Infants’ Understanding of Overarching Goals as Constraints on Action

• In Event: Poster Session 14
  In Poster Session: PS 14 Topic - Social Cognition

Abstract

Many of our everyday actions are driven by overarching goals, such as reaching for a pen to write down a phone number. Infants tend to interpret reaching as goal-directed, expecting an actor who selects one object over another to pursue the same object later on (Woodward, 1998). Contextual factors like the physical environment (Sommerville & Crane, 2009) and the action sequence in which the reach is embedded (Sommerville & Woodward, 2005) can influence infants’ goal attribution. We investigated whether infants consider the context of an overarching goal when interpreting an actor’s selective reach.
Study 1 replicated the finding that infants can attribute an object goal to an actor’s selective reach (Woodward, 1998) using functional objects. 11-month-olds were habituated to an actor reaching for a cup or spoon. When the objects switched locations in test, infants looked longer in the first test trial pair when the actor reached for the new object than the old object (p=.025).
Study 2 asked whether contextual information suggesting an overarching goal would influence infants’ interpretation of the actor’s reach. Infants first saw familiarization events where the actor used each object with a corresponding context object (i.e., the spoon with a bowl to eat, the cup with a teapot to drink). Habituation was identical to Study 1 except that the actor reached for the cup or spoon in the presence of its corresponding context object. In test, the context object was removed. Infants looked equally whether the actor reached for the new or old object (p=.516). Thus, infants did not attribute an object goal to the actor following the change in context.
Study 3 investigated whether infants consider the actor’s construal of the scene. The procedure was identical to Study 2, but in habituation the corresponding context object was hidden from the actor (but visible to infants). In test, infants looked longer when the actor reached for the new object (p=.032), as in Study 1. Thus, infants do not cancel their goal attributions whenever there is a change in context; instead recognising that the actor must have visual access to contextual information for it to influence her behaviour.
Study 4 asked whether infants attribute a person-specific overarching goal, or whether they use their understanding of object function or affordances to interpret the actor’s behaviour. The procedure was identical to Study 2, except that the familiarisation events were completed by a different actor than the habituation or test events. In test, infants looked longer when the actor reached for the new object (p=.009), as in Study 1. This suggests that infants do not simply associate the presence of a bowl with a reach for a spoon, instead selectively attributing an overarching goal to someone who has previously used these objects together.
Our studies show that 11-month-olds no longer attribute specific object goals when contextual information visible to the actor is removed from the scene. Infants may have inferred an overarching goal (e.g., eating) as constraining the actor’s reach (e.g., for the spoon).

Authors

• Michaela Dresel, Victoria University of Wellington
  Presenting Author

• Alia Martin
  Non-Presenting Author