Adjectives that Aren't: An ERP-Theoretical Analysis of Adjectives in Spanish.

摘要

This thesis investigates the syntactic status of adjectives in Spanish through a crossdisciplinary perspective, incorporating methodologies from both theoretical linguistics and neurolinguistics, specifically, event-related potentials (ERPs). It presents conflicting theories about the syntax of adjectives and explores the ways that the processing data gleaned from ERP research can help resolve the conflicting approaches.;Theoretical research has shown that although two surface positions exist for adjectives in Spanish (prenominal and postnominal), the underlying structure for these positions is more complex than the simple merging of an adjective before or after the noun. However, the exact details of this underlying structure are a source of discord among researchers. Whereas there is basic agreement that there are at least two base generated positions for adjectives (e.g., Alexiadou, 2001; Bernstein, 1992, 1993; Cinque, 2010; Taboada, 2010), the location of this base generation within the syntactic structure is a point of conflict. Some authors claim that adjectives participate in specifier merge to at least two different functional projections dominating NP (e.g., Cinque, 2010; Taboada, 2010), whereas others claim that certain adjectives have head status, selecting for NP and projecting their own features (e.g., Bartlett & Gonzalez-Vilbazo, 2013; Bernstein, 1993). Still other authors propose a third position exists for focalized adjectives (e.g., Demonte, 2005; Taboada, 2010), which raise from their base-generated position. Further, the word class status of certain adjectives receives varying analyses, with authors claiming that some (Bartlett & Gonzalez- Vilbazo, 2013; Bernstein, 1993) or all (Cinque, 2010) prenominal adjectives are functional elements. Thus, a consensus analysis of adjectives does not exist within theoretical linguistics literature.;In this dissertation, I approached these theoretical disagreements from a processing standpoint, analyzing the theoretical adjective structures through ERP data. Specifically, I examined the adjectives' (a) potential movement, (b) word class status, and (c) morphosyntactic agreement properties by comparing the electroencephalogram (EEG) of native Mexican speakers of Spanish as they read determiner-adjective-noun or determinernoun- adjective sequences in Spanish. Results indicated that Spanish adjectives were not all processed in the same way. Prenominal adjectives showed evidence of processing distinctions in the correctly agreeing form, with the set of adjectives proposed to be moved elements displaying a sustained LAN compared to a base-generated counterpart. This was taken as evidence to support Demonte's (2005) and Taboada's (2010) that this set of adjectives is, in fact, moved from a base-generated position. The investigation into the adjectives' word class status was inconclusive because a general lexical-functional processing distinction could not be established for the participants. Finally, agreement processing revealed a basic split between prenominal and postnominal adjectives, with prenominal gender agreement violations eliciting a P600 and postnominal ones eliciting a LAN/P600, in addition to an N400-type component for the set of adjectives with multiple semantic meanings. The agreement results themselves did not serve to clarify any particular aspect of the controversial syntactic structure, but they may have implications for the general theoretical concepts of agreement and concord.;Overall, the combination of theoretical and neurolinguistic methodologies helped to clarify some the controversies surrounding the syntactic structure of adjectives, although questions still remain. The present study contributes to an understanding of brain and language by employing a dual-field approach to unresolved linguistic issues, thus elucidating details about competence and processing that increase our knowledge of the structure and mechanisms of human language grammar.