaches cos^2(a,b) as the number of trials approaches infinity)

The individual probabilities never change, no matter what conditions are imposed. If you know a or b or (a,b), it doesn't matter.

However, P(A,B) /= P(A,B|a,b).

So, in this case we no longer have a local theory. We have a global one.

Of course, the instantaneous action that's happening in the global case has nothing to do (as far as anyone knows) with ftl or instantaneous physical propagations or connections. It's simply that when we change the setting, a, then we instantaneously change the setting (a,b), and therefore instantaneously change the joint probability.

 

[RandallB]

.. I don't think that Bell's theorem or experimental violations of Bell inequalities tell us anything about whether nature harbours nonlocal or instantaneous action at a distance forces or connections (or whatever).

No one has claimed that the Bell Theorem or applying it to EPR-Bell experiments helps select a 1)nonlocal or 2) instantaneous action at a distance forces or 3) connections (or whatever) [I assume you include ‘entanglement’ here] or 4) any other QM interpretation of reality.
Only that EPR-Bell experiments applying the Bell Theorem test and question the viability that any Einstein Local [Local and Realistic] explanation of reality might be possible. None of the 1) thru 4) above are compatible with any Einstein Local explanation using Local Realism.

There are very few published Local Realist but there are some, but you would be at odds with them and in agreement with vanesch based on your conclusions here:

…
So, in this case we no longer have a local theory. We have a global one.

Of course, the instantaneous action that's happening in the global case has nothing to do (as far as anyone knows) with ftl or instantaneous physical propagations or connections. It's simply that when we change the setting, a, then we instantaneously change the setting (a,b), and therefore instantaneously change the joint probability.

I take this to mean you agree the evidence rejects Local Realism (my own preference) as unable to explain how joint probabilities P(A,B) can ‘instantaneously change’ with any change in “a” or “b”. And that reality must be defined by what you call a “Global Theory”.
The net of this "Global Theory" in no different than what vanesch have claimed. Namely that the evidence so far says (and some claim it says so conclusively) that no “Einstein Local” explanation can correctly account for EPR-Bell experimental observations with no prefreance towards any ‘Non-Local’ interpretation.
As far as I can tell no one has claimed anything more than that.

 

[vanesch]

I have to say that, for different reasons, I'm occupied with other stuff, and I'm honestly a bit tired of discussing MWI, Bell/EPR stuff and all that. I have the impression I've been writing the same kind of arguments at least a dozen times on these subjects - which remain nevertheless interesting and fascinating. So sorry not to enter the discussion again, right now...

 

[ThomasT]

No one has claimed that the Bell Theorem or applying it to EPR-Bell experiments helps select a 1)nonlocal or 2) instantaneous action at a distance forces or 3) connections (or whatever) [I assume you include ‘entanglement’ here] or 4) any other QM interpretation of reality.

Are you saying that no one has claimed that EPR-Bell stuff implies nonlocality in nature? Of course they have. Even some physicists claim this. But I think they're mistaken, and if one takes the time to sort out the language associated with all this EPR-Bell stuff, then one will find that there's really nothing to get excited about. Some local realistic formulations are incompatible with qm -- that's all.

As for quantum entanglement -- it's an experimental fact and can, I think, be understood from a classical perspective.

There are very few published Local Realist but there are some, but you would be at odds with them and in agreement with vanesch based on your conclusions here:

Except that he, and many others, seem to think that there's a nonlocality problem. But, imho, there isn't.

From what you've written, I'm supposing that we agree that there isn't any sort of nonlocality problem, but there is a problem with making certain classical or realistic formulations compatible with orthodox quantum theory.

I take this to mean you agree the evidence rejects Local Realism (my own preference) as unable to explain how joint probabilities P(A,B) can ‘instantaneously change’ with any change in “a” or “b”. And that reality must be defined by what you call a “Global Theory”.

I agree that the evidence rejects at least one sort of local realistic formulation. But there are lots of them now. Don't some of them actually correctly predict the EPR-Bell correlations?

There's nothing mysterious about why or how P(A,B) changes instantaneously when a or b are changed. This can be understood locally and (somewhat, if not completely) realistically.

If one is using a global observational perspective, then one will need a global theory. I think that when people refer to Bohmian mechanics as being nonlocal, then what nonlocal really means in this context is global. Everything in the universe is, in a global sense (ie., wrt the motion of the universe as a whole), entangled with everything else in the universe. But, this doesn't in any sense mean that some event here on Earth instantaneously causes some event on the other side of the universe.

The word nonlocal has been used where the word global would have been a better choice. This has created a lot of unnecessary confusion (and fantasies of ftl travel and communication) surrounding EPR-Bell questions.

 

[Demystifier]

Bell, EPR, and all that explained through a sex analogy:
https://www.physicsforums.com/blogs/demystifier-61953/sex-quantum-psychology-and-hidden-variables-1477/

 

[jambaugh]

ThomasT,
With regard to Bell's assumption, you are right.

One of my contentions is that Bell's theorem doesn't actually make a locality assumption. If you think it does, then point out where you think it is in his formulation.

But you've reversed the implication when you assert:

If Bell's locality condition isn't, in reality, a locality condition, then Bell's theorem doesn't contradict locality.

Because Locality is one way to assure causal independence in the acts of observation. Thus Bell's "locality" condition (a much broader assumption) is in fact implied by locality and if it is rejected so too must locality.

This having been said it is indeed the "third" assumption classical realism which should be rejected as it too implies either a trivial classical realism (the universe exists and no other objects or objective properties can be well defined) or it implies with or without locality the very same ability factor probabilities under some choice of variables. Any classical correlation matrix can be diagonalized and thus you can always find pairs of random variables which factor.

Locality is just an easy way to assert that you can find a specific pair of variables whose probabilities factor. One could "as easily" assert that say the polarization and momentum of photons are never coupled in a carefully constructed experimental apparatus. Yet they can be entangled prior to the measurements and a violation of Bell's inequality can be demonstrated within QM.

Locality issues are a red herring. It's all about our historic concept of ontological reality getting in the way of understanding the phenomenological behavior of empirical actuality. What happens, happens! And quantum theory describes it quite well. Attempting to "interpret" it in terms of an ontological picture of "reality" breaks down at the quantum level.

I prefer to argue that given science is an epistemological discipline we should excise any non-operational ontological language such as "states" except as tentative shortcuts for classes of empirical phenomena and only when applicable (e.g. when describing the meter reading of a quantum experiment and not the system itself)...(or of course when working within the classical approximation to quantum actuality.)

 

[RandallB]

Are you saying that no one has claimed that EPR-Bell stuff implies nonlocality in nature?

Did you read my post or just react to it?
Perhaps English is a second language so let me repeat: No one has claimed that Bell Theorem applied to EPR-Bell experiments helps in any way to select the best one of many non-local explanations such as your “Global Theory” as better than any other non-local theory.

As I said they are only saying “that EPR-Bell experiments applying the Bell Theorem test and question the viability that any Einstein Local [Local and Realistic] explanation of reality might be possible”.

I suspect you are having a problem with defining the term “Local” and do not understand that redefining a new term for “Global Local” is NOT the same as “Local”. “Local” implies the meaning intended by “Einstein Local”, even Bohm himself acknowledged that BM was non-local with or without the super-deterministic version of ‘local’ that you with “Global Local”; that is nothing new and not “Einstein Local”. I recommend reviewing an https://www.physicsforums.com/showthread.php?t=181904"

I agree that the evidence rejects at least one sort of local realistic formulation. But there are lots of them now. Don't some of them actually correctly predict the EPR-Bell correlations?

There's nothing mysterious about why or how P(A,B) changes instantaneously when a or b are changed. This can be understood locally and (somewhat, if not completely) realistically.

Since when are there lots of local realistic (Einstein Local) formulations? can you name two?
Or just one local realistic formulation that is even close being “somewhat, if not completely” locally and realistically understood to resolve what you call this “EPR-Bell stuff”.
Remember that cannot include a formulation that rejects Local Realism such as the global observational perspective of a “global theory” or a super-deterministic BM interpretation.

I’m not sure if you are looking for a way to support Local Realism, or if you are trying to define the best Non-Local interpretation as one that uses a “Global” / Super-deterministic version of “Local”. If the latter, “Bell” by definition can be of no help to you.

 

[ThomasT]

Locality issues are a red herring.

So are reality issues.

It's all about our historic concept of ontological reality getting in the way of understanding the phenomenological behavior of empirical actuality.

Our understanding of things has to do with our being able to see (or at least visualize) them. So there is a natural desire to render instrumental behavior in terms of deeper causes. Unfortunately, in the case of quantum experimental phenomena this hasn't worked too well.

I prefer to argue that given science is an epistemological discipline we should excise any non-operational ontological language such as "states" except as tentative shortcuts for classes of empirical phenomena and only when applicable (e.g. when describing the meter reading of a quantum experiment and not the system itself)...(or of course when working within the classical approximation to quantum actuality.)

That's a lot of stuff to excise. Meanwhile, a certain amount of time will be taken up deciphering and explaining the semantics of quantum lingo -- with operationalism being the order of the day, hopefully.

Anyway if one were to ask, "What do experimental violations of Bell inequalities have to do with nonlocality in nature?" I feel now that I can answer, "Nothing.", with a certain amount of assurance.

 

[jambaugh]

So are reality issues.

I disagree... else what's left of the implications of EPR?

Our understanding of things has to do with our being able to see (or at least visualize) them.

Yes in so far as our evolutionary method of dealing with our environment. But now that we are looking beyond the scope of say finding food and avoiding tigers and knowing when to plant our corn we must get more formal in the meaning of "understanding of things". In science it can only be measured by our ability to predict. Nothing succeeds like success!...

So there is a natural desire to render instrumental behavior in terms of deeper causes.

Yes exactly... understanding deeper causes and processes rather than a deeper visualization of an ontological reality. Understand the reason behind our desire to visualize... it works at the classical level... then understand when that reason ceases to be applicable... when we push beyond that level. We then must revert to the fundamental epistemological foundation of knowing, our science is based on a epistemology of empirical phenomena (process) not on a Platonic logic of reality (objects).

Unfortunately, in the case of quantum experimental phenomena this hasn't worked too well.

The phenomenological and causal aspects have worked out brilliantly. It's QM conforming to our desire to paint an objective world picture which hasn't worked too well. And that's our failing not the theory's.

That's a lot of stuff to excise.

Yes and no. You needn't excise, just qualify... especially when you get too close to the borderline where the issues begin to become important. In short people should quit trying to (re)"interpret" quantum theory. It's already has its phenomenological (operational) interpretation in the Born probability interp.

Meanwhile, a certain amount of time will be taken up deciphering and explaining the semantics of quantum lingo -- with operationalism being the order of the day, hopefully.

Yes indeed.

Anyway if one were to ask, "What do experimental violations of Bell inequalities have to do with nonlocality in nature?" I feel now that I can answer, "Nothing.", with a certain amount of assurance.

Yes Indeed!

 

[RandallB]

Locality issues are a red herring.

So are reality issues.

I disagree... else what's left of the implications of EPR?

What evidence do either of you use to come to these one of contrary conclusions?
[note: by ‘reality issues’ I assume you guys mean “Realism” as in the realism of a classical reality verses the possible reality of a multidimensional and/or “FTL” wave function or entanglement collapse.]
Neither of you can use Bell or EPR-Bell as evidence as it is only able to address “Local” as understood by Einstein which requires both Locality AND Realism.

Anyway if one were to ask, "What do experimental violations of Bell inequalities have to do with nonlocality in nature?" I feel now that I can answer, "Nothing.", with a certain amount of assurance.

Yes Indeed!

From what do you derive assurance that any solution that may come in the future claiming to be more complete than the Non-Locals (from QM to BM to Strings) should not be required to explain the non-local implications of EPR-Bell as if non-locality means nothing.
As a Local Realist like myself (the Einstein claim) I see that as the exact obligation of any LR explanation. Until a detailed description in LR terms can match the measured EPR-Bell results the Non-Local solution currently in use must be considered at least viable if not most likely complete, regardless of what my or anyone’s personal preference might be.

And IMO any solution that wishes to discredit the current explanation of EPR-Bell results must do so using both locality and realism, in other words find the complete solution Bell himself was originally looking for that demonstrates a more complete hidden variable LR solution as possible.

Arguments trying to decide if we have a misperception in understanding nature because of nonlocality in nature verses nature not based on realism have nothing to do with Bell, as they only address which Non-Local approach is preferable.

 

[jambaugh]

Pardon the long post...

What evidence do either of you use to come to these one of contrary conclusions?
[note: by ‘reality issues’ I assume you guys mean “Realism” as in the realism of a classical reality verses the possible reality of a multidimensional and/or “FTL” wave function or entanglement collapse.]

Contrary to what? Yes I mean realism in this sense. Specifically the modeling of the universe or that part of the universe affecting the outcome of an experiment as a set of points corresponding to states in some manifold or state set. This idea of "the state of the system" or "the state of the system and its environment" be that local or not is all that one needs in essence to derive Bell's inequality. Allow non-local causation if you like. Include any hidden variables you like. You still get outcomes of experiments caused by states of reality and modes of preparation of the system resulting in probability distributions over this set of states. The probability distributions will be positive specifically because in a reality picture negative probability is meaningless. The distributions will be additive over subsets because of the logic of classical reality. Therefore the probabilities for the set difference (XOR) for two sets of experimental outcomes ...
M(A,B)\equiv Pr(A\cap \overline{B}\cup B\cap \overline{A})
will act as a metric with regard to the triangle inequality:
M(A,B) + M(B,C) \ge M(A,C)
This in essence is Bell's inequality. This inequality does not hold under QM...that is if you assert that these sets of outcomes correspond to quantum observables.

Note that you can take a given quantum theory and embed it within a much larger "conspiracy type" classical theory with the same outcomes. One is thereby failing to match up the classical observables with quantum observables. Such is the case in the Bohm pilot-wave type theories where in the wave-functions of the original quantum theory becomes a classical wave unobservable in practice but granting god-like powers observable in principle. However if you in turn "quantize" these again you get a whole new level of Bell inequalities and their violation.

The principle reason I object to these is that they also require non-local causation which in turn means future acting on past. How can we then say the system is in a given state if in the future some cause could revise history and change that state. The state-of-reality picture breaks down by itself without need of invoking QM.

Ultimately I think the order of causation necessarily dictates the time arrow and thus "backwards in time" casuation is both physically and semantically meaningless. The past is the class of phenomena which causally effected our current process of thinking and remembering.

Similarly nearness in space fundamentally is defined by the relative magnitude of causal effect. The sun warms my face more than Alpha Centauri, it is nearer. My fingers feel my keyboard while I have no sensation of yours. It is nearer. The only mystery is how phenomena shake down into events which can be parameterized in 3+1 dimensional space-time. IMNSHO That's what will reconcile quantum theory and Einstein's GR.

Neither of you can use Bell or EPR-Bell as evidence as it is only able to address “Local” as understood by Einstein which requires both Locality AND Realism.

I can speak of locality without invoking classical realism. You are right about Einstein but we can translate his concept of local objective realty into one of locally causal phenomena...what I would call local actuality.

As I explained and speaking only for myself and not ThomasT, I see specifically EPR-Bell as evidence that we must abandon conventional classical reality. As bad as that may sound on first glance it is a proper step toward operationalism. It is not states of reality we see in the lab but phenomena, or as they used to say in the 60's-70's "happenings".

From what do you derive assurance that any solution that may come in the future claiming to be more complete than the Non-Locals (from QM to BM to Strings) should not be required to explain the non-local implications of EPR-Bell as if non-locality means nothing.

Again I qualify that I seen no non-local implications to either QM or EPR-Bell. It is simply a matter of classical being classical (wherein Bell's inequality applies) and quantum being quantum wherein it doesn't. Classical reality is fine for classical theory and wrong for quantum theory. No mysteries no worries.

As a Local Realist like myself (the Einstein claim) I see that as the exact obligation of any LR explanation.

To be repetitive, I don't see any LR explanation possible due to the R issue not the L one. Call me a Local Actualist in contrast to your Local Realism. It is more than a preference. It is a position I've derived from my study of the foundations of QM and its logical structure. Classically probabilities are measures. Quantum mechanically they are squares of measures.
Let me also add that completeness changes meaning in the absence of the concept of state.
QM is quantum complete in a way that CM is not. You have a larger class of observables for a given distinct set of simultaneously distinguishable observable values.

Until a detailed description in LR terms can match the measured EPR-Bell results the Non-Local solution currently in use must be considered at least viable if not most likely complete, regardless of what my or anyone’s personal preference might be.

I'm not clear about what you mean by "Non-Local solution currently in use". If by solution you mean interpretation then the majority of polled quantum theorists adopt the Copenhagen interpretation (which I've been explaining) wherein reality is dropped and locality may then be preserved. If on the other hand you mean by "solution" the actual theory with predictions which match empirical data, QM wins and it is a local causal theory. Pay attention to the meaning of "locality" it is rooted in the causal connection of events and no reference to the states of objects or reality is necessary. It is operationally meaningful in a way that "state of reality" is not.

And IMO any solution that wishes to discredit the current explanation of EPR-Bell results must do so using both locality and realism, in other words find the complete solution Bell himself was originally looking for that demonstrates a more complete hidden variable LR solution as possible.

Again I'm not sure to which "current explanation" you are referring. The current consensus denies what Bell "was originally looking for" and failed to find.

Arguments trying to decide if we have a mis-perception in understanding nature because of nonlocality in nature verses nature not based on realism have nothing to do with Bell, as they only address which Non-Local approach is preferable.

I agree with your words though probably not your meaning. I assert Bell has nothing to do with locality per se. You seem to presume that Bell-EPR automatically imply Non-Locality which blinds you to seeing my point... that it has nothing to do with locality per se.

You need to carefully study the distinction between the lattice of logical propositions about a classical system's state (a lattice of sets with inclusion as the order relation) and the lattice of logical propositions about a quantum system's observables (a lattice of subspaces with sub-space inclusion as the order relation).

The fact that we get a continuum of observables and hence a continuum of distinct logical proposition for a quantum system which can has only a finite spectrum of distinct simultaneous observable values precludes any operationally meaningful "reality" description of that system. The basic laws of classical probability will be violated by the predicted quantum transition probabilities no matter how many hidden variables you introduce...

or you must incorporate a conspiracy theory wherein God knows exactly what future experiments we will make and tweaks and convolutes the initial hidden variables in such a way as to mimic quantum predictions. In either case reality is meaningless or inscrutable to us mere mortals. It should thus IMNSHO be excised from the theory all together.

What happens happens. If we can quantify rules about what happens that predict then we're ahead of the game. Test them empirically and you have science. Hypothesize about the deeper reality behind it and you have mysticism.

Mysticism (from the Greek μυστικός – mystikos- 'seeing with the eyes closed, an initiate of the Eleusinian Mysteries; μυστήρια – mysteria meaning "initiation"[1]) is the pursuit of achieving communion, identity with, or conscious awareness of ultimate reality, the Other, divinity, spiritual truth, or God through direct experience, intuition, or insight.--Wikipedia

 

[RandallB]

Contrary to what?

I must admit I have trouble following how either argument relates to EPR & Bell, but I don’t think I miss read or misquoted the earlier posts.
Is not your position contrary to the ThomasT position that reality issues are a red herring when you disagree to claim that only the locality is a red herring?

Your explanation boils down to your statement:

To be repetitive, I don't see any LR explanation possible due to the R issue not the L one. Call me a Local Actualist in contrast to your Local Realism.

It is more than a preference. It is a position I've derived from my study of the foundations of QM and its logical structure. …. ….

That “Position” of yours is simply an equivalent alternative interpretation of QM CI, no more. If you like you may call QM an equivalent interpretation of your “Local Actualist” view. Just because you slide in the word “local” does not make it the same as the LOCAL be addressed by Bell or EPR-Bell experiments. Local in Bell means Einstein Local and requires both “L” and “R” nothing so far has eliminated “L” as an issue for Bell.

I'm not clear about what you mean by "Non-Local solution currently in use". If by solution you mean interpretation then the majority of polled quantum theorists adopt the Copenhagen interpretation (which I've been explaining) wherein reality is dropped and locality may then be preserved.

First CI does not preserve locality. The currently accepted Non-Local solution presumes that any Non-Local solution including your “Local Actualist” interpretation are equally viable solutions superior to any possible LR solution as Einstein called for; it is that simple.

I agree with your words though probably not your meaning. I assert Bell has nothing to do with locality per se. You seem to presume that Bell-EPR automatically imply Non-Locality which blinds you to seeing my point... that it has nothing to do with locality per se.

In the context of Bell & EPR-Bell “Local” only refers to “Einstein Local” which requires both locality and realism. Therefore the only theory that can be "NOT Non-Local" is Einstien Local LR where locality is clearly significant.
I’m not at all blind to your point: you prefer a Non-Local interpretation you call “Local Actualist” over other Non-Local solutions to EPR-Bell.
You seem to miss the point that EPR-Bell can not differentiate your “Local Actualist” view from any other Non-Local view.

 

[jambaugh]

I'll be brief. What you say w.r.t. Bell invoking "good ole" Einsteinian LR is correct. But remember it is an RAA hypothesis which is negated.

Not(Local and Real) = Not Local OR Not Real.
Thus QM can be interpreted as (Local and Not Real) or (Real and Not Local) or (Not Real and Not Local).
I assert through CI it's Local and Not Real.

I also assert that the Local part is NOT essential to Bell's derivation.

I think you, in holding onto the Real part, are insisiting that locality is negated.
QM is a Local theory. I take issue with your insisting it is not and ask you to expand upon your assertion if you wish to continue making it.

 

[RandallB]

I'll be brief. What you say w.r.t. Bell invoking "good ole" Einsteinian LR is correct. But remember it is an RAA hypothesis which is negated.

Not(Local and Real) = Not Local OR Not Real.
Thus QM can be interpreted as (Local and Not Real) or (Real and Not Local) or (Not Real and Not Local).
I assert through CI it's Local and Not Real.

I also assert that the Local part is NOT essential to Bell's derivation.

I think you, in holding onto the Real part, are insisiting that locality is negated.
QM is a Local theory. I take issue with your insisting it is not and ask you to expand upon your assertion if you wish to continue making it.

Do you not know what a Local Realist is?
Or do you think I don’t know what I’m describing myself as means?
Since I declare myself to be LR how can you possible think I insist that locality is negated?
First unlike your claim I do not “insist” on anything and certainly not that locality has been successfully negated; that is a claim made by Non-Local theory such as yours. As a Local Realist my EXPECTATION is that nature is LOCAL.

Never did I “invoke” Einsteinian LR as being “correct”, I sited it as the definition of what ”LOCAL” means w.r.t. Bell and EPR-Bell. And as you described the different options for what Local means, it gives Four possible discriptions for how nature might work wrt Bell.
First only ONE that is Local where nature is:
#1 (Local and Real)
And THREE that are Non-Local where nature is:
#2 (Local and Not Real)
#3 (Real and Not Local)
#4 (Not Real and Not Local)
[DEF: Real = the realism of a classical reality verses the possible reality of a multidimensional and/or “FTL” wave function or entanglement collapse].

You are correct I do insist that QM is not Local (#1) and I also insist that QM has not been shown to be #2 (Local and Not Real). As you requested I can expand my assertions:
Never has anyone claiming that QM is #2 "Local and Not Real" (or #1 LOCAL for that matter), been able to provide any evidence to support such a claim.
Including where you insist that QM is #2 (Local Actualist but Un-Realistic) but have proved no evidence to support your claim as I requested back in post #120.

Mind you evidence does not need to be experimental; we can accept a description of how Bell & EPR-Bell works in nature using rational logic.

That said I have no doubt that if you start from and hold to a premise that Nature is #2 (Local and Not Real) you will be able to logically conclude that Nature must be “Local Actualist but Un-Realistic”. However that IMO would be Circular Logic and unacceptable proof.

So I in turn take issue with your insisting that QM is #2 (Local and Not Real) and ask you provide some rational and logical evidence without presuming the conclusion, to support your assertion if you wish to continue making it.

 

[ThomasT]

Did you read my post or just react to it?

Yes. I read it, and then I reacted to it.

No one has claimed that Bell Theorem applied to EPR-Bell experiments helps in any way to select the best one of many non-local explanations ...

Many claim that experimental violations of Bell inequalities prove (or at least support) the notion that ftl or instantaneous material propagations are a fact of nature. My contention is that experimental violations of Bell inequalities tell us nothing of the sort.

... such as your “Global Theory” as better than any other non-local theory.

Global is not synonymous with nonlocal. Global refers to an observational perspective.

I suspect you are having a problem with defining the term “Local” and do not understand that redefining a new term for “Global Local” is NOT the same as “Local”.

I use the terms local, Einstein local, and nonlocal in the same way that you do. That is, I think they mean the same things to me that they do to you.

However, I do suspect that you might be using the term global differently than how I meant it, and how it's usually used.

Global refers to an observational perspective. It's only in the global perspective that entanglements emerge. The global perspective correlates the rate of coincidental detection with the angular difference between the polarizer settings. The correlations that emerge, as P(A,B) changes as a function of cos^2(|a-b|), have nothing to do with ftl or instantaneous material influences propagating between a and b, or A and B, or A and b, or B and a. It can all be understood, I think, in terms of local propagations. There's a common cause wrt emission and a common cause wrt the global perspective. There is, of course, an instantaneous connection between changes in a and/or b and changes in |a-b|.

 

[RandallB]

You react by rolling your eyes, because you do not know what the term LOCAL means wrt Bell. And that is made clear when you say:

Global is not synonymous with nonlocal. Global refers to an observational perspective.

I use the terms local, Einstein local, and nonlocal in the same way that you do. That is, I think they mean the same things to me that they do to you.

However, I do suspect that you might be using the term global differently than how I meant it, and how it's usually used.

Global refers to an observational perspective. … There is, of course, an instantaneous connection between changes in a and/or b and changes in |a-b|.

I can allow that “Global” can be taken as an observational perspective, but when you use that perspective to build a “Global Theory” the version “Global local” you create cannot be used in place of what Bell uses for LOCAL as you have been are doing. So NO you are not using those terms the same way I do. You even acknowledge your perspective uses an instantaneous connection between distant settings a & b. Einstein was very clear that is not Einstein Local nor is it local wrt Bell Local.

I already referred you to an https://www.physicsforums.com/showthread.php?t=181904" addressing “BM local” before, your “Global Local” is no different so do read it.

Also, on “how it's usually used” other than your approach here where else is “Global” and especially “Global Local” used by anybody?
As far as I can tell this is a not something found in standard publication, and as a additional non-local interpretation it may be something you’d like to consider for the “Independent Research” area of these forums if youd like to discribe it in greater detail. But it does not refute Bell or "solve" Bell any more so than any other Non-Local and should not be presenteted that way.

 

[ThomasT]

You react by rolling your eyes, because you do not know what the term LOCAL means wrt Bell. And that is made clear when you say:

"Global is not synonymous with nonlocal. Global refers to an observational perspective.

I use the terms local, Einstein local, and nonlocal in the same way that you do. That is, I think they mean the same things to me that they do to you.

However, I do suspect that you might be using the term global differently than how I meant it, and how it's usually used.

Global refers to an observational perspective. … There is, of course, an instantaneous connection between changes in a and/or b and changes in |a-b|."

Which of my above statements (in boldface type) makes it clear to you that I "do not know what the term LOCAL means wrt Bell."?

Here's a quote from Bell (in italics) taken from the thread you provided a link to:

Consider a theory in which the assignment of values to some beables "lambda" implies, not necessarely a particular value, but a probability distribution, for another beable A. (...)
Let A be localized in the space-time region 1. Let B be a second beable localized in a second region 2 separated from 1 in a spacelike way. Now my intuitive notion of local causality is that events in 2 should not be causes of events in 1, and vice versa.

By the term local, then, I mean that the spacelike separated data streams at A and B (in, say, a standard EPR-Bell optical setup) are not causally related to each other. That is, you can do anything you want to a (the polarizer setting at A) and it can have no effect, within a certain time interval (the coincidence interval) on the results at B, and vice versa.

Using this definition, qm is a local theory -- because P(A) and P(B) are always just 1/2.

And, no matter what definition of locality you use, experimental violations of Bell inequalities tell us nothing about whether or not there is a direct causal link between events at A and events at B -- because the data streams at A and B remain random and the rates of detection remain unchanged no matter what is done at one end or the other.

There might be FTL propagations in some medium or interacting media in the deep reality underlying quantum experimental phenomena, however as long as the data streams at A and B are random and the individual detection rates remain the same, then there isn't anything that these experiments can tell us about the reality of nonlocality that qm doesn't already tell us without them. Like the light medium or ether, so long as nonlocality (even if it exists) remains undetectable, then it remains unusable (and effectively nonexistent).

I can allow that “Global” can be taken as an observational perspective, but when you use that perspective to build a “Global Theory” the version “Global local” you create cannot be used in place of what Bell uses for LOCAL as you have been are doing.

Where did "Global local" come from? Did I say that?

Look, the words nonlocal and local refer to the existence or not of FTL causal propagations in nature. The words global and individual refer to experimental designs or observational perspectives. When a theory models, say, the correlations of two or more spatially separated sequences of events, then it is taking a global perspective. When, a theory models, say, the individual setup at A or B in a standard EPR-Bell test, then it isn't taking a global perspective.

You even acknowledge your perspective uses an instantaneous connection between distant settings a & b.

Yes, and the reason I italicized instantaneous connection is because the connection isn't a physical one between a and b. It has only to do with the global observational perspective. We're considering a and b together, not separately. The variable is (|a-b|), so any change you make in a (or b) results in an instantaneous change in the global variable. This is the independent variable in the functional relationship between the angular difference in the polarizer settings and the rate of coincidental detection. The dependent variable, (A,B) is also a composite, or global, variable.

Einstein was very clear that is not Einstein Local nor is it local wrt Bell Local.

Right, the instantaneous connection between changes in a and/or b and changes in |a-b| has nothing to do with locality or nonlocality, but only with the scope of the observational perspective.

I already referred you to an https://www.physicsforums.com/showthread.php?t=181904" addressing “BM local” before, your “Global Local” is no different so do read it.

I enjoyed reading that thread. I think it's possibly a misnomer to call BM a nonlocal theory. Better perhaps to call it simply a global theory, because it doesn't tell us anything about the existence (or not) of FTL causal propagations in nature. It's just based on the idea that everything in the universe is entangled with respect to the gross behavior of the universe. It's also not a realistic theory. It's unique constructions are based on metaphysical speculation. A realistic theory would be one based solely on the behavior of experimental instruments. So, to the extent that a theory is based on instrumental behavior, it's a realistic theory. So, standard quantum theory is a pretty realistic theory, though I don't think that either local or nonlocal necessarily apply to it since it doesn't tell us anything about the existence (or not) of FTL causal propagations in nature.
--------

There's still the question of how to understand the correlations in a standard optical EPR-Bell experiment. Are they any more mysterious than the correlation between polarizer and analyzer in a standard polariscopic setup? I don't think so.

 

[RandallB]

Now you are just getting tiresome if not objectionable.
No one is going to take you seriously if you just keep restating the same false claims.

Using this definition, qm is a local theory -- because P(A) and P(B) are always just 1/2.

That is just plain false, QM does not match the observed results by 1/2 , QM matches observations 100%. It is the current LR descriptions which unlike QM are restricted to “Bell Local" conditions that have only been able to match observations by 1/2. QM is clearly NOT Bell Local (AKA not Einstein Local or Local wrt Bell)

Where did "Global local" come from? Did I say that?

Smile all you want wiseguy, I don’t find you funny just objectionable.
It is you that are designing something “Global” using a version of “local” that is clearly NOT Bell Local. I don’t care if you call it “TommyT Local” but shorting the name to “local” does not make LOCAL WRT Bell.

As I said you can only think that if you do not know what “Local” means when working on Bell. Made all the more clear when you say:

I enjoyed reading that thread. I think it's possibly a misnomer to call BM a nonlocal theory. Better perhaps to call it simply a global theory, because ….

Even Bohm himself made it clear that Local meant more than locality and that his BM was Non-Local! IMO your Global is nothing more than a restatement of Non-Local BM.

Finally you claim the correlations defined by QM are local because as you say:

Are they any more mysterious than the correlation between polarizer and analyzer in a standard polariscopic setup? I don't think so.

That can only mean you do not consider “polariscopic” results “mystic” or “mysterious” but "Local". Thus you should be able to explain the results for 200 photons sent through different configurations of H (horizontal) D (diagonal) and V (vertical) polar filters.
H => 100 pass
V => 100 pass
D => 100 pass
HV => Zero Pass
HDV => 25 Pass
HD => 50 Pass
If these results are not mysterious to you then you should be able to provide a detailed Bell Local (local and realistic), description that specifically describes what is unique about the 25 photons that pass the HDV configuration as compared to the individual descriptions of the other 175 photons.

You and jambaugh both have enough information in this thread to make clear what LOCAL means wrt to Bell, and that it does not just mean locality! At least well enough to understand that only Non-Local descriptions can resolve these polar filter results. If solved in Local terms it would confirm Local Realism not your non-local BM.

There is nothing more I can add to help you know what Local means, so I will unsubscribe from this thread. It is up to you to understand what real scientist mean by Local, until you do you cannot understand Bell, and your in no position to critique Bell.

 

[ThomasT]

That is just plain false, QM does not match the observed results by 1/2 ...

That isn't what P(A) = 1/2 and P(B) = 1/2 means. What P(A) = 1/2 and P(B) = 1/2 means is that for N emissions, then N/2 detections will be recorded at A and N/2 detections will be recorded at B.

Anyway, whether you think so or not, P(A) = P(B) = 1/2 is an expression of the qm prediction for individual detection in a standard EPR-Bell test. It's an expression of the randomness of the emissions.

At the outset of a standard optical Bell test some runs are made with no filters in place. Randomness of the emitted light is assumed. The results of these initial runs, the rates of individual detection at A and B, establish the maximum rate of detection at A and B for this setup. Assuming ideal efficiency, then all of the photons that could possibly be produced per unit of time are produced.

Then the polarizers are positioned, and the rate of detection falls to 1/2 the rate of detection per unit of time without polarizers. If the rates remain constant for many filter settings, then rotational invariance is assumed, the assumption of emission randomness is retained, and with the addition of some coincidence circuitry, then things are in place to do a standard EPR-Bell experiment.

Keep in mind that the emission process hasn't been changed with the addition of the polarizing filters, so we can assume that the same amount of light is emitted per unit of time with the polarizers as was emitted without the polarizers. But, as has been noted, with the polarizers in place only half of the emissions are detected at either A or B. No matter how the polarizer at A is oriented, no matter how the polarizer at B is oriented, no matter how the polarizers at A and B are oriented with respect to each other, the rate of detection at A and the rate of detection at B remain the same and equal to each other, and the sequences of detection attributes in the data streams remain random, during each and every run.

Some think that it's this randomness wrt individual detection that renders any possible nonlocality undetectable and therefore physically meaningless. Of course there's also the possibility that nonlocality simply doesn't exist in nature.

One can refer to any of the interpretations of quantum theory as local or nonlocal, but it really doesn't matter, because it's just a matter of taste.

I choose to call qm a local theory, because the assumption has always been that nature is local and there's no evidence to suggest that it isn't, and the uniquely quantum mechanical constructs tell us nothing about whether nature is local or nonlocal.

Finally you claim the correlations defined by QM are local because as you say:

Are they any more mysterious than the correlation between polarizer and analyzer in a standard polariscopic setup? I don't think so.

That can only mean you do not consider “polariscopic” results “mystic” or “mysterious” but "Local".

Or, it might mean that I consider both to be somewhat mysterious. But the thing about a standard classical polariscopic setup is that it appeals to my physical intuition in a more straightforward and clearcut way in displaying an apparently obvious locality than an optical Bell test setup does.

 

[RandallB]

... it's just a matter of taste.
I choose to call qm a local theory, ...
Or, it might mean that I consider both to be somewhat mysterious.

The terms Local and Non-Local are DEFINED especially wrt what they mean relative to Bell and which theories are Local (LR) or Non-Local (all others).
You do not get to “choose” whatever might appeal to you!

And you “might mean both are considered somewhat mysterious”
Wow if you’re that sure about what to have for dinner you’re going to starve!
Rather than insist you know something others don’t understand, you might want to actually be sure you know what you mean before sticking you two cents in.

I’m not surprised vanesch got tired of this stuff when folks refuse to learn the correct definitions of terms and how to use them when discussing with others.
I’ll double check to be sure the unsubscribe from this thread works this time, this is really rather pointless.