The features of movement in Romanian

Abstract

This dissertation investigates acceptable word order strategies employed in Romanian and the dynamics of movement. It focuses on the forces behind movement and in resulting structural effects, as well as the locus of movement, semantic restrictions, and pragmatic interpretations. Licensing of movement is discussed in terms of feature-checking mechanisms and properties of features. Specifically, it is argued that feature checking is exclus vely overt, but does not always entail movement. We assume two types of formal features which show symmetric behaviour irrespective of whether they are hosted by a lexical item or a functional head: (i) nonselectional features, which check in a less local relationship and do not trigger movement; (ii) electional features, which check in a strict locality relationship. The strict locality relationship involves a specifier-head configuration or head-adjunction, both of which always trigger movement. It is claimed that structural Case is checked in Merge positions in Romanian and proposed that Romanian has a V-type EPP which it checks by verb raising to the Inflectional domain. We show that Spec,IP is therefore not an EPP/Case related position in Romanian. As a result, we propose that the [+ wh] and the [+ focus] feature parasitically incorporate onto I, engendering a syncretic Inflection, with Spec,IP serving as a host to operators which undergo feature-driven movement into the left-periphery of the clause. The nature and composition of functional categories is also discussed and we show that Romanian has minimum structural proliferation. Finally, this dissertation discusses instances of overt movement that are not feature-driven. It is argued that Romanian allows for two types of scrambling: [nu]P-scrambling, which is shown to have A-movement properties and IP-scrambling, which is shown to have A-bar movement properties.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 163 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.