In this paper, based on the sematic and syntactic parallelism between Korean and Spanish we propose that Korean also has a mechanism to distinguish the subjunctive mood from the indicative one: '-ki' is a subjuctive complementizer, while '-ko' is an i...
In this paper, based on the sematic and syntactic parallelism between Korean and Spanish we propose that Korean also has a mechanism to distinguish the subjunctive mood from the indicative one: '-ki' is a subjuctive complementizer, while '-ko' is an indicative one. Korean and Spanish only differ in one aspect: in the former the morphological distinction between the two moods appears in Comp, while in the latter in Infl.
Given this proposal, we try to account for the contrast between Korean and Spanish with respect to 'the disjoint reference effect': the Tense of the embedded subjuctive clauses in Spanish depends on that of the matrix clauses, while in Korean it is independent or referential, and so the binding domain of the embedded pronominal subject does not extend to the higher clauses. The absence of the disjoint reference effect in Korean is due to this.
Finally, we show that the contrast between the subjunctive clauses and the indicative ones with respect to NPI licensing can be captured by proposing that in the latter the maximal projection of a functional category, CP₂(or Proposional Phase), which intervenes between CP and IP, renders the NPI movement illicit.