Evolution of prosocial behavior in multilayer populations
Qi Su,Alex McAvoy,Yoichiro Mori,Joshua Plotkin
DOI: https://doi.org/10.21203/rs.3.rs-85624/v1
2020-10-22
Abstract:Abstract Human societies include many types of social relationships. Friends, family, business colleagues, online contacts, and religious groups, for example, can all contribute to an individual's social life. Individuals may behave differently in different domains, but their success in one domain may nonetheless engender success in another. The complexity caused by distinct, but coupled, arenas of social interaction may be a key driver of prosocial or selfish behavior in societies. Here, we study this problem using multilayer networks to model a population with multiple domains of social interactions. An individual can appear in multiple different layers, each with separate behaviors and environments. We provide mathematical results on the resulting behavioral dynamics, for any multilayer structure. Across a diverse space of structures, we find that coupling between layers tends to promote prosocial behavior. In fact, even if prosociality is disfavored in each layer alone, multilayer coupling can promote its proliferation in all layers simultaneously. We apply these techniques to six real-world multilayer social networks, ranging from the networks of socio-emotional and professional relationships in a Zambian community, to the networks of online and offline relationships within an academic University. Our results suggest that coupling between distinct domains of social interaction is critical for the spread of prosociality in human societies.