Talk:Bell's theorem/Archive 5

Archive 1

←

Archive 3

Archive 4

Archive 5

Archive 6

Archive 7

→

Archive 10

Let's have another poll. How many supporting exclusion of Dr. Christian's work are physicists?

I would like to know how many of you supporting exclusion of Dr. Christian work on Bell's theorem are actually physicists? And are experts on Bell's theorem? I suspect that Dr. Christian is the only physicist involved in this current dispute. The article on Bell's theorem is a physics article and Bell's theorem is specifically about the physics of the EPR-Bohm scenario. Editing of this article should be done by a professional physicist that is an expert on Bell's theorem. Is there any other qualified physicists here on this talk page besides Dr. Christian? Fred Diether (talk) 05:59, 2 June 2012 (UTC)

In response to Fred Diether, that is not how Wikipedia works. Per Wikipedia policies a group of 12 year old students can vote and a Nobel laureate can be outvoted. That is how the "free market" economy of Wikipedia works. You may want to try Citizendium which does not have a free market model. They have less than 20,000 articles vs the 4 million of Wikipedia. But then you would need to convince the experts there to be included. And some of them may have the habit of looking through Physical Review. But please feel free to try that. In any case, I think this suggestion was in effect an acknowledgement of the fact that in the section above clear consensus has been reached for the exclusion of Christian's item. I will hence move to declare consensus in the above discussion. History2007 (talk) 07:35, 2 June 2012 (UTC)

The suspicion is wrong (and not all that relevant as History2007 points out above). Count Iblis (talk) 15:07, 2 June 2012 (UTC)

Your request to find out who here are physicists shows a deep misunderstanding of the underlying issue here. It's not about whether Christian's physics are solid or not - it's about Wikipedia policies on non-mainstream views and especially the policies of WP:UNDUE and WP:FRINGE - both of which clearly apply here. You freely admit that mainstream physics is locking out Christian's work - and that's why we're excluding it from the article. What you see as an injustice that needs correcting, Wikipedians see as a fact about the real world that we are mere observers to. It's not our job here to right wrongs of an intellectual nature. SteveBaker (talk) 16:01, 2 June 2012 (UTC)

As others have stated, it's a moot point, but I will gladly out myself as a physicist. a13ean (talk) 17:55, 2 June 2012 (UTC)

And me too. IRWolfie- (talk) 13:47, 3 June 2012 (UTC)

I think On the Internet, nobody knows you're a dog is relevant here, and Wikipedia is pretty hot on forbidding attempting outing of dogs or whatever are tapping away out there. Dmcq (talk) 22:34, 2 June 2012 (UTC)

Time for a compromise for "disproof" dispute

I think it is time to attempt to work out a compromise for this dispute. Which I am sure that Wikipedia would most appreciate a compromise of some kind so that all parties might be happy and not attempt more edit warring in the future. I would suggest that mention of Dr. Christian's work be allowed with a disclaimer of some kind that it is not mainstream opinion. Fred Diether (talk) 06:12, 2 June 2012 (UTC)

As Sławomir Biały stated above, if suitable WP:Secondary WP:RS sources of note are found to provide a review of the opposition to Bell's Theorem they may be included later, provided a number of conditions regarding WP:Due, etc. are met. But that is a separate issue from the discussion above, which has now reached consensus on exclusion. History2007 (talk) 07:40, 2 June 2012 (UTC)

No “opposition to Bell's Theorem” is needed. Bell’s theorem is dead. All one needs to do to see that is to be able to understand the four or five lines of calculation in this paper: http://arxiv.org/abs/1103.1879 -- Joy Christian — Preceding unsigned comment added by 86.148.6.36 (talk) 08:56, 2 June 2012 (UTC)

In any case, Joy Christian was upfront about it, accepted the consensus in the above discussion, and it has now ended. Thanks. History2007 (talk) 08:45, 2 June 2012 (UTC)

I would actually support mentioning someting about this issue, but I doubt that would be something that Joy or you would like to see in this article. The bottom line is that the judgement of the physics community is negative about these attempts to find a local deterministic framework for QM. Then that has to be the context in which anything about this is going to be mentioned. I believe that we should say something along the lines of the physics community having given up on local determinsitic theories, that there are only a few attempts made in this direction, and that these have been judged to be flawed or otherwise problematic. Then one can mention Joy and some others in the footnotes with some comments.

By saying this, one illustrates better that such attempts only have an infinitesimal amount of support. If we don't say anything, a reader may still have read about a discussion involving Joy somewhere else, and think that the Wiki article hasn't been updated to include this. Count Iblis (talk) 15:28, 2 June 2012 (UTC)

I agree. No mentioning anything will make the reader wonder given that web searches produce some results to that effect. But the context for it has to be that the vast majority of physicists do not buy it. The issue remains proper sourcing of course and the guideline to follow is WP:RS/AC if that type of overview statement is to be made. And I think if a "list of suggested improvements" is made now, that should fit into a restructuring of the article. History2007 (talk) 15:37, 2 June 2012 (UTC)

I agree, too. One could mention Accardi's "chameleon" model (built on the detection loophole), Hess and Phillip's micro time model (built on an error in their notation), de Readt's event simulation modes (built on the detection loophole and on the coincidence loophole), Caroline Thompson's chaotic ball model (built on the detection loophole), Joy Christan's model (built on an algebraic error and on redefining the definition of correlation). But I think it is more important to get the first part of the article clear and clean. People have doubts because they're not convinced by some lines of calculus and formula manipulation. I think the article needs some more modern, more direct, more intuitive proofs of Bell's theorem centering around counterfactual definiteness and simple logic. A 12 year old can understand a good proof of Bell's theorem. But not the original or traditional proofs.

But first of all we need to cut away a lot of the dead wood which has accumulated over the years.Richard Gill (talk) 09:46, 3 June 2012 (UTC)

Actually the way I usually do cleanups is to improve the peripheries first, then the main topic. The article on Superdeterminism is no gem and needs work, as does Principle of locality, etc. Once those are in really good shape the top level can refer to them. So my suggestion would be to clean those surrounding articles first, build a basis and then rework the main. However, as you said there is plenty of deadwood that just needs to be zapped here. History2007 (talk) 09:56, 3 June 2012 (UTC)

We've had the debate - we've tested the consensus - the results are clear. No compromise is necessary - and indeed, any kind of compromise would also fail to get consensus because the grounds that almost all of the "Support exclusion" people gave is that Wikipedia does not permit this kind of fringe view to clutter up articles about mainstream science. The couple of "Object to exclusion" !votes talk only of physics and "fairness" - which really aren't relevant to the discussion. I understand your frustration - but the issue isn't with this article - or about the people who are telling you "No" - it's about the core policies here at Wikipedia. You can try to take up the cudgel at WP:FRINGE or WP:UNDUE - but those are well-established principles at the core of Wikipedia and they are unlikely to change. However, the debate here is over - and it's time for you to drop the stick and back slowly away from the horse carcass. SteveBaker (talk) 16:09, 2 June 2012 (UTC)

Well, you all can continue your posturing or take this issue to mediation, arbitration or whatever but I do believe Wikipedia would prefer to see a compromise attempted first. This section is about what Christian and Gill will accept to stop their warring. Comments that don't pertain to that are not necessary. How many here have actually looked at the content of the article that is in dispute? Here it is,

"However, no principle of physics can ever be proved absolutely beyond question; some theorists argue that experimental loopholes or hidden assumptions refute the theorem's validity,[9][10][11]" Reference 10 points to Dr. Christian's one page "disproof" on arXiv.

So the question is what will Christian and Gill accept for changes to that to stop the warring? Let's work on that. Thank you. Fred Diether (talk) 16:32, 2 June 2012 (UTC)

No, I am sorry Fred Diether, it does not work that way. As SteveBaker said the discussion took place, consensus was reached, and J. Christian agreed that consensus was achieved. In other words: the party is over. Should you continue to argue against consensus, the issue of WP:TE will arise. Regarding your reference to edit wars, let me assure you that Wikipedia does have policies to quickly deal with edit wars and unending discussions against consensus. History2007 (talk) 16:40, 2 June 2012 (UTC)

I am sorry, but you missed Dr. Christian's sarcasm. He did not accept consensus. I am sure Wikipedia would in fact prefer a compromise to this dispute so that it is settled for all time. Or take it to mediation or arbitration if you wish. The fact is, compromise is most always a better solution. Enough said; let's get to work on a compromise. Fred Diether (talk) 17:03, 2 June 2012 (UTC)

Consensus was reached. Clearly. Now, in this edit I left you a note about WP:REHASH regarding "repeating the same argument over and over again, without persuading people." Please avoid that. History2007 (talk) 17:31, 2 June 2012 (UTC)

We should take Dr. Christian at his word; otherwise we might rationally assume that his papers were parody.

However, if we could find published commentary on previous (failed) attempts to disprove Bell's Theorem, that would be appropriate to include. — Arthur Rubin (talk) 17:34, 2 June 2012 (UTC)

This suggestion of a "compromise" is laughable. It's as if someone wanted to cite a paper in the global warming article that claims to prove that pirates cause global warming. When the community responds overwhelmingly in an RfC that the work has not been discussed in any reliable secondary sources, the proponents then demand the addition of a line like "Some individuals have suggested that pirates cause global warming, although this is not a mainstream opinion". Not only is such a statement expressly forbidden by the WP:NOR policy (we commit original research with our assertion that it is "not a mainstream opinion"), but even this small mention places a gross overemphasis on a thoroughly fringe viewpoint (WP:WEIGHT). Sławomir Biały (talk) 00:58, 3 June 2012 (UTC)

I agree. The lager point is that if the door is opened to allowing WP:RS to be compromised here, eventually statements like the pirate example will come in elsewhere and then it is just a question of time before Elvis is found to be alive, and living a block from WMF headquarters. History2007 (talk) 08:21, 4 June 2012 (UTC)

• UPDATE: To make it more clear why compromise for this dispute is a goood idea here is an update that I probably should have put at the very start of this section. I contend that since Dr. Christian is an expert physicist on Bell's theorem as evidenced by funding from FQXi (FQXi is not going to fund someone that is not an expert) on this very topic that Dr. Christian himself can be considered a reliable source for the disputed content according to Wikipedia policy. And for the same reason the disputed content is not fringe. So no Wikipedia guidelines or policies are being violated in my viewpoint of the dispute. Compromise is usually a good solution to have a stable article with a well balanced viewpoint. Thank you. Fred Diether (talk) 04:12, 4 June 2012 (UTC)
• WP:SPS states that, self-published expert sources may be considered reliable when produced by an established expert on the topic of the article whose work in the relevant field has previously been published by reliable third-party publications. There are none; arxiv.org and Dr. Christian's self-published book are not "reliable third-party publications". — Arthur Rubin (talk) 04:49, 4 June 2012 (UTC)

Bell's theorem is pretty old and there is not much new content that could be written and published about it other than an alternative viewpoint or Bell test experiments. And we both know that due to extreme prejudice that exists for this topic that it is totally impossible for any kind of "disproof" to get published in a peer-reviewed journal of stature. Dr. Christian has been published in Physical Review D on another topic and in other journals. The Wikipedia policies and guidelines are just guides and don't really have to be taken to the "letter of the law" in every single case. And should not be in this case. I just don't understand why no one wants to even try to compromise on this dispute. It is ridiculous and supports my supposition of extreme prejudice. Fred Diether (talk) 05:38, 4 June 2012 (UTC)

I am sorry if you feel the whole world is against you. But Wikipedia is not the venue for dealing with those grievances. History2007 (talk) 07:45, 4 June 2012 (UTC)

I am glad that "Dr. Christian has been published in Physical Review D on another topic"; but on this topic, I fail to understand his point, see User talk:Tsirel#My model for EPR -- Joy Christian. Why do you think that "due to extreme prejudice that exists for this topic that it is totally impossible for any kind of "disproof" to get published in a peer-reviewed journal"? You did not try a valid disproof yet; and an invalid disproof is naturally rejected; no prejudice. Boris Tsirelson (talk) 08:25, 4 June 2012 (UTC)

"there is not much new content that could be written and published about it" -- really? See this very selective bibliography. Boris Tsirelson (talk) 09:09, 4 June 2012 (UTC)

Now it is time to get to the real work needed here

There is a lot of work needed to get this page up to speed. We all know it is weaker than it should be for a topic so important. I personally plan to work on removing excess that has little to do with Bell itself. In addition, I would like to update some of the references. For example, there is no reference to the important Innsbruck experiment, "Violation of Bell’s inequality under strict Einstein locality conditions".

I take it as a given that this decision applies to other similar attempts to insert fringe and/or unreliable sources on the Bell page. I have removed, for example, the Nagasawa references as being completely unsuitable. So hopefully Interintel will honor the results. I will be reverting any attempt to re-insert them.

My question is about the Jaynes and 't Hooft references. I see no value in the Jaynes comment, other than as a minor historical point. There have been numerous other challenges to Bell which have also been dismissed. I rarely see mention of him except by members of the local realist fringe. 't Hooft is questionable too. Superdeterminism is rightly a concept that could apply to all experiments done, having nothing whatsoever to do with Bell. For example, 't Hooft could similarly assert that relativity does not exist, that instead all tests of the value of c are a result of the initial conditions of the universe and superdeterminism. These authors are respected, but their work on Bell is not considered an important part of the literature. So I would propose to remove these.--DrChinese (talk) 17:46, 2 June 2012 (UTC)

I would support a mayor rewriting effort by you. Also removal of the text that refers to Jaynes and 't Hooft (I think 't Hooft is only mentioned in one sentence anyway and that could be moved to the main article on superdeterminism we have here).

The only thing I disagree with is your argument about 't Hooft argument, but then that's not so relevant for editing this article. While I personally don't believe in superdeterminism, 't Hooft does motivate in one of his articles why superdetermistic effects would be relevant. When the counterfactual reasoning involves the same microstates then it's not plausible that a different experimental set-up would not have affected that. You need to start with some well defined class of initial states and then use whatever degrees of freedom are left when considering changes in the present state.

So, it is wrong to just assume that changing the setting of a polarizer in one place would not have affected the state somewhere else, even though the theory is locally deterministic. Local determinism refers merely to the rules by which you can calculate the future state from the past state. Now, it may be that this objection can be adressed by modifying the proof of some Bell inequalities so that you are always explicitely sampling from a set of physically acceptable initial states (of course, this will involve some assumptions, but then these would be more plausible in the context of such theories).

Proving 't Hooft wrong may thus well be possible in a reasonable physical setting, but of course not if you take superderminism to mean that "you can always invoke superdeterminism". Count Iblis (talk) 20:12, 2 June 2012 (UTC)

I strongly object to David (Dr. Chinese) to edit or re-write this article. He is not a professional physicist and has an extremely strong bias about Bell's theorem. This article needs attention from a professional physicist that would hopefully be neutral. Perhaps Dr. Christian or Gill could help in the enlistment of such a professional? Fred Diether (talk) 20:55, 2 June 2012 (UTC)

My friend, let me break the news gently here: Wikipedia can be edited by any 12 year old. That is policy. There are only two requirements for editing Wikipedia articles: a heartbeat and a modem. You may really prefer the model used in Citizendium, as mentioned above. Yet I should quickly add that the Wikipedia model has produced a wealth of knowledge so far - Citizendium has not. In my experience problems are not caused by the 18 year old editors - they are usually still idealistic, hopeful and constructive. Problems come from discussions like this which ignore that now forgotten cliche... The time wasted here could have been put to much better use. Please follow consensus and avoid WP:TE as mentioned before. Thanks. History2007 (talk) 21:23, 2 June 2012 (UTC)

Your attempts at intimidation are pretty lame. I suggest that you stop. Please read the very first part at the top of the article. Wikipedia is requesting attention for the article from an expert. That means a professional physicist that has peer-reviewed published work about Bell's theorem to me; not from some software developer that has no peer-reviewed published work about Bell's theorem. Nor from some 12 year old kid. Thank you. Fred Diether (talk) 21:51, 2 June 2012 (UTC)

Fred, I am not asking your permission. And I will not respond to your baiting. The pool of folks here is easily strong enough to manage this task.--DrChinese (talk) 23:04, 2 June 2012 (UTC)

David, sorry that you think I am baiting you; I am not. I am just trying to prevent future disputes from happening and I am suggesting that we comply with Wikipedia's request to have expert attention to this article. Your unpublished internet articles do not make you an expert on this topic. Really, compromise is usually the best solution. Fred Diether (talk) 00:00, 3 June 2012 (UTC)

I'm sorry, but as you've been told, this is not how Wikipedia works. We don't check credentials at the door. We are the free encyclopedia that anyone can edit. Sławomir Biały (talk) 00:51, 3 June 2012 (UTC)

It seems that you missed the phrase "I am just trying to prevent future disputes from happening...". Is anyone actually paying attention to the problem here? Is there anyone here that is really interested in making this article stable? It seems to me that the best way is for some kind of compromise to accomplished. Your example above in the compromise section is laughable; it does not apply here however since Dr. Christian himself can be considered a reliable source for this topic given his credentials. So once some of you can get past that and are willing to compromise then we can get this article to be stable and well balanced. Thank you. Fred Diether (talk) 02:18, 3 June 2012 (UTC)

There are a number of items throughout the article that have been thrown in to undermine the Theorem. Example: "Accordingly, experimental results to date cannot be regarded as definitive proof that locality or counterfactual definiteness must be rejected." (All loopholes have been closed, but not all in a single experiment.) There are also an unusual number of references to the loopholes. In addition, it is not clear that either locality OR realism (aka CD or hidden variables) must be jettisoned. This has the effect of implying a Bohmian type interpretation is a consequence, when it is not. For example: "...if Bell's conditions on local hidden variable theories are correct, then the results which are in agreement with quantum mechanical theory appear to evidence superluminal effects.". I plan to remove those as well.--DrChinese (talk) 23:04, 2 June 2012 (UTC)

DrChinese, let me help you out in your noble task. Wikipedia is seeking help from an expert to clean up the article on Bell’s theorem. Let me help you and Wikipedia to do that. Here are my credentials: I did my PhD under the world authority on Bell’s theorem, Professor Abner Shimony, the “S” in the Bell-CHSH inequality, and learned about Bell’s theorem partly from John Bell himself (yes, from John Bell himself). Another great pioneer of Bell's theorem, the Nobel Laureate Eugene Wigner, was my intellectual grandfather (mentor of my mentor Abner Shimony). Mike Horne (the first “H” in the Bell-CHSH inequality) is my intellectual brother (the first and the most famous physics student of Abner Shimony). Young Lucien Hardy (of the famous Hardy’s version of Bell's theorem) is one of my best friends (I was the Best Man at his wedding). I have known several other Bell pioneers like Greenberger, Zeilinger, and Gisin for many years, since my student days at BU in the 1980’s. I wrote my Master’s thesis on Bell’s theorem in 1985, and PhD thesis in 1991. I can go on, but let me instead mention that I have much wider knowledge of physics, philosophy, and mathematics than the four lines of Bell inequalities. After doing my doctorate in Foundations of Quantum Theory under the supervision of the renowned philosopher and physicist Professor Abner Shimony (the “S” in the Bell-CHSH-inequality), I received a Research Fellowship from the Wolfson College of the University of Oxford in foundations of physics, where I have remained affiliated both with the college and a number of departments of the university. I am an invited member of the prestigious Foundational Questions Institute (FQXi), and have been a Long Term Visitor of the Perimeter Institute for Theoretical Physics, Canada. I am well known for my contributions to the foundations of quantum and gravitational physics, including quantization of Newton-Cartan theory of gravity, generalization of Special Theory of Relativity to incorporate the objective passage of time, and elimination of non-locality from the foundations of quantum physics. A partial list of my publications can be found here: http://arxiv.org/find/all/1/au:+Christian_Joy/0/1/0/all/0/1. Considering these humble credentials, would you allow me to help you clean up the article on Bell’s theorem DrChinese? -- Joy Christian

Joy, you are welcome to edit this article, however, until that time that your theory gets a better reception in the scientific community, we cannot include it (at least not in the way it was proposed). Look, if 't Hooft wanted to include details about his determinsitic models in this article, we would also reject that, so the opposition here has nothing to do about your scientific credentials. Count Iblis (talk) 02:11, 3 June 2012 (UTC)

I agree with Count Iblis, and per WP:REHASH this repetitive "resume posting" by J. Christian needs to stop, given the consensus above. History2007 (talk) 07:42, 3 June 2012 (UTC)

It does have everything to do about Dr. Christian's credentials. Wikipedia allows an author himself to be a reliable source if they are an expert in the subject. You all are trying to disallow that. Wrongly, I might add. Compromise or I expect this dispute will be going to mediation or arbitration. Fred Diether (talk) 02:27, 3 June 2012 (UTC)

Lev Vaidman would not be allowed to rewrite the Aharanov-Bohm effect article based on his recent article, at least not yet. At most a brief mention might be possible (I would support that), but even that would probably not be supported by most of the editors there. Count Iblis (talk) 02:55, 3 June 2012 (UTC)

You too can knock off the posturing. We both know that compromise can lead to a stable article with a well balanced viewpoint. Fred Diether (talk) 05:05, 3 June 2012 (UTC)

No, compromise would unbalance the article as has already been thoroughly explained, the material just doesn't belong in this article. I suggest you do not push this issue further due to your clear conflict of interest. IRWolfie- (talk) 12:48, 3 June 2012 (UTC)

From the WP:COIN notice board, "Furthermore, accusing another editor of having a conflict of interest in order to gain the upper hand in a content dispute is prohibited and may result in sanctions against you." I strongly suggest that you also back off the posturing. Your intimidation tactics are also lame. I guess we are going to end up at arbitration with this if not willing to compromise. Fred Diether (talk) 20:28, 3 June 2012 (UTC)

That is a noticeboard, not a policy page, and in any case that requires an intent to gain an upperhand in a content dispute. Wolfie has made only 2 edits to the page and I have never edited the page, so that angle is closed. And there is absolutely no question that J. Christian has a WP:COI. No question whatsoever, given that he is discussing his own book. And from what I have seen on the SPI discussion, you have your own theories about the subject. Do you, or don't you? In any case, this will probably be on WP:AN before arb, just due to the established consensus that the ideas are WP:Fringe. Wikipedia policies do not allow compromises on WP:Fringe items. So if you want to compromise that "the earth is semi-flat", else will go to arbitration, that is a no-go approach - pun intended. History2007 (talk) 20:44, 3 June 2012 (UTC)

I too would like to help improve the article in the coming months (am on vacation right now). I think a separate page is needed on loopholes, that is the place where super determinism can be discussed. Not just the proposal, but also the reasons why it is not generally accepted. I think the reference to Jaynes can go. I would like to see mention of Masanes, Acin and Gisin's work http://arxiv.org/abs/quant-ph/0508016 showing that no-signalling (no action at a distance) plus non-locality (by which they mean violation of Bell inequalities) implies non-determinism. Regarding secondary or tertiary sources on constant fringe attempts to de-fuse Bell, there is a paper by myself with an overview of fringe theories some years back "Time, finite statistics and Bell's fifth position", http://arxiv.org/abs/quant-ph/0301059 . This was published in a conference proceedings volume. Andrei Khrennikov has organized a sequence of conferences on the topic where main-stream and fringe speakers mingled amicably. But I should not play any role in this editing as far as it involves my own work. I'm not aware of other secondary and tertiary work on fringe theories collectively, only individually. Richard Gill (talk) 08:50, 3 June 2012 (UTC)

• Should we try to write a new lede first? What is now in the article is hardly satisfactory, since a non-involved reader can not understand anything.--Ymblanter (talk) 09:55, 3 June 2012 (UTC)

Yes, in this case a top down approach may work, given the clutter in the sections below it. And please see my comment above about cleaning up the peripheral articles as well. History2007 (talk) 09:58, 3 June 2012 (UTC)

I completely agree regarding the overall presentation. The lead is poor, and it never gets better. Many points are repeated, often in sections that are focused on something else. Although a lot of information is presented, yet there is no real sense of perspective. I think starting with the lead works best, in part because that is the biggest takeaway for many readers. I.e. good payback on time, although no way could we stop there.

Ultimately, I think that Bell is best presented (somehow) as a member of the trilogy: EPR, Bell and Aspect. That is really how it is understood by the scientific community at large, even though technically it stands on its own. I am including the following link to one of my Bell pages in case anyone here is interested in seeing an example of how that approach is presented. My page is not intended to be anywhere near as comprehensive as this article needs to be, it is shorter and hopefully sweeter: DrChinese Presents Bell's Theorem In addition, for a long time the following link was a reference on this article, although it was removed at some point. EPR, Bell and Aspect: The Original References (My site is currently also the host for the PDF version of the original paper in the article.)--DrChinese (talk) 15:44, 3 June 2012 (UTC)

Given that "DrChinese Presents Bell's Theorem" is not a wiki, I propose a correction here: "That a single theory could bring us to the end of a such an important debate" --- probably, "single theorem" rather than "single theory"? Boris Tsirelson (talk) 12:40, 4 June 2012 (UTC)

Boris, thanks for this. You are so right, I have made that correction.--DrChinese (talk) 15:53, 4 June 2012 (UTC)

My favourite proof of Bell for precocious 12 year olds is User talk:Gill110951#Bell's theorem on my talk page. It's the Mermin approach (built on counterfactual definitenrss and simple arithmetic) applied to CHSH and tightened up a bit. It came out of giving a lecture on Bell to art students, and discussing Bell on various Internet fora in recent months. (This is where I get banned for promoting OR). I think it's common sense and elementary arithmetic.Richard Gill (talk) 11:38, 4 June 2012 (UTC)

And here's another excellent source: http://en.citizendium.org/wiki/Entanglement_(physics), and http://knowino.org/wiki/Entanglement_(physics) . Both by Boris Tsirelson. Richard Gill (talk) 11:51, 4 June 2012 (UTC)

In response to both of the comments, a reliable secondary source is required for the proof. I am not sure if citizendium falls afoul of the open wiki's are all non-reliable since it isn't exactly open; but it is definitely not up their with peer review but maybe it has a role in the lede by providing an overview since it's a tertiary source like wikipedia. From here: Wikipedia:Reliable_sources/Noticeboard/Archive_7#another_wiki.3F it looks like citizendium does have issues with reliability, i.e it can be used as an external link but not a source. IRWolfie- (talk) 12:37, 4 June 2012 (UTC)

Both Tsirelson and my day jobs are writing primary and secondary sources in mathematics. Boris could publish his Citizendium text on arXiv.org and his home page and then it slowly becomes a reliable source. I have done the same with new (pedagogical) ideas I get while editing Wikipedia. Other editors can use or ignore such spin-off on their own merits, later. Richard Gill (talk) 04:29, 5 June 2012 (UTC)

What is it about? My text on Citizendium contains no proofs. After a general introduction, it presents two examples with statements and comments, not proofs. Comments and explanations are more or less mine, but examples and statements are not. They are borrowed from the cited literature (quite reliable sources). There, proofs can be found. In particular, try this:

Buhrman, Harry; Cleve, Richard; Massar, Serge; de Wolf, Ronald (2010), "Nonlocality and communication complexity", Reviews of Modern Physics, 82 (1): 665–698 (also arXiv).

Boris Tsirelson (talk) 06:40, 5 June 2012 (UTC)

Ok, I have re-written the lead. I realize that every person has their own spin on things. I would recommend that unless what I have written is incorrect or misleading in some way, we leave as close to this as possible. It is intended to serve only as a lead that would be useful to someone who wants to get a high level summary quickly. Instead, I would suggest that we use the detail sections to provide a greater wealth of background rather than placing here. If the lead does not follow usual guidelines in your opinion, let me know and I will try to edit with that in mind.

After we get a decent lead, it should help us to work on the sections. We may need to change a few section titles in the process.--DrChinese (talk) 23:12, 4 June 2012 (UTC)

Between the first sentence and the second, I would like to see mentioned the fact that quantum mechanics is non-deterministic. There will be readers who are not into physics and who need to be told explicitely that quantum mechancis makes statistical predictions, and that the question then arises if there exists a hidden variable theory that would explain the outcome of individual measurements. Count Iblis (talk) 23:33, 4 June 2012 (UTC)

Good suggestion, I think I can write something that will make sense of that. I will add something once I get a few more suggestions. My goal is to make sure that we don't create a Frankenstein article by adding one thing here after another. So I am going to try to make sure the net effect of each change does not cause that. BTW, to all: I am not trying to control anything, just want to see a good result. If you look at Shimony's Stanford article, you can see the upside to a degree of consistency. Of course, that one is quite technical in comparison to what we need here. I would like to assist in some of the historical background and overall perspective. I hope someone else will come up with a solution to how to best present the proof itself. I have several different proofs that I present on my own website. Although they get the same basic answer, all the proofs that are usually cited are different than's Bell's own. The issue is that Bell had a different situation than we do today. He originated the idea, and wasn't worried about presentation to anyone but technical readers. Most of the easy to understand proofs (like the ones I use) do not strictly follow Bell in his particulars.--DrChinese (talk) 00:27, 5 June 2012 (UTC)

I would suggest two proofs are given: 1) Bell's CHSH proof in his "Bertlmann's socks" paper. This is Bell's own final word on the proof of his theorem, it is connected to the famous experiments, the CHSH inequality is most often nowadays thought of as the paradigmatic Bell inequality. 2) Mermin's proof with the three angles (this is a two party, three measurements per party, two outcomes per measurement experiment, it corresponds to a " generalized" Bell inequality for that situation). This connects to the modern work investigating "all Bell inequalities", GHZ, CGLMP and so on. Mermin's proof uses counterfactual definiteness, arithmetic and logic. Bell's uses simple calculus / formula manipulation. Different people find different styles of proof useful. That's why I think these two are a good contrast. (PS I like the lead so far. EPR has a Wikipedia page, must add the link). Richard Gill (talk) 04:15, 5 June 2012 (UTC)

Thanks Richard, good suggestions. We may want to include more than one. When I think about it, Bell doesn't actually require a generalized proof. The minimum requirement is a (i.e. any) single counterexample (such as Mermin's 0/120/240) in which you cannot construct a dataset which reproduces the QM expectation. Mermin's approach then reduces to an explanation of WHY such a dataset cannot be constructed for that particular choice of angles. The way I see it, once Bell threw out his particular counterexample, he really exposed the crack in the foundation of Einstein's realistic viewpoint. After that, we saw how to construct a variety of roughly equivalent proofs. So I teeter between embracing Bell's original one (which I think can be cryptic in some ways) or embracing the newer ones (which are far easier to convey).--DrChinese (talk) 15:10, 5 June 2012 (UTC)

The proof via CHSH is also a proof via a single counterexample: the correlations 1/sqrt 2, 1/sqrt 2, 1/sqrt 2, -1/sqrt 2. It has yet more significance: for a 2 party, 2 measurements per party, 2 outcomes per measurement experiment, the set of CHSH inequalities is necessary and sufficient for local realism. In other words, any 2x2x2 counterexample is actually a proof by CHSH. It connects to the experiments (Aspect, Weihs, ..). It's related to EPR: two complementary observables per party. Finally, when the proof is written out using counterfactual definiteness rather than the factorization and hidden variables as in User talk:Gill110951#Bell's theorem, it is as simple as Mermin's.

On the other hand, there is already a (not very good) article CHSH inequality. Maybe best to improve that one, first. Richard Gill (talk) 18:40, 5 June 2012 (UTC)

Yes, that's an issue too. We are going to have to pick a few of the key related articles and re-work them too. Part of the issue is: how much do we duplicate (i.e. the text is here AND it is there)? Hmmm.--DrChinese (talk) 20:31, 5 June 2012 (UTC)

As a wry aside (re Shimony's Stanford article) : I missed the reference to his former student's "groundbreaking" work on that page.--DrChinese (talk) 00:27, 5 June 2012 (UTC)

Cyclic debate: Time for WP:AN?

I really do not remember how many times I have seen the long paragraph about J. Christian's lifestory/thesis adviser being the “S” in the Bell-CHSH inequality etc., etc. on this talk page. It was posted again now, just above. Could we have an agreement that we all know that paragraph by heart now, and we do not need to read it again and again? As for compromise on including J. Christian's views, that has been discussed and soundly rejected above per WP:Fringe. There is no need to re-run that debate.

When User:Salix alba suggested a topic ban due to WP:COI two days ago, I felt it was not necessary. But I am beginning to agree with him. Should this cycle continue, I think his suggestion should be followed on WP:AN and a topic ban on J. Christian and Fred should be sought. There are enough users here for that to happen, given the clear existence of WP:COI and the general agreement that the views are subject to WP:Fringe. I see no need for a ban on Richard Gill, however. It would take 2 days of discussion and wasted time, but we are already spending that effort here anyway. Wikipedia can not be Fringe/COI powered. History2007 (talk) 08:03, 3 June 2012 (UTC)

It seems we are getting a lot more SPAs coming along though as well. IRWolfie- (talk) 12:46, 3 June 2012 (UTC)

I was trying not to say the same thing. If you want to ask for a quick check on those 2 or 3 user accounts and IPs as possible WP:SPI sleeper cases it will not hurt. All that can happen is that we may get an answer that none relate to each other. History2007 (talk) 12:50, 3 June 2012 (UTC)

I don't think it's a case of sock puppetry but meat puppetry. I'll launch the SPI investigation anyway, I may be the wrong/partially wrong. IRWolfie- (talk) 12:59, 3 June 2012 (UTC)

In fact, Sławomir hit mark here with a good search. History2007 (talk) 14:18, 3 June 2012 (UTC)

There was a comment above by Fred Diether to IR Wolfie to the effect that "either compromise or we go to arbitration". Given that there is a clearly established consensus above that the ideas are WP:Fringe and lack WP:RS sources (and the direct acknowledgement that mainstream scholarly sources refuse to publish them), Wikipedia policies do not allow compromises on them. As stated above there can be no compromise that "the earth is semi-flat" once it has been established that the flat-earth theory is subject to WP:Fringe. This may well be one more reason for going over to WP:AN. History2007 (talk) 21:00, 3 June 2012 (UTC)

Yes, something needs to be done alright. Your and IR Wolfie's intimidation tactics don't scare me and I am probably going to report your behavior and disruption of this dispute discussion to WP:ANI. Let me clue you in on something; consensus only works if all parties agree to it. Same with mediation. That is why I believe this content dispute will end up in arbitration if you are not willing to compromise as I have repeatedly requested. Fred Diether (talk) 21:28, 3 June 2012 (UTC)

Please be my guest. I have absolutely breached no policy whatsoever. FYI: arbitrators will have to agree to take the case. Not likely here. You will have to go to try another venue first. There were 13 people who argued against you, not just myself and IR Wolfie. And given that you have a WP:SPA account with a total of 30 edits in the past 5 days only to this talk page and never to another Wikipedia page, the situation is somewhat ironic as you argue against the overall trend. And per WP:LOCALCONSENSUS editors can not locally compromise to breach policy by using non-WP:RS fringe sources. And per WP:CON, consensus against you does not require your agreement. History2007 (talk) 21:32, 3 June 2012 (UTC)

(To Fred Diether:) Your notion that the editors opposed to inclusion are somehow obliged to compromise is ... strange. Consensus for exclusion has already been established. Before you report alleged disruption to WP:ANI, please read WP:BOOMERANG.  --Lambiam 11:05, 4 June 2012 (UTC)

Fred Diether: "consensus only works if all parties agree to it" - this is another misunderstanding you have of Wikipedia policy. Check out WP:CONSENSUS - right there in the first paragraph, it says "Consensus on Wikipedia does not mean unanimity". SteveBaker (talk) 13:34, 4 June 2012 (UTC)

This has truly gone too far. Wikipedia operates by consensus. We gave the proponents of Christian's work a long debate and a proper consensus check (a !vote). Not only do we have consensus - but not one of the minority arguments swayed a single majority editor in the least degree - which is a strong indication that views have now solidified and that further debate won't alter the consensus. This is how you know that "it's over".

Hence, there is now a stunningly clear consensus involving many experienced editors who understand Wikipedia policies and guidelines on such matters - and that should be the end of it. Further demands for compromise and circumlocutions of arguments that have already been soundly defeated have now become a clear case of WP:DISRUPT. I strongly suggest that Fred Diether reads at least the WP:DISRUPT#Summary and realize that he has fallen squarely into that category - then observe the predictable progress through the list of actions at WP:DDE is now being engaged in - which will inevitably result in serious consequences for those who continue to assault the deceased equine's battered carcass.

With a solid consensus and no movement on either side in response to debate, I can tell you from L-O-N-G experience that this means that there is absolutely zero chance of any of Christian's work getting the slightest mention in this article without some new secondary and tertiary sources showing up as evidence for mainstream support of his views...in which case, we would of course reconsider and re-open the debate.

The editorial and content discussion is clearly over - all that remains is for the disruption to spiral off to the point where those who continue to POV-push in the face of consensus dig themselves into a deep enough hole for admin sanctions to be applied. WP:DEADHORSE strongly applies here.

SteveBaker (talk) 13:22, 4 June 2012 (UTC)

• I agree with Steve. In any case user:Interintel has now been indef-blocked, and on the SPI case the closing admin (user:AGK) commented that these accounts are tag-teaming to subvert consensus and recommended further blocks. I think there has been enough discussion here to end this matter. Hence I will stop looking at this page for a week or two, and hope that those involved in editing will continue with the clean up, given that there are now a good number of people familiar with the topic here. Thanks. History2007 (talk) 14:09, 4 June 2012 (UTC)

Counterfactual definiteness

There was a strange reference to a 1973 article by Eberhardt which apparently said Bell's theorem could be proved without hidden variables. I'm guessing this means that we should talk about CFD instead. Hidden variables implies CFD. The converse is also true if one takes the hidden variables to be all outcomes of all measurements, I put in a reference to counterfactual definiteness but also had to edit that article. Wrong definition, no references. I did not add references yet. What would be the best? Richard Gill (talk) 17:11, 5 June 2012 (UTC)

I also did some rewriting of the leads of Bell test experiments and CHSH inequality. Neither article has been much touched since about 2005. Both articles need a whole lot more work. They are biased and out of date. Richard Gill (talk) 18:44, 6 June 2012 (UTC)

The article Loopholes in Bell test experiments seems on the other hand to be in fairly decent shape.I added it in the "See also" list. Richard Gill (talk) 07:19, 7 June 2012 (UTC)

"See also" also has Quantum mechanical Bell test prediction. Seems to me that article could be deleted. Richard Gill (talk) 16:57, 8 June 2012 (UTC)

Or maybe merged into other articles, if it contains something useful. Boris Tsirelson (talk) 16:28, 9 June 2012 (UTC)

At a quick look it contains nothing which is is not in "expected" articles. (BTW Boris, I really like your Citizendium article! I will log in and improve your English - delete about 50% of the occurrences of "the").Richard Gill (talk) 18:01, 9 June 2012 (UTC)

Yes, please! Indeed, writing on a wiki I am less bothered with my English: someone can correct it. (And strangely, on Wikipedia this happens very rarely, even though I wrote a long article "Space (mathematics)".) Boris Tsirelson (talk) 19:38, 9 June 2012 (UTC)

Can the 3-4 of you guys agree on what it would take to fix this article within the next 30-60 days to the point where all the tags can come off? I do like the new lede, but Stapp needs a link, etc. This is a "key result", and now that the book reviewers are no longer with us, there should be no major contention and a collegial atmosphere can prevail. Based on my personal theory, having key articles in good shape improves the stature of Wikipedia and increases donations, and helps the encyclopedia in the long term. So help in getting this (and the surrounding articles) in good shape in the next couple of months will be appreciated. History2007 (talk) 22:40, 9 June 2012 (UTC)

Probably what they will change and add will require more tags. Good luck with 60 days; in a year from now the article will probably be a mess again. — Preceding unsigned comment added by 68.6.106.100 (talk) 02:20, 10 June 2012 (UTC)

No, be positive. Be positive. As we have seen when a key article needs help the community responds. If tags are "applied without proper reasons" and against consensus, further action will take place. This is a key article and unilateral disruptions will not go far. I encourage improvements without fear of undue disruption. The community will respond and reason will prevail. History2007 (talk) 06:47, 10 June 2012 (UTC)

30-60 days seems a bit optimistic to me - but it's a nice summer project. Richard Gill (talk) 18:41, 10 June 2012 (UTC)

Well, in the computer business, one always estimates the time, then multiplies by 3, so I guess I was doing that as a matter of habit... History2007 (talk) 21:32, 11 June 2012 (UTC)

About link to Stapp: it is rumored to be here (but I did not check). Boris Tsirelson (talk) 20:20, 10 June 2012 (UTC)

Well, as they say... History2007 (talk) 21:32, 11 June 2012 (UTC)

Archive 1

←

Archive 3

Archive 4

Archive 5

Archive 6

Archive 7

→

Archive 10

Probability Calculation in Original Bell's Inequalities

I made this edit last year and it was immediately undone, but either I'm right or the assumptions made behind these calculations need to be explicitly stated. We have 3 statistical coin flips, which I am interpreting as independent events. There is a 99% chance that A = B and a 99% chance that B = C. It's the next line where there's an issue. The article adds 1% chance that A and B mismatch and 1% chance B and C mismatch to get 2% chance A and C mismatch. This violates basic probability rules because it ignores the possibility that A = C and neither equals B. The possibility that A = -1, B = +1 and C = -1. If that is an impossible situation then it needs to be stated and if that is a possible situation then the probability calculations needs to be changed to my edit from last year. — Preceding unsigned comment added by Slick023 (talk • contribs) 14:06, 13 October 2013 (UTC)

The point is not to calculate a probability (under some assumptions about independence etc) but to give a guaranteed bound for the probability, a bound guaranteed irrespective of possible dependencies. Boris Tsirelson (talk) 14:44, 13 October 2013 (UTC)

That's not how probability works. There is no such thing as a 100% confidence interval and nothing is "guaranteed". If you were to flip a coin, the bounds of what percent of the times you can get heads are 0% and 100%. Any result is possible but probability distributions dictate a band of numbers that are most likely to occur. The point is to calculate the probability using probability distributions correctly and to demonstrate that the deviation between those and the experimental results are statistically significantly.Slick023 (talk) 13:58, 26 October 2013 (UTC)

Right; but you write about relations between (theoretical) probabilities and (empirical) frequencies, while Bell inequalities are for probabilities only. True, confidence intervals and all that appear when we turn to experiments. But this is a separate matter; and in this matter, there is nothing special in Bell inequalities. What is somewhat special (but not unique) is, theoretical bounds on probabilities. It does not contradict to what you write. It is a different aspect of the problem. Here, probabilities are not derived from experiments. Intervals for them are not at all confidence intervals. They are theoretical bounds. Afterwards they should be compared with experimental data via usual statistical procedures, which is unproblematic. Boris Tsirelson (talk) 18:06, 26 October 2013 (UTC)

You misunderstood me; if theoretical bounds existed then so would 100% confidence intervals. You can calculate the number of times A & C and theoretically expected to be the same but it provides no guarantees. However, I think I've discovered my grievance with the calculation here. I was interpreting it that A & B are the same 99% of the time as a probability whereas I am supposed to interpret it as A & B are always the same exactly 99% of the time? Because that is the only way addition doesn't break basic probability rules. Slick023 (talk) 23:39, 26 October 2013 (UTC)

Also right; it is not easy to me to understand what do you really mean; but I try.

Generally, for a pair of events A and B, we have the joint distribution consisting of 4 probabilities that form a 2×2 array. Three degrees of freedom (since the sum is 1). If they are independent, only two degrees of freedom remain. If instead A implies B, another two degrees of freedom. There are a lot of possible special cases and one general case.

For three events A, B and C we have a three-dimensional array of size 2×2×2=8, and (generally) 7 degrees of freedom. Now imagine that we know that A and B are the same at least 99%. (This is just probability; whether you interprete it as the limit of frequence, via time etc., or not, is irrelevant for now.) Then our 8 numbers are restricted by an inequality (containing the constant 0.99). Knowing the same for B and C restricts our 8 numbers by a similar inequality (also with 0.99 inside). It appears that these two restrictions (combined) imply another inequality, with 0.98 inside, for A and C. And moreover, the 0.98 is tight, in the sense that it is reached by some 2×2×2 array satisfying the former two restrictions. (But, of course, it is technically much easier to do other way, without 2×2×2 array.) This is not (yet) related to frequences; just (theoretical) probabilities. Also not related to any independence; quite the general case, with the most general dependencies allowed. Boris Tsirelson (talk) 06:00, 27 October 2013 (UTC)

That's a good way of putting it. Thanks. — Arthur Rubin (talk) 06:12, 3 November 2013 (UTC)

The entire section, titled "Original Bell's inequality", which talks abut these 99% coin flips, is confusing and misleading. It smells like some novices attempt to somehow rationalize the inequalities in some intuitive way, but it fails to actually do that. Rather, it appears to be founded on some unjustified assumptions, which, taken literally, lead to the above discussion. Perhaps the entire section should be cut or completely re--written? User:Linas (talk) 03:16, 18 November 2013 (UTC)

I am new to this theory, but the whole discussion of coin flip correlations seems to be looking at the wrong problem. An attempt to measure a quantum variable where the probability of the correct result relies on the angle of measurement is a different problem to series of random binary decisions. If the probability of a matching result is cos^2(theta), then 5.7 degrees gives 99%, and 11.4 degrees gives 96%. There is no reason at all to suppose the A+B correlation and the B+C correlation should somehow correspond in a linear way.Natty Stott (talk) 19:09, 9 December 2013 (UTC)

Hope you understand that "the whole discussion" follows carefully refereed scientific articles, discussed a lot by very very competent, and highly motivated, and very astonished experts of different kind, all over the world, during decades. Boris Tsirelson (talk) 20:22, 9 December 2013 (UTC)

Let me try to explain the A, B, C story. Suppose you have a pair of particles which you can measure at distant locations. Suppose that the measurement devices have settings, which are angles. eg you measure something called spin in some direction. You choose the direction. For each particle, separately. Suppose the measurement outcome is binary (eg spin up, spin down). Suppose the two particles are perfectly correlated in the sense that whenever you measure them both in the same direction you get identically the same outcome (ie both spin up or both spin down). The only way to imagine how this works is that both particles leave their common source with somehow encoded in them both, what outcomes they will deliver when measured in any possible direction. How else could particle 1 know how to deliver the same answer as particle 2 when measured in the same direction? (They don't know in advance how they are going to be measured...).

Start with both settings equal to one another, say both at 0 degrees to some common reference direction. All the pairs of particles give the same outcome (each pair is either both spin up or both spin down). Now increase Alice's setting to +1 degree leaving Bob's at 0 degrees. A small fraction of the pairs, say f, now give different outcomes. If instead we had left Alice's setting at 0 degrees but decreased Bob's to -1 degrees, then again a fraction f of the pairs of particles turn out to give different outcomes.

It should not be hard to convince yourself that if Alice's setting is put at +1 degree and Bob's at -1 degree, at most a fraction 2f of the pairs can give different outcomes! Richard Gill (talk) 21:19, 17 December 2013 (UTC)

I put a version of my explanation into the article. Richard Gill (talk) 12:30, 22 December 2013 (UTC)

Cut text

I removed the following text from the article:

In the usual quantum mechanical formalism, the observables X and Y are represented as self-adjoint operators on a Hilbert space. To compute the correlation, assume that X and Y are represented by matrices in a finite dimensional space and that X and Y commute; this special case suffices for our purposes below. The von Neumann measurement postulate states: a series of measurements of an observable X on a series of identical systems in state $\scriptstyle \phi$ produces a distribution of real values. By the assumption that observables are finite matrices, this distribution is discrete. The probability of observing λ is non-zero if and only if λ is an eigenvalue of the matrix X and moreover the probability is

$\|\operatorname {E} _{X}(\lambda )\phi \|^{2}$

where E_X (λ) is the projector corresponding to the eigenvalue λ. The system state immediately after the measurement is

$\|\operatorname {E} _{X}(\lambda )\phi \|^{-1}\operatorname {E} _{X}(\lambda )\phi .$

From this, we can show that the correlation of commuting observables X and Y in a pure state $\scriptstyle \psi$ is

$\langle XY\rangle =\langle XY\psi \mid \psi \rangle$

We apply this fact in the context of the EPR paradox.

There are multiple issues here:

• What the heck are X, Y?
• Why should I assume they commute?
• lambda is introduced in a previous section as a hidden variable. Here it is used as an eignevalue. That's confusing.
• E is introduced in previous sections as an expected value. Here it is used as a projection. This is misleading. Also, no one but no one ever uses E to stand for projection.
• The last few sentences don't make sense...

So I cut the thing. User:Linas (talk) 03:51, 18 November 2013 (UTC)

Good. I have cut a whole lot more repetitious material, also cut a lot of nonsense. Richard Gill (talk) 12:31, 22 December 2013 (UTC)

Hot steaming mess....

Wow. Not a bad article, but certainly not a good one. I'm not surprised, given the contentious nature of the topic. Issues:

• Section "Importance of the theorem" is interesting, but has a number of non-sequiters in it, while also pointlessly repeating content from the intro.
• Mentions of "faster than light" come out of nowhere, and are not really appropriate. The FTL nature of wave function collapse is interesting, but not really germane to the argument.
• Section "Original Bell's inequality" appears to be some kind of strange, flawed attempt at a simplified explanation. It encourages wild misunderstandings, see talk section immediately above. It needs to be cut or entirely re-written.
• Section "Bell inequalities are violated by quantum mechanical predictions" starts talking about observables X and Y. This is the first occurance of X and Y. What are they? Why should they commute? Most everything important to Bell's thm do NOT commute! Worse, it uses lambda as an expectation value; but the previous section defines lambda as a hidden variable! The symbol E is used as expected value earlier, yet here its used as a projector. That makes no sense. Big WTF here. OK, I cut the offending part, see below.
• ...an overall lack of flow, organization. The whole thing is clearly a compendium of random facts inserted by random authors, Ugh.

What to do? User:Linas (talk) 03:09, 18 November 2013 (UTC)

Agreed, quite a mess.

However mentions of faster than light (superluminal) communication are not so inappropriate: after all, the conclusion of Bell's theorem is that (assuming that QM predictions do indeed fit reality to a sufficient degree) one of three formerly uncontroversial assumptions about the nature of reality has to be discarded: either we must discard locality (in favour of superluminal communication), or we must discard freedom (to choose measurement settings how we like, AKA no conspiracy), or we must discard something called realism roughly meaning that outcomes of measurements which were not actually performed can also be considered part of reality (aka counterfactual definiteness). Sorry I am using some somewhat technical language here. The whole Bell story started with people being unhappy that a wave function of one particle in one place would instantaneously collapse when something is measured on another particle far away. But this is only a problem if you think that wave particles are real things. The genius of Bell (and his predecessors EPR) was that he showed that weird things happened not only at the level of wave functions (which after all might be considered just some computational device, not as something physically real) but also at the level of hard outcomes of real lab measurements.

The section original Bell is indeed quite a mess though one can recognise one of the oldest versions of Bell's theorem here, and one which is used in many (not bad) popular explanations too.

The commuting observables were any of Alice's and any of Bob's (spin on one particle in one direction, spin on the other particle in some other direction). The non-commuting observables are any pair of Alice's spins, or any pair of Bob's spings.

What to do? Good question... Richard Gill (talk) 21:34, 17 December 2013 (UTC)

How about going through the article removing material which is not properly referenced and which is of specialist importance? Maybe that way we can recover a kind of viable living core to the article. Then after that, people who want to add specialist or controversial material can do so, but hopefully in a balanced way. Richard Gill (talk) 21:45, 17 December 2013 (UTC)

Apart from being a big mess, the article is filled with subtle caveats which, though much discussed in the philosophy of science, distract from the big picture and represent a particular point of view, which though legitimate, is not main-stream. Because of this, the basic argument does not come across. Richard Gill (talk) 14:29, 20 December 2013 (UTC)

I have cut a lot of the crap, and tried to improve what is left. Someone else, please help too! Richard Gill (talk) 12:32, 22 December 2013 (UTC)

For a proper understanding of the importance of the theorem (in my opinion, so this is open for discussion) all resulting cases should be treated, even if in short paragraphs (actually recommendable in short paragraphs as too much information may be too much). The superluminal communication case can be referenced with this experiment that places a lower bound on the speed http://www.nature.com/nature/journal/v454/n7206/full/nature07121.html?free=2 ; http://arxiv.org/abs/1303.0614 Alma (talk) 11:08, 23 December 2013 (UTC)

Nice references! I was aware of the Gisin et al. work. I too think that a discussion needs to be added about the metaphysical consequences of Bell's theorem - i.e., about "all resulting cases". What are they?

Many present day writers point out that (supposing a sufficiently - loophole free experiment indeed gets done in a few years, ie that Nature is convincingly seen to take the side of quantum mechanics) we will have three basic choices. After all, Bell's theorem can be formulated as stating that quantum mechanics is incompatible with the conjunction of *three* basic principles: realism (more precisely and technically: counterfactual definiteness), locality (more precisely and technically: local relativistic causality) and freedom (no-conspiracy principle or no super-determinism). So if we believe that Nature agrees (if only to a decent enough approximation) with quantum theory, then we are logically forced to abandon at least one of the triple: locality, realism, and freedom. The presently most popular choice is to abandon realism, in the sense of admitting fundamental, irreducible quantum randomness as a fact of Nature. However there is a strong school of people who prefer to abandon locality (in particular, the Bohmians; and Bell himself belonged to this category). There are just a few people who prefer to abandon freedom - most notably, Gerard 't Hooft (Nobel prize laureate).

There is a smaller but noisier category who disagrees with the validity of Bell's theorem and therefore have no need to make a choice of which beloved principle to abandon: they can keep hold of all.

Since I have written quite a lot about this myself, and also have a declared preference - abandon realism - I should not take the lead in writing such an overview. However, for what it's worth, my latest is http://arxiv.org/abs/1207.5103. I also like very much the writings of Boris Tsirelson on these topics. He has an excellent article on Citizendium: http://en.citizendium.org/wiki/Entanglement_(physics). And recently I came across http://arxiv.org/abs/1310.3288, which is about using photons from distant galaxies to further ensure freedom: the setting choices are determined outside of the backwards lightcone of the source. Richard Gill (talk) 15:22, 23 December 2013 (UTC)

Thank you Richard! I'm not sure you've seen this. Regarding the consequences of the theorem, I honestly see no point to treat the fourth case. A theorem is a theorem and no amount of denying will change that. As to whom should write about them, this article needs attention from an expert and you are the expert. The Wikipedia guidelines ask for the articles to be presented in a neutral manner - which you already did above by covering each consequence. I have already asked involvement for cleaning some articles in the Physics portal talk page, but so far received almost no reply, so there is little hope that someone else with your level of expertise will help with this article. About Bohm, do you think pointing the loophole of his theory would go in here? (correct me if I'm wrong - the issue is 'no backreaction', the particles don't influence the pilot wave). Alma (talk) 16:28, 23 December 2013 (UTC)

About the fourth category: a mathematical theorem is a mathematical theorem. But physicists might like to deny that the mathematical concepts which are used in the theorem correspond correctly to the physical or metaphysical concepts which are needed in the real world. So you could agree with the theorem as a piece of pure mathematics, but you could deny it has any relevance to physics. Some of the fourth category people are of this kind. Actually they do serve a useful purpose, namely to keep us sharp, refine our concepts, improve our theorems (weaken the assumptions / strengthen the conclusions). Richard Gill (talk) 17:47, 23 December 2013 (UTC)

About Bohmian mechanics: yes I think it fits in here. Bell himself was very attracted to it. There are a smallish number of very serious people doing serious work within that framework. Richard Gill (talk) 17:49, 23 December 2013 (UTC)

The usual picture

The article includes the usual picture of a (negative) cosine curve (quantum mechanics: the singlet correlations) and a piecewise linear curve (local realism). The suggestion is made that local realism can only give us the second of the two. But actually, many many different curves are possible under local realism. Perhaps, in some sense, the piecewise linear curve is the best that local realism can do. But in what sense? Has anyone actually proved something about this? Richard Gill (talk) 12:30, 22 December 2013 (UTC)

I understand your concern with the strictness of the representation. There are two things here. On one hand, the picture is intended to provide visual aid for the general public ("it looks something like this"). On the other hand this article is of interest for (and was presumably mostly elaborated by) physical theoreticians. Here I will use the very good point that Simon Singh made in his book on the demonstration of Fermat's Last: while mathematical proofs remain true for all time, scientific theories and demonstrations can never be as flawless and can afford that because they can be tested. Based on this I issue the following opinion which is mine and may or may not be applicable: I think a risk/benefit analysis should suffice (how difficult is to find the proof if it exists; how fundamentally wrong is the representation - showing all the possibilities of local realism would contradict the current representation? hopefully not because wouldn't that disprove the theorem?; what is the impact on the reader; is it enough to place a caveat that the representation is not necessarily strict but a good approximation for a number of cases etc). Not all facts in all encyclopedias are rigorously proved and the effort to create a flawless article may not be proportional to the benefit. Alma (talk) 11:47, 23 December 2013 (UTC)

I have edited the legend to this picture so that it is hopefully both true and useful. In the meantime I just try to find out if any editors of wikipedia are aware of any results in this direction.

Fortunately, the correctness of Bell's theorem (as a mathematical theorem, not as a statement about physical reality!) is independent of the picture. And on further thought, I realized that the picture does illustrates several possible proofs of the theorem. The usual proof goes by showing that quantum mechanics allows the correlations of +/- 0.70... at the angles pi/2, 3pi/2 etc while under local realism one can go no further than +/-0.5. The picture illustrates the proof because the red curve does represent the best that local realism can do in this particular respect. Another proof of Bell's theorem, also sketched in the article, works by showing that the QM correlation is smooth (and hence flat) at its peaks and valleys (at the angles 0, pi, 2pi...) while any local realist correlation which achieves the same summits and depths is necessarily "pointed" at those points. The picture does indeed illustrate this too, since the red curve does illustrate this necessary feature of local realist correlations under the same constraints. So I added to the legend of the picture some remarks pointing out these connections between the text of the article, and the picture itself.

Actually I think the article - as it was up to a few days ago - was mostly written by amateurs and/or non specialists, together with people with strong personal opinions that the conventional understanding of Bell and all that, was wrong. Hence it was an incredible hodgepodge of conflicting material, repetitions, technical niceties; and moreover it was rather out of date, representing the field as it was maybe ten years ago; not as it is today. There has been an awful lot of progress both theoretical and experimental, and this has changed the relative importance of various issues. Richard Gill (talk) 14:58, 23 December 2013 (UTC)

Much better than before; thanks to Richard Gill. Boris Tsirelson (talk) 17:45, 23 December 2013 (UTC)

I posted my questions about what the picture could be telling us (in what sense is the saw-tooth the best approximation to the cosine) as an arXiv preprint http://arxiv.org/abs/1312.6403v1. I suspect that Steve Gull's proof of Bell's theorem using Fourier analysis could provide some good tools. See also http://arxiv.org/abs/1307.6839 by Kent and Pitalua-Garcia. Richard Gill (talk) 09:28, 24 December 2013 (UTC)

Check if you doubt

Hi Sławomir, the references you undid because you 'doubt' they are free belong to an open source collection. Please check before you doubt (exmaple here https://archive.org/details/IntroductionToQuantumMechanics_718). After convincing yourself if they are opensource or not, please put them back. Regarding the compatibility, I fixed that, just before you undid everything. I was still working on it when you started working on it as you can check by the timestamps. Alma (talk) 15:05, 24 December 2013 (UTC)

That linked work contains a standard copyright notice: "No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise, except as permitted under Sections 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher." That seems pretty clear cut to me, and the internet archive is violating this notice. Other than that, the only things I undid was to put the resources back into templates (you had removed the templates, along with in some cases important reference information like page numbers, titles, and doi codes) and restored the named references. For information on the fields in citation templates (including the URL field), see Template:Citation. It's better to use these templates than to format the references in an ad hoc manner, since they will then have a uniform appearance, are easier for automated processes to parse and keep up to date, and often contain more complete bibliographical details. Sławomir Biały (talk) 15:26, 24 December 2013 (UTC)

Sławomir, isn't there a way to check whether content belongs to public domain or not? Leggett's incompatibility theorem is stored by an university and part of a course. The notice on Griffith QM limits reproduction only and does not refer to storage or retrieval, should I have assumed these? And yes, I was still working on the templates, at the same time with you - why not ask me to undo the changes? Alma (talk) 15:38, 24 December 2013 (UTC)

The Griffiths text is clearly not in the public domain, unless there is explicit evidence that the publisher has released it. The primary reason I interrupted this sequence of edits as I did is that, if allowed to continue as it seemed, it was not only removing the references from the templates, but also appeared to be removing vital bibliographic data. I opted to "revert" rather than deal with this by normal editing largely because it would be quite awkward to recover this deleted data. For example, the edit to the Legget reference removed the title of the article, the page numbers of the article, the volume and issue information, and the doi. The edit to the Griffiths reference removed the publisher and the year. The edit to the EPR paper removed the doi, the volume, issue, and page number, and so forth. It's not clear to me that your intention was to restore this necessary information in the future (or to put things back into templates). If that was indeed the case, why remove the information in the first place? It would simply make more work for later. Sławomir Biały (talk) 15:59, 24 December 2013 (UTC)

I wasn't getting the look I wanted for the references before saving the page (there was no way to check how it will look, I don't know why -before saving it looked in a way and after saving it looked different) and was working with both the page and a word sheet. If you would have asked, you would have give me the chance to explain that. Working on references is tedious and does not massively change the articles, so someone who is not into details will probably do something else. On Griffith's book I take your word that it's not in the public domain, how about Leggett's? And Eberhard? http://inis.iaea.org/search/search.aspx?orig_q=RN:8282046 Alma (talk) 16:10, 24 December 2013 (UTC)

Griffith's book is currently sold in hardcover for an awful lot of money by Pearson publishers, it is in print. You can get it from Amazon too, or Barnes and Noble. No-one offers an eBook version. Students and other poor people will know how to find a free electronic copy on internet. Wikipedia should refer to the book at the publisher's webpage and leave the rest to the creativity of the individual person who wants to follow up the reference.

The Eberhard article is published in an out of print, difficult to find conference proceedings. The present link to an electronic version seems to me to be fine. Especially since Eberhard's inequality has recently been used in new experiments (Giustina et al.) closing the detection loophole.

Leggett's article can be bought from Springer, publisher of the journal in which it appears, for a lot of money. Google Scholar gives links to a dozen copies at university researcher's web sites, the citeseer archive at Penn State University, and other university archives. If Leggett had published this article today he would have posted a preprint on arXiv and also on his personal web-page. Most journals nowadays grant authors the rights to do just this. Seems to me that Wikipedia should link to the Springer source; the resourceful reader will track down other copies. Richard Gill (talk) 09:46, 26 December 2013 (UTC)

Anyway, I was thinking to add online references for everything that's available. We should probably coordinate somehow, what do you think, Sławomir? Alma (talk) 16:33, 24 December 2013 (UTC)

I agree: good references which moreover are easily available online are really valuable. Richard Gill (talk) 09:30, 26 December 2013 (UTC)

Wording

This wording looks a bit over-complicated and circumlocutious to me: '...one is forced to reject locality, realism, or the possibility of nondeterminism (the last leads to alternative superdeterministic theories, none of which has yet replicated the predictions of QM...'.

We have a double negative, '...forced to reject...the possibility of nondeterminism...'. Would it not be better to say something along the lines of, '...forced to assert (super)determinism'?

'Realism' also seems a bad word to use in this article. It is somewhat ambiguous and might be considered a bit biased against QM. Martin Hogbin (talk) 15:20, 23 December 2013 (UTC)

The sentence you quote is a bad one, I did not yet attempt to improve it.

But: realism is a technical term in this field. It is not biased against QM. And the metaphysical consequence of Bell's theorem is that assuming the predictions of QM are roughly true and that Bell's theorem is essentially correct (correct mathematical implementation of the key concepts), then we must reject at least one of realism, locality, or freedom. Personally I like to believe in QM and I am happy to reject realism. I think this is also the mainstream opinion, at least, among people who know what we are talking about...

There are alternative labels to give to these three concepts, most of them rather technical and unhelpful for general readers. Richard Gill (talk) 15:29, 23 December 2013 (UTC)

I cleaned up that bad sentence (from the intro). There is as yet no superdeterministic theory which expains QM. Gerard 't Hoofd seems to be the only notable physicist who thinks that one is possible and who made some first (as yet very incomplete) steps. The theory has to explain somehow how experimenters who choose measurement settings by tossing coins or using pseudo random numbers with seed equals their wife's birthdate, observe extraordinary correlations between these settings and between measurement outcomes of measurement made on distant pairs of photons, which somehow can only be explained by the distant photons somehow knowing in advance what the measurement settings on the other side of the experiment were going to be. The problem about superdeterminism is that it is essentially ludicrous. It might explain some important part of physics at the Planck scale but it doesn't (and can't!) scale up. Richard Gill (talk) 15:39, 23 December 2013 (UTC)

I also defined the three concepts. Richard Gill (talk) 15:48, 23 December 2013 (UTC)

Here is what Boris Tsirelson has to say, on http://en.citizendium.org/wiki/Entanglement_(physics)#Nonlocality_and_entanglement:

The words "nonlocal" and "nonlocality" occur frequently in the literature on entanglement, which creates a lot of confusion: it seems that entanglement means nonlocality! This situation has two causes, pragmatical and philosophical.

Here is the pragmatical cause. The word "nonlocal" sounds good. The phrase "non-CFD" (where CFD denotes counterfactual definiteness) sounds much worse, but is also incorrect; the correct phrase, involving both CFD and locality (and no-conspiracy, see the lead [of the Citizendium article "entanglement"]) is prohibitively cumbersome. Thus, "nonlocal" is often used as a conventional substitute for "able to produce empirical entanglement".

The philosophical cause. Many people feel that CFD is more trustworthy than RLC (relativistic local causality), and NC (no-conspiracy) is even more trustworthy. Being forced to abandon one of them, these people are inclined to retain NC and CFD at the expence of abandoning RLC.

However, quantum theory is compatible with RLC+NC. A violation of RLC+NC is called faster-than-light communication (rather than entanglement); it was never been observed, and never predicted by the quantum theory. Thus RLC and NC are corroborated, while CFD is not. In this sense CFD is less trustworthy than RLC and NC.

Richard, I think your definitions of the terms are excellent; clear and concise. It is just the word 'realism' that I find problematic. Without reading your definition of what the term is intended to mean in this article, the reader may be puzzled as to whether you are referring to Scientific realism, Philosophical realism, or something else. Intuitively it could be taken by some readers to suggest that QM is in some way 'unreal' or that it does not represent reality (whether that is true or not remains to be seen, of course). Bearing in mind Boris' and your comments, which seem to say that no simple term is actually the correct one, is there a another word or phrase that we could use. Martin Hogbin (talk) 12:28, 24 December 2013 (UTC)

"Realism and freedom are part of statistical thinking on causality: they relate to counterfactual reasoning, and to the distinction between selecting on X=x and do-ing X=x, respectively.", as per his paper here http://arxiv.org/pdf/1207.5103v2.pdf. Alma (talk) 13:26, 24 December 2013 (UTC)

I'm afraid that "realism" is the term which is very widely used in this field (foundations of quantum mechanics...) and it has a precise (though over the decennia, evolving) meaning, which I hope, is now clear in the article. It goes back to the famous Einstein, Podolsky, Rosen paper of 1932 (?), perhaps the most cited paper in science ever, and arguably one of Einstein's most significant contributions. The authors (who were "realists"!) gave a criterium by which one can sometimes determine that something is an "element of reality" (if one is a realist, at least). The meaning of the term is probably different from the major definitions of realism in philosophy. Some people nowadays equate realism (in the present context) with determinism. If you want a more precise technical phrase, say "counterfactual definiteness". But then no-one has any idea anymore what you mean, whereas people seem to have an instinctive understanding of "local realism" (locality + realism). Richard Gill (talk) 17:38, 25 December 2013 (UTC)

(1) I find the term rather loaded (as indeed it was intended to be by EP and R). Would it be possible to rephrase the section to use something like 'local hidden variable theory'?

(2) Could you also explain Boris' comment above. Is there a more accurate term that we should really be using? Martin Hogbin (talk) 00:21, 26 December 2013 (UTC)

(1) Local hidden variables = locality plus realism. It's important to be able to separate the two parts of "local realism", since if we decide that we have to reject "local realism" we have a choice: reject locality, or reject realism. Bell and the Bohmian's took the first choice. I think that most experts nowadays take the *second* choice (to name some big names: Gisin, Zeilinger, Tsirelson, ...). The crowd follows the authorities. Further from the field, most physicists (not QM specialists) think that the interpretational problems of QM are solved by the many worlds theory and that there are no problems with Bell because measurement is not real - there are only wave functions. No collapse. Richard Gill (talk) 02:52, 26 December 2013 (UTC)

(2) Which of Boris' comments do you want explained? And I already said: a more accurate term is "counterfactual definiteness". Most of the literature uses the word "realism". The term is loaded and is meant to be loaded. Richard Gill (talk) 02:52, 26 December 2013 (UTC)

Use of 'realism' for CFD suggests that CFD is how the world 'really' is or must be. Whilst this may have been the view of ERP and Bell (who obviously did intend the term to be loaded) it is certainly not a universally held view. Using the term 'realism' here to mean CFD suggests that WP supports the view of EPR, and Bell. Why not use CFD here, it is a neutral term which does not appear to support either side of the philosophical argument.

Because the term CFD hardly appears in the literature, while the term "realism" appears throughout. Using the term "realism" in an article on Bell's theorem does not suggest that anyone is supporting any particular person's point of view. Richard Gill (talk) 07:11, 28 December 2013 (UTC)

I was talking about Boris', 'The phrase "non-CFD" (where CFD denotes counterfactual definiteness) sounds much worse, but is also incorrect; the correct phrase, involving both CFD and locality (and no-conspiracy, see the lead [of the Citizendium article "entanglement"]) is prohibitively cumbersome' [my italics]. Am I right in thinking that Boris is not saying that CFD is the wrong term for what you have called 'realism', but that it cannot be use to describe CFD + non-locality + no-conspiracy? That seems fair enough to me. Martin Hogbin (talk) 19:25, 27 December 2013 (UTC)

Boris is saying, among other things, that CFD is not the correct term for CFD+locality+no-conspiracy. (Not non-locality). Richard Gill (talk) 07:10, 28 December 2013 (UTC)

This article is written as if the theorem has not been proven

From "De Broglie-Bohm theory":

Decades later John Bell proved Bell's theorem (see p. 14 in Bell[28]), in which he showed that, if they are to agree with the empirical predictions of quantum mechanics, all such "hidden-variable" completions of quantum mechanics must either be nonlocal (as the Bohm interpretation is) or give up the assumption that experiments produce unique results (see counterfactual definiteness and many-worlds interpretation). In particular, Bell proved that any local theory with unique results must make empirical predictions satisfying a statistical constraint called "Bell's inequality".

Why is the article written in such a mealy-mouthed fashion, making it perfectly unclear to the reader what the accepted stance toward this theorem is? Having browsed the many discussions appertaining to article's discussion, perhaps it needs to be entirely re-written, top-down, so that it has at least some value for the reader. In any case, it is not "fixed;" nor is it getting any "fixeder." — Preceding unsigned comment added by 173.25.13.90 (talk) 09:30, 25 January 2014 (UTC)

One problem is that "Bell's theorem" is not a theorem in the standard mathematical sense but what is described in the Stanford Encyclopedia of Philosophy as, '...the collective name for a family of results, all showing the impossibility of a Local Realistic interpretation of quantum mechanics'. It is not quite clear whether 'results' refers to mathematical results which, as you say, were proved by Bell, or to predicted experimental results, which have been fairly well verified but not conclusively so.

I have reworded my own contribution above to try to make this distinction clearer. Martin Hogbin (talk) 10:40, 25 January 2014 (UTC)

Firstly, various authors have written mathematically careful treatments of the mathematical core of Bell's theorem. There is no debate as to whether or not those mathematical results are true or false. The mathematics is pretty trivial! Some people (not many) do criticize the appropriateness of the background assumptions.

Secondly, experiments over the years have confirmed quantum mechanical predictions in more and more stringent settings. But no experiment to date, on its own, is thought to be conclusive (the loopholes question). The first "loophole free" confirmation of Bell's theorem is expected in about five years from now.

Thirdly, please note that Bell's theorem says that QM predictions are incompatible with the conjunction of *three* things: locality, realism, freedom. So believing the theorem, and believing that QM predictions are essentially corrent, one still has three options: reject locality, reject realism, or reject freedom. There is no single "accepted stance" on this matter. Different subcommunities in physics tend to hold different views about it. Richard Gill (talk) 18:46, 28 January 2014 (UTC)