Learning from, and about, others

Other people are a hugely rich source of information: we can use their testimony to learn about things we haven't directly experienced, we can use their actions to draw inferences about what they believe, and we can form beliefs ourselves about whether someone else is knowledgeable, trustworthy, reliable, or informative.

Even in early childhood, human beings are sophisticated social learners, understanding that not all information is equally valuable, but there is also substantial change in how children and adults use social information to guide their beliefs and actions. Using a Bayesian framework, I show the evolution of children's and adults' social reasoning, particularly in how they evaluate testimony. My work has shown that substantial developments in children's ability to form more sophisticated representations of others' beliefs and the evidential justification for those beliefs take place between childhood and adulthood.

More broadly, I am interested in how children's and adults' ongoing conceptual development supports our ability to obtain and organize new knowledge, and how these lessons might be applicable to developing machines that can socially learn in much the same way that we do, for example by developing the capacities for theory of mind and pedagogical reasoning.