Language Development in Children with ASD: A Longitudinal Study of Grammar and Lexicon

Background: Within language development research, there is an ongoing debate about the relationship between lexicon and grammar acquisition. While lexical items must be learned, grammatical rules may be part of a domain-specific system (Pinker, 1999) or may be learned through domain-general mechanisms as a continuation of lexical learning (Bates & Goodman, 1999). Children with autism spectrum disorders (ASD) provide a unique perspective: research findings are mixed as to whether children with ASD have intact grammar (Tager-Flusberg, 1994) or show impairments (Eigsti et al. 2007). Could children with ASD show a dissociation of these areas of development? If so, are lexicon and grammar simply delayed or do they have deviant trajectories? We address this using a longitudinal data set and growth curve analyses (Singer & Willett, 2003).

Objectives: We will explore the trajectories of lexical and grammatical development in young children with ASD and typically-developing (TD) children.

Methods: The current study followed 28 children with ASD and 30 TD children, assessing their language comprehension and production at six visits spaced four months apart, and again at an outcome visit two years after visit 6. The groups were matched on language ability at the first visit (at visit 1, TD children: Mental Age = 20 months, mean CDI =118; children with ASD: Mental Age = 33 months, mean CDI = 94). Through parent-child play sessions, we assessed lexical development (type-token ratio; TTR) and grammatical development (mean length of utterance; MLU).

Results: We analyzed children’s TTR for nouns and verbs, in order to explore lexical development with a production measure, and found significant group differences: whereas the TD children’s noun TTR increased as they learned more types, their verb TTR decreased, suggesting that they are using the verbs they do know more flexibly. In contrast, the children with ASD showed increases in both their noun and verb TTRs, suggesting that they were not increasing their verb flexibility and instead used different types over many situations. However, MLU (tapping grammatical development) showed significantly different slopes across the groups, with the ASD group progressing more slowly over time than the TD group.  At the outcome visit, children with ASD do score significantly lower than TD children on both lexical and grammatical measures (Test of Auditory Comprehension of Language vocabulary: t(28)=7.05, p<.01; TACL syntax: t(28)=8.55, p<.01).

Conclusions: These results illustrate an interesting pattern: while TD children show a characteristic pattern of increase in their MLU, TTR is more complicated and dependent on whether we look at noun or verb development. Children with ASD show continual improvement in some lexical aspects (nouns) but delayed grammatical development. This may point to a dissociation in the lexical and grammatical development of children with ASD, which bears on theories of language development overall.