Bilingualism, Language Contact and Change: The Case of Bengali and English in India.

Abstract

This dissertation examines the nature of bidirectional influences in Bengali and English in the understudied bi/multilingual setting of West Bengal, India. Close contact between the two languages has resulted in extensive bilingualism in a section of the population. In order to get a better understanding of language change and development in this community, the following research questions are investigated: i) What types of contact effects are observable in Bengali today? ii) Which grammatical features in bilingual Bengali-English have changed under contact and which ones result from language-internal developments? iii) How can we ascertain that new properties found in bilingual Bengali-English are contact-induced? iv) What are the grammatical differences between monolingual and bilingual Bengali? These questions are addressed from both diachronic and synchronic standpoints. Two corpora of spoken monolingual and bilingual Bengali, collected and compiled through fieldwork in India, were compared with nineteenth-century Bengali plays for a set of linguistic features, following the methodology of Thomason (2001). Further, a corpus of English speech was collected from Bengali-English bilinguals and analyzed to evaluate what types of changes occurred in English. This investigation has also been complemented with a quantitative analysis of the distribution of linguistic features in the corpora, following a variationist methodology (Poplack & Levey 2010). The findings indicate that English influence has led to extensive code-switching, heavy lexical borrowing in Bengali, and morphosyntactic changes in the domain of bilingual Bengali-English complex verbs, formed with English nouns or verbs and the ‘do’ verb from Bengali. Change was also identified in Bengali equational sentences, where a new meaning and position of the hocche ‘be’ verb, which was unattested in the nineteenth-century Bengali plays, were argued to emerge from multiple causation. The English data reveals slight divergences in the use of articles and progressive forms, which have resulted partially from interference from Bengali, but are mostly conditioned by bilingual speakers’ different proficiency levels in English. These findings provide evidence for overall stability in the grammar of both languages (specifically for the subject pool that was tested in this dissertation), despite the presence of extensive bilingualism in the community.