Scientific report on the Workshop “Rich cognitive models for policy making”, January 12-16, Lorentz center, Leiden.


Organisers: V. Dignum, C. Jonker, W. Jager



The aim of the workshop was to derive a perspective on how much cognition is required in social simulation models to make them useful in practical policy issues, such as the conflicts in Afghanistan and Georgia, the contrasting views on energy issues and the diffusion of the iPhone, to name just three very different examples

            The invited speakers presented contrasting views on this matter. B. Silverman presented a simulation framework which was very rich and incorporated many cognitive concepts such as culture. N. Gilbert took the opposite view, and took the stand that cognition is usually not necessary to model higher order phenomena. Here two different approached were contrasted, one which reflected a more engineering view, aimed at building complex models, and one focussing primarily on a Occam’s razor approach, trying to explain social phenomena from simple models. This fuelled discussions during the rest of the week on the trade off between the complicatedness of models versus the transparency of outcomes, a critical issue for understanding and modelling complex systems. Another approach was illustrated by J.M. Bradshaw, who focussed on man-machine interactions. This presentation raised the issue on the level of agent cognition needed for the interaction with people.

            During the workshop a total of 19 attendants gave short presentations on their own work, which further fuelled vivid discussions, as the common interest in cognitive rich agents in social settings were addressed in many different ways. Hence these presentations were both informative for the audience, being confronted with different approaches, as well as for the presenters, getting feedback from different perspectives.

            An important part of the workshop was devoted to work in subgroups. During the first day, attendants joined different topical groups on the basis of their research interest, such as environmental policy, migration in Europe, transportation and transitions in energy. Here discussions emerged on what tools and methodologies could and should be used in developing simulation models that would provide a perspective on policy making in complex environments. During the second day groups were formed along theoretical methodological interests. Here the groups focussed on questions relating to norms, second order cognition (the representation I have of what another thinks of me) and social networks.

            Over the week discussions were aimed at developing ideas for joint research projects. Several ideas for proposals (often aimed at the EU FP7 program) emerged, and discussions are continuing after the workshop, indicating that a number of working groups will actually submit proposals. Also it was decided to organise a special session during the upcoming ESSA conference on rich cognitive models.

            Finally we observed that the discussions during the day and the evening were very vivid, indicating that the workshop contributed to strengthening the links between researchers from different disciplines. The many positive comments we received after this workshop confirmed our belief that this workshop reached it’s aims.