On Korean “ Case Stacking ” : The varied functions of the particles ka and lul ∗

Korean exhibits a phenomenon that has been dubbed case stacking, in which the “nominative” particle ka is optionally suffixed to the subject in addition to the dative particle eykey. Numerous authors have claimed that “stacked” ka represents the realization of structural nominative case assigned to subject position, independent of inherent dative case. Under this view, ka-stacking provides prima facie evidence that structural licensing (a.k.a. abstract Case) can have a direct reflection in morphological case marking. In contrast, I give numerous arguments that stacked ka is not a case morpheme at all, but a discourse particle that can mark focus, and therefore provides no evidence for this view of licensing. I show that stacked ka patterns closely with other uses of “case” particles that reflect not case but discourse status. In particular, accusative lul, like ka, can be stacked on other case particles to mark focus. I suggest how other noncase uses of these particles, including the so-called “ECM” construction in Korean, can be unified while maintaining the widespread crosslinguistic generalization that inherent case, as found on dative subjects, blocks the checking of nominative case (and phi-features) between a subject and an Infl head. Similar “schizophrenic” uses of case morphemes in unrelated languages are also discussed. In the appendix, I consider the possi-