In the first part of his talk, Roland Muhlenbernd will introduce the model of cultural frames (Gibbons, LiCalzi & Warglien, manuscript). This model approaches the idea that agents distill the complex world around them into simpler situations. More precisely, a cultural frame defines how an agent partitions a space of interactive events (games), whereby she recognizes each partition as a frame game, which entails the average utility values over all games within that partition. The game space under investigation is a two-dimensional parameter space that reproduces a continuum between common-interest games and zero-sum games (cf. Morris & Ui, 2004). In such a setting it can be studied under which circumstances agents are better or worse off than they were able to perceive the reality in full detail.

In the second part of his talk, Muhlenbernd will present results of his current research: population dynamics within cultural frames. Muhlenbernd developed different types of
population dynamics which are related to imitation dynamics (cf. Fudenberg & Imhof, 2006) and which he calls approximation dynamics. For a given group of agents with different cultural frames, Muhlenbernd simulated and analyzed the behavior of the population under different approximation dynamics. He studied and elaborated behavioral characteristics of the population, such as: under which dynamic do the agents’ cultural frames coincide/drift apart, under which dynamics do agents approach a local or global maximum/minimum of the game space, or under which dynamics do agents become primarily cooperators/defectors?

References
Fudenberg D. & L. A. Imhof (2006). Imitation processes with small mutations. Journal of Economic Theory, 131: 251–262.
Gibbons R., M. LiCalzi & M. Warglien (manuscript). Interaction under cultural frames.
Morris S. & T. Ui (2004). Best response equivalence. Games and Economic Behavior, 49 260–287.