Talk:Bell's theorem/Archive 6

Archive 1

←

Archive 4

Archive 5

Archive 6

Archive 7

Archive 8

→

Archive 10

Many worlds interpretation resolution (?) of non-locality question

Martin, you removed the little paragraph about the many worlds interpretation. This is a very important topic in the literature on Bell theorem. The many worlds interpretation is much favoured by many physicists and its proponents claim that it succesfully resolves the non-locality question, through its very radical stance on what is actually real. Namely "reality" as we know it is not really real at all, since only one insignificant path among so many branching possibilities: according to them it is the collective of possibilities which is real, not any particular one of them. In my humble opinion this is all nonsense but I'm afraid it keeps being brought up in the "reliable sources" hence cannot be omitted from the article. Richard Gill (talk) 15:53, 27 December 2013 (UTC)

Yes, the MWI supporters are rather vocal and make many extravagant claims for their interpretation of QM including that it solves problems that other interpretations cannot.

The MWI interpretation may well be mentioned in the literature but, as I am sure that you know, the generally accepted view is that none of the QM interpretations contains any new physics. Supporters of other interpretations would say that the problems that the MWI claims to solve do not actually exist. We could possibly have a section on philosophy showing how the various QM interpretations would deal with confirmation (or falsification) of Bell's theorem but this would duplicate what we have elsewhere in WP. I do not think WP should present philosophical musings as science. Martin Hogbin (talk) 18:58, 27 December 2013 (UTC)

Correct me if I'm wrong, but the neutrality guideline requires that all notable details should be presented. The paragraph is not an exact duplication of content but an explanation as to how does many world apply here - and it's only 3 rows long. Whether some scientists who support one interpretation or the other are vocal or not is of less interest for the articles, otherwise we may want to present our opinions on how does this interpretation fail to be consistent. Alma (talk) 19:42, 27 December 2013 (UTC)

By that token we should have a paragraph on how other interpretation deal with Bell's theorem. There is nothing special about MWI. Martin Hogbin (talk) 20:26, 27 December 2013 (UTC)

Martin, those sentences which you want to be rid of are in a section called "theoretical challenges" which is *exactly* about how some interpretations can dispose of Bell's theorem. (In my opinion, using false arguments, but these are arguments which are commonly reproduced in the literature). Moreover, MWI is rather special in this respect, because it has a strikingly different picture of "reality" from any other interpretation I know, and disposes of Bell in such a grand way that, in my opinion, rather underlines the fact that MWI is quite empty - it is just playing with words. However, among un-informed physicists, the general opinion is that MWI has resolved all interpretational problems of quantum mechanics. It's some kind of comfort blanket.

An article about Bell's theorem does not just belong to "hard science" but also to foundations/philosophy of science. It is often argued that Bell's philosophical musings have led directly on to the present day field of quantum information which is both about stunning experiments and and revolutionary new technology. Richard Gill (talk) 20:51, 27 December 2013 (UTC)

Regarding other interpretations one can be brief. According to a strict Copenhagen view, talking about counterfactual outcomes of non-performed experiments is meaningless, anyway. Realism must go, and we can keep locality. On the other hand, the Bohmians have a richly developed theory of non-local hidden variables which reproduces the "surface" predictions of QM - if you want hidden variables, you can have them, but you must accept non-locality. Superdeterminism is mentioned in the article, nobody has got very far with it. Logically it might be thought of as an option, but in my opinion it's a ludicrous option. Richard Gill (talk) 20:57, 27 December 2013 (UTC)

It seems that we agree on much. I would have no objection to a section 'Bells theorem and interpretations of QM' in which we discuss the ways in which various interpretations of QM deal with CFD (although I do no think it is necessary). My objection was to discussing only MWI, thus giving the impression that it resolved all interpretational problems of quantum mechanics, as you put it.

I have no objection to mentioning philosophy so long as it is distinguished from physics. Once it becomes experimentally testable, as in Bell's case, then it become physics.

I think you are missing the point. Bell's theorem is interesting because of its *metaphysical* implications. Bell himself pointed out that there were at least five different attittudes to take to his results, five different sets of conclusions which one might draw. The literature on Bell's theorem belongs to the foundations of physics, or if you prefer, the philosophy of physics. Bell's work was directly inspired by the Einstein-Bohr conflict concerning the *interpretation* of quantum mechanics. In fact, Einstein used a Bell type experimental set-up as a thought experiment in order to show that quantum mechanics must be incomplete. Its experimental predictions might be fine, but could only be a reflection of a deeper and more *realistic* theory. There is a separate wikipedia article on Bell theorem experiments. Finally: the section "theoretical challenges" could be renamed "objections from philosophy of physics".

That experiments have confirmed quantum mechanical predictions is not a surprise. The interesting thing is that they are in conflct with "local realism" which is a metaphysical concept. People call this area "experimental metaphysics". Richard Gill (talk) 08:51, 28 December 2013 (UTC)

I still think it would be a good idea to separate physics from metaphysics. I realise that they are both part of the interesting history of this subject and are likely to be part of its future but I think we should indicate which theories make experimentally testable predictions and which do not. I am not insisting on separate sections here, just that we try to keep the distinction in the text. Bells theorem does sit on the dividing line between the two as the EPR paradox was considered untestable until Bell found a way to do it.

I do still object to using the term 'local realism' in the article. As you have pointed out it comes from the history of the problem (and may be used in current literature, which is intended for people who understand the problem) but I do not think it is the best language to use here.

EPR, and Bell used the term 'realism' because they strongly believed that that was how the world should be. If Bell's theorem is conclusively verified it will mean that they were wrong and that QM accurately describes the real world in which we live. 'Realism' will then describe an imagined world which has been proved not to be real. For the general public, this seems an unhelpful use of language. Martin Hogbin (talk) 13:25, 30 December 2013 (UTC)

"Realism" is actually some kind of idealism. QM does accurately describe the real world in which we live, and nobody doubts that! A loophole free Bell test experiment (five years down the road?) won't change anything - everybody knows it will again confirm QM. However the discussion as to what are the correct metaphysical interpretations will go on for many many years to come. And probably the general public will never quite get the point, either. Richard Gill (talk) 17:38, 2 January 2014 (UTC)

Determinism

Determinism is defined in my dictionary (Collins 1994) as [my emphasis], 'The philosophical doctrine that all events including human actions and choices are fully determined by preceding events and states of affairs...'. We should state clearly that this is a physically valid interpretation of QM. The fact that this is philosophically unpalatable to some is another matter. Martin Hogbin (talk) 23:13, 27 December 2013 (UTC)

"Determinism" is not an interpretation of QM. It is in conflict with QM (that is what Bell is telling us - unless you want to take even more unpalatable routes). The reason there has been so much controversy about Bell is because *non-determinism* is unpalatable to most physicists.

Bohmian mechanics is an interpretation of QM which is deterministic. Unfortunately, however, disturbingly non-local, and hence also unpalatable to most physicists.

Bell leaves us only with unpalatable choices, or as some writers have said, with the necessity for a paradigm shift. Richard Gill (talk) 06:47, 28 December 2013 (UTC)

Maybe we need an "arguments" page in which to discuss subtle issues whether of philosophy or nomenclature, and concentrate here on editorial matters concerning the article. For instance: many passages need better or more citations. Richard Gill (talk) 07:16, 28 December 2013 (UTC)

No, an arguments page will not help here. This discussion is directly about the content of the article and how to improve it.

There still seems to be some disagreement about what 'determinism' means. This is critical to the article. If words are used with one meaning in the article but understood by readers to have another then the article will be confusing and will need improvement. I am not sure that we disagree about any physics, maths, or philosophy but I think that using terminology which might mean different things to different people is not the best way to write an encyclopedia. That is also my point with 'realism'. I know what you mean by it (and understand its historical usage) but some readers may take it to mean something different. There are many public misunderstandings about QM and the more we can do here to set matters straight the better.

I missed out the last bit of my dictionary definition of 'determinism' because I thought it was not necessary but it seems that it is. Here it is in full [with my emphasis again]: 'The philosophical doctrine that all events including human actions and choices are fully determined by preceding events and states of affairs, and so that freedom of choice is illusory'. This is the idea that the future is absolutely fixed, destiny if you like, the idea that all decisions made by humans and all random numbers, however generated, are known (by God or nature or whoever) in advance. Can we not agree on a term to use in our discussion for this concept? How about 'absolute determinism'?

It is a generally accepted fact that absolute determinism is fully compatible with QM and Bell's theorem. Do you not agree? Martin Hogbin (talk) 10:19, 28 December 2013 (UTC)

I suggest to link the determinism article on Wikipedia. The definition of determinism seems to be a very complex subject, see Hawking's definition in paragraph 2 and the problem of decay.

Martin, maybe you can help improve the determinism article? For my understanding, how is 'absolute determinism' fundamentally different from superdeterminism?

This article seems contain a relevant approach http://cms.unige.ch/gap/quantum/wiki/_media/publications:bib:stochqdynrel.pdf Alma (talk) 11:08, 28 December 2013 (UTC)

'Absolute determinism' was a term that I proposed so that we could discuss things here without talking to cross purposes. By this I mean that the future is fixed and that our human decisions, however random and off-the-cuff we may think them to be are all predetermined. There are people who hold this view. This position is not very interesting from a physics point of view because any theory (if such a thing is even possible) would be ridiculously complex. It would have to explain the connection between the choice of a measurement to make, maybe based on an experimenter's spur of the moment decision to use his wife's birthday as a seed for a (pseudo)random number generator, and the result of the experiment.

Looking through the Determinism article, I cannot find any term that exactly and unequivocally matches what I am describing. Nomological determinism looks close, Necessitarianism looks closer. Predeterminism and Fatalism also seem to describe the concept well but seem a bit theistically based. Superdeterminism seems to refer specifically to local hidden variable theories and thus presupposes that such a theory might be possible. I claim no expertise in this philosophical quagmire and it is not a subject in which I wish to be involved. All I am asking is that we explain as unambiguously as we can to our readers what exactly we mean. In particular there is a common misconception that QM is in some way in disagreement with absolute determinism. Nomological determinism says: 'nomological determinism is the notion that the past and the present dictate the future entirely and necessarily by rigid, all-encompassing natural laws, that every occurrence results inevitably from prior events', but the Determinism article says of the same term, 'Quantum mechanics and various interpretations thereof pose a serious challenge to this view'.

The fact is that QM is fully compatible with absolute determinism. Some philosophers and religious people may not like the idea of absolute determinism but that is a different matter. Martin Hogbin (talk) 12:33, 28 December 2013 (UTC)

Hopefully I understood this and to make sure I will rephrase so you can correct me. Absolute determinism is equivalent to history dependence (the entire chain of events leading to 'this event') as opposed to a stochastic process that has the Markov property (is memoryless, depending only upon the present state, not on the sequence of events that preceded it), describing a Markov chain. If so, this may be the 'root of all evil' so to say. QM depends on probability distributions. The even more restrictive Markov chains can be used in QM as a way to compute (example http://arxiv.org/abs/1208.0764). The problem is discreteness vs continuity: Markov chains are discrete but QM has a continuous time evolution, and therein lies conflict. Is the history dependence what you would like to see clarified? Alma (talk) 14:10, 28 December 2013 (UTC)

How come this is a fact? Could you give me some authoritative references? The usual conclusion drawn from Bell's theorem is that we should reject absolute determinism. The exceptions being (a) Gerard 't Hooft, lonely proponent of super-determinism; (b) many worlds people who in my opinion mostly talk nonsense; and (c) Bohmians who are happy to work with non-local hidden variables and forced to assume that there is a special reference frame, in contradiction with the message of relativity theory, that the laws of physics are invariant under change of inertial reference frame. So one can certainly argue from the point of view of philosophy or logic, that QM is fully compatible with absolute determinism, but from the point of view of physics, there are extremely good arguments why absolute determinism should be relinquished. Richard Gill (talk) 13:54, 28 December 2013 (UTC)

I suspect that "absolute determinism from the point of view of physics" is ill-defined anyway. Indeed, if I accept absolute determinism then I loss any interest in physics (and anything else, probably). Indeed, it may be that all articles that I send to journals are predetermined, but do they describe results of my experiments? Not at all. The conspiracy of my (and our) atoms makes me sure that I observed A, no matter that really I observed B. You surely can continue this sought much further. Not only Einsteins locality means nothing under absolute determinism; but also any intellectual activity does. Boris Tsirelson (talk) 14:58, 28 December 2013 (UTC)

Sorry, I did not write it clearly. I'd say, there are two kinds of absolute determinism: with conspiracy, and without conspiracy. The latter is quite acceptable for classical physics; "hidden variables" that predetermine coin tossing are micro degrees of freedom (coordinates of atoms etc); they are a vast reservoir of pseudo-random numbers, and so, probabilistic thinking is OK. But in this case we cannot observe Bell inequalities violated. Conspiracy means that the reservoir contains not at all a noise, but a giant fine grained plan of fooling us all. In this case of course we can (believe that we) observe Bell inequalities violated... as well as any other physical law... Boris Tsirelson (talk) 15:11, 28 December 2013 (UTC)

Richard, if you want an authoritative reference we could start with Bell himself, who said: 'There is a way to escape the inference of superluminal speeds and spooky action at a distance. But it involves absolute determinism in the universe, the complete absence of free will. Suppose the world is super-deterministic, with not just inanimate nature running on behind-the-scenes clockwork, but with our behavior, including our belief that we are free to choose to do one experiment rather than another, absolutely predetermined, including the "decision" by the experimenter to carry out one set of measurements rather than another, the difficulty disappears. There is no need for a faster than light signal to tell particle A what measurement has been carried out on particle B, because the universe, including particle A, already "knows" what that measurement, and its outcome, will be'.

Boris, I agree with you that absolute determinism may not be very interesting from a physics point of view, indeed it could be argued that it makes physics impossible but the fact remains, as Bell says above, that absolute determinism is fully consistent with QM and Bell's theorem. You may not like it and I may not like it but that does not alter the facts. Absolute determinism is not going to be a fruitful area of study for sure but all we need to do is accept the fact that that it does 'escape the inference of superluminal speeds and spooky action at a distance' and then put it to one side as of no interest and look for more interesting physics. We should mention it here though. Martin Hogbin (talk) 16:44, 28 December 2013 (UTC)

OK; but do not forget to inform the reader that this solution solves not only this specific problem (with Bell...) but all kinds of physical (and probably other) problems at a single shot. :-) Boris Tsirelson (talk) 17:15, 28 December 2013 (UTC)

Yes you are quite right, absolute determinism answers all questions, but rather badly. The reason that I want to make the facts clear is that there are people who, for their own philosophical or religious reasons, believe in absolute determinism. It is quite common then for other people to claim that QM, and in particular evidence for Bell's theorem, disproves this belief. This is incorrect and, in my opinion, brings physics into disrepute. I think we should briefly and clearly set the record straight here. The point I really want to make is that subjects like absolute determinism are beyond the scope of physics. Martin Hogbin (talk) 10:43, 29 December 2013 (UTC)

More seriously, Martin, I agree with you that "If words are used with one meaning in the article but understood by readers to have another then the article will be confusing and will need improvement." In particular, Laplace determinism should not be confused with cryptodeterminism. (I am not sure about the terms, but probably you understand me.) Boris Tsirelson (talk) 17:56, 28 December 2013 (UTC)

Yes, the subject of determinism seems to be a quagmire of vague and overlapping historical and philosophical terminology. The best solution, in my opinion, is to explain exactly what we mean by terms in the article where necessary. Martin Hogbin (talk) 10:43, 29 December 2013 (UTC)

I think it is important to distinguish between logical possibilities and reasonably held beliefs. On wikipedia we have to report what reliable sources say. We have to report current concensus opinions. So logical possibilities may be mentioned, if reliable authorities have put them on the table, but if they are minority opinions, then this has to be made clear too. We are a portal to "the real literature" in specific fields, so we have to make use of (and explain, of course) the terminology used in those fields, or we are no use as portal.

The big problem with these policies is that some topics belong to different fields, and there are different communities in which different concensus' reign, and in which different terminology is used. Even to say that Bell's theorem belongs to physics does not solve this problem since within physics there are different communities in which different views are held. In fact the majority of physicists hold opinions on precisely this topic, which, in my humble opinion, are logically self-contradictory ... but that's just the way it is. Not everyone is equally well informed and ideas percolate only slowly from one sub-field to another - roughly with the speed of human reproduction: it is never the case in science that new ideas are proven superior to old ideas; it is always the case that old ideas simply die out, literally. Fortunately scientific generations follow one another a bit faster than human generations. From PhD student to PhD supervisor takes maybe 10 years on average, or less. Richard Gill (talk) 12:41, 29 December 2013 (UTC)

Actually I avoided defining terms exactly in this article, since by doing so one adopts a position. In my own scientific works I'll happily define what I want terms to mean. But the topic of this article is a topic where people do not agree on the definitions. Great! it's a living part of science! It has not been fossilized and forgotten. It is on the move. It's a moving target. Richard Gill (talk) 12:44, 29 December 2013 (UTC)

I'd say, there is determinism that underlies statistical physics (and therefore everyday probabilistic intuition about "wife's birthday as a seed for a (pseudo)random number generator" etc), and determinism that undermines it. Boris Tsirelson (talk) 20:49, 29 December 2013 (UTC)

Nicely put! (and a nice paradox too). Richard Gill (talk) 11:22, 30 December 2013 (UTC)

Are you chaps happy for me to add something to clarify the current statement, 'Non-determinism becomes a fundamental property of nature'? Martin Hogbin (talk) 14:20, 30 December 2013 (UTC)

I'm happy for you to try. Especially if you can add something which can be supported by a literature reference or link to another wikipedia article. Richard Gill (talk) 07:19, 31 December 2013 (UTC)

I would like to do it in a spirit of cooperation. I think we are now all agreed that absolute determinism is consistent with Bell's theorem (and everything else for that matter) but the kind of determinism being considered in the above quote is not absolute determinism but one in which a free choice of the experimenter is assumed.

I suggest that we explain somewhere near the start that the Bell's theorem is consistent with absolute determinism (we could use the quote from Bell as a reference) but word 'determinism' is used in this article to refer to the case where the experimenter is assumed to have a free choice. Maybe you could suggest an appropriate term for this. Having made these facts clear it is then 'business as usual'. What do you think? Martin Hogbin (talk) 17:54, 31 December 2013 (UTC)

Sure, free choice of the experimenter is sufficient for Bell theorem to work. But not necessary. It is sufficient if the choice is made by a random generator. Moreover, it is still sufficient if it is PSEUDOrandom generator, provided that its correlations with other components are negligible for all practical purposes. This is what I mean by "underlies, not undermines". Indeed, to exclude the experimenter from the deterministic universe is rather strange. Fortunately, not really needed. Boris Tsirelson (talk) 21:50, 31 December 2013 (UTC)

Yes, I guess that all is needed is for one process, human or otherwise, not to be completely deterministic although most people like to think that if there is only one non-deterministic element it is us.

Complete absolute determinism of everything (just being sure) is a religious/philosophical position in which everything is fixed in advance and therefore correlated. Its truth is not decidable by physics. Martin Hogbin (talk) 11:13, 1 January 2014 (UTC)

I would like to know what problem we are trying to solve. The article on Bell's theorem is an article in the field of physics, not religion. "Complete absolute determinism" has not been used in any serious scientific work to somehow "resolve" quantum non-locality - with two possible exceptions: (1) MWI. Popular and taken seriously by many physicsts, so I think it needs to be mentioned, even though IMHO it is nonsense. Yet Martin wanted the material on MWI removed. (2) 't Hooft's superdeterminism. Being ignored by just about everyone, and (IMHO) for good scientific reasons. The article is about physics, not about religion, but Martin wants to make some point about an extreme religious/philosophical position which has not had any impact on the literature on Bell's theorem. Richard Gill (talk) 11:47, 1 January 2014 (UTC)

"everything is fixed in advance and therefore correlated" - why "therefore"?? It does not follow. One may believe that the pseudorandom numbers are fixed in advance but still useful for all practical purposes. The other may believe that everything is correlated. Not "therefore". These are the two kinds of determinism: that underlies and that undermines. Boris Tsirelson (talk) 12:56, 1 January 2014 (UTC)

Indeed, "fixed in advance and still uncorrelated" (just like pseudorandom numbers) is a prominent idea in the classical physics, namely, on the interface between mechanics and statistical physics. This combination is probably the form of determinism most important in physics. This is why the "therefore" above is utterly unacceptable. Boris Tsirelson (talk) 18:13, 1 January 2014 (UTC)

(Edit conflict) Some quotes:

"there has recently been a considerable amount of work by philosophers defending the thesis that chance and determinism are consistent."

"It does not apparently require fundamental indeterminism to have a randomised trial."

Stanford Encyclopedia of Philisophy: "Chance versus Randomness". Boris Tsirelson (talk) 19:01, 1 January 2014 (UTC)

Some more quotes:

"one should keep the mind open to a concrete and testable proposal regarding the mechanism of the suspected conspiracy"

"one should not give the broad and unspecific proposal that a conspiracy exists such high prior probability that the concrete hypothesis of the correctness of Quantum Mechanics is debarred"

"This strategy actually is implicit in ordinary scientific method."

Stanford Encyclopedia of Philisophy: "Bell's Theorem". Boris Tsirelson (talk) 21:36, 1 January 2014 (UTC)

Richard, I am not trying to solve any problem. As you have said above many physicists, I would say most, would not see a problem at all if Bell's theorem were verified. My point is that absolute determinism is compatible with both the verification and the disproof of Bell's theorem. This very point was made by Bell and this is very relevant here. To turn this round the other way, all I am saying is that whether absolute determinism exists in our universe is a question that is outside the scope of physics. The problem that I see is that some people claim that verification of Bell's theorem would disprove absolute determinism. This is not the case and we should say so here. I have always understood this to be a generally accepted fact and not at all contentious. Martin Hogbin (talk) 18:53, 1 January 2014 (UTC)

Boris, I am not quite sure of the point you are making. Maybe we were talking to cross purposes so my 'therefore' was wrong.

My point is this. It makes no difference how you set up your Bell's theorem test or what the results are, they are still consistent with an absolutely deterministic universe. Do you agree? Martin Hogbin (talk) 19:50, 1 January 2014 (UTC)

Really, I think this is a wrong question. Bell theorem deduces a prediction from several assumptions. If the prediction contradicts to experiment, then at least one assumption fails.

Now, what exactly does it mean, "results are consistent with X" where X is an assumption? In mathematics it could mean existence of a model that satisfies the assumption X and predicts these results. But Bell theorem is not this kind of statement. It can be formulated as "results are INconsistent with Z" where Z is the conjunction of several assumptions. But not as "results are consistent with X".

Maybe a reasonable interpretation of the phrase "results are consistent with X" is that "Bell theorem does not imply that results are inconsistent with X"? Let us think logically. Bell theorem implies that results are inconsistent with Z. Therefore, also with any X such that X implies Z. And only with such X. Thus, the claim "Bell theorem does not imply that results are inconsistent with X" is true if and only if X does not imply Z.

Absolute determinism does not imply the Bell assumptions. In this sense it is indeed consistent with the results. Well, but can we imagine experimental results that contradict to absolute determinism? Or to its negation? Or to existence of free will (alone, not in combination with several other assumptions)? No, we cannot. Roughly, ANY philosophical assumption is consistent with ANY experimental result.

But then, what is indeed the point? Is it a notable statement? Will it be understood correctly by the reader? I am afraid, not. Yes, I know, you can find in the literature statements of this form and insist that they should be included. Really? Even if we do not know how to interpret them (and so are unable to explain it to the reader)?

Boris Tsirelson (talk) 20:39, 1 January 2014 (UTC)

What I want to say in the article is that Bell's theorem does not disprove absolute determinism. If I understand you correctly we agree that it is not possible to imagine an experiment that can disprove absolute determinism, that is why I say that it is beyond the scope of physics. So in this respect Bell's theorem is just like any other result in physics, no results can disprove absolute determinism, there is nothing special about Bell's theorem. However there are people who either believe in absolute determinism or wonder about it. I do not care about that, people can believe what they like. What I do care about is that some people claim that Bell's theorem disproves absolute determinism. This is wrong and is an abuse of physics. There is no special mathematical or physical connection between Bell's theorem and absolute determinism. The connection is made by ignorant people and I think we should set them straight in this article, especially as this very point was mentioned by Bell. Martin Hogbin (talk) 23:06, 1 January 2014 (UTC)

This is itself logically OK with me (as long as you mention that no results can disprove absolute determinism, there is nothing special about Bell's theorem). But I am afraid some readers will interpret it as "Bell theorem does not disprove the determinism that underlies...". Significantly, it does! This is why I want to balance your claim by the claim that "chance and determinism are consistent" (see the quote above), and that "Bell theorem disproves the determinism combined with chance and locality". (Just because determinism implies CFD.) Boris Tsirelson (talk) 06:54, 2 January 2014 (UTC)

Clarification. By "determinism combined with chance" I mean deterministic dynamics in concert with random choice of initial conditions (of the Universe, at Big Bang, something like that). A small initial randomness amplifies afterwards in unstable processes, leading to (pseudo)random coin tossing, and does not amplify in stable processes, leading to (almost) deterministic planet orbits. Just the widespread picture of classical mechanics combined with classical statistical physics. Much more important for physics than extravagant forms of absolute determinism. Boris Tsirelson (talk) 07:22, 2 January 2014 (UTC)

That is all understood. It is why I would rather work with you and Richard in writing something. I would like to ensure that there is no possibility that what I say could be misunderstood. Confirmation of Bell's theorem certainly does disprove every form of local determinism except the the most extreme form that I am referring to, that is why it is so important to physics. I do not want to dilute that important fact.

I do hope that you understand my point though. Making invalid claims for Bell's theorem brings physics into disrepute and causes people to lose faith in it. Martin Hogbin (talk) 09:52, 2 January 2014 (UTC)

Happy editing. Boris Tsirelson (talk) 12:47, 2 January 2014 (UTC)

I think the problem here is the use by Martin of the word "prove". In mathematics we have proofs. Not in physics. Bell was a physicist, not a mathematician. He presented an argument, not a theorem. One can encapsulate the abstract argument as a theorem stating that one thing is incompatible with the conjunction of three others. We now try to apply this argument, this piece of abstract logical reasoning, to the real world. When we do that we make use of a great deal more of our knowledge/understanding of the real world. We don't prove things, but we can make some things very plausible. I think that one can make a strong argument from Bell plus (plus an awful lot) that irreducible randomness is a basic ingredient of the workings off the physical world. This is nothing to do with philosophical arguments about free will or absolute determinism. This is sensible physics.

The next thing is that we have to realize we are working on wikipedfia, not expounding our personal opinions. I believe that there is a fairly broad consensus among those working in this field, that what I just said is the responsible current scientific position to hold. However it is certainly far from universal. Depending on which scientific sub-community you look at, the consensus might be quite different. For instance, there are an awful lot of physicists who believe that all foundational issues in QM have been settled by the MWI which is an absolutely deterministic theory of everything. Including famous people like Stephen Hawking. And of course if he says so, the majority will go along with that.

I don't see that the article makes invalid claims about Bell's theorem. The issues are subtle, easily misunderstood. The article has to help newcomers get orientated in the field. Any snappy one-liner does not do justice to the topic. Richard Gill (talk) 17:05, 2 January 2014 (UTC)

Richard, I did not call it Bell's theorem; that seems to be its standard name. I agree that 'theorem' is an unusual word to use in physics but I am just following what seems to be common terminology in this particular case. For clarity by "Bell's theorem" I am referring to the case where what it says at the start of this article, "No physical theory of local hidden variables can ever reproduce all of the predictions of quantum mechanics", has been experimentally confirmed. If you want to suggest a better terminology that is fine with me.

You also seem to be misunderstanding what I am asking for. I do not want to use the word 'prove'.

There is a very old and well known philosophical/religious position which states that future is absolutely fixed, in every respect. Every thought and every action are fixed. This implies no free will or choice of any kind. The future is totally mapped out for us all. There may be inconsistent terminology concerning this concept but I am using the term 'absolute determinism' to describe it. What you, or I, think of this belief is irrelevant; it exists.

The generally accepted view on absolute determinism is that it cannot be proved or disproved by any conceivable experiment. This belief is outside the scope of physics, maths, and science. Belief or disbelief in absolute determinism is purely a matter of personal preference. No theorem or experimental result can prove or disprove absolute determinism. Richard, would you kindly indicate here whether you agree are disagree with this statement. I want to ensure that we do not argue about something that we actually agree on.

I think that QM and in particular Bell's theorem and the experiments which (though imperfect) have confirmed it, provide cogent reasons to reject absolute determinism. Or to put it positively: cogent reasons to believe in an inherently stochastic world. But obviously, no proof. There is no "proof" in science. There is only proof in mathematics, where it is another word for "tautology". Richard Gill (talk) 17:07, 3 January 2014 (UTC)

Richard, it seems to me you try not to see people outside science. They do exist. Reasons cogent to you (and me) need not be cogent for them. It is not quite about "proofs", it is about scientific method. Why not write that Bell theorem (being a part of Science) is valid within a general scientific paradigm, and is not intended for use outside it. This is fair, isn't it? For people that come for science it says: do not bother, this remark is not for you. For people that do not, it says: do not bother, this article (and Bell theorem, and science) is not for you (unless you wish); be happy with your beliefs or whatever. Boris Tsirelson (talk) 19:36, 3 January 2014 (UTC)

Well I think this article should first of all reflect current understanding in the field it comes from. After that we can explore whether we can make it more accessible to people from more distant fields. That is a hard job. Every time you read an article by a science journalist about quantum entanglement, Bell and all that, you can see that they have not got the point. Even many hard-core experimentalists working in the field have not got the point (I can give you many amusing examples). So yes: let's also try to cater for people who are interested in religion or new-age mysticism, but let's realise that nobody ever managed before to be both scientifically responsible and accessible to people "outside science". Even "inside science" is a tall order. Richard Gill (talk) 13:20, 4 January 2014 (UTC)

Richard, you are simply wrong in your answer here. Absolute determinism is outside the scope of physics, there are no conceivable experiments that can prove, disprove, or even produce any evidence for or against it. It is purely a matter of personal choice and would be completely irrelevant here but for the fact that some people seem to promote your view, which is that conclusive verification of Bell's theorem would provide cogent reasons to reject absolute determinism.

There are also no conceivable experiments that prove or disprove quantum theory or relativity or .... yet these are considered part of physics. And physicists today "believe" them in the sense of finding them highly adequate frameworks for describing the world as we know it. I don't buy your argument. You don't understand mine. But I don't object to adding some well sourced remarks on the relation between Bell and absolute determinism, as long as it does not "interfere" with the physics content of the article, which is that Bell and all that provides good reasons for "believing in" non-determinism. (Not proof). Richard Gill (talk) 13:20, 4 January 2014 (UTC)

If course there are conceivable experimental results which would disprove relativity (within its realm of applicability). Einstein famously said that it would take one experiment to prove him wrong. So far there have been no such results but one day there could be. QM, relativity, and the rest of physics is subject to experimental verification (not proof) and experimental disproof. That is how physics works.

Absolute determinism is completely outside the scope of physics. Experimental results can never provide any evidence for or against it. It is not even possible to imagine an experiment that would verify or disprove absolute determinism. It can never be verified or disproved by experiment. Martin Hogbin (talk) 22:31, 4 January 2014 (UTC)

All other forms of determinism, in which the experimenter us taken to be able to make a free choice which is not determined in advance (that is to say in which free will is taken to exist) are subject to experimental verification and disproof by experiment. In particular experimental verification of Bells theorem would provide strong evidence against non-absolute determinism. Martin Hogbin (talk) 22:31, 4 January 2014 (UTC)

This subject was discussed by the ancients who concluded that there is absolutely no way to confirm or disprove AD. That has been the accepted view ever since having been confirmed relatively recently by by Bell. I will look for sources which might convince you. Martin Hogbin (talk) 12:18, 4 January 2014 (UTC)

My opinion is voiced; I cannot do more (and has nothing to add). It seems, Richard treats scientific approach as cogent to everyone, and will optimistically insist on this anyway. I do not. Boris Tsirelson (talk) 15:07, 4 January 2014 (UTC)

Now we come to the connection with this article. There of often discussions and arguments about absolute determinism in other places. In such places it is not uncommon to see it claimed that QM, and in particular Bell's theorem, disprove (not my word choice) absolute determinism. This incorrect. No experiment can prove or disprove this belief.

It may well be that a participant in this argument, not me or you, comes to this article to get the truth about the subject. As it is now, it is not at all clear whether the article says that Bell's theorem disproves absolute determinism or not, thus the article fails to elucidate the reader in this respect. All I want to do is to state somewhere words to the effect that QM and Bell's theorem do not disprove absolute determinism. In view of the fact that this is a contentions topic in other places to which we can provide and answer here and the fact that Bell himself commented on this same subject, I think a few brief words in the article are required. Nothing else changes. Martin Hogbin (talk) 18:05, 2 January 2014 (UTC)

Hmmm, I really did not know that somewhere outside physics "determinism" means more than "the past determines the future". As for me, this is (at least) a third very different "determinism". Martin, where can I see it? Boris Tsirelson (talk) 19:39, 2 January 2014 (UTC)

Oops, I've asked a stupid question. Just here: Predestination. Of course, I know that "Everything is foreseen; yet free will is given" according to Rabbi Akiva. Also here: History of the Free Will Problem (I enjoy the word "determinisms" there). Now I understand Martin's position better than before. Boris Tsirelson (talk) 21:12, 2 January 2014 (UTC)

Also interesting source: John Stewart Bell. Boris Tsirelson (talk) 21:29, 2 January 2014 (UTC)

I recall that physicists told me more than once: No, we do not investigate the nature "as is" (this would be too good); we investigate (physical) models of the nature.

Thus maybe we should write: when claiming that Bell theorem establishes a property of the nature itself (rather than of quantum theory or any other theory of the nature) one really means that Bell theorem establishes a property of every possible physical model of the nature (rather than of quantum theory or any other specific theory). Outside physical models, outside the scientific method, Bell theorem does not apply. (This is about the end of the article: "What is powerful about Bell's theorem is that it doesn't refer to any particular physical theory. What makes Bell's theorem unique and powerful is that it shows that nature violates the most general assumptions behind classical pictures, not just details of some particular models.") Boris Tsirelson (talk) 06:53, 3 January 2014 (UTC)

Bell's theorem does not force anyone to any particular metaphysical position, in particular, regarding determinism or non-determinism. But together with everything else we know it does provide a strong argument for irreducible randomness in nature. Now, If you want, you can imagine all the quantum coin tosses which will ever be needed to have been performed in advance, before the big bang, by God. Or you can imagine them being performed one at a time when needed by Nature. There is no difference in the observable consequences, between these two pictures. It's a matter of taste. But the second picture is more economical, so Occam's razor might lead us to prefer it. So on the one hand, it is fine by me that we tell philosophically inclined readers that physics does not *prove* anything. On the other hand philosophically inclined readers do need to be told that Bell's theorem provides support for the idea that indeterminism is a fundamental property of nature, not an "emergent phenomenon". Quantum randomness is not *merely* the reflection of lack of knowledge and lack of control at some underlying more detailed (more complete) level of description. The punch-line is: plausibly, there is no underlying level. (If you want proof, go to religion, not physics). Richard Gill (talk) 10:00, 3 January 2014 (UTC)

Nice. But this is a talk page. Now some of these valuable thoughts should appear in the article. Boris Tsirelson (talk) 11:30, 3 January 2014 (UTC)

Yes. But these nice thoughts have to be attributed to "reliable sources". I just came across http://arxiv.org/abs/1307.0611 by Jose Latorre "Quantum Will: Determinism meets Quantum Mechanics". Richard Gill (talk) 11:43, 3 January 2014 (UTC)

Boris, I strongly agree with what the physicist said to you. Physics is about describing the world in a human-understandable way; no more than that. Many philosophical/religious beliefs are simply outside the scope of physics; things like the existence or otherwise of God and absolute determinism. Attempting to answer these questions using physics is futile. Martin Hogbin (talk) 12:25, 4 January 2014 (UTC)

Philosophical/religious belief is also about understanding the world in a human-understandable way. My scientific knowledge can and does influence my religious belief. Physics does say something about absolute determinism. It does not give proof. But then it doesn't give proof about anything! Finally we just have to figure out what works for ourself. "Proof" does not play much of a role here. But elegance, harmony, beauty, logic, parsimony, predictive power do. Richard Gill (talk) 13:33, 4 January 2014 (UTC)

But to get back "on topic": if anyone wants to add material on the connection or not between Bell's theorem and absolute determinism let them go ahead and try! Preferably with references to decent sources. Bell's original statements for instance, and the new paper by Tarach which I cited. I will look for some more good sources. I do not want to write this myself: it is a huge subject, I think it is tangential to the real issues raised by Bell, it is very subtle, and no doubt contentious as well. Physics has always led to metaphysical speculations and metaphysical speculations have often led to physics. Bell's theorem belongs to the foundations of quantum physics which belongs at the interface between physics and philosophy. That's exactly why it is so interesting to so many people. Bell's theorem and quantum entanglement keep popping up in popular culture. This particular wikipedia article was a terrible mess for many years because of editors with strong personal opinions on the subject who did battle with those with more conventional views. I would like to see the article as neutral as possible. So I would like to avoid controversial issues of interpretation. Concentrate on reporting the facts so that readers can make up their own minds. Richard Gill (talk) 18:56, 4 January 2014 (UTC)

Richard, you are making far too much of this, I think because you have not appreciated the crucial conceptual difference between absolute determinism and other, less strong, forms of determinism. Absolute determinism is a belief that is completely outside the scope of physics and the scientific method. This is not something that I have just made up, it is part of generally accepted mainstream physics.

If you are not convinced, I challenge you to come up with any experiment (you can make up any results that you like) that would provide evidence either for or against AD. Martin Hogbin (talk) 23:26, 4 January 2014 (UTC)

I think I do appreciate the difference. I am not against someone writing something on this topic in this article. I am just telling you that I am not going to do it myself. I told you my reasons. Personally I think this is a pretty irrelevant side-issue. Go ahead and find some good references and write something. I already mentioned a recent publication by Latorre. It looks good. He knows what he is talking about. For me it is important that the article is not put off-balance by going off-topic. I am concerned that adding these platitudes will distract from the main message from the article, namely that Bell gives good reasons to believe in non-determinism. (But of course, no proof). Richard Gill (talk) 09:21, 5 January 2014 (UTC)

I certainly do not wish to distract from the main message of the article, that conclusive experimental verification of Bell's theorem would provide very strong evidence against non-absolute determinism. I only want to write a sentence or two. Bell's own comment seems a perfectly good reference to me.

I do think that this is essential though. Even your last sentence above would seem, on the face of it, to imply that Bell's theorem provides evidence against absolute determinism. Appearing to promote this misconception is very damaging to the authority of WP and to faith in science. Martin Hogbin (talk) 13:36, 5 January 2014 (UTC)

Yes! Bell's theorem provides evidence against absolute determinism. That's what I keep saying and that's what you keep ignoring. It is not a misconception. The statement need to be promoted by Wikipedia since it belongs to current scientific consensus. (Reliable souces...). I am *not* saying that Bell's theorem provides (conclusive) proof against absolute determinism. Don't you understand my English? Evidence is not the same as proof. Evidence can point more or less strongly in one way or another. There can be conflicting evidence. Evidence never leads to certain knowledge - neither in science nor in law. But people do need to make up their minds (at least provisionally), both in science and in law. Richard Gill (talk) 16:02, 5 January 2014 (UTC)

Richard, I understand you perfectly. I have understood the distinction you make for many years. You will note that I have never claimed that anything is proved by Bell's theorem.

Now perhaps you can listen to what I am saying. Conclusive experimental verification (not proof) of Bell's theorem would provide very substantial evidence (not proof) against non-absolute determinism. However it would provide no evidence whatsoever for or against absolute determinism. It cannot possibly do so because absolute determinism is outside the scope of physics, the scientific method, and experiment.

If you do not agree with my last assertion then I challenge you to propose any experiment and its results that could provide any evidence (not proof) against absolute determinism. Martin Hogbin (talk) 16:14, 5 January 2014 (UTC)

I do not agree with your last assertion. My total life experience provides me evidence (not proof) against absolute determinism. I therefore choose to disbelieve it. Similarly I choose to disbelieve in the existence of God. I believe that both of these choices are quite rational. My experience tells me to trust Occam's razor: I choose to believe that our experience of life is not the result of a huge conspiracy. Bell's theorem, and the experimental support for QM, adds support to my choices.

I also don't see the point of paying attention to a nutty extremist religious or extremist purely philosophical point of view in an article on quantum physics. Nuts will gain support for their nutty views from everything, they are not going to desist because of some kind of official health warning on wikipedia web pages. Richard Gill (talk) 16:46, 5 January 2014 (UTC)

We are not talking about what you choose to believe, based on your life experience, we are talking about evidence provided by verification of Bell's theorem. Bell's theorem provides no evidence for or against absolute determinism theorem. If you disagree then explain to me how.

The problem that I am fighting is actually your view; that Bell's theorem does provide evidence about the truth or otherwise of absolute determinism. This view is damaging to science. Martin Hogbin (talk) 17:27, 5 January 2014 (UTC)

I think that I am being rational in rejecting absolute determinism, and the totality of my knowledge, including Bell's theorem, supports me in that. But you think it is rational to believe in absolute determinism? Anyway, don't waste your time arguing with me. Study the literature on Bell's theorem and help improve this and related articles. By the way, experiments don't verify Bell's theorem. They do tend to confirm quantum mechanics. Bell's theorem says that QM is incompatible with locality+realism+no-conspiracy. So experiment suggests we should reject one (at least) of those three. The absolute determinists reject no-conspiracy. It's a logically allowable option. But it flies in the face of the rest of physics. And has zero predictive power. And gives zero aid to understanding. So a rational person is inclined to reject it especially considering there are much more attractive (fruitful, powerfull) options around. QM (in view of Bell) makes absolute determinism even more ludicrous than it already was. Occam's razor. Richard Gill (talk) 22:34, 5 January 2014 (UTC)

You really are making things much harder than they need be. It does not matter whether you reject absolute determinism or not, that is your choice. What you cannot say though is that any physical experiment can produce any evidence whatsoever that supports or opposes absolute determinism. The article cannot make statements like the one you made above, ' QM (in view of Bell) makes absolute determinism even more ludicrous than it already was'. QM (in view of Bell) has nothing whatever to say about absolute determinism and to claim that it does makes physics look silly.

The reason that I am continuing to discuss this with you is that I do not want to end up edit warring in the article. If you are happy to let me add the mainstream physics view to the article that will be fine. I donot need to study any literature to make this improvement. Martin Hogbin (talk) 00:10, 6 January 2014 (UTC)

I already said that I am happy if anyone adds something relevant and well sourced on this subject. The article makes statements (or at least: should make statements) which represent the general view in physics about these things, and whose content is well sourced. For instance, that is why it has some material on the Many Worlds Interpretation even though I personally think that it is complete nonsense. On this talk page I say what I think. I can say what I did say, and I meant it. I did not say the article had to say it. I earlier suggested an "arguments" page for philosophical discussions, so that the "talk" page can concentrate on editorial matters. Richard Gill (talk) 08:31, 6 January 2014 (UTC)

MWI is not equivalent to other interpretations of QM, this has been shown by David Deutsch. The difference only shows up when contemplating experiments where the observer can be isolated from the environment. In practice that would mean implementing the observer using some huge quantum computer that would simulate the brain of an observer. These practical issues are not relevant from the point of view of theory, so one is allowed to ignore such problems. Basically the idea here is that if you perform a measurement with two or more possible results, the laws of physics allow you to undo the measurement while still keeping the memory of the mere fact that a measurement was carried out. Such a time evolution will still be unitary. In the Copenhagen interpretation, the state would have collapsed after the measurement was made, it is then impossible even in principle to undo the measurement. Count Iblis (talk) 13:03, 11 January 2014 (UTC)

Are you sure that the Copenhagen interpretation treats this quantum computer as an observer? I am not. I do not say it states the opposite. Maybe this remains controversial. Probably the Copenhagen interpretation does not define "observer" but takes for granted that an observer can be described classically. Also, before Deutsch the matter appeared in Wigner's friend paradox and Schrödinger's cat paradox (is a cat an observer? is a friend an observer??) Boris Tsirelson (talk) 13:44, 11 January 2014 (UTC)

Yes, I agree that this remains controversial. But then, one way or another, you can't have an equivalence between a theory that says that non-unitary processes exists and one that says that they don't exist. Count Iblis (talk) 21:48, 11 January 2014 (UTC)

Do not be sure; it depends on the equivalence relation used. It may happen that both predict exactly the same probabilities for observable phenomena. Boris Tsirelson (talk) 06:56, 12 January 2014 (UTC)

Copenhagen has an (undefined) "observer" that can cause wavefunction collapse. This means that one should at least in principle be able to detect non-unitary time evolution. Boris in a box measuring the x-component of a spin that was initially polarized in the z-direction leads to a different state in MWI compared to the CI and that difference is in principle detectable. Count Iblis (talk) 10:38, 12 January 2014 (UTC)

I agree with Boris. Saying that a 'measurement' causes wavefunction collapse is better that invoking an 'observer'. The assumed distinction between a human observer and some undefined concept of a quantum computer that can make measurements is not supported by any experimental evidence and can only be regarded as speculation. Martin Hogbin (talk) 13:17, 12 January 2014 (UTC)::::::

Postulating that there exists any process at all that causes a real wavefunction collapse introduces new physics relative to the MWI. Since measurement is presumably not some supernatural phenomenon, it should be either described by the Schrodinger equation, or one has to postulate that the Schrodinger equation isn't exactly valid. While one can say that it is speculation to postulate an absense of new physics, it should be clear that Copenhagen and MWI are cannot be considered the same physical theory, the difference is more than just meta physics, even if for all practical purposes they do make the same predictions. Count Iblis (talk) 12:48, 21 January 2014 (UTC)

I am not sure, but it seems to me that Copenhagen interpretation in its "ancient" form (maybe now people use this term a bit differently) was reluctant to say "there exists a process that causes a real wavefunction collapse". It was not in fashion to speak about "real processes". It was in fashion to say that our mathematical objects are intended to represent our knowledge about nature, not the state of the nature itself (and moreover, that we now understand that the nature does not admit a closer, "literal" description). Boris Tsirelson (talk) 16:40, 21 January 2014 (UTC)

Not wrong but misleading

No, I am not happy with the lead edited by Martin Hogbin on Jan 6. The statement is logically correct. However the reader can readily interpret it as "now physics has three possible ways: to reject locality, realism, or to accept absolute determinism". This is wrong (I think so). What is true is that "now thinkers have three possible ways: to reject locality, realism, or to accept absolute determinism". Two of these ways are possible for science; the third way means rejection of scientific method. This should be clear to the reader. In principle, absolute determinism can be that "underlies", and so, compatible with science (its negation being compatible, too); but this one does not sustain Bell theorem; or that "undermines", by admitting deep conspiracy in seemingly random events; this one ruins science as whole, thus, Bell theorem (as any other scientific claim) loses any meaning and relevance. Boris Tsirelson (talk) 09:22, 6 January 2014 (UTC)

Boris, thanks for your comment. I tend to agree with you although I do not quite agree with your remark, ' '...this one ruins science as whole, thus, Bell theorem (as any other scientific claim) loses any meaning and relevance', but that is a philosophical point that is irrelevant here.

Feel free do discuss it, anyway. Not necessarily irrelevant. --Boris

I did start this thread in the hope that we could work together on some suitable wording. I started with what I thought would be the minimum change to the article. You are right though in that I do not want to project absolute determinism as an alternative scientific stance, which my wording appears to do. I will try to improve it. Martin Hogbin (talk) 09:38, 6 January 2014 (UTC)

Happy editing. Boris Tsirelson (talk) 11:25, 6 January 2014 (UTC)

So Martin wrote: "While this does not demonstrate QM is complete, one is forced to reject locality, realism, or to accept absolute determinism in which there is no free will and the future is fixed." The point Boris and I are making are that for a physicist, only options 1 and 2 make any sense. Option 3 is not acceptable. It is a logical option for philosophers, but it is not an option for physics. And the article is about physics. Bell mentioned this option because he was methodical and logical. He was not promoting it - he was in fact ridiculing it.

Please note that while both the non-locality and non-realism options correspond to well-developed interpretations of quantum mechanics, and both have many supporters, the absolute determinism option does not correspond to any physics and does not have any adherents - except on the one hand, one notable physicist in glorious isolation, 't Hooft, and on the other hand, the many worlds people who argue that reality is not real and the only thing that is real is the wave function of the universe. IMHO the latter is merely a comfort blanket for people who don't want to think and who are happy just to get good predictions. It has not led to any notable physics. IMHO it merely leads to noise. Richard Gill (talk) 12:29, 6 January 2014 (UTC)

Richard, I know perfectly well what is physics and what is not. Bell's theorem is physics and absolute determinism is not. That is exactly why this article must not say or imply that experimental verification of Bell's theorem would provide evidence against absolute determinism. So long as this article cannot be understood to support that fallacy then I am happy. The easiest way of doing this that I can see is to simply state the facts (that verification of Bell's theorem would not provide evidence against absolute determinism), exactly as Bell did. As a physicist, I, like Bell, am neither supporting nor opposing absolute determinism, because it is outside the scope of physics. Your views on absolute determinism, like mine, are irrelevant to this article. Martin Hogbin (talk) 15:32, 6 January 2014 (UTC)

Bell ridiculed absolute determinism. He did not "support it". Good, so please can we make this point clear also in the article? That option 3 is not an option for physics; options 1 and 2 are. Richard Gill (talk) 08:20, 7 January 2014 (UTC)

You are going too far Richard. I have never said that Bell supported AD but neither did he ridicule it. He merely mentioned it as a possible way to avoid the spooky action at a distance. That is all I want to do here. Is my new wording better in your opinion?

I do not understand what you mean when you say AD is not an option for physics. It is independent of physics. It is a philosophical stance that is neither for nor against physics. I completely agree that AD cannot be used as a tool for the advancement of physics; it is not a concept that is in any way helpful to physics, but it is not against physics. Martin Hogbin (talk) 10:37, 7 January 2014 (UTC)

Let me also say that I am quite happy to restore the original wording and mention AD elsewhere in the article if you prefer. Martin Hogbin (talk) 17:51, 7 January 2014 (UTC)

As a result of a discussion with Martin on my talk page I came to such proposal:

I could support (in the article) a disclaimer like this:

Bell theorem deals with classes of physical models rather than a single model; but it does not mean crossing the boundary between physical models of the nature and the nature itself. Philosophical problems concerning relations between the nature itself and its scientific models are a separate matter, not related to Bell theorem and therefore not touched in this article.

Martin's reaction:

The language you use in your disclaimer is too complex to address the disinformation in the intentionally unreliable sources that you have found. We need to say something simple. It need not be long but it needs to make clear to the layman that what your unreliable sources say is nonsense. It may not be physics but it is an important piece of information for the general public.

However, I am afraid, Wikipedia is not a good place for effective polemic with bloggers. I am afraid, it will be very hard to find a formulation both clear to layman and correct, given the very subtle nature of the problem (too subtle even for many physicists, as Richard wrote). Maybe one should instead post to these blogs, saying: look, "Philosophical problems...are...not related to Bell theorem" and in particular, in simple words, Bell theorem cannot refute absolute determinism (treated as a claim about the nature itself, not about its physical models). Boris Tsirelson (talk) 20:13, 7 January 2014 (UTC)

But wait; Martin, you probably has another option: to edit the article Superdeterminism (redirected from "Absolute determinism"). Maybe your claim fits there better than here? Boris Tsirelson (talk) 22:04, 7 January 2014 (UTC)

(And by the way, that article contains a nice phrase: "Superdeterminism has also been criticized because of perceived implications regarding the validity of science itself.") Boris Tsirelson (talk) 22:13, 7 January 2014 (UTC)

Wikipedia is a place to set the record straight. I people ignore us there is not much more that we can do. We have agreed on the facts, that what I (and as it turns out Bell) have called have called absolute determinism is compatible with Bell's theorem. I can see no reason not to add to the article a clear statement to that effect. I am happy to not let it detract from the rest of the article and to try to ensure that our statement is worded in a way that is acceptable to all but to exclude a point clearly made by Bell himself (and agreed to be factually correct by all of us) from an article on Bell's theorem is perverse. What about using Bell's words in the article. This makes rather more of the matter than I would like to but I cannot see that anyone here has a right to argue with them. Martin Hogbin (talk) 22:47, 7 January 2014 (UTC)

Martin, you are quite insistent. You do not seek a compromise. You reject a good option to edit the article "Superdeterminism". Not nice. I am afraid that Richard will be equally insistent in the opposite direction. The problem is that the best service to one category of readers is a disservice to another.

Well, what about such formulation:

Bell theorem deals with classes of physical models rather than a single model; but it does not mean crossing the boundary between physical models of the nature and the nature itself. Philosophical problems concerning relations between the nature itself and its scientific models are a separate matter, not related to Bell theorem. In particular, Bell theorem does not refute absolute determinism (treated as a claim about the nature itself rather than its physical models), as emphasized by Bell himself (the reference)".

Boris Tsirelson (talk) 06:25, 8 January 2014 (UTC)

I think we are beginning to get somewhere now. As a sign of good faith I have reverted my change. I really do not see how clearly refuting a common misconception (which usually uses the word 'determinism' in the way defined in the dictionary) using language that Bell himself used is a disservice to anyone. I do not want to say anything incorrect but I do want to make sure that the average reader understands what is being said.

Yes, we are in some trouble understanding each other, since you represent here interests of those living on one side of "the boundary between physical models of the nature and the nature itself" while I (and Richard) - of those living on the other side. :-) --Boris

I am happy with your suggestion above apart from the bit is brackets, 'In particular, Bell theorem does not refute absolute determinism (treated as a claim about the nature itself rather than its physical models), as emphasized by Bell himself (the reference)'. I understand what you mean but wonder if there is some better way of saying this.

What do you mean? --Boris

I agree that AD is a philosophical stance about nature itself, but what else could it be? The concept does not really make sense in physics. Oddly enough I would be happier with a longer discussion of this point at the end of the proposed section rather than as a bracketed comment in the middle of it. See my suggestion below. Martin Hogbin (talk) 13:19, 8 January 2014 (UTC)

As an alternative we could have your suggestion followed by the quote from Bell. Martin Hogbin (talk) 08:58, 8 January 2014 (UTC)

In fact you proposal above does not make clear a crucial aspect of AD that we discussed at some length. The requirement that determinism extends to human free will or freedom of choice. Without this aspect being made clear your statement is misleading. Could we use something from Bell's statement, 'Suppose the world is super-deterministic, with not just inanimate nature running on behind-the-scenes clockwork, but with our behaviour, including our belief that we are free to choose to do one experiment rather than another, absolutely predetermined' ? Martin Hogbin (talk) 09:11, 8 January 2014 (UTC)

As for me this is possible, but surely not in the lead. Rather, in Section "Final remarks". The lead should be a summary of central points of the article, and this is a detailed discussion of a marginal point.

Richard, do you like to voice your opinion?

Boris Tsirelson (talk) 09:30, 8 January 2014 (UTC)

Sorry for my silence, I was spending some time thinking about other things! I like this discussion. I especially like Boris' remark that absolute determinism actually makes science meaningless and hence makes Bell's theorem irrelevant. I am sure that Bell himself would have liked this! And actually, I have always thought that that was actually what he was saying in those sentences of his which Martin quoted. I think those sentences were written "tongue in cheek", in such a way that both physicsts and philosophers would be happy, and each would think that he meant the opposite to what the other thought. However, this is just my personal interpretation. But having read every published word which he wrote (in the field of quantum foundations) I think I am able not only to read the lines but also before them. Bell has a strong sense of humour! Richard Gill (talk) 16:06, 8 January 2014 (UTC)

I am happy for the statement not to go in the lead. Martin Hogbin (talk) 12:57, 8 January 2014 (UTC)

A suggestion

In particular, Bell's theorem does not refute absolute determinism, which was described by Bell himself as '...not just inanimate nature running on behind-the-scenes clockwork, but with our behaviour, including our belief that we are free to choose to do one experiment rather than another, absolutely predetermined'. Although absolute determinism does allow an escape from Bell's theorem it cannot be regarded as part of physics but more a philosophical statement about nature. Martin Hogbin (talk) 13:19, 8 January 2014 (UTC)

I would add to this: absolutely predetermined in such a way that the outcome observed by Alice measuring one photon in one particular way (which she determined by tossing a coin) depends in a subtle way on Bob's measurement choice (which he made using a pseudo random number generator with random seed set to his wife's birthdate) and Bob's outcome. Physicists and philosophers have tended to be absolute determinists for the last two thousand years:

Omar Khayyam (1048-1131): "The first day of Creation wrote what the Day of Judgement will read"

Baruch Spinoza (1632-1677): “God could not have made things in a different way or in a different order”

Gottfried Wilhelm Leibniz (1646-1716): "One sees then that everything proceeds mathematically - that is, infallibly - in the whole wide world, so that if someone could have sufficient insight into the inner parts of things, and in addition has remembrance and intelligence enough to consider all the circumstances and to take them into account, he would be a prophet and would see the future in the present as in a mirror"

but as scientists they would all have difficulties taking the "absolute determinism" route in order to explain Bell's theorem. It requires more than determinism, it requires deep time entanglement between physical systems of completely different nature which - except in the quantum optics lab - never ever exhibit the kind of dependence which the singlet correlations of quantum physics exhibit. So sure, God could have made the universe that way, but I don't think that Omar Khayyam, Spinoza or Leibniz would have accepted such a *conspiratorial* (senseless) predeterminism. They also thought that the universe ran according to beautiful laws which God himself could not escape from. Richard Gill (talk) 16:28, 8 January 2014 (UTC)

Richard, you are making far too much of this. Bell explained very well exactly what he meant, and as it happens he used the same term as I proposed. Absolute determinism. There are not different grades of this, we are both talking about the strongest, hardest form of determinism possible. The entire future is absolutely fixed, including the fall of a coin, the outcome of a pseudo-random or even a genuinely random number generator or even the fact that I am typing this. You have called it '*conspiratorial* (senseless) predeterminism' and Boris has called it 'malignant'. Both of you are entitled to your opinions on the subject but they are not relevant to this article. It is not a difficult concept to understand and it has been around for centuries. It has no connection whatever with physics or physicists.

You don't take account of Bell's sense of humour, and ability to write things which are acceptable both to physicsts and to philosophers yet contain different messages for the two audiences! Richard Gill (talk) 09:03, 9 January 2014 (UTC)

Which of the philosophers you quote believed in absolute determinism is debatable but some people, for sure, do. All I want to say is that Bell's theorem does not prove them wrong. You might think that they are wrong anyway but that does not matter. Martin Hogbin (talk) 17:42, 8 January 2014 (UTC)

Ah, I see: there is a logical possibility of the "absolute determinism" route in order to explain Bell's theorem. But only logical possibility. "...we never speak of logical modalities beyond the context of logical scholarly discourse. If we say in the course of ordinary discussion that something is (really) possible or (really) impossible, we do not mean anything like logically conceivable or logically inconceivable." Quote from: František Gahér "LOGICAL, SCIENTIFIC AND REAL POSSIBILITY" Katedra logiky a metodológie vied FiF UK, Bratislava [1]. Boris Tsirelson (talk) 20:42, 8 January 2014 (UTC)

Another quote: "In their most common usage, logical possibilities are invoked as weapons of philosophic debate to give plausibility to eccentric views which would be dismissed as absurd if judged by the regular criteria of science and common sense. Thus from bare assertions of the logical possibility of solipsism, universal doubt, universal lying, the prediction of another's action, or the feeling of someone else's pain, we may be inveigled into concluding that there is some evidence that these oddities are empirically or practically possible, because after all they are logically possible." F. Rinaldi, "Logical Possibility", Philosophy and Phenomenological Research, Vol. 28, No. 1 (Sep., 1967), pp. 81-99 [2]. Boris Tsirelson (talk) 21:26, 8 January 2014 (UTC)

*All* the three *scientists* I quote believed in absolute determinism. As all scientists have done for the last 2000 years - up to and including Einstein but excluding Bohr. Right now there is no concensus. But Bell's theorem might well make intellects of the caliber of Omar Khayyam, Spinoza and Leibniz come to accept the existence of true randomness. It doesn't have anything to do with "free will": a universe could run according to *stochastic* rules and our illusion of free will still be merely an illusion, from the point of view of physics (which is a different point of view from that of law, and a different point of view from that of morality).

Incidentally, the pre-Socratic philosophers did not have any hang-up about chance. They saw it as a positive, creative feature in the workings of the world. However the later Greek philosophers didn't buy it and the Judeo-Islamic-Christian tradition has a big problem with it, for obvioius reasons. Hence chance became a monster which needed to be banned from science.

By the way, it doesn't matter what is true and what is false, it doesn't matter what I think or not. What matters on Wikipedia is the consensus in relevant authoritative publications. The scientific consensus is that the kind of absolute determinism needed to "fix" Bell's theorem is not worth putting on the table.

Here is a new pre-publication about experimental test of super-determinism: http://arxiv.org/abs/1401.0286 "Testing Superdeterministic Conspiracy", Sabine Hossenfelder: Tests of Bell's theorem rule out local hidden variables theories. But any theorem is only as good as the assumptions that go into it, and one of these assumptions is that the experimenter can freely chose the detector settings. Without this assumption, one enters the realm of superdeterministic hidden variables theories and can no longer use Bell's theorem as a criterion. One can like or not like such superdeterministic hidden variables theories and their inevitable nonlocality, the real question is how one can test them. Here, we propose a possible experiment that could reveal superdeterminism.

By the way Martin I am not fundamentally disagreeing with you. I'm saying that things are more subtle than you seem to realise. The history of the "popularization" of Bell's theorem has been one long consistent history of *over* simplification. I would hope that this article would not follow that route. Richard Gill (talk) 08:49, 9 January 2014 (UTC)

It seems that we still cannot even agree on the meaning of terms that we use here. When I talk of 'absolute determinism' I mean the complete absence of free will, that the future is immutably fixed, there is no such thing as chance. This is the same meaning that Bell gave to the term in my quotation from him (whether you somehow know he meant something completely different is irrelevant). Can we please agree to use one term here for one concept only, even if this is not standard usage elsewhere, otherwise we will go on arguing forever. If you would prefer to use a different term than tell me what term you would like to use and we will all use that. With my meaning there is no room for different kinds of absolute determinism, there is only one kind. If you think there is still some ambiguity please say what it is and we can discuss it and agree some terminology. At the moment it seems we are still all talking about different things. Martin Hogbin (talk) 11:18, 9 January 2014 (UTC)

I agree with the terminology. I'm saying something which might be difficult to grasp: even if there is no free will and the future is immutably fixed (which is the standard working hypothesis of classical physics), Bell's theorem still has something interesting to say if one approaches it from the point of view of physics. I am not talking about philosophical word games and I'm not talking about logical possibilities. I'm talking about scientifically plausible pictures of how the world works. Richard Gill (talk) 09:31, 11 January 2014 (UTC)

I think we are very close to full agreement, although I do remain puzzled by your statement that AD is the starting hypothesis for classical physics. I would agree that scientists accept that AD is possible, but they then ignore that possibility and carry on with their work. I personally do not think that the acceptance of AD in any way breaks physics in general or makes it pointless, I am indifferent to it, but that is a discussion for elsewhere.

But I tried to explain you this point; you did not react; I also invited you to explain me the opposite point; you did not try. --Boris

On the other hand, accepting AD as the explanation for Bell's theorem achieves nothing for the physicist. In fact it reduces what is a startling and powerful theorem to pointlessness and I am most certainly not promoting that point of view and I am am perfectly happy to ensure that what we write in the article is not seen to do so. My hope is that Bell's theorem and its experimental confirmation confirmation (or disproof) will help physics move forward. (We may possibly disagree on whether any 'explanation' at all is required for Bell's theorem at all but again that is a discussion for elsewhere).

Regarding this article, my point is that there is a common misconception that Bell's disproves AD. I think we have agreed that this misconception is indeed incorrect. The only question is how can we say this in the article without promoting alternative misconceptions. Do we agree on this point?Martin Hogbin (talk) 10:58, 11 January 2014 (UTC)

If we do agree on that last point then we can work together on improving the article. We should take any irrelevant philosophical disagreements elsewhere (or drop them if you are not interested). Martin Hogbin (talk) 11:08, 11 January 2014 (UTC)

"Very close to full agreement"?? While for one of you logical possibilities are important while for the other they are ridiculous? It seems, the only possible agreement is, to agree to disagree and leave to the reader to choose an attitude to logical possibilities. Boris Tsirelson (talk) 15:01, 11 January 2014 (UTC)

Indeed, "does not disprove" means "leaves a possibility for"; and now the question is, what is meant by "possibility". Boris Tsirelson (talk) 15:37, 11 January 2014 (UTC)

I do not understand what you are saying. When I say that Bell's theorem does not disprove AD I do mean that it allows the possibility that AD is true. I thought that we had already agree this fact. If Bell's theorem were conclusively verified that would not affect whether AD was true or not, it would provide no evidence at all for or against AD, it would leave open the possibility that AD was true or that AD was false. Do you agree? Martin Hogbin (talk) 23:43, 11 January 2014 (UTC)

And now the question is, what is meant by "possibility". (Please follow the link if you did not before!) What is unclear in these words? "If words are used with one meaning in the article but understood by readers to have another then the article will be confusing" (Martin Hogbin).Boris Tsirelson (talk) 06:14, 12 January 2014 (UTC)

I mean by possibility the normal English meaning of this word. My dictionary gives for 'possible', 'capable of existing' and 'capable of being true'. I refer to a modality that allows me to attach an expressions of belief to a statement (for your reference, as a Bayesian, I choose to attach a value of 1/2 to the probability that AD is true). AD is an unfalsifiable belief like, for example Roman Catholicism. People choose to either believe or not believe in it. The only point I want to make in this article is that conclusive experimental verification of Bell's theorem would not provide any evidence for or against AD, just as it would not provide any evidence for or against RC.

Considering that both you and Richard have stated your agreement with the above fact and that Bell himself made this very same point, the onus is on you to give a good and clear reason why we should not say this in the article rather that for me to give a reason why we should. I do not want to get bogged down discussing in minute detail the meaning of one word. If you think the agreed facts (but not your opinion of them) can be better expressed then suggest some better wording. Martin Hogbin (talk) 10:23, 12 January 2014 (UTC)

In such a case you would yourself say: your personal way of treating the word "possibility" (as well as your subjective probability 1/2 to every logical possibility) is irrelevant. If you think it is the mainstream, try first to edit accordingly such articles as possibility and logical possibility. And in addition, note the two quotes about logical possibilities that I put several paragraphs above. Boris Tsirelson (talk) 14:27, 12 January 2014 (UTC)

I really cannot see what you are getting at. I have looked that the articles that you suggest and I see no particular problems in using the word 'possibility'. No logical contradictions follow from the use of this word. It might be easier if you were simply to tell me what problems you see. Martin Hogbin (talk) 16:08, 12 January 2014 (UTC)

I am sorry, but it seems to me that you simply skip my words whenever they disagree with your understanding. Everything was already said, and most of arguments were said 2-3-4 times. Boris Tsirelson (talk) 19:54, 12 January 2014 (UTC)

If you are just going to be awkward I will go back to editing the article. I reverted my original edit as a sign of good faith and because I wanted to work with you rather than against you

We have all agreed that AD is unfalsifiable. It therefore cannot be falsified by Bell's theorem. This an agreed true fact with an impeccable reference (John Bell). There really is no reason why we should say this in the article.

Maybe you believe AD is an eccentric view which would be dismissed as absurd if judged by the regular criteria of science and common sense but you have not proved that point and many people would disagree with you. It is accepted by many people. In any case that is irrelevant for reasons that I have already given. Martin Hogbin (talk) 23:31, 12 January 2014 (UTC)

:-( Boris Tsirelson (talk) 07:25, 13 January 2014 (UTC)

Boris, I really do want to discuss this issue with you and Richard sensibly. There is clearly some point that you are making that I am not getting, maybe that is because I am too stupid to understand it, maybe it is because you are not making yourself clear; we do not need to argue about that. It really would make things easier if you were simply to tell me (yet again) what your objection is to writing words to the effect that Bell's theorem does not falsify AD in the article. Feel free to cut and past your earlier comments if you like. Martin Hogbin (talk) 19:55, 13 January 2014 (UTC)

(Unindent) A well-known joke:

— What’s the chance of meeting a dinosaur on the street?

— 50%!

— Really! Why?

— You’ll either meet one or you won’t!

Your words:

as a Bayesian, I choose to attach a value of 1/2 to the probability that AD is true

People choose to either believe or not believe in it

The number of people that believe in AD could be relevant in the article about AD (and why really not go there?) but not here.

Richard spent a lot of words arguing why he disagrees to 1/2 in this case. He emphasizes that the correlations (of seemingly uncorrelated values) needed in order to reproduce the entanglement are so much incredible that even a person of the past that believed in AD, but also in a logical order in the nature etc (Richard says it much better than me) could get into doubts.

Thus, your 1/2 makes impression (hopefully wrong) that for some reason (ideological?) all Richard's arguments do not enter your mind. Not even a very little bit (otherwise maybe you would say 0.49 rather than 0.5?).

Well, I do copy-paste this:

Some more quotes:

"one should keep the mind open to a concrete and testable proposal regarding the mechanism of the suspected conspiracy"

"one should not give the broad and unspecific proposal that a conspiracy exists such high prior probability that the concrete hypothesis of the correctness of Quantum Mechanics is debarred"

"This strategy actually is implicit in ordinary scientific method."

Stanford Encyclopedia of Philisophy: "Bell's Theorem".

Once again, it seems to me that the only possible compromise is, to inform the reader about these controversies, and let him/her choose for himself/herself. Boris Tsirelson (talk) 21:26, 13 January 2014 (UTC)

I wish only to add that Bell's theorem does not falsify AD to the article but I cannot see what your objection is to this and how it relates to what you say above. You are leaving me to guess this part.

Is your objection based on the assertion that AD would be dismissed as absurd if judged by the regular criteria of science and common sense? Martin Hogbin (talk) 10:54, 14 January 2014 (UTC)

I am afraid we enter the same loop again.

I mean that Bell's theorem makes AD considerably less believable (at least for people that share some scientific ideas). In Bayesian terms, it decreases its probability considerably. But still not to zero. It is not about the value of this probability. It is about the ratio: (probability after Bell)/(probability before Bell).

I worry that "Does not falsify" (if not commented) may be understood as "does not considerably decrease its probability". Likewise "I am not sure you will not meet a dinosaur on the street" (if not commented) may be understood as "this is a real possibility (in my opinion)", while in fact a logical possibility was meant.

Boris Tsirelson (talk) 16:16, 14 January 2014 (UTC)

Indeed, for me, before Bell, AD was completely acceptable. It was, of course, AD without conspiracy. (Do not blame me for the term, it is not mine, see the quote from Stanford Encyclopedia.) After Bell this option is dead. And AD with conspiracy is for me just a logical possibility. I share the attitude to it expressed in the same quote from Stanford Encyclopedia. I guess, I am one of many... Boris Tsirelson (talk) 16:55, 14 January 2014 (UTC)

Bell's theorem makes AD considerably less believable

We were in a loop because I thought from our previous discussion that you had agree that the above statement is wrong. Conclusive experimental verification of Bell's theorem would give us no information whatsoever on the truth of AD. Bell himself made this very point, although Richard seems to be claiming that he really meant something completely different.

If you dispute this please tell me how. Please note that words like conspiracy and probability have no real meaning within AD. Everything that happens does so with a probability of 1; that is the meaning of AD. There can be no conspiracy because the future is certain. Martin Hogbin (talk) 09:55, 15 January 2014 (UTC)

Surely I disagree. No, I cannot tell you how. I already told you everything I could (rather repetitively), but somehow all words that seem to be quite convincing to me do not enter your mind. This is a strange experimental fact (for me). Therefore I must wait for someone else in order to present my arguments. (I have something to say against your phrase "words like conspiracy and probability have no real meaning within AD", but I do not say, for two reasons; first, the experience tells me that nothing can convince you; second, my objection is in fact contained in my previous messages.) Boris Tsirelson (talk) 11:27, 15 January 2014 (UTC)

I have looked at the article in the Stanford encyclopedia and it does not seem anywhere to make clear that it is referring to AD as we have previously defined it. Bell's theorem does, of course, provide strong evidence against local determinism in the normal sense of the word, where the freedom of action of the experimenter is assumed, however, nothing in the article clearly refers to AD, where even the decisions and choices of the experimenter are completely fixed in advance.

It actually seems to me quite obvious that no experiment can give us any evidence about AD. Whatever the results are the response can always be, yes that was in the original plan.

You and I are free to have our own, and differing, opinions on the subject but, as we have a very clear statement from Bell himself confirming that AD is permitted by his theorem and no source or authority saying that it is not, we must accept that the standard mainstream position on this subject agrees with me. Martin Hogbin (talk) 11:44, 15 January 2014 (UTC)

Yes --- on the level of logical possibilities. Yes, I did agree that AD is not falsifiable. But it was not about the same subjective probability. It was about non-zero subjective probability. Yes, everything can be in the original plan. But some versions of the original plan astonish me more than others. Yes, this is rather non-modest for a mortal. But a mortal is created (if created) in order to think... Boris Tsirelson (talk) 12:47, 15 January 2014 (UTC)

By logical possibility you mean that it is not absolutely ruled out but that it is extremely unlikely, like for example the possibility that you and I would both pick the same random number from the set of natural numbers. Martin Hogbin (talk) 13:42, 15 January 2014 (UTC)

Not necessarily "extremely unlikely". Some logical possibilities are more unlike than others. Also, different people assign different subjective probabilities. Just something (A) known not to be provably impossible, and (B) not (yet?) known to be more than that. Once again,

(A) "one should keep the mind open to a concrete and testable proposal regarding the mechanism of the suspected conspiracy"

(B) "one should not give the broad and unspecific proposal that a conspiracy exists such high prior probability that the concrete hypothesis of the correctness of Quantum Mechanics is debarred"

"This strategy actually is implicit in ordinary scientific method." (Stanford encyclopedia)

This is well-balanced: a reasonable lower bound and a reasonable upper bound. For now we have only "the broad and unspecific proposal that a conspiracy exists". But we "keep the mind open to a concrete and testable proposal regarding the mechanism of the suspected conspiracy". Boris Tsirelson (talk) 17:04, 15 January 2014 (UTC)

I now understand what you are saying but I disagree with you. I can see nothing indicating that your quotes apply to AD. In AD there is no chance. A coin comes up heads because it is destined to do so, a radioactive atom decays when it is its time to do so, I write this now because that is what must happen. The arguments in Stanford do not apply to AD.

Furthermore I have to say that the thoughts of various philosophers do not represent facts or even the opinion of any particular body of thought. They are entitled to their opinion but no more than I am entitled to mine and you to yours. Martin Hogbin (talk) 19:02, 15 January 2014 (UTC)

I did not expect that you'll agree. (This was a logical possibility.) Happy editing. Boris Tsirelson (talk) 19:07, 15 January 2014 (UTC)

New proposal

I propose to add this to the lead. It is brief and uses 'falsify', which I understand to be acceptable to you all.

Bell's theorem does not falsify absolute determinism, which was described by Bell himself as '...not just inanimate nature running on behind-the-scenes clockwork, but with our behaviour, including our belief that we are free to choose to do one experiment rather than another, absolutely predetermined'.

I would have no objection to a longer philosophical discussion of this subject in the body of the article so long as philosophical opinions on the subject are not described as physics. Martin Hogbin (talk) 10:30, 16 January 2014 (UTC)

Sorry it's been a busy week. Lot's of interesting stuff written here. Hope I will come back to it later this weekend. Richard Gill (talk) 08:19, 18 January 2014 (UTC)

But I don't like the new proposal for many reasons. A mathematical theorem does not falsify a philosophical position. Nor does a perfect and succesful physical experiment. I think the sentence is misleading and off-topic for the reasons which were put forward by Boris and myself and which we did not succeed in communicating to Martin. And we were putting forward physical (not philosophical) reasons, which are much discussed in the physical literature. We were not talking about anyone's philosophical opinions but about a scientific discourse within physics.

The next thing I want to say is that in order to conclude from Bell's theorem plus its corroboration by nature, that decent physical theories of reality must either be non-local or non-realist, we have to assume that Nature can effectively generate pseudo randomness (we need to assume this can be done to determine settings on measurement devices). This is not a philosophical opinion but a statement about our physical understanding of the world. We do understand why completely deterministic systems like coin-tosses can for for practical purposes be thought of as unpredicable. This assumption does not give us free will or anything like that. The conclusion (if you don't want to discard "realism") is that at the quantum level genuine randomness is generated as part of the underlying physics of nature. This conclusion again does not suddenly give a space in physics for free will. The universe can be (philosophically) just as well be thought of as blindly following set rules, when those rules include perfect (unpredictable) randomness.

So there is a kind of bootstrapping going on here: assume effective randomness (an uncontroversial part of classical physics), observe a close correspondence between QM predictions and experimental outcomes, and then (if we want to keep "locality") we are led to conclude that "true" randomness exists as well - at the interface between the quantum world and the macroscopic world of day to day reality.

If we follow this argument we have not made the world a more cosy place to live in because we now have free will. No, we have simply come to the conclusion that the world evolves according to laws which include an irreducibly non-determistic element.

And any philosopher can always come along with the logical possibility that actually the correlations between the outcomes of two coin tosses and two photo-detection events are subtly constrained because all four systems actually are only doing what they were *all* told to do at the big bang. Doing physics is a waste of time because we are being cheated so as to see apparently beautiful laws which are actually a total illusion.

Note: if we only look at three of the four systems in a Bell type experiment we don't see anything out of the ordinary at all. Only the third-order correlation is remarkable. But it is remarkable in the opposite way to how things are popularly presented. Knowing three of the four outcomes tells us *less* about the fourth than classical physical mechanisms could possibly allow. The four together have *more* possible joint behaviours than classical physics can allow. The fourth system is *less* constrained given the other three, than classical physics allows. The popular account has that quantum correlations are stronger than classical but this is a false picture. Classical correlations put stronger constraints on joint behaviour than quantum correlations. Richard Gill (talk) 08:57, 18 January 2014 (UTC)

You say, 'A mathematical theorem does not falsify a philosophical position' and I agree and for that reason I was considering changing the wording to start 'Conclusive experimental evidence of Bell's theorem....'.

You go on to say, 'Nor does a perfect and succesful physical experiment'. I agree again, would you be happy with, 'Conclusive experimental verification of Bell's theorem does not falsify absolute determinism.'? That is pretty much the content of your opening remarks.

I also agree with 'And any philosopher can always come along with the logical possibility that actually the correlations between the outcomes of two coin tosses and two photo-detection events are subtly constrained because all four systems actually are only doing what they were *all* told to do at the big bang'. Yes, that is what AD is. It is only a philosophical position. It can 'explain' everything and everything but has no usefulness whatever. None of this, however, negates what we have already agreed on, that 'a perfect and succesful physical experiment' (Conclusive verification of bells theorem) cannot falsify a philosophical position (AD). I must make clear that I do not want to say or imply any more than this. I do not wish to make any judgement about whether AD is a desirable, tenable, or sensible, just to note that it cannot be falsified. Neither do I want to suggest that there is any special connection between Bells theorem and AD; that special connection is made by others, not me. We agree that no experiment can falsify AD.

Your comment, 'Doing physics is a waste of time because we are being cheated so as to see apparently beautiful laws which are actually a total illusion', represents what seems to me to be a misunderstanding of physics. Physicists do not claim to be discovering laws of nature and can make no demands that nature be beautiful. Physicists simply describe the world in ways which are attractive and useful to humans and consistent with experiment. The truth or falsity of AD makes no difference to this.

Imagine a physicist thinking about and working on QM and Bell's theorem. A shaft of golden light comes down from the heavens and a voice from above tells him, 'Inanimate nature is running on behind-the-scenes clockwork, and your behaviour, including your belief that we are free to choose to do one experiment rather than another, is absolutely predetermined'. The physicist reaction is to say, 'Thanks for letting me know.', and to then continue with his work. Martin Hogbin (talk) 10:27, 18 January 2014 (UTC)

I would also ask (if permitted), whether or not seemingly independent events are indeed independent for all practical purposes (of a mortal), even though YOU know everything beforehand. In case of answer "no, they do conspire" I would leave any science forever. Boris Tsirelson (talk) 15:36, 18 January 2014 (UTC)

I'd say, the article is now somewhat unbalanced. It does not say that Bell theorem breaks the (possible before) harmony between underlying determinism and observable randomness. It would be nice to balance the edit of Martin with something like that:

However, Bell's theorem reduces drastically the class of acceptable deterministic theories. Absolute determinism itself is consistent with the observable (even if illusory...) phenomenon of randomness (for instance, statistic of tossing coins follows well-known probabilistic rules). By Bell's theorem, (absolute) determinism compatible both with quantum predictions and the observable randomness must admit bizarre violations of randomness that manifest themselves in presence of entanglement (only, as far as we know).

Is there an appropriate source? Boris Tsirelson (talk) 15:28, 18 January 2014 (UTC)

Does the article not say something along the lines of, '...Bell's theorem reduces drastically the class of acceptable deterministic theories'. If not I agree that it should.

I do not understand 'By Bell's theorem, (absolute) determinism compatible both with quantum predictions and the observable randomness must admit bizarre violations of randomness that manifest themselves in presence of entanglement (only, as far as we know)'. Martin Hogbin (talk) 21:34, 18 January 2014 (UTC)

Proposal 3

Taking account of comments so far, how about?

Conclusive experimental ,verification, evidence of the violation of Bell's ,theorem inequality would drastically reduce the class of acceptable deterministic theories but would not not falsify absolute determinism, which was described by Bell himself as '...not just inanimate nature running on behind-the-scenes clockwork, but with our behaviour, including our belief that we are free to choose to do one experiment rather than another, absolutely predetermined' . Martin Hogbin (talk) 18:52, 21 January 2014 (UTC)

I have made a change to the wording. Better or worse? Martin Hogbin (talk) 10:22, 25 January 2014 (UTC)

No! You surely mean "experimental VIOLATION of Bell's inequality". Boris Tsirelson (talk) 19:30, 25 January 2014 (UTC)

Yes, of course you are right. Even then it is still a bit ambiguous. We need to clearly distinguish between the mathematical facts and experimental verification of the predictions of QM. I will rewrite my proposal. Martin Hogbin (talk) 14:25, 26 January 2014 (UTC)

How about that? Martin Hogbin (talk) 17:22, 26 January 2014 (UTC)

Does either of you think it is misleading now? Martin Hogbin (talk) 09:20, 27 January 2014 (UTC)

I think it is still misleading because even if we were an absolute predeterminist we would not likely believe that the outcome of a coin toss in Amsterdam is equisitely correlated with the click of a photo-detector in Zanzibar in order to generate subtle third-order correlations which do not occur anywhere else in Nature and which do not have a satisfactory physical explanation.

My point being that one *must* distinguish between sensible kinds of absolute predeterminism (after all, Laplace, Leibniz, Spinoza, and all those other great thinkers were eminently *sensible* people) and the kind of absolute predeterminism which makes a mockery of all science. Bell certainly understands this distinction and his intended readers understood it too. It's obvious to the intended readership which kind of absolute predeterminism Bell is talking about here - the ridiculous kind; the useless kind. But it won't be obvious to the reader on wikipedia who has come to the wikipedia pages on quantum entanglement hoping to find support for some New Age thinking. Richard Gill (talk) 12:06, 1 February 2014 (UTC)

I agree with you that we must distinguish between absolute determinism as described by Bell and which is the only form of determinism that I am talking about and other weaker forms of determinism. To my mind the word 'absolute', which is a word that I proposed before I noticed that Bell used the same word, is pretty clear and unequivocal but if you think it needs making even clearer I would suggest starting a section on the body of the article on the subject, where I would be quite happy to make it even clearer what we mean.

As a matter of personal philosophical belief, and thus not strictly relevant to this page, I do not agree with your statement that absolute predeterminism makes a mockery of all science. In my view, and that of many others, it makes no difference whatsoever. Martin Hogbin (talk) 13:41, 1 February 2014 (UTC)