Abstract

All theoretical approaches to syntax recognize a distinction between A- and A’ – dependencies. In this talk, I develop the idea that the two types of movement are distinguished solely by the features that drive them, and not by the position they target (e.g. Chomsky 1981, 1995; Mahajan 1990; Déprez 1990; Miyagawa 2010). This proposal eliminates the notion of distinct A- and A’-positions and requires only that all movement is established by Agree and Merge. In addition, I show how such an approach can account for the properties associated with the A/A’-distinction, including differences in locality, pied-piping, reconstruction for Principle C, Weak Crossover, anaphor binding, and the licensing of parasitic gaps.

Finally, I investigate a prediction of this featural view of the A/A’-distinction. In particular,if we can find movement types that are driven both by features usually involved in A-movement and features usually involved in A’-movement (for example, Wh and φ), the resulting movements should be associated both with the properties of A-movement and those of A’-movement. I will show that is systematically true in the Nilotic language Dinka Bor (South Sudan), so that all phrasal movement displays a mix of A/A’-properties: movement can be long-distance, but, at the same time, is always visible for case assignment and binding.