Korean First Phase Syntax as Non-Voice-bundling

This paper investigates the question of whether Korean possesses two distinct syntactic heads for v and Voice (i.e., non-Voice-bundling) or they are bundled into a single syntactic head (i.e., Voice-bundling). The diagnostic provided by Pylkkänen (2002; 2008) favors the former, while that offered in Tubino Blanco (2011) and Harley (2013) suggests the latter. In this study, I argue that in Korean the verbalizing layer (i.e., vP) is separated from the external-argument-introducing layer (i.e., VoiceP), based on the ambiguous syntactic behaviors exhibited by the HI construction as an inchoative and as a passive. That the derivation of the verb structure involving the HI suffix can either represent a structure at the vP level or project an additional VoiceP motivates a classification of Korean as a non-Voice-bundling language. The inchoative/passive ambiguity of the HI construction is ascribed to the interplay of the teleological capabilities of the participating arguments. The predictability of the previous diagnostics for non-Voice-budling is evaluated and their relationship with the newly proposed diagnostics is considered.