Bound Noun Pa in Korean

This paper addresses the syntatic aspect of a bound noun pa in Korean. Woo (2006) extensively discusses the bound noun pa in terms of syntax, semantics, and grammaticalization. Woo (2006) argues that there are two different syntactic structures of pa: a relative-clause structure with an ellipsis operation and a noun-complement-clause structure. However, the two structures appear to have empirical problems. Unlike normal relative clauses, pa cannot be separated into two different clauses, and the elided parts cannot have full pronunciation. The problem of the noun complement clause is that pa cannot be modified by adjectives. Given that pa can be replaced by a bound noun kes in general (Woo 2006), we may assume Kim’s (2016) N-to-n movement of kes. However, we find a difference between pa and kes in terms of an adjectival modification pattern. In this paper, I argue that pa directly merges in small n, and it takes an NP complement which is headed by a null noun. This analysis has more explanatory power than Woo (2006) in that it accounts for the related phenomena seamlessly, and its grammaticalized nature is directly reflected in the structure.