The Acquisition of Flexible Word Order and Case-Markings in Korean Children with Autism Spectrum Disorder

Background: Researchers generally agree that the language impairments of children with ASD include pragmatic deficits (Naigles & Chin, 2015); however, less consensus exists concerning whether they also have morphosyntactic deficits. Children’s productions have revealed omitted morphemes and sparse use of questions (Eigsti, Bennetto, & Dadlani, 2007; Roberts, Rice, & Tager-Flusberg, 2004); however, their comprehension data generally indicate intact, if somewhat delayed, grammatical understanding. However, most of the previous studies were conducted with English-speaking children with ASD, and morphosyntactic features existing in other languages such as flexible word order or case-markings have been barely studied. In Korean, case markers provide the cues to the relations between agents and patient and multiple word orders are permissible (Kim, 1997). Learning how to use both word order and case markings to interpret syntactic relations may be challenging not only for TD Korean learners, but also for Korean children with ASD.

Objectives: We used Intermodal Preferential Looking (IPL: Naigles & Tovar, 2012) to investigate Korean children’s comprehension of SOV and OSV word order presented with nominative and accusative case-markers; we compare TD preschoolers with those with ASD.

Methods: Monolingual Korean-speaking TD children (n=16, MA=51.19months, Leiter-R Brief IQ=45.75, expressive language AE of 54~59months) and children with ASD (n=13, MA=70.47months, Leiter-R Brief IQ=48.00, expressive language AE of 54~66months) participated. Children viewed IPL videos, whose test sentences followed SOV ((a) A-Nom B-Acc Verb) or OSV ((b) B-Acc A-Nom Verb) order. The videos showed familiar actions (e.g. pushing) with agent A-patient B on one side paired with agent B-patient A on the other side; baseline trials had no directing audio. Children’s eye movements were coded and two dependent variables were calculated: ‘Percent looking to match’ and ‘Percent looking to the 1^st NP as agent’ during the test vs. baseline trials. Children who understand the audios will look longer to the matching scene during both SOV and OSV test trials; however, children who simply treat the 1^st NP as the agent regardless of case marker will look longer to the nonmatch during OSV trials(Naigles, Kelty, Jaffery, & Fein, 2011).

Results: The TD children looked longer to the match during the test trials compared to baseline trials for SOV order (t(15)=2.02, p=.03), but not OSV order (t(15)=-.41, p=.34). With the percent looking to 1^st NP as agent measure, though, the SOV/OSV comparison yielded a significant Frame effect (F(1, 15) = 8.78, p = .01). In contrast, children with ASD looked significantly away from the matching video during SOV trials compared to the baseline trial (t(12)= -2.27, p=.02). For OSV sentences, no significant effect was found (t(12)= .08, p=.47).

Conclusions: In sum, TD children demonstrated successful comprehension of SOV orders; moreover, they seem to pay attention to case-markers because they differentiated the OSV trials from the SOV trials. In contrast, language-matched children with ASD did not appear to make use of the case markers at all, and seemed not to understand either the canonical SOV nor OSV word order. This study suggests that Korean morphosyntax places unique challenges for children with ASD.