Raising of major arguments in Korean and Japanese

The question of whether languages like Korean and Japanese possess genuine instances of Subject-to-Object Raising (SOR) has been a matter of debate since Kuno (1976), as a number of the properties of the putative SOR construction in the languages differ from those found in languages like English, while others are shared between the languages. I argue in this paper that the paradoxical properties begin to fall into place once we posit that what undergoes movement in SOR in these languages is not the embedded subject, but the embedded Major Subject. The Major Subject is the initial Nom-marked DP in a Multiple Nominative Construction. It is shown that if we posit that the Major Subject raises in SOR, the unexpected properties of SOR can be accounted for. Under this analysis, SOR in Korean and Japanese conform to known constraints on A-movement taking place from the highest A-specifier (Major Subject) position of the embedded clause. It is the coindexation of the Major Subject with the predicate-internal position that gives rise to the illusion of non-locality. I then compare the analysis with an alternative base-generation analysis. While the two are roughly equal in terms of coverage, only the Major Subject raising analysis is able to account for properties of the raised nominal that could only have been determined in the embedded clause.