Online Processing of Svo Word Order in Monolingual- and Bilingual-Exposed Chinese Preschool Children with Autism Spectrum Disorder

Background: Previous studies in monolingual-exposed children with autism spectrum disorder (ASD) have witnessed their intact knowledge to comprehend Subject-Verb-Object (SVO) sentences in English or Chinese (Swensen et al., 2007; Zhou et al., 2017; cf. Su & Naigles, 2018). Moreover, Reetzke et al. (2015) and Dai et al. (2017) reported equivalent general language abilities in bilingual-exposed children with ASD via standardized scales. However, in oral format, dialects were differentiated from Mandarin for their syntactic differences (Reetzke et al., 2015). Therefore, a refined investigation needs to be conducted on the understanding of SVO word order in Chinese bilingual-exposed children with ASD.

Objectives: Using the Intermodal Preferential Looking, this study attempts to examine early knowledge of SVO word order in bilingual children with ASD exposed to Mandarin and dialect, compared with monolingual peers merely exposed to Mandarin.

Methods: 28 preschool children with ASD in Hunan province of China were recruited to be categorized into 2 groups, i.e., monolingually exposed children with ASD (ME, N=14) and bilingually exposed children with ASD (BE, N=14), well-matched on age (AGE=44.21±5.90 vs. 46.64±8.04 months), autistic behavior scores (ABC=63.79±20.22 vs. 66.14±18.65), vocabulary production size (PCDI=297.50±311.95 vs. 260.29±217.89 words) and syntactic complexity scores (MLU=2.11±2.04 vs. 2.68±2.13), all ps>.05, Coh’s ds<0.35. Children listened to simple reversible SVO sentences paired with two visual scenes, only one of which matched the test stimuli, e.g., distinguishing between ‘the bird pushing the horse’ and ‘the horse pushing the bird’.

Results: A repeated measure analysis of variance (2 groups×5 verbs×2 trials) only yielded main effect of trial F(1, 10)=5.42, p=.04, η²=.35. One-tailed T-tests revealed that children with ASD in both groups exhibited their sensitivity to SVO word order, by looking longer at the matching screen during test trials than control trials (ME: Test: 52.98%±14.87% vs. Control: 45.83%±10.35%, t(13)=1.93, p=.04; BE: Test: 52.44%±9.84% vs. Control: 43.60%±11.86%, t(13)=2.54, p=.01). However, two groups differed in their optimal performance shown either in the 2^nd half (ME: 2^ndTest: 56.82%±16.85% vs. Control: 45.83%±10.35%, t(13)=2.66, p=.01) or 1^st half (BE: 1^stTest: 54.12%±10.13% vs. Control: 43.60±11.86%, t(13)=3.05, p<.01) of test trials. Time course confirmed that the ME group showed a preference of looking at the matching scene only after the object NP was presented (fig. 1), while the BE group could give an immediate response with the presence of the subject NP (fig. 2). Correlation data revealed that monolingual-exposed children with higher severity of autism were slower to locate matching screens (r=.55, p=.02). Bilingual-exposed children with larger vocabulary production size looked more to the match during the 2^nd half of test trials (r=.530, p=.03).

Conclusions: Therefore, grammatical strength in children with ASD won’t be challenged with bilingual exposure. Both monolingual- and bilingual-exposed children with ASD succeed in mapping the subject and object noun phrases in SVO sentences with the thematic agent and patient roles around 4 years old, corroborating “the universal constraint on the form-meaning mapping” (Franke et al., 2013). Moreover, the efficiency of online processing in children with ASD is affected by multiple factors including language environment, severity of autism and vocabulary size.