The complexity of social groups and social systems described by graph structures

This paper presents methodology which permits the complete ranking of nondirected graphs (NDG's) on an attribute labelled ‘complexity.’ The technique applies to both small and large systems as might arise in studies of group or organization behavior. The methodology extends to cover the complexity of directed graphs (DG's) and permits the detailed specification of individual and group behavior.For the NDG an abstract automaton representing the participants' interaction or communications function is sited at each node. Each automaton is constructed so its internal complexity is sufficient to realize the minimal social action (e.g. transmission of a rumor and the path followed by the rumor) within the framework of the NDG. It is shown that the complexity of each node automaton depends upon the order of the graph, the degree of the node and the longest path parameter of the graph. The combined complexity of node automata constitutes the complexity of the NDG. The complexity of a DG is specified as a composition of complexities computed for the associated NDG and logical devices which produce the observed behavior. Illustrative examples pertaining to the committee-subcommittee problem and to organizational structures are presented.