An Agree-based Analysis of Extraction out of Nominals in Korean

This paper reports a novel observation on extraction out of nominals in Korean and offers an analysis of the correlation between NP-movement and the appearance of overt classifier morphology. To be specific, for extraction out of floating numeral-classifier constructions in Korean an Agree relation must be established between the NP and the overtly realized classifier head. If no overt classifier morphology is realized, stranding just numerals, extraction out of the nominal domain is prohibited. I argue that the observation can be explained straightforwardly under Reeve’s (2019) recent proposal that links the availability of movement out of nominals to the overt realization of agreement morphology on the head noun.