Back to All Events

Sapientia Lecture Series: Polarization, Forgetting, and a Computational Approach to Social Epistemology

Lecture by ICE Fellow Daniel Singer

103 Thornton Hall, Dartmouth College
Standard epistemological methodology uses models of agents, reasoning, evidence, arguments, belief, and knowledge but usually in an implicit way. In this talk, Daniel Singer argues by example that much is to be gained by making those models explicit and using formal techniques to analyze them. It is common, especially among psychologists, to see polarization as the product of human irrationality. Using an agent-based model, Daniel argues that the persistent disagreement seen in political and social polarization can be produced by rational agents, when those agents have limited cognitive resources. The main argument for this comes from computer simulations of the model, which show that groups of agents using a rational coherence-based strategy for managing their limited cognitive resources tend to polarize. Daniel then introduces an extension of the model to argue that individual memory limitations are often more important than they are typically assumed to be in social epistemology. How much we can remember and how we forget have large effects on whether groups we're in achieve optimal epistemic outcomes. Forgetting, Daniel concludes, should be a topic of central epistemic importance in social epistemology. But more generally, these two cases show that being explicit about our modeling assumptions and analyzing those models formally can help us better understand the implications of our theories and starting assumptions in epistemology. 

Free and open to all. Reception follows.
The Sapientia Lecture Series is funded by The Mark J. Byrne 1985 Fund in Philosophy.