Andrea Beltrama
MindCORE Postdoctoral Fellow
University of Pennsylvania

Imprecision and speaker identity: How social cues affect meaning resolution

Speakers’ descriptions can be associated with a varying margin of imprecision: an actual time of 2:57 could be reported precisely as “2:57″, or less precisely as “3 o’clock” (Lasersohn 1999 i.a.). This variability introduces an element of indeterminacy in meaning interpretation, making the evaluation of a description contingent on the margin of precision tolerated in a context. In this talk, I explore whether social information serves as a cue for  the resolution of imprecision – that is, whether, listeners interpret an utterance with a different degree of tolerance depending on the expectations generated by the speaker’s identity. Evidence from a picture selection task (N=71) suggests  that (i) reasoning about precision induces a processing cost, especially when it leads to rejecting a statement; and (ii) this cost is modulated by the social identity of the speaker. In particular, it’s easier for listeners to interpret utterances strictly when the speaker belongs to a social group whose members are expected to speak precisely. More broadly, these findings suggest that meaning interpretation can be modulated by social information about the speaker, highlighting a novel angle on the investigation of semantic processing.

Rista Plate
MindCORE Postdoctoral Fellow
University of Pennsylvania

How children make choices when information sources diverge

When making decisions, children often have access to multiple sources of information and must reconcile instances in which sources disagree. Three experiments investigated how children weight divergent information in a dynamic learning context. Participants (4-9-years-old) searched for rewards and saw cues that varied in how well the cue located the reward. Participants were told that the cues were from another player (Experiment 1, N = 102) or because of a computer glitch (Experiment 2, N = 89). Children did not optimally discount social cues that diverged from the rewarded locations, and younger children weighted social cues more heavily than older children when the cues were only sometimes helpful (an effect that was unique to the social, as compared to the “computer glitch,” cue). In Experiment 3 (N = 129), participants received social cues that unexpectedly changed in level of disagreement with the underlying reward distribution over time. All children updated how heavily they weighted the cues after the change. However, younger children were more heavily influenced by their initial experience with the cues. Taken together, these experiments provide evidence to suggest that children are skilled in detecting the usefulness of information sources. However, younger children may have more difficulty disengaging from social information in uncertain learning environments.

A pizza lunch will be served at 11:45am. The seminar will run from 12:00pm – 1:30pm.