Humans are uniquely capable of reaching consensus within large, hierarchically structured societies. Yet the pathways by which consensus emerges, especially under constraints imposed by social organization, remain poorly understood. We use an agent-based model to explore how marriage structure, social group nesting and decision-making norms can shape a group's ability to reach consensus. In our model, simulated agents are embedded in multi-level social networks and possess noisy information. Decisions are spread via three different cascades, each with different interaction norms. We find that grouping of individuals into families via marriages impedes consensus by slowing the rate of information diffusion and elevating informational entropy, especially when nested further into kin groups. By contrast, increasing the size of nested subgroups in a multi-level network reduces redundant social ties and promotes consensus. Finally, decision-making norms that rely on formation of coalitions or representative bodies lead to faster group decisions by bypassing early-stage clustering of information within families. These results offer insights into how consensus dynamics are shaped by social structure and provide a theoretical bridge between research on network topology, collective intelligence and human social evolution. This article is part of the theme issue 'The evolution of collective intelligence'.
Vadavalli et al. (Thu,) studied this question.