The primary purpose of this paper is to show the theoretical and empirical problems of Case theory suggested in GB and the Minimalist Program. The secondary purpose of the paper is to criticize the notion of Case feature, arguing that the feature is not necessary for computation but for interpretation at the interfaces. From a pure ontological view, the notion of Case feature is not necessary for computation. The morphological Case manifests the relationship between nominals and functional heads.

Korean ECM constructions seem to give rise to many interesting problems with the minimalist program. Of these problems, we are mainly concerned with the following: (A) How is the ECMed NP/DP assigned accusative Case despite the fact that the clause containing it is supposed to be finite? (B) How can the ECMed NP/DP be affected by the accusative Case assigner though finite clauses are phases? Accessing answers to these problems, we will find out that some unknown properties of the embedded predicates in Korean ECM constructions are involved in the transitivity restriction, and that there is a derivational mechanism for determining whether given movement is A or A`.