Subject/non-subject movement asymmetries in Late Archaic Chinese

This paper accounts for the inability of a non-subject DP to move over the subject in Late Archaic Chinese (LAC) by proposing that DP movement could only be licensed by checking a case feature. Consequently, movement into the C/TP layer was generally only possible for the subject. An object could be topicalized, but only if it was base generated in the left periphery and resumed by a pronoun in argument position. Likewise, focused objects could undergo dislocation but they moved no further than the edge of vP. This paper further proposes that the restriction of DP landing sites to positions where they could check a case feature is in turn a consequence of how Labeling takes place. I propose that features like topic, focus, [Q], etc. require overt marking in order to participate in Labeling. In the absence of such marking, the only feature a bare DP can share for the purposes of Labeling is its case feature. LAC did allow non-subjects to move into the C/TP layer in order to form a relative clause, but such relative clauses were nominalized, and the subject valued genitive case with an external determiner. Because the subject was licensed external to the clause with genitive case, it did not need to check the case feature on C/T inside the relative clause, thereby leaving this feature available for the object.