UC Berkeley

"Little kids are like sponges," goes the saying. While young children are commonly believed to be naturally good language learners--and have often been used in second language research as a homogeneous comparison group for older learners--this dissertation paints a more complex picture. It examines the process of learning English as a second language for young children in their first year of preschool and takes a practice perspective (Bourdieu, 1977, 1992) to understand how language learning relates both to the social context and to the classroom identities that students developed in their classrooms.

This ethnography takes place in a mostly English-speaking, former Rust Belt city, where a new and growing population of resettled refugees has made many teachers into de facto ESL teachers. After briefly investigating this changing landscape, including the historical and economic factors that led to the present demographics, the study zeros in on one Head Start classroom that reflects these city-level changes. It follows parents, teachers, and the class of children--including 11 Nepali speakers and one Turkish speaker--across the school year.

Data were collected one full day per week for 27 weeks, through field notes, video recordings, and interviews. Data were coded thematically to understand parents' and teachers' histories and perspectives on language and learning; then, through a combination of coding and discourse analysis, I used the data to explore how four focal students were positioned as more or less competent and authoritative within the classroom. Finally, video data were analyzed to construct a linguistic corpus for each focal student in order to analyze their English language growth.

Findings show that students in the classroom who were seen as socially and academically competent were also seen as linguistically competent--by teachers, by peers, and even by me--regardless of their actual level of English. This was because what mattered in day-to-day interaction was what students could do with language. Competence meant the ability to assemble linguistic resources in real-time in order to accomplish social tasks, to be listened to, and to be taken seriously. Within the classroom, language was understood and assessed as a social practice. When at the end of the year, I compiled corpora of student talk and analyzed them for growth in vocabulary and syntax, a different picture emerged: The most socially successful student had learned much less language than her teachers (or I) had thought, and one of the most peripheral students had in fact learned much more.

By accounting for language both as social practice and as a system of vocabulary/syntax, this study shows that the two versions of success do not always align and it raises questions about what should count as successful language learning in schools, as well as what kinds of interactions support this learning. One implication of this work is that teachers must understand that how one defines language changes how one measures success. Cultivating teachers' ability to see language more than one way may provide a much-needed balance to the tendency of schools to measure success as the individual accumulation of words and structures. Additionally, since the same social conditions that made some students successful in language-as-social-practice were those that kept them from having the comparable success in acquiring language-as-a system, it is equally important to help teachers to understand how different conditions can lead to different kinds of growth.