Nonclassical total probability formula and quantum interference of probabilities (Q868270)

Quantum joint probabilities include an ``interference'' term foreign to classical mechanics, and while Hilbert space analysis provides an ``elegant'' derivation of the quantum statistics, this ``rather loses links with traditional concepts of probability theory'' (p. 50). The aim of this paper is to give an account of the quantum probabilities within Kolmogorov's traditional analysis by providing a ``simple stochastic model of measurement'' for quantum joint probabilities. At heart is the claim that all properties are ``contextual'', a claim backed up by Kolmogorov's own view that every probability is conditional. In addition ``by changing context we change mathematical representation (Kolmogorovian random variable) of a property'' (p. 51). Systems are ``quite sensitive to perturbations induced by transformations of contexts'' (p. 52, author's stress) so that choice about describing one property can change our descriptions of others. A simple example consists of a system with two properties (shape and colour) each with two values (cube, sphere; red, blue). The contextual nature of properties in quantum systems means that after measurement ``some cubes can become spheres and vice versa some red bodies can become blue'' (ibid.). The simple example is developed formally in terms of a pair of two-valued independent random variables over a statistical ensemble. In section 3 ``Nonclassical total probability formula'' the stochastic analysis is generalised to sequences of pairwise independent random variables, and a limit theorem for such random variables is proved (p. 54), apparently extending earlier work of Khrennikov. Five ``remarks'' follow to elaborate aspects of the theorem and to demonstrate analogies with quantum probabilities, for example (Remark 4) between ``contextual'' and quantum interference terms. This is not the first paper to argue that quantum probabilities can be explained in a Kolmogorov framework (this reviewer did so in 1984 and work of Khrennikov is referenced). The real interest of the paper lies in the detailed development of a stochastic model based on the sequences of random variables, and the theorem that helps show this does yield frequencies with appropriately quantum characteristics. This paper deserves a wide readership so it is a pity that glaring typographical errors ``for example in the date referenced for Kolmogorov's opus (p. 50) and variable names in the example'' (p. 52) have been allowed to mar an otherwise thoughtful paper.