Infants Can Use a Newly Learned Object for Scene Segregation

Abstract

In the present study, we tested whether prior experience of an object allows infants to segregate a display when competing cues operate against this interpretation. 4.5-month-olds were shown three exemplars of a novel object. Infants were familiarized with a test display consisting of a new exemplar of the novel object abutted against a complementary object. The infants then viewed one of two events in which 1)the objects moved together as a single entity or 2)moved independently indicating their separation. Infants looked longer at the move-together event (M=30.35 sec) than at the move-apart event (M=14.72 sec),t(17)=3.41, p.003 indicating that the novel object was segregated from its complement. Infants in a control experiment who were not familiarized with the novel object, showed no differences in looking times, (M(Move-together)=17.76 sec, M(Move-apart)=17.92 sec),t(16)=0.04, p≤.97. For the first time we show that a small amount of prior experience can exert an effect that overpowers other cues for grouping. Furthermore, 4.5- month-old infants can use a newly learned object for scene segregation despite the fact that the grouping cues in the test stimulus operate against this interpretation. Finally, we show that object learning is robust enough to transfer from 3D presentation with real objects to a video display.