Abstract
It is shown that all those theses of traditional logic which were rejected by Russell in terms of a preferred interpretation of 'all' and 'some', in fact lead to inconsistency in any formal system of traditional logic satisfying certain minimal conditions. Hence, Russell's refutation is ultimately independent of his interpretation. Further, the derivation of each of the refutable theses depends crucially on the Bochenski/Lukasiewicz postulate 'Some _A are _A'. If this postulate is removed, the theses which remain are exactly those which translate into theses of quantification theory