Individual Submission Summary
Share...

Direct link:

Infants’ understanding of information-seeking

Fri, April 9, 10:15 to 11:15am EDT (10:15 to 11:15am EDT), Virtual

Abstract

Ranging from eye-movements to sophisticated experimentation, information-seeking is ubiquitous in the human realm, both as a sub-goal of everyday tasks and as a goal in itself. Although infants can frequently observe others gathering information, it is an open question whether and how they make sense of such activities since their mental causes and intended consequences are heavily underdetermined by the available evidence: there is an infinite set of potential variables that an observed agent might be interested in, and the very same actions can be deployed to achieve a variety of informational goals. We propose that a naive theory about what generates uncertainty and subsequent information-seeking could play a constraining role in solving this problem of action interpretation. The presented study was put forward to test the hypothesis that the goals of others' information-gathering actions are initially understood as sub-goals of maximizing utility in instrumental tasks.

We tested 14- to 15-month-olds' (n = 24) ability to attribute information-seeking goals and uncertainty to agents’ in a situation where, from the perspective of the observed agent, the occurrence of an action-relevant state of the world fell within a range of possibilities. First, infants observed an agent repeatedly approaching a duck rather than a second object, in different positions. Then, participants were also familiarized with a scene where the objects could hide in one of two boxes, and where the content of these boxes could be seen from a platform reachable by climbing a ladder. Finally, in the test phase, the objects hid in the boxes either visibly (certain condition) or invisibly (uncertain condition) from the agent’s perspective. In both conditions, the agent then used the ladder to arrive at the platform where it gained visual access to the objects’ location (information-gathering detour). We paused the animation and measured infant’s looking time from this point. Pupil dilation was also measured from the point that the agent started to climb the ladder. We reasoned that in the uncertain condition, when the agent had no visual access to the objects’ actual location, infants should represent the costly information-gathering detour as justified by its contribution to overall expected utility, while in the certain condition the detour should be seen as inefficient. Therefore, assuming that infants can differentiate the two events, we expected them to display different looking and pupil size patterns, looking longer and having higher pupil size at the certain condition, where the detour was unexpected.

Preliminary analyses show that the results are in line with our expectations, both in case of looking time (p<0.01) and pupil size (p<0.05). Although this evidence is compatible with the hypothesis that infants are able to attribute information-seeking goals when these serve higher-order instrumental goals, further studies are planned to rule out leaner interpretations of our findings. To our knowledge, in addition to a single verbal study with preschoolers (Huang, Hu, & Shao, 2019), this is the first attempt to investigate how children start to understand the information-directed nature of others’ behavior.

Authors