Model Agents: Social Behavior Through the Formal Lens

The course will offer a dive into the most prominent models for studying social behavior. As the saying goes, such models are wrong (i.e., simplifications of reality), but remarkably useful. We will study ways to model segregation, the spread of an idea, the dynamics of beliefs and opinions, the emergence of cooperation and coordination.

The course will mainly follow Paul Smaldino’s recent book Modeling Social Behavior, and will consist of weekly discussions around each chapter. The book doubles as a an introduction to agent-based modeling, but no coding experience is needed, or assumed, for the course. Our primary focus will be the models themselves and what they tell us about reality.

Announcements

By popular demand, we will be starting the classes at 18:00 sharp, rather than with the customary 15 minute delay.

Lectures

Week 1 (October 15, 2024)

We start out by introducing ourselves, followed by a breakdown of the logistics of the course. Details are found on the slides, here.

We followed with a discussion on models in science, and a glimpse of the Netlogo modeling language. Slides are here.

Week 2 (October 22, 2024)

We covered Chapters 1 and 2 of Smaldino’s book. The goal was to arrive at a working model for flocking in birds. We saw how building a model involves starting simple (in Smldino’s words, doing ‘violence to reality’) and building things up to the desired effect. We also saw how just a few rules can already deliver believable features from the real world. The slides are here.

Historically, one of the first examples of a working flocking algorithm was the Boids model. For a quick and intuitive overview of this model, watch this YouTube video.

Beatrice pointed out that this is similar to the Vicsek model studied by physicists.

There is a great deal of research on how various animals (insects, fish) coordinate their movement. Sumpter (2010) is a textbook-style work that covers a lot of this. For brief overviews, see Couzin (2007, 2009). For a nice glimpse into the research going on this topic in nearby Konstanz, check out this talk by Iain Couzin himself.

Week 3 (October 29, 2024)

We study Chapter 3 of Smaldino (2023), which covers the Schelling model of segregation. This is a model that is more relevant to human social behavior, and provides a mechanism for generating segregated communities based on local, individual actions rather than top-down decisions imposed by an authority.

Slides are here.

If you want to play around with the Schelling model, check out Nicky Case’s implementation here.

Week 3 (November 5, 2024)

We study Chapter 4 of Smaldino (2024).

Bibliography

1. Couzin, I. D. (2007). Collective minds. Nature, 445(7129), 715.
2. Couzin, I. D. (2009). Collective cognition in animal groups. Trends in Cognitive Sciences, 13(1), 36–43.
3. Sumpter, D. J. T. (2010). Collective Animal Behavior. Princeton University Press.
4. Smaldino, P. (2023). Modeling Social Behavior. Mathematical and Agent-Based Models of Social Dynamics and Cultural Evolution. Princeton University Press.