A uniform syntax for phrasal movement: A case study of Dinka Bor
, Coppe van Urk 2015
This dissertation examines the question of why there should be different types of phrasal movement, with different syntactic and semantic properties. I develop the hypothesis that all instances of phrasal movement are the result of Agree and Merge (Chomsky 1995, 2001), and that the existence of different types of movement derives solely from variation in the properties of the feature involved in the Agree relation.
I first apply this view to the A/A¯ -distinction (Chapter 2). I argue that all of the differences between A- and A¯ -movement come from the features that drive them, and that the notion of distinct A/A¯ -positions should be eliminated. I then provide an argument for this approach based on the absence of a clear A/A¯ -distinction in the Nilotic language Dinka Bor (Chapter 3 & 4). I show that, in Dinka, long-distance movement must be accompanied by '-agreement. In accordance with the idea that it is the Agree relations involved that matter, the resulting movements combine properties of A- and A¯ -movement.
I also discuss the difference between intermediate and terminal movement steps of successivecyclic dependencies (Chapter 5). I offer several arguments from Dinka and other languages that intermediate movement, like all other movement, is always triggered by an Agree relation (Chomsky 1995; McCloskey 2002; Abels 2012b). Along the way, we will see evidence that clauses universally decompose into a CP and vP phase (Chomsky 1986 et seq.). Finally, Chapter 6 looks at patterns of pronoun copying in Dinka and other languages, and presents a novel argument for the Copy Theory of Movement (Chomsky 1995 et seq.).