The co-evolution of individuals and institutions: a group-based approach

Humans are unique in their ability to cooperate as groups to address large-scale collective action problems. This capacity emerges from the co-evolution of cultural psychology and group-level behaviors, resulting in multiscale dynamics from smallgroups (e.g., households) to large institutions (e.g., countries, firms). This thesis develops a framework for modeling the co-evolution of individuals and institutions from a group-based perspective. Drawing on insights from complex systems, philosophy, and cultural evolution, we show how traditional network dynamics are qualitatively transformed by adopting group-based formalisms that jointly represent institutional dynamics and individual behavior. As a first step, we recount the tumultuous history of group ontology, showing how assumptions about the existence of groups have shaped research methodologies throughout the twentieth century. Building on this foundation, we introduce a typology that maps conceptual views of groups onto modeling assumptions from the physics of higher-order interactions. This typology illustrates how progressively stronger ontological commitments—from momentary multi-way interactions to persistent and potentially misaligned group structures—enable richer representations of group dynamics. To formalize these ideas, we propose a group-based master equation framework that captures both enduring group structures and the co-evolution of cultural group traits. We then present two case studies in group-based modeling that reinterpret diffusion dynamics through a group-level lens. The first integrates institutions into contagion models, revealing a counterintuitive dynamic: higher infection rates can lead to smaller utbreaks due to a call for action–given how a contagion spreads, we might see varying institutional responses. The second examines the emergence of new skills in research groups, using programming in the humanities as a case study. Here, groups face a trade-off between accelerating their computational transition—potentially at the cost of overburdening individuals–or adapting more slowly to accommodate learning curves. Finally, we present preliminary empirical investigations grounded in our framework. One project models the co-evolution of COVID-19 contagion and government policy responses; the other examines the computational turn in the humanities, focus-ing on tensions between group-level incentives and individual costs of acquiring new skills. We conclude by reflecting on the theoretical challenge of integrating collective intentionality–how shared perceptions among individuals shape group behavior–into formal group-based models.