Spontaneous Syntactic Complexity In Preschool Children with ASD

Background:  Abnormalities in language and communication are hallmarks of the Autism Spectrum Disorders (ASD). In addition to well-established weaknesses in language (e.g. early language delays, weaknesses in high-level discourse, and abnormalities in pragmatic abilities), deficits in syntactic ability (combining words into phrases and using morphemes) have also been reported (Eigsti et al., 2007, 2009).  Previous studies have offered contradictory results regarding the acquisition of grammatical morphemes in ASD; several reports have concluded that individuals with ASD use grammar commensurate with their cognitive abilities (Fein & Waterhouse, 1979; Tager-Flusberg et al., 1990 ), while others have noted significant deficits in syntactic development in ASD that go beyond cognitive deficits (Eigsti, Bennetto, Dadlani, 2007; Bartolucci, 1982; Bartolucci, Pierce, & Streiner, 1980).

Objectives: To clarify the debate regarding syntactic development in ASD, we sought to replicate an earlier finding of delayed syntactic development in young children with ASD.  In addition, we sought to examine the persistence of any grammatical deficits by examining language in children who are older than those previously studied. 

Methods:  Language samples were collected from children with ASD (n = 19) and a group of age- and Performance IQ-matched children with typical development (TD; n = 24) ages 4.0 – 7.6 years. The first 100 utterances of language samples were analyzed using the Index of Productive Syntax (IPSyn) to quantify  grammatical complexity of four language domains: Nouns, Verbs, Questions/ Negations, and Sentence Structure. Standardized measures of language and cognitive skills were collected for all children. 

Results: Preliminary results for 24 children are reported here.  Groups did not differ in overall IPSyn score, or subscale scores for Noun phrases; Verb phrases; or Sentence Structures. Interestingly, the ASD group earned significantly higher scores (p = .001) on the Questions/Negations subscale, producing significantly more questions on 6 of 7 items. Correlational analyses of IPSyn data and standardized measures of cognition, language, and autism severity will be presented.

Conclusions:  A subset of children with ASD produced language that was similarly syntactically complex as a group of age- and IQ-matched TD peers; further analyses will take into account history of language delay and autism severity.  These results differ from a previous study using IPSyn for a younger group of children with ASD (Eigsti et al., 2007). Results could indicate that initial syntactic delays in ASD remediate relatively early. Alternatively, it is noteworthy that the current spontaneous language sample was drawn primarily from the ADOS, which is an adult-directed interaction.  As such, results may suggest that the TD group was sensitive to the social context of their interactions, in that they waited passively for adults to lead the interaction. Participants with ASD may have felt less constrained by the social interaction, and felt uninhibited in asking questions.