Gordon Watson wrote:Joy Christian wrote:***
The following is my latest response to L.J. at Retraction Watch:
All any local-realistic model of the singlet state is required to produce are the correlations
E(a, b) = (1/N) Sum_k { A(a; h_k) B(b; h_k) } = -cosine(a, b)
using the local functions A(a; h_k) and B(b; h_k) defined in eq. (1) of Bell’s 1964 paper.
Nothing else is either required by the known experimental facts or demanded by realism.
***
Thanks Joy,
I like this (very much) and would be happy to learn of any objections or improvements ...
… though I prefer the notation E(AB) since the expectation is over the conjunction of the outcomes A and B.
PS: The respective detector-settings (a, b -- 'reasonable' unit-vectors in 3-space) are already included in the respective definitions of A and B (each of which -- per Bell -- can take only ±1 as values).
Gordon
The singlet correlations are not functions of A, B, or AB. They are functions only of a and b. Therefore the notation E(AB) is incorrect. The correct notation is E(a, b).
***