|Title:||Construction-based Approaches to Flexible Word Order||Add Dissertation|
|Author:||Andrew Wetta||Update Dissertation|
|Email:||click here to access email|
|Institution:||University at Buffalo, Department of Linguistics|
|Abstract:||This dissertation investigates surface-oriented, construction-based approaches to flexible and verb second (V2) word order. Generally, syntactic structure is analyzed as binary branching structures, which often form complex embedded phrases. Such analyses are problematic for flexible word order which allows many alternate yet equally valid linearizations of elements. Positional constraints like V2, where the finite verb is required to appear in the second position, contribute further difficulties. Often extraction processes and complex mechanisms are required to account for these properties. This dissertation argues against this type of underlying structure which cannot be empirically detected. Instead, linguistic evidence suggests that syntactic structure is flatter than usually assumed. Phenomena that were previously attributed to the form of underlying structure are readily seen in surface structure or better described by non-syntactic processes. The word order properties of six V2 languages including Breton, German, Ingush, Kashmiri, Karitiâna, and Yiddish are explored, and it is shown that simple linearization constructions appropriately license flexible word order across languages in a surface-oriented manner, and that extraction is not the appropriate way to derive V2 order.
Furthermore, having argued against the use of unobservable underlying structures and mechanisms, this dissertation provides an extensive analysis of German clause structure with flat constructions. It is proposed that a clause with flexible word order and other positional constraints is licensed by the combination of two general classes of constructions: those which license the linear placement of elements and those which license argument saturation. These constructions capture observable patterns at the clause level and underspecify the positions of flexible elements. This flatter approach to flexible word order and V2 is not only more consistent with observable data but also accounts for problematic phenomena in other approaches.