Modelling the Dynamics of Collective Cognition: A Network-Based Approach to Socially-Mediated Cognitive Change

A number of studies in the network science literature have attempted to model the effect of network structure on cognitive state fluctuations in social networks. For the most part, these networks use highly simplified models of both cognitive state and social influence. In order to extend these studies and provide the basis for more complex network science simulations, a model of socially-mediated cognitive change is presented. The model attempts to integrate ideas and concepts from a number of disciplines, most notably psychology, evolutionary biology and complexity science. In the model, cognitive states are modelled as networks of binary variables, each of which indicates an agent’s belief in a particular fact. The links between variables represent the ‘logical’ dependencies between beliefs, and these dependencies are based on an agent’s knowledge of the domain to which the beliefs apply. Drawing on the psychological notion of cognitive dissonance, it is further suggested that agents are under internal pressure to adopt highly consistent belief configurations, and this identifies one source of cognitive dynamism in the model. Another source of dynamism derives from the structure of the social network. Here, the existence of network ties creates a dependency between the belief systems of connected agents. Cognitive change in such ‘coupled belief systems’ is modelled using Kauffman’s NK(C) model of co-evolutionary development in biological systems. As a final source of cognitive dynamism, the model incorporates the notion of an aggregate belief system (or cultural model), which represents the dominant set of beliefs associated with specific agent sub-groups. By explicitly incorporating the notion of an aggregate belief system into the model, the model supports the analysis of cognitive state fluctuations at the individual (psychological), social and cultural levels. It also provides the basis for future network science simulations that seek to study the complex interactions between these various levels.