Large-scale transitions in societies are associated with both individualbehavioural change and restructuring of the social network. These two factorshave often been considered independently, yet recent advances in social networkresearch challenge this view. Here we show that common features of societalmarginalization and clustering emerge naturally during transitions in aco-evolutionary adaptive network model. This is achieved by explicitlyconsidering the interplay between individual interaction and a dynamic networkstructure in behavioural selection. We exemplify this mechanism by simulatinghow smoking behaviour and the network structure get reconfigured by changingsocial norms. Our results are consistent with empirical findings: Theprevalence of smoking was reduced, remaining smokers were preferentiallyconnected among each other and formed increasingly marginalised clusters. Wepropose that self-amplifying feedbacks between individual behaviour and dynamicrestructuring of the network are main drivers of the transition. Thisgenerative mechanism for co-evolution of individual behaviour and socialnetwork structure may apply to a wide range of examples beyond smoking.
展开▼