Observe, Mimic, Repeat: Evaluating Theories of Order in Social Interaction

• In Event: Middle-Range Theorizing and Its Empirical Tests

Abstract

How do dynamic social systems come to display regular and predictable patterns of interaction? This paper evaluates three explanations drawn from classical and contemporary social theory, operationalized into competing statistical network models. A Tardean model represents the recursive reproduction of action through imitation and repetition; a \textit{Goffmanian} model captures the sequential expectations and turn-taking that support smooth interaction; and a Simmelian model includes triadic closure and group-forming tendencies. Using relational event models applied to more than a hundred empirical event networks across ten settings, I assess each model’s ability to reproduce the observed order and stability of real-world interactions. All three capture existing tendencies, but the Tardean model consistently generates the most accurately-structured patterns, with the Goffmanian model close behind. A fixed-effects baseline model that accounts for individual heterogeneity performs notably worse than social models. These results suggest that interactional regularity arises primarily from the cumulative reproduction of past acts, placing repetition and imitation at the center of social order.

Author

• Ben Gibson, RAND Corp