Can Children with ASD Learn Language through Overhearing?

Background: Much of the literature on language learning in children has focused on 'child-directed speech' – language directly addressed to a child.  However, cross-cultural research has indicated that children do not REQUIRE child-directed speech in order to learn new words (Ochs & Schieffelin, 1984), and laboratory-based research has confirmed that young children learn new words via overheard speech (Akhtar et al., 2001).  This is a valuable observation, because it highlights the extent to which children are very skillful at 'mining' their environment for useful language input, making good use of information that may not be provided explicitly for their benefit.

Children with autism spectrum disorder (ASD) often have language delays, but some children with ASD do not (Kjelgaard & Tager-Flusberg, 2001). We hypothesize that the ability to learn language through overhearing may help to explain why some children develop language faster than others: a child with ASD who is capable of learning via overheard speech has access to much larger set of useful language input than a child with ASD who is not capable of learning via overheard speech. Given previous findings associating quantityof adult language input with language and communication development for both typically developing children (Hart & Risley, 1995) and those with ASD (Warren et al., 2009; Bang & Nadig, 2015), these invididual differences may help to explain the variability in language ability for children with ASD.

Objectives: This pilot study serves as a first step in exploring the role of overhearing in language learning for children with ASD; the immediate objective was to address whether overheard language is ‘accessible’ for children with ASD.

Methods: This pilot study included two preschoolers with ASD (aged 4;3 and 4;6); diagnosis was confirmed using the Module 2 of the ADOS-2 (both children were speaking in phrases).  Each child completed a within-subject quasi-experimental design (from Akhtar et al., 2001), comprised of two word-learning training conditions and subsequent test trials.  In the ‘addressed’ training condition, children were directly addressed by an experimenter, who introduced a novel label for a new toy.  In the ‘overheard’ training condition, the experimenter instead addressed a second experimenter, offering a second new label for a different toy (while the child with ASD was seated across the room). After the training conditions were complete, the child’s comprehension of each novel label was tested.  The primary question was whether the children were able to learn the new word in the ‘overheard’ training condition.

Results: Both children were successful at learning new words in the ‘addressed’ condition, as well as in the ‘overheard’ condition.

Conclusions: These preliminary results suggest that some children with ASD are capable of monitoring the conversations of others in order to learn new words.  The extent to which children are able to do this could be a variable of interest in explaining heterogeneity of language development.  In the upcoming months, we will enroll additional participants (estimated N=36) and include additional measures of social communication and language to better understand the correlates of this language-learning capability.