Interpersonal links in sociometric structures are interpreted as input-output channels for the transmission of influence. Magnitudes of influence flow generate objective indices of cohesiveness, which serve as the basis for the identification of cliques. This stands in contrast to the factor-analytic rationale of similarity of incoming and outgoing choice patterns. The model departs from the classical sociometric tradition by permitting links to have fractional and/or negative strength, and by taking simultaneous account of direct and indirect linkages. The cohesiveness index is useful for inter-group and diachronic comparisons of structures.