Clitics at the edge: Clitic placement in Western Iberian Romance languages.

摘要

This dissertation focuses on the distribution of pre- and postverbal clitic alternations in both matrix and finite embedded environments, in three Western Iberian Romance languages: Galician, European Portuguese, and Asturian. The analysis of these clitic alternations in Romance has a long tradition within the generative enterprise, and different analyses have capitalized on different triggers to account for those alternations.;In this study, I show that the inclusion of Asturian raises interesting issues for analyses dealing with clitic placement alternations. In short, while pre- and postverbal clitic patterns in all Western Iberian Romance languages are subject to the same conditions in the matrix environment, crosslinguistic differences arise the moment one turns to the finite embedded one: In Asturian, unlike in Galician and European Portuguese, postverbal clitics arise obligatorily after a Topic in finite embedded contexts as that in (1).;(1) Repitote [que yo dexelo aquel diecisiete de mayu] Ast repeat1SG-youCL that I left1SG-IND -itCL that seventeenth of May "I repeat to you that I left it that May seventeenth" [de Pablo, Memoria];Furthermore, speakers of a variety of Asturian which I refer to as Conservative Asturian (CAst) report data as that in (2), where both a post- and a preverbal clitic can be found. Interestingly, each pattern correlates with a different interpretation.;(2) a. Digo [qu'ayudame] CAst say1SG that-help 3SG-IND-meCL b. Digo [que me ayuda] say1SG that meCL help3SG-IND "I say that s/he helps me out " [From Viejo (2008)];In this dissertation, I argue that pre- and postverbal clitic alternations in Western Iberian Romance languages may be captured as follows: in Western Iberian, Finiteness° (cf. Rizzi (1997)) is a phase-head (cf. Chomsky (2008)) which (i) is responsible for the different clitic patterns, and (ii) is the locus of crosslinguistic variation in the finite embedded context. Under this analysis, I show that both similarities in clitic alternations in the matrix context and the noted variation in the finite embedded one in this group of languages can be easily captured. Moreover, I claim that this analysis can naturally explain the interpretation differences I observe between a postverbal and preverbal pattern in (2a) and (2b).