Dependencies across Phases: From sequence of tense to restrictions on movement

Abstract

This dissertation deals with two differences between the grammars of Russian, on the one hand, and English or Dutch, on the other hand. The first lies in the sphere of Sequence of Tense (SOT). English and Dutch are "SOT languages": in these languages past tense in complement clauses embedded under matrix past is interpreted as denoting simultaneity with the matrix event(uality). In contrast, Russian is a "non-SOT language", which uses present under past for a simultaneous interpretation. The second difference concerns locality of movement: in Russian, long-distance movement out of indicative complement clauses is not allowed, in contrast to languages like English where it is grammatical.

The dissertation demonstrates that these facts stem from one difference in the structure of the two types of languages. This difference is defined in terms of Phase Theory. In English, C (complementizer) is a phase head, while in Russian, the phase head is T (tense) and not C. One consequence of this is that in English, C and T share a tense feature. The temporal feature on C participates in an Agree chain between the matrix and embedded T heads which enables a simultaneous reading of past in SOT languages. In Russian C does not carry an instance of the tense feature, consequently such a chain cannot be formed. In addition, the Phase structure of Russian blocks successive cyclic movement.

The dissertation is of relevance to researchers working on syntax, in particular on grammatical tense, movement, or the grammar of Slavic languages.