A Raising Analysis of the Relative Head in Korean

In this paper I argue for the head raising hypothesis of relative clause in Korean. I first claim that there are relative operators in Korean, which are relative temporal affixes. Then, I argue against the approaches of non-movement and OP-movement to Korean relative clauses. Next, I present idiom chunks, reconstruction effects, lower reading, and internally-headed relative clauses as the supporting evidence for the raising analysis. Finally I investigate how the relative head is raised in Korean. While doing so, I suggest 1) that the complex NP containing a relative clause in Korean is DP, not NP, 2) that Korean relative clause is IP/TP, not CP, and 3) that the complement of Det in Korean is the relative clause, not the relative head as in English. I propose that the relativized NP which is [+rel] moves to SpecTP to satisfy Spec-head agreement with the relative operator T, which is [+rel]. I claim that this movement is A-movement, the adjoined position of SpecTP being an A-position. I present the unavailability of lower reading in Korean as a consequence of this claim. I suggest that after this movement the relative clause TP moves to SpecDP, leaving the relativized NP behind to be a relative head.