Left periphery and Chinese yes-no questions

Assuming the cartographic approach (Rizzi 1997, 2004; Cinque 1999), this study attempts to map four Chinese yes-no question markers onto the left periphery, including ke/shifou/A-not-A in Mandarin Chinese and kam in Taiwan Southern Min. It is found that while these four markers all represent yes-no questions, they are not exactly alike on syntactic grounds. The syntactic behavior of ke is particularly different in that it is not able to lead an embedded null-subject question whereas kam/shifou/A-not-A are able to. Also, ke in its own right cannot license the focus interpretation of its following NP or clause whereas kam/shifou/A-not-A (MEpi-not-MEpi) can. Given these differences, this study proposes a topography and attributes them to different projections of the question markers in the fine structure of split CPs, where ke is merged in Fin0 and raised to Int0 in LF, while kam/shifou/A-not-A(MEpi-not-MEpi) are merged in SpecFocP and raised to SpecIntP in LF. VP-not-VP forms, however, are not base-generated in CP, but within TP/IP. They are subsequently forced to raise to SpecIntP in LF to check the feature [+Q].