Developmental Science Seminar | Dr. Deon Benton

Tuesday, September 21, 2021 1:00pm

Please join us this TUESDAY (9/21) for our next Developmental Science Seminar featuring Dr. Deon Benton, director of the Causality, Mind & Computational Modeling Lab and co-host of the 'It's Innate Podcast'. All are welcome!

Date: Tuesday September 21st
Time: 1:00-2:00 pm
Location: Tobin 521b or via Zoom (Meeting: 916 8523 2011, Email Jen McDermott <jmm@umass.edu> for passcode)

Rethinking Innateness: Using computational modeling and behavioral experimentation to examine causal learning and social evaluation in infants and young children

The abilities to detect causal relations and to evaluate others based on their social actions are cornerstones of cognitive development. Although there is broad agreement about when these capacities emerge developmentally, there is much less agreement among developmental scientists about how they emerge; that is, it is an open question what the mechanisms and processes are that support the emergence of causal learning and social evaluation. According to one perspective, infants and young children possess innate, specialized, and rational mechanisms that allow them quickly to learn about causality and to evaluate others on the basis of their social behaviors (e.g., Gelman, 1990; Gopnik et al., 2004; Hamlin, 2013; Leslie, 1995; Premack, 1990; Spelke & Kinzler, 2007; Xu, 2019). In contrast, my theoretical approach stresses a domain-general framework in which I argue that causal learning and social evaluation are based more on perceptually-based cues than on abstract, innate knowledge and that domain-general rather than domain-specific processes are sufficient to account for their emergence. To this end, I will begin this talk by showing that one specific kind of domain-general learning mechanism—namely, second-order correlation learning—can support causal learning in older infants and young children (Benton, Rakison, & Sobel, 2021; see also Rakison & Benton, 2019). I will then demonstrate the domain-generality of this learning mechanism by showing that it can explain certain aspects of social evaluation in preverbal infants (Benton & Lapan, invited revision, Cognitive Development). In particular, I will outline a novel, mechanistic explanation for infant social evaluation and then demonstrate that when this account is instantiated in a computational (connectionist) model it can explain how infants come to prefer prosocial beings over antisocial beings and make concrete predictions that can be tested in future studies. Finally, I will conclude the talk by discussing a promising future direction that will involve exploring the possible connection between social evaluation and causal reasoning in infants and young children.

Research Area: 

Developmental Science