The Wh-Questions Comprehension in Korean Children with Autism Spectrum Disorders

Background: Producing wh-questions is frequently problematic for English-learning children with ASD¹; comprehension emerges earlier but is nonetheless delayed relative to overall language level². Researchers disagree on whether these difficulties are attributable to pragmatic vs. grammatical sources. Pragmatically, the desire to communicate with others is impaired in autism; therefore, children may not know when and how to ask questions appropriately³. Grammatically, English wh-questions deviate from the standard SVO word order (‘What did John eat?’ is OSV); therefore, it is possible that children with ASD who have acquired declarative SVO sentences have difficulties learning another order for questions. Korean provides a test of the grammatical perspective, because in Korean wh-questions the canonical word order (SOV) does not change⁴(‘John-NOM what-ACC eat-did-CM?’)

Objectives: The current study tests whether Korean-learning children with ASD have advantages in wh-question acquisition. If grammatical complexity contributes to English-learning children with ASD’s delay in wh-question acquisition, Korean-learning children with ASD might demonstrate earlier comprehension. In contrast, if the pragmatics of questions is the key source of wh-question difficulties in ASD, children learning Korean should show a similar delay to children learning English. 

Methods:  Typically developing(TD) Korean 2-year-olds (n=15, MA=28.35 months), and 4-year-olds (n=15, MA=51.68 months) participated. Korean children diagnosed with ASD and matched with the TD 2- and 4-year-olds on raw NVIQ scores (Leiter-3) were also recruited: 11 children with ASD (M_age=56.72 months) matched the TD 2-year-olds (TD_Mean=26.73; ASD-low_Mean=23.92) and 9 children with ASD (M_age=70.47 months) matched TD 4-year-olds (TD_Mean=45.88, ASD-high_Mean=45.85). Language levels were also comparable by group: (TD 2-year-olds M=267 words, ASD-low M=211.50 words; TD 4-year-olds M_{expressivelanguageage}=54.9 months, ASD-high M_{expressivelanguageage}=54.9 months). Children viewed a wh-questions video^{2, 6} in which pairs of familiar items (apple, flower) appeared side-by-side in the baseline trial. Then the items appeared in transitive events (apple hitting flower); finally, they were again shown side-by-side, paired with object-what-questions, subject what-questions, or where questions: (1) (Sa-gwa-ga-Mu-eos-eul-Chyeo-seo “What did the apple hit?”) (2) (Mueosi-Kko-cheul-Chyeo-seo “What hit the flower?”) (3) (Eo-di-e-Kko-chi-I-seo “Where is the apple/flower?”). Children’s eyegaze was coded offline. Percent looking to the named item (‘apple’ in (1)), were analyzed; children who understand the audio were expected to look longer to the correct items (compared to the named items). 

Results: TD 4-year-olds demonstrated significant wh-question comprehension for both subject and object what-questions (ts>4.43, ps<.001); these children correctly looked at the named item less during the ‘what’ than ‘where’ trials. TD 2-year-olds only demonstrated this comprehension for the subject what-questions (t=3.78, p<.001). The ASD-high group demonstrated significant wh-question comprehension for the object what-questions (t= 2.43, p=0.02), but not for subject what-questions (t= 1.08, p>.05), while the ASD-low group did not demonstrate significant comprehension for either question type. 

Conclusions: TD Korean 4-year-olds showed robust comprehension of wh-questions; however, the other groups all demonstrated some difficulties, with TD 2-year-olds succeeding only on subject wh-questions, ASD-high children succeeding only with object-wh-questions, and ASD-low children not succeeding on either. Therefore, the grammatically simpler form of Korean wh-questions do not appear to facilitate earlier acquisition of these forms, leaving open a role for pragmatics.