A Processing-Based Account for the Preferred Ordering of the Korean Classifier Structures

This study explores whether native Korean speakers have a preference between the two numeral classifier structures, N+NUM+CL and NUM +CLgen+N. According to Hawkins’s (2004) domain minimization account, the N+NUM+CL structure would be preferred over the NUM+CLgen+N structure because the former considerably has a bigger IC-to-word ratio than the latter. To test this prediction, we conducted an experiment in which native Korean-speaking adults completed two acceptability judgment tasks, one written (n = 67) and one spoken (n = 46). The results of the two acceptability judgment tasks indicate that native Korean speakers prefer N+NUM+CL over NUM+CLgen+N, compatible with the prediction of the domain minimization account.