Abstract

To address the universality of the sentence processing strategies across different languages, various experimental results of Korean sentences were reviewed and examined. The predictions from such parsing strategies as the minimal attachment and the minimal chain were compatible with the word-by-word reading times of Korean ambiguous sentences such as embedded structures and dative noun phrases. The late closure strategy itself might not be appropriate for Korean, however, we argued that more inclusive parsing principle like locality could handle the ambiguity of Korean relative clause attachment. We confirmed that Korean parsing operations were performed incrementally, and argument information of verbs might be used in making expectation for a forthcoming structure. We showed that reparsing processes of Korean seem more compatible with diagnostic model, Suggestions for further studies were made.