No objections to Backward Control
This chapter addresses two main counterarguments put forth in Landau (2007) against the movement analysis of Control, and especially against the phenomenon of Backward Control (BC), as proposed for Tsez by Polinsky & Potsdam (2002). The first objection concerns the Case properties of the subject chain in Tsez BC. Landau argues that if this could bear two distinct cases, ergative and absolutive, one would expect the merging of a second DP in the matrix clause to be possible. The second objection is that BC is very rare. In Tsez only two verbs display BC. These are aspectuals (/<i>begin/, /continue/, /stop</i>/), which also have a standard raising analysis, thus casting doubts on the idea that the two BC constructions really involve Control rather than Raising. We show that, unlike the situation described in Tsez, Landau’s objections do not hold for Greek and Romanian. First, BC is not rare in these languages, as all obligatory control (OC) verbs exhibit BC. Second, data involving quirky subjects clearly suggest that Control chains are Multiply Case Marked Chains. Our results thus provide stronger empirical support for an approach to Control in terms of Movement, as defended in Hornstein (1999) and subsequent work. We then explore the differences between Tsez and Greek/Romanian and argue that Tsez is basically a FC language, which explains why OC is limited to aspectuals. Finally we show that the availability of BC is related to the extensive availability of agreementassociate relationships of the clitic doubling type in a language.