In this paper, I critically review Takahashi’s (2014) and Sato’s (to appear b) analyses on the null argument in Mandarin Chinese. I point out some problems with these analyses, and then propose an alternative analysis according to which
null arguments are derived either by argument ellipsis (LF-copy) or by movement of null operators. I claim that the Extended Projection Principle of tense T plays a crucial role in explaining availability of strict and sloppy readings
in null argument positions in Mandarin Chinese.
page 25
