First language attrition in late bilingualism: lexical, syntactic and prosodic changes in English-Italian bilinguals

Abstract

Among the most notable findings in recent bilingualism research is that the two languages are constantly active, thus giving rise to complex interactions, in the bilingual mind. One of the natural consequences of these interactions is the non-pathological attrition of the firstly acquired language, which is here understood as the manifestation of selective and reversible changes in the way the first language is processed and produced by speakers of a second language. In the last forty years, these changes have been investigated mainly in long-term residents immersed in a second language environment (usually anglophone countries) rather narrowly, by focusing only on a single domain in the first language. What happens to other types of late (i.e., post-puberty) bilingual speakers in different language domains at the same time, alongside what happens to their second language, is mostly underexamined.

Hence, this thesis aims to fill the aforementioned research gaps by investigating the attrition of a firstly acquired language (English) in three different language domains (the lexicon, syntactic interfaces and prosody) of two types of late bilingual speakers (British classroom-based learners of Italian at UK universities as well as British long-term residents in Italy), by also considering what happens in their second language (Italian) within the same domains. The results of this research show, firstly, that similar types of lexical retrieval interference are experienced by both classroom learners and long-term residents, with only a slightly higher degree of interference being experienced by long-term residents. Secondly, syntactic interface structures are selectively affected by first language attrition only for long-term residents, even though both bilingual groups resolve syntactic ambiguities in the second language similar to a control group of first language speakers of Italian (despite some processing delays). Thirdly, while prosodic changes in the first language are not reported for either bilingual group – at least in initial confirmatory analyses – both groups do seem to have acquired the prosodic patterns of their second language, with long-term residents being virtually indistinguishable from Italian controls.

Taken together, the findings of this thesis lend support to the idea that non-pathological attrition of a first language is a natural by-product of bilingualism, as it is found to affect the processing and production of different first language domains for both late bilingual groups examined. Crucially, these research findings reveal that the extent to which attrition is experienced varies according to the language domain and the specific structure under investigation. Interestingly, an inverse relationship between the second language and the first language is found for some bilingual speakers in certain areas, but not conclusively in others (e.g., for long-term residents in some aspects of the lexicon and syntax, but whether this is also the case for their prosody is not yet clear). This work thus points to the need for further research into first language attrition and second language acquisition, to fully understand the complex relationship between the two languages in the bilingual mind.