The Keynote on Simulating Collective Agency. Emergence and Stability of Autonomy in Horizontal Collective Actors will be given by Prof Johannes Marx.

Team-Reasoning is a theory proposed to explain cooperative behaviour in social settings of strate-gic choice, even when classical game theory fails. It takes its starting point on the well-observed discrepancy between human actor behaviour and judgments in simple one-shot games and the results predicted by game theory. The theory of circumspect team reasoning (Bacharach 1999) ex-tends classical game theory by introducing the idea of frames to game theory. Simply put, an actor engaged in team reasoning is not asking, 'What is the best I can receive in a particular situation?' but 'What is the best we can get in that situation and what is my part in making that happen?'. Address-ing this question enables rational actors to solve coordination games and even action dilemmas characterized by a tension between egoistic self-interests and a Pareto-efficient collective output.

However, the conditions under which team reasoning arises and stabilizes in complex social struc-tures still need to be explored; the results in the analytic literature on the viability of team reason-ing from an evolutionary perspective show a mixed picture and theoretical research on the deter-minants of the emergence of team-reasoning is rare. To address this lacuna, we developed a model for simulating team reasoning in complex social settings, its stability, and performance in different network structures. We find team reasoning to be generally viable and stabilizing under favourable conditions but dependent on social structure. Simulation results also suggest temporal stability of locally clustered team reasoning even in mixed-game settings. We take this as an indicator of the emergence of collective agency. In the last step, we discuss whether the current framework might be extended to model more complex collective actors by enabling those clusters of stable team reasoners to gain a team pay-off, e.g., the average of the individually achieved pay-off, instead of individual pay-offs to stabilize the cluster further. With this project, we hope to contribute to a bet-ter understanding of the emergence of horizontal collective actors on the macro level by modelling the interaction of individuals in social networks and their reliance on team reasoning.