How hidden are hidden variables?
By 'hidden variables' we usually mean classical variables for beables that satisfy a deterministic mechanics and allow quantum mechanical predictions to be reproduced.
Since I consider expectation values and correlation functions to be beables, they are hidden variables in this sense.
On the other hand they are of course old friends from statistical mechanics and thus in a certain sense not hidden. But while they are there interpreted in orthodox statistical mechanics as mean values over many instances, they are interpreted as irreducible variables in my thermal interpretation. In this sense they did not previously exist, and they are therefore hidden.
That this double interpretation is possible and meaningful is guaranteed by the compatibility of the new interpretation with the formalism of quantum mechanics.
Arnold Neumaier (Arnold.Neumaier@univie.ac.at) A theoretical physics FAQ