Why are Bell's inequalities violated?

DrChinese

That is what we are doing, and it makes perfect sense. QM is incompatible with local realism, predictions as defined above, and that was Bell's discovery. It is not often that such clear disagreements occur with such fundamental ideas.

Giving up realism for QM + locality is a good trade, in my opinion.

nanosiborg

As regards Bell's theorem, locality or localism refers to the particular form in which Bell has expressed it in his lhv model of quantum entanglement. Since that form is necessarily realistic (ie., expressed in terms of hidden variables), then BI violation can't entail the option of keeping either locality or realism in a model of quantum entanglement. As far as Bell's lhv formulation is concerned locality and realism are inseparable.
Keeping in mind that it's only locality and realism as formalized by Bell in his lhv model of quantum entanglement that are relevant.

morrobay

According to this PDF paper : Resolution of the nonlocality puzzel in the EPR paradox.

The definition of Bell/EPR realism is the problem:
Realism defined by observers in the classical world requires outcomes before measurement.
But there are physical systems that are beyond the scope of the EPR definition of reality.
Their realism ( for spin 1/2 particles ) is a system with phases associated with spin rotations
( a geometric phase ). With no objective reality to the outcomes before measurement.
The actual outcome is related to the phase varible.

nanosiborg

We're only concerned with realism as formalized in Bell-type hidden variable models of quantum entanglement. Bell writes A(a,λ)=±1, B(b,λ)=±1 , denoting that individual results are determined by unit vectors, a and b, and an underlying parameter, λ. That's Bell realism.

Nugatory

That is very true.

At first glance, this might appear to be a limitation of Bell's model, but in fact these models do cover just about anything that an intuitive layman (and probably Einstein, Podolosky, Rosen and kindred spirits) would accept as "non-weird".

Thus, the real importance of Bell's inequality and its observed violations is that we're stuck with quantum weirdness. Post-Bell, we don't talk about whether the world is weird, we talk about how to deal with that weirdness.

nanosiborg

I don't think I'd use weird to describe Bell tests, though they are incompletely understood. I think the real importance of Bell's theorem is the experimental and interpretational innovation that's happened because of it.

bohm2

It is interesting to read Bell's thoughts on this issue:
Speakable and Unspeakable in quantum mechanics
http://www.futuretg.com/FTHumanEvolu...0Mechanics.pdf

morrobay

Its been close to 50 years and why Bells inequalities are violated has not been explained.
So first , is it possible to have ' spin rotations ' and ' geometric phase' as taken from the PDF paper I referenced ? If not then Admin can delete this post.
But if so then A(aλ)=±1 where λ is a phase variable related to entangled two photon spins
from a Calcium atoms' 6_s level can be considered. And this table:
A________B
xyz______xyz
+++______---
++-______--+
+-+______-+-
+--______-++
-++______+--
-+-______+-+
--+______++-
---______+++

And this P[x-z+]≤ P[y+x-] + [x+z+] being violated could be explained by the above table not having fixed values but with ' rotating spins ' and it would be like an 8 level slot
machine set in motion. The challenge would be to explain why the spins at two equal angular settings are always opposite.
Im only taking the initiative here because the question is not being answered when limited
to EPR/Bell realism

nanosiborg

As I mentioned above, we're only concerned with realism as formalized in Bell-type hidden variable models of quantum entanglement.

BIs are based on a linear correlation between θ and rate of coincidental detection, which is due to the form that Bell's locality condition requires his lhv-supplemented qm expectation value formulation to take, ie., that the probability distribution be factorizable into the functions that determine individual detection.

I mentioned in an earlier post that A(a,λ)=±1, B(b,λ)=±1 are Bell realism. A(a,λ) and B(b,λ) are also explicitly local. As opposed to the explicitly nonlocal A(a,b,λ) and B(a,b,λ), A(a,λ) and B(b,λ) specify that A doesn't depend on b, and B doesn't depend on a.

The intensity of light (or photon flux) transmitted by the analyzing (or second) polarizer in sequenced two polarizer (local) setups is always a nonlnear function of the angular difference of the polarizer settings. (In the two polarizer Bell test setups both polarizers are the analyzer, and rate of coincidental detection is intensity.)

BIs are (must be) violated because a necessarily linear correlation expectation is being applied to a setup that must necessarily (even if nothing nonlocal is happening, as in local sequenced setups) produce nonlinear correlations.

nanosiborg

Thank you bohm2.

ZealScience

I don't quite understand your question. Because Bell's equation follows from Hidden variable and statistics which is orthogonal to QM prediction.

And it is later EXPERIMENTALLY proven to be violated. Maybe the only thing we could ask is the validity of the experiment rather than the reason...

Personal opinion

DrChinese

Asked and answered, morrobay. They are violated because local realism is untenable. And no one knows the answer to that any more than anyone can answer why c is the specific value it is. Further, QM explains why spins are opposite as mentioned.

bohm2

I just realized that these are not the only options. Another possibility is backward causation, where future apparatus settings can affect system in past. I think the Transactional Interpretation and Aharonov presented such models. I'm guessing that neither non-locality or anti-realism is required. And of course, the MWI, which denies that the results of measurements have definite outcomes (e.g. measurement outcomes are relative to a branch).

DrChinese

I think of retro-causal as being non-realistic. That is because realistic implies PRE-existing hidden variables. If the hidden variables are in the future, then it is not realistic.

bohm2

I've seen retro-causal interpretations also described as being non-local. In fact, that's how it's typically described but I've also read what I wrote above (e.g. backward causation does not imply non-locality) so I'm a bit confused.

DrChinese

I think it comes down to your (or perhaps my) definition. The time symmetric (TS) and retrocausal interpretations do not have any effects propagating directly faster than c. But obviously you do have correlations and indirect effects that exceed c. I call that non-realistic, you might call it non-local.

I call anything non-realistic if the interpretation has as adjunct that there are no values for counterfactual measurements - i.e. there is a dependency on the observer. I call anything local if there exists a light cone bounded by c which limits propagation of effects. So by that, TS is local non-realistic. MWI is the same. And to me, Bohmian class theories are non-local AND non-realistic (because there is always a measurement context to consider).

By contrast: I have seen Relational Blockworld (a TS class theory) described by one of its authors as both local and realistic. MWI is often called local realistic. And Bohmian is often described as non-local realistic. Yet by the definitions of EPR, I think my viewpoint is just fine. I don't think it matters all that much, the essential points seem to come out the same in the end.

bohm2

Yes, I think the paper by Wood and Spekkens summarizes a lot of the problems with these definitions. On pages 16-18:

Superluminal causation: One option for explaining Bell correlations causally is to assume that there are some superluminal causes, for instance, a causal influence from the outcome on one wing to the outcome on the other, or from the setting on one wing to the outcome on the other, or both. In the most general case one allows hidden variables that can causally influence the measurement outcomes.

Retrocausation: "Retrocausation" refers to the possibility of causal influences that act in a direction contrary to the standard arrow of time. It has been proposed as a means of resolving the mystery of Bell-inequality violations by purportedly saving the relativistic structure of the theory: rather than having causal influences propagating outside the light cone, they propagate within the light cone although possibly within the backward light cone.

The authors also discuss some of the difficulties in distinguishing retrocausality from superluminal causation:
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

To be honest, I've always found Gisin's description as quantum correlations lying *beyond* spacetime as the most interesting suggestion. At first it didn't make sense to me but then, when one thinks about the early "creation" of matter and space, it seems that it appeared out of something pre-spatial/temporal. So, why can't a remnant of that "pre-spatial stuff" still be with us at some level and play some role in physical laws. I understand this is mere speculation. But others have suggested this:
Against 3-N Dimensional space
http://spot.colorado.edu/~monton/Bra.../Articles.html