Exploring The Role of Referential Context in Young Children's Cross-Situational Word Learning
MetadataShow full item record
A fundamental question in early language acquisition concerns how young children learn new words despite often hearing them in visually complex and noisy environments. One potential mechanism that has emerged in recent years is "cross-situational" learning, which refers to young children's ability to track word-object co-occurrences across several ambiguous naming instances to learn new words. Often in cross-situational learning studies young children's learning is assessed using their eye-gaze and pointing behaviors in response to a forced-choice comprehension question (e.g., “can you point to the bosa?”). Although these testing procedures reveal that some learning has occurred because of the cross-situational exposure, they are not informative about the nature of the word representations underlying that learning. In the current research, 4-year-olds first learned eight new word-object pairs in a cross-situational task and then had their event-related potentials (ERPs) recorded in a picture-word matching paradigm. In this task, pictures of objects were followed by words that were either associated with the object during training or with a different object. If participants learned the word-object associations, then when these were violated during test, I expected to see modulations of ERP components in early stages of word processing. If participants assigned meaning to the words, then I expected to see the early differences as well as differences on the later N400 component of the waveform, a negative peak at approximately 400 msec that indexes violation in meaning. The ERP findings suggest that, (1) young children encoded the new word-object pairs and assigned meaning to the new words when they experienced familiar object labeling before training; (2) young children were less likely to assign meaning to the new words when they did not experience familiar object labeling before training; and (3) there is preliminary evidence that while not all children learned well without a familiarization routine, those that did showed evidence of forming semantic representation for the new word meanings. I discuss the implications of these findings for the study of the cross-situational learning and for broader theories of early word learning.