Social signals scaffold learning from novel cues during infancy

Infants are bombarded with a bewildering array of events to \nlearn. In such an environment, referential cues (e.g., gestures \nor symbols) highlight which events infants should learn. \nAlthough many studies have documented which referential \ncues guide attention and learning during infancy, few have \ninvestigated how this learning occurs. The present eyetracking \nstudy provides clear evidence for a social scaffolding \nprocess: When preceded repeatedly by communicative signals \n(i.e., a face addressing the infant), 9-month-olds learned that \narbitrary cues predicted the appearance of an audio-visual \nevent. Importantly, the arbitrary cues continued to guide \nlearning of these events, even after the face disappeared from \nthe screen. A control condition confirmed that learning from \narbitrary cues alone was unsuccessful, and that eventual \nsuccess was not just due to extended practice. These results \nare discussed in terms of a theory of cue scaffolding.