This paper investigates how the language comprehension system tackles syntactically/semantically different types of grammatical constructions, with their differences due to some abstract properties. To do so, in this paper we concentrated on Case alternation in the three different types of single/double nominative construction (SNC/DNC) in Korean using the ERP experimental paradigm. We found a significant P600 in the family relation-type DNC, a marginal P600 in the DAT-type DNC, and a significant left anterior negativity (LAN) plus a significant P600 in the property-type DNC. Since the P600 in our experiment was detected at posterior regions, we take this component to indicate that the three types of DNC are all recognized as syntactically marginal or degraded, though they vary in degradedness. We also take the LAN component in the Property type to signal the property-denoting nominal’s licensing relation with the property-‘possessing’ one. Thus, this study points toward the interconnection of the theory on different types of grammatical constructions with their aspects during sentence processing.
1. Introduction
2. Setting the stage
3. Experiment
4. Discussion
5. Conclusion