Functional explanations seek to account for the intrinsic randomness of events whose efficient cause cannot be ascertained. In contrast to causal in- quiries which focus on determinate relationships, function statements map the limits of randomness. Social systems form the setting for functional inquiries. Yet, the practical difficulties of resarch on complex structures deny success to attempts to treat systems as the point of departure, as the original datum of information, and functions as its derivatives. Diffuse social systems are not ready subjects of observation; they can only be construed as networks of dependeney relations. Such models provide the matrix for generalizations. Not all parts of a system are interdependent and they do not form an in- terlocking ensemble. Appropriate system models must disclose not only re- ciprocal relations but also positions of relative autonomy. Autonomy or rela- tive independence is a structurally circumscribed condition of indeterminacy in a system. The tempting assumption that all components of a social system which has stood the test of survival are consistent and interdependent can ex- plain change only as a consequence of external invervention. Such a schema may fit stable archaic societies whose alternative to stability is anomie. If we are to understand intrinsic changes which occur in complex and diffuse so- cieties(, ) we have to assume that the consitituents of a system posses varying measures of autonomy which permits non-repetitive and atypical modes of interaction. Ernst Manheim - Bibliographie Monographien Zur Logik des konkreten Begriffs. Miinchen: C. H.