Elicited Production of Pronominal Clitics in French-Speaking Children with ASD: A Comparative ASD/SLI Study

Background:  

While deficits in pragmatic aspects of language have been well studied in children with Autism Spectrum Disorders (ASD), much less is known about formal aspects of their language, especially in languages other than English, although it is acknowledged that many children with ASD show impairment in these areas. Indeed, the prevalence of impaired phonology and morphosyntax in children with ASD has led to much discussion about the extent, the nature and the implications of this intersection between ASD and Specific Language Impairment (SLI) (see Whitehouse et al., 2008).

Objectives:  

We studied the language of children with ASD and SLI, focusing on exploration of complexity of syntactic computation. Children with SLI avoid syntactic complexity, favoring structures which entail simpler derivations (Jakubowicz 2005; Prévost et al. 2010). Our aim was to investigate whether children with ASD behave similarly.  We elicited production of pronominal clitics in French because object clitics involve movement of the pronoun from the canonical (postverbal) object position to a preverbal position, as in (1b), as opposed to subject clitics (1c), which do not. Avoidance of object clitics has been argued to be a clinical marker of SLI, tied to the computational load demanded by the movement operation entailed in their derivation (Jakubowicz & Tuller 2008).

(1)     a.   Pierre prend le livre             b.  Pierre le prend __         c.  Il prend le livre

               Peter   takes the book              Peter   it takes                   he takes the book

               ‘Peter is taking the book’         ‘Peter is taking it’                ‘He is taking the book’

Methods:  

We tested 20 6- to 12-year-old children with ASD (diagnosed via ADI-R, confirmed by ADOS, Lord et al. 1989), with PIQs ranging from 48 to 108, and with a minimum MLU of 2.5, and 20 age-matched children with SLI. Standardized tests assessed expressive phonology, syntax and receptive vocabulary; an experimental probe elicited pronominal clitics, nominative, accusative, and reflexive (32 items in all). Raven’s Matrices were administered to see whether levels of non-verbal reasoning might be related to language performance. Control groups included 50 typically developing (TD) children (4-year-olds, 6-year-olds, 8-year-olds, and 11-year-olds).

Results:  

Standardized test results revealed that a large proportion of children with ASD, like the children with SLI, had difficulties in both phonology and morphosyntax. Elicited production of clitics showed striking similarities between the ASD and SLI groups: Nominative clitic rates were significantly higher than reflexive clitic rates, which were significantly high than accusative clitic rates, yielding a pattern which mirrors the acquisition sequence observed in TD children. The ASD and SLI groups did not display inter-group differences for these rates, and non-clitic responses were generally similar.  High inter-subject variation was not correlated to non-verbal level, in either group.

Conclusions:  

Structural language deficits in French-speaking children with ASD are widespread, confirming results of studies of English-speaking children.  These include specific difficulties with complex syntactic operations, widening and refining observed similarities with language in SLI.  Difficulties do not appear to be related to general cognitive disability, suggesting that impaired formal language is a separate component of ASD, for many children.