AbstractsLanguage, Literature & Linguistics

In this thesis, I study use of grammatical gender agreement, a problematic feature of both L2 and heritage language studies. Here I work with a group of Mexican bilingual children in order to assess their performance with this linguistic object at a local school. Though grammatical gender agreement has been previously studied in bilingual children, this project affords a new view with respect to the following contributions. First, I aim to investigate a diverse group of heritage language users who participate in multiple domains, some accessing Spanish in their social network and school classrooms, while others fail to access input in either class or home network. Secondly, here I address the topic of grammatical gender from a multiple methods approach, using a concord task in Chapter four; a story retelling task commonly used in the acquisition literature in Chapter five; sociolinguistic interviews, a modern day Spanish version of Labov’s sociolinguistic projects with African American Youth (Chapter Six); and finally incorporate a qualitative ethnographic focus to studying local language use (Chapter seven). Throughout this multi-methods project, one surprising result was that children living in this ethnic enclave area were found to be highly proficient in their use of gender agreement in Spanish, with all groups averaging from 98% to 100% during the sociolinguistic interview. Children also performed well in the story retelling segment, though by comparison the sociolinguistic interview relayed even more proficient data. This project underscores the role of input in language acquisition. Some surprising results including finding the overwhelming mastery of gender concord in this group of bilinguals, and the high levels of performance on a sociolinguistic interview compared to an acquisition task. These differences prompt the reader to consider the role of methodology in heritage language studies. Between group differences suggest assets of an additive bilingual program, such as dual language education. During the project, I focused on gender agreement, however, as I developed my data set I discovered that gender was only one possible linguistic object to consider, and data collection with bilingual children provides a wealth of information about their i-knowledge of language structure. Advisors/Committee Members: Cameron, Richard (advisor).