# Talk:Bell's theorem

WikiProject Mathematics (Rated B-class, High-priority)
This article is within the scope of WikiProject Mathematics, a collaborative effort to improve the coverage of Mathematics on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
Mathematics rating:
 B Class
 High Priority
Field: Mathematical physics
WikiProject Physics (Rated B-class, High-importance)
This article is within the scope of WikiProject Physics, a collaborative effort to improve the coverage of Physics on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
B  This article has been rated as B-Class on the project's quality scale.
High  This article has been rated as High-importance on the project's importance scale.

## Correction to the substitutions from CHSH (4 term) inequality to the simplified (3 term) version of Bell's inequality

I'm not sure how to perform/commit the correction, so I'll just post it here, in the hopes that a "guru" can fix the main article. The correct substitutions for transforming Eq(1) in §CHSH inequality

$\quad \rho(a,b) + \rho(a,b') + \rho(a',b) - \rho(a',b') \leq 2$

into the (un-numbered) equation at the top of §Original Bell's inequality

$\rho(a, c) -\rho(b, a) - \rho(b, c) \le 1,$

are: b+π ← a, a+π ← a', a ← b, c ← b'. Also, might be useful to note that in order to reproduce the version of the CHSH inequality in this article from that in the main article on the CHSH Inequality,

$S = E(a, b) - E(a, b^\prime) + E(a^\prime, b) + E(a^\prime, b^\prime)$

one needs to perform a 180° rotation of b', that is: b'+π ← b'.

ProfessorJohnFrink (talk) 01:42, 20 January 2016 (UTC)
Here is my understanding of this matter.
Treatment of the parameters in CHSH and Bell inequalities is different.
In CHSH inequality, there are two settings $a,a'$ on Alice side, and two settings $b,b'$ on Bob side. Nothing special (like perfect anticorrelation) is assumed; thus, it is pointless to ask whether $a=b$ etc. The parameters are just (labels of) settings of measuring devices (possibly, of quite different type), not necessarily angles; thus, one should not try to rotate them. CHSH is purely informational. Accordingly, for proving it we introduce observables $A,A',B,B'$ such that $\rho(a,b)=E(AB)$ etc; assume they all commute (according to the local realism); and get $\rho(a,b) + \rho(a,b') + \rho(a',b) - \rho(a',b') = E(AB + AB' + A'B - A'B') \le 2$.
Labels $a,a'$ are assigned to the two Alice settings arbitrarily. Equally well we may swap them, getting $\rho(a',b) + \rho(a',b') + \rho(a,b) - \rho(a,b') = E(A'B + A'B' + AB - AB') \le 2$. There are four such inequalities. In addition, labels $+1,-1$ are assigned to the two possible outcomes arbitrarily, and may be swapped for (say) Alice but not Bob, which changes the sign: $\rho(a',b) + \rho(a',b') + \rho(a,b) - \rho(a,b') \ge -2$ (and therefore $|\rho(a',b) + \rho(a',b') + \rho(a,b) - \rho(a,b')| \le 2$). These are symmetries of the CHSH inequality.
In Bell inequality, there are three angles (not just "settings") $a,b,c$; two of them, $a,b$, are used by Alice, and two of them, $a,c$, are used by Bob; note that $a$ is used by both. Denoting the two corresponding observables of Alice by $A,B$ we see that the first observable of Bob is $-A$ due to the perfect anticorrelation. The second observable of Bob may be denoted by $-C$ or by $C$; both options are legitimate. Let us choose the former one. Then $\rho(a,c)=E(A(-C))=-E(AC)$, $\rho(b,a)=E(B(-A))=-E(BA)$, $\rho(b,c)=E(B(-C))=-E(BC)$, thus, assuming commutativity (as before) we get $\rho(a,c)-\rho(b,a)-\rho(b,c) = E(AB+BC-AC)$.
In order to reduce Bell inequality $\rho(a,c)-\rho(b,a)-\rho(b,c) \le 1$ to CHSH inequality, we subtract $\rho(b,b)=E(B(-B))=-E(B^2)=-1$, thus rewriting Bell inequality as $\rho(a,c)-\rho(b,a)-\rho(b,c)-\rho(b,b) \le 2$, that is, $E(AC-AB-BC-BB) \le 2$. Taking into account the symmetries of CHSH we see that the latter is its special case.
Boris Tsirelson (talk) 19:32, 20 January 2016 (UTC)
In the presence of such august editors, I should perhaps not say too much. But in response to Editor Tsirel's comment "assume they all commute (according to the local realism)", I feel it could be mentioned that 'local realism' is an item of Bellspeak. At stake here are two things: a metaphysical principle, and a religious principle. The metaphysical principle is about causality. The religious principle is that Albert Einstein is a silly old goat, not nearly as clever as his critics. The causality thing is translated by his Bellspeaking critics into a matter that is exactly stated by classical mechanics, which is not a metaphysical principle, but something downstream from there, a physical theory. We don't need Bellspeakers to tell us that classical mechanics is unable to account for the existence and behaviour of atoms. That was apparent by 1920. The Bellspeakers muddle the metaphysical principle with the physical theory, and crow that because the physical theory is inadequate, they have overthrown the metaphysical principle and can thereby see in the dark and walk on water; well, at least they will be able to do so given one more research grant. But anyway, they are cleverer than Albert Einstein; the religious principle stands.Chjoaygame (talk) 22:47, 20 January 2016 (UTC)
No comment. :-) Boris Tsirelson (talk) 06:14, 21 January 2016 (UTC)
On the metaphysical front, a wise move. :-)
I may add that the actual discussion on the formulas offered here by Editor Tsirel is helpful and illuminating.11:51, 21 January 2016 (UTC)
Thanks for the compliment. If your "august" was toward me, you are (to the best of my knowledge) the first to call me this way. Another very generous compliment "great magician" is here. What a collection! Boris Tsirelson (talk) 13:21, 21 January 2016 (UTC)