On the nature of the conducting state of polyaniline