Bell-type inequalities for bivariate maps on orthomodular lattices (Q493670)

In non-quantum physics, when a measurement does not have to change the measured state, it is natural to define conjunction ``\(a\) and \(b\)'' of two properties \(a\) and \(b\) as their intersection (meet), i.e., as the largest property \(c\) which is contained in both \(a\) and \(b\). Under this interpretation, for each state \(m\), we can defined the probability of the conjunction ``\(a\) and \(b\)'' as the probability \(m(a\wedge b)\) of the ``yes''-answer to the meet query \(a\wedge b\). In quantum physics, measurements change the state. As a result, for incompatible properties \(a\) and \(b\), the probability \(m(a\wedge b)\) corresponding to the meet \(a\wedge b\) no longer has the clear physical meaning of the probability to observe both \(a\) and \(b\). This physical meaning can be more adequately captured by the probability \(p(a,b)\) that when we first measure \(a\) and then measure \(b\), we get ``yes''-answers in both cases. In terms of this function \(p(a,b)\), the probability \(m(a)\) of a ``yes''-answer to a query \(a\) corresponds to \(p(a,a)\). Such a function \(p(a,b)\) -- with natural properties -- can be defined for general orthomodular lattices; the resulting functions \(p(a,b)\) are known as \textit{s-maps}. Since \(s\)-maps better capture the physical meaning of conjunction, it is reasonable to replace meets with \(s\)-maps in the formulation of Bell-type inequalities. For example, the inequality \(m(a)+m(b)-m(a\wedge b)\leq 1\) is replaced by \(p(a,a)+p(b,b)-p(a,b)\leq 1\). It turns out that for the resulting inequalities -- in contrast to the versions with meet -- the basic inequalities are always satisfied. The satisfaction of other \(s\)-map versions of the Bell-type inequalities is equivalent to the possibility to extend \(s\)-maps to functions \(p(a,b,c,\dots)\) of three or more variables -- which correspond to several consequent measurements.