Learning two languages simultaneously: mechanisms recruited for dealing with a mixed linguistic input

The present thesis aims to investigate the cognitive processes involved in learning two languages simultaneously. Early bilingual exposure might induce changes in specific domain-general or domain-specific mechanisms, which might have an impact on how bilingual language acquisition proceeds and also on other domains of cognitive development. Learning the vocabularies and the grammatical rules of two languages can be maximally efficient if young learners successfully separate and distinctively represent the two languages. In the first series of studies we investigate the mechanisms that bilingual infants may recruit to efficiently deal with a bilingual input. We explore the hypothesis that exposure to a bilingual signal affects the development of cognitive abilities involved in monitoring two languages early on. Previous research has suggested that bilingual speakers have enhanced cognitive control due to the extensive use of control while speaking one language and inhibiting the other (Bialystok et al., 2004). In contrast, here we ask whether processing two languages at a preverbal age would enhance the development of domain-general executive control abilities (Experiments 1-4) in the absence of any production. Bilingual infants may monitor and switch attention between the representational sets corresponding to the two languages well before they start to speak. This will serve an efficient acquisition of the two languages and will also result in an acceleration of executive control development. Well-developed executive control might lead in turn lead to an advantage in dealing with conflicting linguistic and nonlinguistic representations. The second series of experiments explores how monolingual and bilingual infants learn conflicting regularities from bimodal speech-like stimuli. In seven experiments (Experiments 5-11) we investigate the mechanisms involved in extracting regularities from an input that contains multiple data sets, such as adjacent and nonadjacent repetitions (AAB and ABA patterns), or a repetition-based pattern and a diversity-based pattern (ABA and ABC patterns).