Intact Grammar in HFA? Evidence from Control and Binding

This study contributes original results to the topical issue of the degree to which grammar is intact in high-functioning children with autism (HFA). We examine the comprehension of binding and obligatory control in 26 HFA children, mean age=12;02, compared with two groups of younger typically developing (TD) children: one matched on non-verbal mental age (MA), mean age=9;09, and the other on verbal MA, mean age=8;09. On the binding task, our HFA group showed a good performance on reflexives on a par with TD matched children, in line with recent reports of intact knowledge of reflexive binding in higher but not lower-functioning children with autism. Their comprehension of personal pronouns was somewhat poorer, with no difference observed between the groups, again supporting the existing literature. Results on the control task, which probed mastery of syntactic relations never previously examined in autism, revealed that both HFA children and the two matched TD groups were at ceiling on single-complement subject control (try) and object control (persuade). However, a considerably poorer attainment on double-complement subject control (promise) was present equally in the HFA group and the verbal MA-matched TD group but not in the non-verbal MA-matched group. Performance on promise correlated with age only in the verbal MA-matched group, whilst in HFA it correlated with general cognitive and language abilities. These novel findings demonstrate that regular obligatory control and reflexive binding are preserved in HFA. We contrast these results with previous literature that has demonstrated deficiencies with passives and raising in HFA populations. The emerging bifurcation suggests different analyses for the principles underlying these constructions: whereas the latter incorporate movement, control and binding do not. The poor performance on promise supports all previous literature on this lexically and syntactically idiosyncratic construction. Its breaking of locality, which in turn results in a conflict between lexical and syntactic requirements, is exceptional and introduces an extra step of learning. This step appears to be related to maturation in TD children, and to stronger language and cognitive skills in HFA children.