Search
Browse By Day
Browse By Time
Browse By Panel
Browse By Session Type
Browse By Topic Area
Browse Posters
Search Tips
Register for SRCD23
Personal Schedule
Welcome Letter
Program Guide
Change Preferences / Time Zone
Sign In
Infants generate basic expectations about their physical and social environment. This early knowledge allows them to identify opportunities for learning, preferring to explore and learn about objects that violate their prior expectations (Stahl & Feigenson, 2015). However, expectancy violations may differentially affect how infants interact with the physical versus the social world. As compared to objects, people have additional value in that they intentionally transmit information. If infants have a basic repertoire of principles to understand others, and understanding others’ actions is critical for social learning, then expectancy violations about agency may indicate situations to avoid learning from others, rather than opportunities for learning. In contrast, expectancy violations about people’s actions may signal an opportunity to learn something new about agents and their actions, as previously found in the object domain (Stahl & Feigenson, 2015). To test these hypotheses, Study 1 focused on manipulating the principle of rationality, the expectation that people choose efficient means to obtain their goals (Gergely & Csibra, 2003). Eighteen-month-old infants saw an agent who acted either efficiently (expected) or inefficiently (unexpected) and then labeled an object. Infants’ learning was then evaluated in an audiovisual matching paradigm. In the label trials, infants saw the target object paired with a distractor while either the familiar label or a novel label was played. In the voice trials, infants saw two speakers side by side, the target and a novel agent, while either the target’s voice or a novel voice was played. As hypothesized, infants in the efficient condition, but not in the inefficient condition, recognized the label-object association – they look longer at the target object when hearing a familiar relative to a novel label. However, there was no evidence that infants learned the voice-speaker association in any of the conditions. These results suggest that expectancy violations about people’s actions signal a situation to avoid learning from them. In a follow-up study (Study 2), we are testing whether infants avoid learning from people who are aberrant in any way, or, alternatively, they selectively consider expectancy violations of intuitive psychology. To address this, we focused on manipulating the principle of solidity (Spelke et al., 1992). Previous studies found that infants are surprised when seeing an arm passing through a solid wall (impossible action) rather than around the wall (possible action) (Saxe et al., 2005). In Study 2, 18-month-old infants (N = 58; expected sample = 80) saw an agent who performed either a possible or impossible action as expected by the principle of solidity, and then labeled an object. In the test, infants’ recognition of the label-object association was evaluated. If expectancy violations in both the physical and psychological domains hinder social learning, we would expect infants to learn the label-object association only in the possible condition. Otherwise, infants should also learn the association in the impossible condition, and seeing an unexpected physical behavior could even enhance their social learning.