Some Remarks on the Licensing Condition on the Null Argument in Korean (and Japanese)

This squib examines Oku's (1998) and Takahashi's (2011, in press) ellipsis analysis of the null argument in Japanese and Korean, and the accompanying licensing condition on it. Oku proposed that the elliptical null argument in these languages is literally unrealized in overt syntax, but is reconstructed in covert syntax to meet the 'weak' selectional feature of a verb. Takahashi also proposed to attribute the delaying of the structure building for the null argument to the absence of agreement/phi-features. Despite their brilliant endeavors to account for the parametric differences in the availability of the null argument, they are shown to confront some non-trivial problems. We explore an alternative analysis of the null argument, proposing that it is not a DP-substitute but an NP-substitute. Based on this proposal, we provide an account not only of the reason for the phonologically null form of argument and the lack of verbal agreement in Korean and Japanese, but also of the parametric variation in the availability of the null argument.