# Nonlocality - fact of nature

1. Apr 21, 2008

### confusedashell

I just wonder to the doubters after so many experiments proving nonlocality, what type of "Proof" would be "proof" enough for the people out there who still think a local interpretation could ever describe reality?

2. Apr 22, 2008

### ahrkron

Staff Emeritus
A macroscopic one, involving daily-life-size objects... which pretty much precludes any QM effects to show clearly,

The reason locality is so strongly ingrained in our interpretation of the world is that it works so well for our interaction with all of the objects in our usual surroundings.

3. Apr 22, 2008

### confusedashell

Well, according to every person who advocates nonlocality there is no macroscopic nonlocality

4. Apr 22, 2008

### akhmeteli

The last time I checked, and it was not long ago, there were no "experiments proving nonlocality". For some mysterious reason, each and every one of them had one or more "loopholes", such as "fair sampling", "locality", etc. If my information was wrong or something radically new has happened since then, please advise. Meanwhile, independent of whether I "doubt" nonlocality or not, or whether I "still think a local interpretation could ever describe reality", nonlocality has not been proven experimentally.

5. Apr 22, 2008

### confusedashell

I thought Bell in essence proved that if any theory are to fit the emperical evidences of Quantum Theory it has to be nonlocal?
So either you must believe quantum theory is wrong, or accept nonlocality?

6. Apr 22, 2008

### akhmeteli

Not really. As far as I understand, it is believed that Bell proved that a local theory cannot be compatible with predictions of quantum theory. Predictions, but not existing empirical evidence. In other words, experimental results incompatible with local theories have not been obtained yet. And I suspect there are reasons to doubt such results will ever be obtained. I am not an expert in the Bell inequalities, but my reasoning (it follows nightlight's posts) is as follows. Unitary evolution under quantum theory does not seem to suggest nonlocality. To prove nonlocality, you need the theory of measurements of quantum theory (projection postulate). However, you can include your instruments (and the observer, if you wish) in the system, and this extended system, according to quantum theory, will undergo unitary evolution and thus would not suggest nonlocality. So attempts to prove nonlocality experimentally may well be an uphill battle. Again, I am not an expert, so the above reasoning may be faulty.

If what I wrote above is correct, it seems that the only thing you need to reject nonlocality is to reject the projection postulate. However, this postulate is, strictly speaking, at odds with quantum theory anyway (as the latter suggests unitary evolution). The situation reminds me the situation with classical mechanics (though replacing it with quantum mechanics does not change the following) and thermodynamics. Mechanics underlies thermodynamics, but it cannot provide irreversibility, however thermodynamics with its irreversibility still gives reasonably correct experimental predictions.

7. Apr 22, 2008

### Staff: Mentor

Make that "a local and realistic theory." Here "realistic" means that variables such as position or spin orientation have definite (although possibly "hidden") values at all times, even before a measurement of those variables.

As far as I know, the experimental evidence so far is consistent with the predictions of QM and therefore inconsistent with the predictions of any local realistic theory (per Bell's Theorem). Therefore, a viable theory to replace QM must be either non-realistic or non-local, or both.

8. Apr 22, 2008

### strangerep

One thing always bugs me about the proof(s) of Bell's theorem...

Take (for example) eq(5) on p36 of "Speakable and Unspeakable in QM":

$$P(\hat a, \hat b) ~= \int d\lambda ~ \rho(\lambda) A(\hat a, \lambda) B(\hat b, \lambda)$$

where "$\lambda$" denotes all the hidden variables.

My problem is that such an integral is only defined for finite-dimensional sets
"$\{\lambda\}$", or sets of countably-infinite dimension.

This relates to an elementary theorem in general topology which says that finite
product spaces, and product spaces of countably-infinite dimension, are
metrizable, but product spaces of uncountably-infinite dimension need
not be.

Function spaces are generally uncountably-infinite, so maybe this is
why Bohmian mechanics got as far as it did - its "hidden variables" are
such a function space.

So it seems to me that Bell's theorem doesn't go through in such cases,
and therefore says nothing about local theories with an uncountably-infinite
number of hidden variables.

Or am I missing something?

9. Apr 22, 2008

### akhmeteli

Well, both I and OP were too lazy to mention that :-). Another reason, as far as I was concerned, was that THAT Bell :-) believed nonlocality is immanent to quantum theory as well.

I am afraid this is a non sequitur. Experimental evidence consistent with the predictions of QM is not "inconsistent with the predictions of any local realistic theory (per Bell's Theorem)" as long as it does not break the Bell inequalities. And as far as I know, no existing experimental evidence breaks those inequalities.

10. Apr 23, 2008

### Staff: Mentor

The entire point of the Aspect and other EPR-Bell type experiments is that they confirm the predictions of QM in regions where the Bell inequalities are broken.

11. Apr 23, 2008

### Demystifier

You misunderstood something.

12. Apr 23, 2008

### peter0302

Show me information transferred faster than the speed of light and then I'll believe in non-locality. Until then, what you call the "loopholes" look to me like symtoms of the fundamental problem, i.e., that locality is real and that realism is wrong.

13. Apr 23, 2008

### akhmeteli

I agree, that was the entire point of those experiments. The problem is, strictly speaking, they failed to make their point. Abner Shimony (and it seems he knows what he is talking about and he is no fan of local realistic theories of QM) wrote the following in his article on the Bell theorem in the Stanford Encyclopedia of Philosophy (http://plato.stanford.edu/entries/bell-theorem/#3 ):

"The incompatibility of Local Realistic Theories with Quantum Mechanics permits adjudication by experiments, some of which are described here. Most of the dozens of experiments performed so far have favored Quantum Mechanics, but not decisively because of the “detection loophole” or the “communication loophole.” The latter has been nearly decisively blocked by a recent experiment and there is a good prospect for blocking the former."

Therefore, as I said, until recently (the article is marked "Copyright 2004"), there have been no experiments demonstrating (without additional, more or less arbitrary assumptions, such as the "fair sampling" assumption) a violation of the Bell inequalities. Again, maybe something radically new has happened since then that I am not aware of. In such case I would appreciate if somebody more knowledgeable than I could enlighten me.

14. Apr 23, 2008

### akhmeteli

Certainly, I value your opinion and would appreciate if you could explain.

15. Apr 23, 2008

### ZapperZ

Staff Emeritus
I posted this just 2 days ago:

https://www.physicsforums.com/showpost.php?p=1698898&postcount=67

Zz.

16. Apr 23, 2008

### akhmeteli

Thank you very much for the reference.

For the benefit of others: as far as I could understand, the experiment in the cited paper (I guess there is a version in arxiv as well) demonstrates the Bell inequalities with the detection loophole closed. The locality loophole is still there big way though: to close it, they need to increase the spatial separation from 1 m to 15 km. Thus, it looks like the general conclusion still stands: there has been no experimental proof of nonlocality so far (44 years since the Bell's article), and I don't hold my breath waiting for such proof. And not just because of experimental difficulties or because local realism is any relative of mine, but because, for reasons outlined in my post #6 in this thread, such proof would mean that unitary evolution of quantum theory is not universal. And this may be too big for me to swallow.

17. Apr 23, 2008

### peter0302

Well, there've been other experiments in which the distances were sufficient to close the locality loophole. I know you want both loopholes closed in one experiment but I think you can syntehsize the two results and deduce what would happen...

18. Apr 23, 2008

### ZapperZ

Staff Emeritus
At some point, you need to step back and look at the body of evidence here.

There's a whole slew of experiments that have closed the locality loophole. All of them involved photons.

There's another slew of experiment that have closed the detection loophole. All of them involved some form of particles, from protons, neutrons, muons, etc.

Now, you are telling me that, somehow, each one of the still "open" loophole conspires to make themselves be THE factor that can still account for the apparent violation of Bell inequality. Don't you just find that to be rather a very unlikely coincidence?

There's another glaring aspect of this. All, and I mean 100%, of the experiments on Bell inequality all make the same claim of the violation. Now, one would think that someone who isn't convinced of this, or who is claiming that such-and-such a loophole is responsible for the apparent violation, would at least be able to conduct his/her own experiment, show the data, and argue conclusively that others who have performed the experiment have erroneously analyzed the data, and that the <insert favorite loophole here> loophole is there, in the data. Now, can you find me such an experimental report? I have found none. Don't you find this rather curious? Why is that?

I will tell you why. In the detection loophole, for instance, even without a 100% efficiency in photon detection, people who do such experiments have to first of all learn about the behavior of their detectors. Everyone who depends on any form of photodetectors have to do this, including high energy experimentalists. We need to know how these instruments behave, what they can do, and more importantly, what they can't do. We need to know when the data we have is reliable, and when we are over-reaching.

So when experiments involving entangled photons are performed, even without a 100% efficiency, we have an excellent idea of the performance of the detectors to say with reasonable confidence of what the actual data are. To me, it is why you have never, ever seen such experiments that contradict the conclusion of violation of local realism so far. It is because once you learn and understand the behavior of such detectors, you'd never pay attention to the weak "detection loophole" argument. The argument against the validity of a data set can only be made by experts not only in the physics, but also in the detection scheme. When Talayerkhan claimed to detect fusion in his bubble fusion experiment, his detractors were not someone who have no clue on the experimental method he was doing. In fact, many of them were world-renowned experts in neutron detections, and they pointed out exactly where the device he's using and the method he adopted can easily produced faulty results. This then threw a lot doubt in the data and subsequently the conclusion (I haven't yet mentioned the fact that others who tried to reproduce the experiment did not get the same result). I have never seen that done with any of the Bell experiments done so far. Considering that there have been plenty of such experiment, and with 100% agreement on the conclusion, I find the lack of contradicting experimental results to be a very obvious shortcoming of those who claim otherwise.

So to me, the continued stubbornness in proclaiming that local realism is still valid because this loophole is still open, or that loophole is still open, has nothing to do with not having convincing experiments. That's like saying Evolution isn't true simply because there are still "gaps" in our knowledge, or that QM isn't right because it still can't be reconciled with GR. It rings hollow because of what they all CAN do already, whereas the alternative have done nothing. All the experiments have produced ONE very convincing argument in favor of violation of non-local realism per the Bell theorem. That is what all these papers have argued and concluded. The paper that I had recently cited simply tried to start hammering down the last nail in the coffin - by being the first to attempt at closing BOTH locality and detection loophole simultaneously. The alternative, being local realism via the non-violation of Bell inequality, have ... er .... zero experimental evidence!

Zz.

19. Apr 24, 2008

### akhmeteli

Earlier in this thread you wrote the following:

I am somewhat at a loss trying to understand your point of view: so are "loopholes" symptoms of a fundamental problem or "you can synthesize the two results" and thus eliminate the issue of loopholes?

20. Apr 24, 2008

### akhmeteli

I don't know anything about any conspiracy. There is just no experimental evidence of violation of the Bell inequalities (the assumptions of the inequalities require sufficient spatial separation). What am I supposed to do? Consider the entire body of experimental evidence demonstrating no violations of the inequalities and conclude that the inequalities are violated? With all due respect, this is no easy task. The OP asked: how local realists can be so obstinate? It does not matter whether I am a local realist or solipsist or whatever. I just tried to explain that local realists' life is not as hard as OP believed. I hope my explanation was not absolutely useless as it looks like some part of it was new for the OP.

I readily admit that I don't know much about experiments testing the Bell inequalities. But my understanding is there is a consensus among knowledgeable people (including you), no matter what they think of the interpretation of QM, on what has been demonstrated experimentally and what has not (actually, you are not saying that I misrepresented the experimental situation. If you believe I did, please advise). So I don't see any need to perform any special "local realism inspired" experiments. I hope the bulk of the existing experiments were conducted by knowledgeable people, and I don't care whether they were local realists, positivists, solipsists, whatever. There is no disagreement on what experimental results were obtained. But everybody is free to draw their own conclusions, as long as such conclusions are consistent with the established experimental data.
Look, you know a lot about photodetectors, I know next to nothing about them. Does this mean that I am not in a position to have my own opinion on the violations of the Bell inequalities? I don't know if people who first raised the issue of the detection loophole knew a lot about photodetectors. What I know is that this issue is generally recognized as such. Otherwise why all these attempts to close the loophole? Why publish the results of such attempts in PRL? Are you telling me I must accept the fair sampling assumption just because I don't know anything about photodetectors? Am I supposed to believe in god just because I have not read the bible or the quran? Then in which god am I supposed to believe - the god of the bible or the god of quran? I just know that the fair sampling assumption is not generally recognized, so I am free to accept or to reject it. I choose to reject it, and I am left with no experimentally observed violations of the Bell inequalities.

OK, forget about local realism for a moment. In my posts I was trying to explain why I don't admit (as long as it is possible without a contradiction with experimental data)that there is any experimental evidence of violations and why I don't expect any such evidence to appear. You might have noticed that my reasoning had little to do with local realism. Furthermore, I actually swore by quantum theory. You see, you cannot prove the Bell inequalities without the projection postulate (please advise if I am wrong). The projection postulate introduces irreversibility, whereas unitary evolution of the quantum theory allows no such thing. You cannot have it both ways - unitary evolution and the projection postulate. You have to choose. I choose the unitary evolution, which is quantum theory, pure and simple. You may say: but we have to accept both mechanics (classical or quantum), where there is no irreversibility, and thermodynamics with its irreversibility. Yes, but we understand that the irreversibility of thermodynamics is just a very good approximation, as the underlying mechanics does not allow any rigorous irreversibility. You need some "manual" interference, however subtle, to obtain irreversibility. And it seems that the analogy between mechanics vs. thermodynamics, on the one hand, and unitary evolution vs. projection postulate, on the other hand, may be deep enough, as arXiv:quant-ph/0702135 (Phys. Rev. A 64, 032108 (2001), Europhys. Lett. 61, 452 (2003), Physica E 29, 261 (2005)) demonstrates, using a rigorously solved model, how the projection postulate evolves as a result of thermodynamic irreversibility.