Comment on “Separability of quantum states and the violation of Bell-type inequalities”

The statement by Loubenets [ Phys. Rev. A 69 042102 (2004)] that separable states can violate classical probabilistic constraints is based on a misleading definition of classicality, which is much narrower than Bell’s concept of local hidden variables. In a Bell-type setting, the notion of classicality used by Loubenets Phys. Rev. A 69 042102 (2004)] corresponds to the assumption of perfect correlations if the same observable is measured on both sides. While it is obvious that most separable states do not satisfy this assumption, this does not constitute “nonclassical” behavior in any usual sense of the word.