People can infer the magnitude of other people's knowledge even when they cannot infer its contents

Inferences about other people's knowledge and beliefs are central to social interaction. However, people's behavior is often consistent with a range of potential epistemic states, making it impossible to tell what exactly they know. Nonetheless, we are still often able to form coarse intuitions about how much someone knows, despite being unable to pinpoint the exact contents of their knowledge. Here, we sought to explore this capacity in humans, by comparing their performance to a normative model capturing this type of broad epistemic inference. We evaluated this capacity in a graded inference task where people had to make inferences about how much an agent knew based on the actions they chose (Experiment 1), and joint inferences about how much someone knew and how much they believed they could learn (Experiment 2). Critically, the agent's knowledge was always under-determined by their behavior, but the behavior nonetheless contained information about how much knowledge they possessed or believed they could gain. Our results reveal that people can make graded inferences about how much other people know from minimal behavioral data, but, interestingly, will sometimes achieve this through simpler approximations to the normative model that get the broad inferences right. Altogether, our paper reveals that people can make quantitatively precise judgments about the magnitude of an agent's knowledge from minimal behavioral evidence.