Why are Bell's inequalities violated?

kith

Usually, we have a system S and an observer O measuring some observable of the system. As soon as we consider the combined system S+O as a physical system (which may be observed by another observer O'), we acknowledge that the current state of S and O influences the evolution of the combined system. This evolution may of course include interactions between S and O. What's wrong with this kind of thinking?

nanosiborg

I agree with this. The assumption of superdeterminism is useless for understanding why Bell's lhv formulation is inappropriate (ie., why it produces incorrect predictions) for modeling Bell tests.

harrylin

Why would you do that? Or was it meant as a personal attack?

kith

Do you think the observer obeys the laws of nature, i.e. can be considered as a physical system?

nanosiborg

Sure. But the OP is concerned with why Bell's inequalities are violated. How will an ad hoc metaphysical assumption, such as superdeterminism, inform regarding that?

kith

Do you agree that in a deterministic theory, the behaviour of a physical system is determined by its past or current state? Do you agree that in such a theory, the behaviour of the observer is determined by its past or current state if we treat him as a physical system?

Bell's theorem makes the assumption that the probability distributions for Alice and Bob are independent. If you don't make this assumption, you can't derive the inequality in the first place. So discussing this assumption seems relevant to me. Personally, I haven't completely wrapped my mind around superdeterminism and want to understand the arguments more deeply.

bohm2

Just to be clear, when you are using the term "anti-realism", do you mean: no pre-existing properties (non-counterfactuals). I'm asking because this stuff is a bit confusing as there are problems even with what is meant by "realism" also. For instance Wood and Spekkens write:
The lesson of causal discovery algorithms for quantum correlations: Causal explanations of Bell-inequality violations require fine-tuning
http://arxiv.org/pdf/1208.4119v1.pdf

Perhaps, Maccone's definition of "realism" seems as one of the more clearer ones:
Simplest proof of Bell’s inequality
http://arxiv.org/pdf/1212.5214v1.pdf

But I always have trouble understanding this. If something is not pre-existing, would not the Wood and Spekkens argument above hold?

DrChinese

It can apply to any theory equally! It certainly applies to special relativity, that is it's very reason to exist is to account for observer reference frames. There is absolutely no reason to suspect that there is anything special in that regards about QM. And not the slightest evidence to suspect that the observer's PAST influences the outcome (only the observer's choice of measurement setting).

DrChinese

I don't attack people, sorry if anything I said implied otherwise.

However, I think it is clear that you value a specific view that is usually excluded by Bell+experiment. You are certainly not the only one. However, it is very difficult to discuss the subject meaningfully when you assume that which you seek to prove.

Of course, for all I know you might say the same thing about my viewpoint.

Dan Fitzgibbon

Suppose there are two sorts of time, one in which the entangled particles in Aspect's experiment both change state at the same time and one (the usual one) in which their change of state should be separated by the amount of time it takes for the effect to travel between the two events at the speed of light, but is not observed to do so, hence our problem. Would that help?

nanosiborg

I agree with what DrChinese said. It's irrelevant.

I believe that superdeterminism is clutching at straws, and that it will not help us to understand the incompatibility between the lhv formalism and experiment. On the other hand, Bell's formulation of the independence assumption (his locality condition) is relevant, and some people (eg., Jarrett) think that a component of it (which, re Jarrett's analysis, doesn't necessarily inform regarding locality/nonocality in nature) might be the effective cause of BI violation.

Here's another way to approach the OP question. What is it about a basic Bell lhv model that produces a linear correlation (which is incompatible with the nonlinear one produced by qm and experiment) between θ and rate of coincidental detection?


I should have said that I think the answer has to do with some aspect of the realism of Bell's formulation, which includes the separable functions A and B (which describe individual detection) expressed in terms of λ, and the expression of the independence (locality) assumption in terms of the functions A and B -- with the result of the Bell lhv program being against realism or "anti-realism" in that only nonrealistic models of quantum entanglement, such as in standard qm or MWI, are allowed (unless you assume ftl or instantaneous action at a distance, such as in dBB).

I don't know. I don't understand the Wood and Spekkens article. Maybe you can explain it?

I don't see how it would. There's no particularly compelling reason to posit effects traveling between the two events. Paired detection events aren't correlated with each other except when the analyzer settings are aligned, in which case a local common cause explanation suffices.

bohm2

MWI is a "realistic" interpretation.

nanosiborg

Ok, then change ...
"with the result of the Bell lhv program being against realism or 'anti-realism' in that only nonrealistic models of quantum entanglement, such as in standard qm or MWI, are allowed (unless you assume ftl or instantaneous action at a distance, such as in dBB)"
... to ...
"with the result of the Bell lhv program being against realism or 'anti-realism' in that only models of quantum entanglement which don't employ hidden variables, such as in standard qm or MWI, are allowed (unless you assume ftl or instantaneous action at a distance, such as in dBB)".

So my answer to your question ...
"Just to be clear, when you are using the term "anti-realism", do you mean: no pre-existing properties ... ?"
... would be that when I'm using the term "anti-realism", I mean no hidden variables.

harrylin

Thanks for that link - regretfully they handle the assumption that it "is meaningful to assign a property to a system (e.g. the position of an electron) independently of whether the measurement of such property is carried out." However I see no reason to believe that this must always be applicable, and I thought that also Bell did not impose such a requirement to "realism" (or did I overlook it somewhere?).

DrChinese

Yes, sadly Bell does not highlight this point. I guess he thought it was obvious. It is the spot after his (14) where he says, "It follows that c is another unit vector". The idea is that a, b and c are simultaneously real (i.e. elements of reality). So this is the spot where that assumption takes place, and without it he would not be able to continue to derive his conclusion.

harrylin

I never realised that, and will have to check it out - thanks!

DrChinese

Yes, and note that this is the very first equation with a, b and c in it. Without all 3, of course, you don't have realism expressed as a testable assumption.