Bilingualism as a first language: language dominance and crosslinguistic influence

Abstract

Even though research on bilingual first language acquisition (2L1) could be conceptualized as monolingual acquisition (L1) of two individual languages, the fact that in 2L1 acquisition there is exposure to input from two languages has consequences in terms of how the two language systems interact in the mind of the bilingual. This century has seen two important developments in this respect. First, a consensus seems to have been reached on the idea that the two systems are differentiated from the early stages (e.g. Genesee, 1989; De Houwer, 1990; Genesee, Nicoladis & Paradis, 1995; Köppe & Meisel, 1995; Genesee, 2003). The second development is related to how the 2L1 language faculty compares to the L1 language faculty and the consideration that the grammatical processes and operations in both bilingual and monolingual speech must be accounted for in the same terms (MacSwan, 2000; Liceras, Spradlin & Fernández Fuertes, 2005; Liceras et al., 2008, among others). However, while it is unquestionable that L1 and 2L1 acquisition share similar mechanisms and processes, there are core issues such as language dominance, crosslinguistic influence and code-mixing that are specific to simultaneous bilingual acquisition. In this chapter, we address these three language contact phenomena by analyzing spontaneous and experimental data from the simultaneous bilingual acquisition of English and Spanish by two identical twins in Spain (FerFuLice corpus in CHILDES) as it compares to data from other 2L1 and L2 children and adults. We conceptualize language dominance in terms of the computational value of grammatical features in a given language. And so, the dominant language is the one that provides the functional category whenever that category is highly grammaticized. Crosslinguistic influence between the two languages of a bilingual is analyzed in the case of sentential subjects and copula predicates and we propose that the occurrence as well as the directionality of influence is linked to lexical specialization. Therefore, the presence of two sets of subjects (i.e. overt and null) and two sets of copulas (i.e. ser and estar) in Spanish leads to a lack of negative influence from English into Spanish. However, a facilitation effect appears in bilingual English as seen in bilinguals’ lower copula omission rates and lower null subject rate. In terms of code-mixing patterns between Determiners and Nouns, child and adult spontaneous production data differ from experimental data in that while the former show a preference for the Spanish Determiner (the category which is more grammaticized), the latter prefer the English Determiner. We propose constructs such as the Grammatical Features Spell-Out hypothesis or the Analogical Criterion to account for these patterns. The analysis of these language contact phenomena provides an insight on how language properties shape bilingual production.