This study investigates the encoding of the count/mass distinction in Korean, a generalized classifier language (GCL), through an ERP study that examines neural responses to numeral-classifier and plural marking constructions. Previous research has debated whether classifier languages such as Korean inherently encode the count/mass distinction in their grammar or if such distinctions are represented solely through semantic and pragmatic interpretations. Linguistic theories have proposed varied accounts, suggesting that classifiers might explicitly categorize nouns as count or mass at a syntactic level, whereas plural markers, like Korean -tul, may reflect more variable morphosyntactic behaviors.
We conducted two ERP experiments with native Korean speakers who performed grammaticality judgments on sentences presented via rapid serial visual presentation. Experiment 1 investigated numeral-classifier-noun constructions, manipulating classifiers to be either compatible or incompatible with the mass/count status of the noun. Results showed significant P600 effects when classifiers mismatched the noun type, indicating that classifiers in Korean indeed encode count/mass distinctions syntactically. Experiment 2 examined plural marking constructions involving -tul with either count or mass nouns. In contrast to Experiment 1, Experiment 2 did not yield significant ERP effects, suggesting that the plural marker -tul does not exhibit a strong morphosyntactic constraint based on the count/mass status of the nouns. These findings support a view of morphosyntactic optionality in Korean plural marking, indicating that -tul is only partially grammaticalized and highlighting the complexity and variability inherent in Korean morphosyntax.