In 1961 Robert C. Tucker argued, contrary to the then prevailing assumption of the uniform nature of totalitarian systems, that such systems could be classified into several different types for purposes of analysis. Subsequently, H. Gordon Skilling applied interest group theory to his study of Communist politics and, by doing so, also called into question the case for regarding totalitarian governments as a single category of states possessing unique attributes. Skilling asserted that Communist states cannot be considered “conflictless,” as is sometimes assumed, but can be more adequately understood in terms of the competing social forces commonly found in non-Communist societies. Because of the special, but varying, limits imposed by a central leadership elite on the public expression of conflict in the several Communist-run countries, he added, Communist political parties could play special and quite diverse roles. His thesis contrasts with that of Carl J. Friedrich, which stresses the uniformity of the party's role under totalitarianism. According to Professor Friedrich, in his discussions of “the unique character of totalitarian society,” the presence of a single mass party is a common feature of all totalitarian politics, and is “typically either superior to, or completely commingled with the bureaucratic organization.” While Friedrich in his later work, written jointly with Zbigniew K. Brzezinski, acknowledges that “within the broad pattern of similarities, there are many significant variations” in totalitarian dictatorships, the authors' emphasis is on the novelty and uniqueness of these dictatorships. They state: