Retour vers liste

Détail de la contribution

Auteur: Letitia NAIGLES

Grammatical strengths and weaknesses in children with autism: Multiple comparisons of comprehension and production across development

Abstract/Résumé: Controversies exist regarding grammatical knowledge in children with ASD: Some researchers proposed that one subgroup manifests deficits akin to children with SLI whereas others countered that children with autism have a more pervasive grammatical deficit. The origins of grammatical abilities are also debated; some suggest a pragmatic basis while others promote a general cognitive basis. This project addresses these controversies via a longitudinal study of grammatical production and comprehension, during which time pragmatic and cognitive abilities were also assessed. Children are tested every four months for 6 visits. At onset, children with ASD (n=17; MA=33 months) had language scores comparable to the control group of 20-month-old TD children (n=18). Children’s comprehension of four components of grammar was assessed using Intermodal Preferential Looking (IPL) : SVO Word Order, Syntactic Bootstrapping, Wh-Questions, and Aspect. Dependent variables included children’s latency of first look to the matching scene and percent of time looking to the matching scene, during test compared with baseline trials. Spontaneous speech was collected at each visit. Utterances were coded for Brown’s 14 grammatical morphemes, wh-question complexity, MLU, and word types. Spontaneous speech analyses revealed two subgroups of children with ASD: one group, HFA, produced grammar and morphology at similar frequencies and developmental trajectories as the TD children. The second group, LFA, produced many fewer grammatical items overall and had a flatter developmental trajectory for the grammatical items produced. The HFA group also had average standardized test scores; the LFA group had standardized test scores significantly below average. The IPL tasks revealed a different picture: 90% of children followed SVO word order, 87% demonstrated syntactic bootstrapping, and 67-71%% understood the ‘ing/-ed distinction. Wh-question comprehension changed over time: 47% demonstrated understanding at visit 3 and 67% by visit 6. Some LFA children—not always the same subset across tasks—showed reliable comprehension when their production levels were extremely low. Robust correlations were found between the children’s grammatical production and comprehension and their proclivity to engage in joint attention; in contrast, few correlations were observed with their performance on nonverbal IQ tests. We propose that basic grammatical knowledge is generally intact in children with ASD. Difficulties with early grammatical production and comprehension seem more related to pragmatic challenges than to strictly linguistic or cognitive impairments.