Remarks on the Restrictions on the Adnominal Right Dislocation

Abstract

This paper discusses certain restrictions on the adnominal Right Dislocation (RD) in Korean. I suggest that the adnominal appendix tends to be associated with an overt host NP in the host clause, and that the association process is a processing-related one in nature, operating on the overt, surface form. The association process is further restricted by some factors such as closeness. In certain contexts, the workings of the factors give rise to the embedding effect, and when the factors become ineffective, the embedding effect is obviated/weakened. I also suggest that the association process is compatible with the ellipsis analysis since it applies to the surface form that results from the application of ellipsis. Due to the general processing mechanism, the overt host NP in the host clause is chosen over the elided/null NP in the appendix clause to lessen the processing load. I demonstrate that the ellipsis effect is observable in certain contexts