Cues to Word Learning in Autistic Spectrum Disorders

Background: Cognitive theories of autistic spectrum disorder (ASD) predict difficulties in acquiring new words either because of deficits in attending and/or processing social cues, or deficits in using linguistic context to infer word meaning. However, studies of word learning have focused exclusively on ‘fast-mapping’ abilities; we know little about learning the phonological and semantic features of new words over time. In addition, we cannot be sure why children with ASD fail to interpret social or linguistic cues because previous studies have only been able to focus on the end product of learning rather than the process itself.

Objectives:

1)     to investigate learning of novel words over multiple trials and over time in children with ASD relative to peers
2)     to determine whether words are more easily learned in social versus linguistic contexts
3)     to determine how children with ASD utilize social or linguistic cues to word learning using eye-tracking techniques

Methods: Our participants included 13 children with ASD, 13 children with language delay and 13 typically developing children (aged seven years). Children saw three novel objects on a computer screen and clicked the photo that matched a spoken sentence. In the social cue condition, a female gazed at the target item. In the linguistic cue condition, information in the sentence biased a particular interpretation. We recorded children’s eye-movements as they completed the task. Immediately after the experiment and approximately four weeks later, we assessed word learning via word recognition, definition and naming tasks.

Results: In the recognition task, all participants identified more words learned with social cues than linguistic cues. Similarly, all groups provided more semantic information in definitions for items presented in the social condition, even though semantic information was explicitly stated in the linguistic condition. In the naming task, there was an interaction between group and cue type such that participants with ASD were better at recalling phonological information for words presented with social versus linguistic cues, whereas type of cue did not affect performance in the comparison groups. We are currently analysing eye-tracking data, focusing on the hypothesis that children with ASD are able to devote more processing resources to phonological information in the social cues condition because they do not spend as much time studying the social cue (i.e. the face) as much as peers.
Conclusions: Our results indicate that social cues such as eye gaze and head turn are particularly salient cues for word learning, even for ASD participants. More semantic information was recalled in the social cue condition, suggesting the possibility that social cues are mapped quickly, leaving more time to encode visual features of novel objects. The most notable finding of this study is that children with ASD were better than peers at phonological aspects of word learning, especially when words were presented with social cues. Our eye-tracking analyses will enable us to determine whether success on this task is the result of devoting more attention to sound than meaning. We consider these findings as an alternative mechanism for acquiring vocabulary in ASD.