Epistemic Landscapes and Optimal Search - a podcast by MCMP Team

Jason McKenzie (LSE) gives a talk at the MCMP Colloquium (30 April, 2014) titled "Epistemic Landscapes and Optimal Search". Abstract: In a paper from 2009, Michael Weisberg and Ryan Muldoon argue that there exist epistemic reasons for the division of cognitive labour. In particular, they claim that a heterogeneous population of agents, where people use a variety of socially response search rules, proves more capable at exploring an “epistemic landscape” than a homogenous population. We show, through a combination of analytic and simulation results, that this claim is not true, and identify why Weisberg and Muldoon obtained the results they did. We then show that, in the case of arguably more “realistic” landscapes — based on Kauffman’s NK-model of “tunably rugged” fitness landscapes — that social learning frequently provides no epistemic benefit whatsoever. Although there surely are good epistemic reasons for the division of cognitive labour, we conclude Weisberg and Muldoon did not show that “a polymorphic population of research strategies thus seems to be the optimal way to divide cognitive labor”.

Further episodes of MCMP – Epistemology

Further podcasts by MCMP Team

Website of MCMP Team