UIUC Quant Brownbag

Thinking with a Community at Multiple Levels of Analysis

Babak Hemmatian - University of Illinois Urbana-Champaign

Humans overcome their individual processing limitations to effectively manipulate a complex world mainly by relying on knowledge they do not possess but have access to through their communities (Sloman & Fernbach, 2017). Two sets of strategies for outsourcing and collaboration support interfacing with this resource, both adapted to work well with very shallow representations focused on knowledge access rather than content (Hemmatian & Sloman, 2020). A complete understanding of this process is only possible through simultaneous modeling of cognitive processes in the individual and the broader community. In this talk, I will outline a research program that has been laying the groundworks for such an integrated analysis. I will start with a theoretical framework based on Dual Systems theory. I will then present novel behavioral paradigms that identify outsourcing strategies with the intuitive system and examine computational modeling approaches for quantifying them. Afterwards, I will explore natural language processing pipelines for analyzing the community context for outsourcing and recent neuroscientific projects that clarify individuals’ computing constraints in interfacing with it. I will end my talk with an overview of how the different levels of analysis may be combined, and what the implications would be for the brain and psychological sciences.