[Author Login]
Title:Phrase Structure Directionality: Having a few Choices
Authors:Alex Zepter
Comment:Rutgers University dissertation, 2003. In order to properly read the file (including all symbols), one needs to have WordPerfect software installed on his/her computer.
Abstract:The thesis claims that the linear organization of specifier, head and complement in a phrase and throughout a syntactic tree is determined by a conflict between general violable constraints on X-bar-structure. The adopted framework is Optimality Theory (cf. Prince and Smolensky 1993; in syntax, cf. Grimshaw 1997). The proposed constraint system explains why phrase structure directionality is mostly uniform and why only some non-uniform cases exist, while other logically possible kinds of mixed directionality are unattested.

Central to the dissertation is the idea that head-initial oriented languages have a greater structural conflict to resolve inside their lexical projections than head-final languages: The combination of a general preference for [head - complement]-order and for a left-peripheral specifier bars the lexical head from surfacing at an edge of the phrase. The combination of a general preference for [complement - head]-order and for a left-peripheral specifier still allows alignment at one edge. This greater conflict can be resolved in different ways, which leads to slightly more variation among head-initial oriented languages: Not only do we find uniform SVO-languages, but we also find VOS-languages, VSO-languages and head-initial oriented languages with a head-final verb phrase. On the primary examples of the Mayan VOS-language Tzotzil, the 'strict' VSO-language YosondĂșa Mixtec, and the 'mixed-headed' languages German and Persian, I show that mixed directionality is not arbitrary in its ways of deviating from uniformity. The proposed system derives various implicational universals capturing the persistently systematic nature of phrase structure directionality. The predictions made about the verbal domain have systematic correlations in the domain of all other categories.
Article:Part 1
Part 2
Part 3
Part 4