Attractor models provide a generalized way to represent processes found thr
oughout science. A fuller articulation of the attractor framework requires
that it be addressed qualitatively and conceptually as a nonlinear mathemat
ical order residing between cyclical and random processes. Many significant
nonlinear social processes have been identified and analyzed in classical
social theory. These include the circulation of the elites (Pareto), cultur
al dynamics (Sorokin), social differentiation (Durkheim) and rationalizatio
n in modern institutions (Weber). The present discussion develops a qualita
tive consideration of such classical social processes as attractor systems,
and discusses possible applications of such models in computational social
science.