This research explores the relationship between infants' evaluations of others' social behaviors and their reasoning about intentional agency more generally. Study 1 considered the types of evidence used to determine that a novel entity is an intentional agent, capable of mental representations that are about its environment. 12-month-old infants and adults observed a novel entity respond contingently to a confederate experimenter, whose actions were either social or non-social. Intentionality attribution was assessed by the extent to which infants subsequently followed the faceless entity's implied gaze and by adults' use of psychological terms in describing the event. Both age groups limited construal of the entity as an agent to conditions where it had participated in a contingent social interaction. This result demonstrates that the mechanism that attributes intentionality following observed contingency is a context-sensitive inferential process, and is continuous across development. The influence of social context on this process provides new evidence for a rich, integrated concept of intentional agency by the first birthday: intentional agents are seen as inherently social beings.

Studies 2 and 3 documented the emergence of the ability to view one person's social gaze towards another person as a target-directed, goal-driven action. Study 2 used a habituation method to explore infants' abilities to distinguish between presentations of two people engaging in mutual and averted gaze. 10-month-old infants, but not 9-month-olds, looked longer to test presentations of averted gaze, indicating that they were encoding at least one of the actors' looks as directed towards or away from the other. Study 3 used a violation-of-expectation method to investigate infants' expectations for social gaze between conversational partners. After witnessing an actor have a conversation with a hidden person, 10-month-old infants, but not 9-month-olds, looked longer to displays where the second person appeared in a location inconsistent with the direction of the first actor's gaze. The simultaneous onset of the ability to encode social gaze and expectations governing its use suggests that this development reflects a newfound appreciation for how gaze can be used to achieve the social goal of interacting with another person.