On the Edge

Conclusion

Despite the great complexity and abstractness of the data, there is rich evidence that long-distance relations have a successive-cyclic character. This often gives rise to what appear to be edge effects, and can be taken as a strong argument for the importance of a special area at the edge of each cycle. The phase-based model gives the underpinnings of an explanatory account for these phenomena, rooted in a conceptually appealing model of syntactic structure. What I have tried to do here is to explore the extension of the phase-based theory to the nominal domain, and to explore the interaction of the nominal structure with the clausal structure. The interactions mainly express themselves not in edge effects, but in the successive opacity of different nominal domains, blocking relations from the clausal heads into spelled out projection of the extended DP. I have suggested a few points of parametric variation, relying mainly on the presence or absence of attractors of different types in different heads.

Many thanks to David Adger, Gillian Ramchand, and especially Tarald Taraldsen for carefully reading and discussing earlier drafts. Thanks also to Noam Chomsky and an anonymous reviewer for comments on an earlier draft of what, after extensive revision, has become section 2. I also benefitted immensely from the presentations and discussion at the Peripheral Positions conference in York in 2000 and the Displacement workshop on the MS Polarlys in 2002. Thanks, finally, to the editors for their hard work and easy mercy. None of the persons mentioned or alluded to above can be implicated in any of the shortcomings of this chapter.