Polarity in L3 English: the initial stages and beyond

The main goal of this dissertation is to understand how L3/Ln acquisition unfolds and how studying it transitions to the bigger field of non-native acquisition as well as theoretical proposals for grammatical properties. In an attempt to do so, the compilation of four studies help us answer the three over-arching questions that guide this doctoral dissertation, as presented below:I. What can the study of multilingualism tell us about the cognitive processes underlying the initial stages and beyond of any instance of non-native acquisition?II. What do methodological practices in the field of L3/Ln acquisition tell us about the variability found in the literature?III. How can the study of multilingualism help us to understand the nature of certain linguistic domains?In summary, the first study examines the knowledge of Negative Concord Items and Differential Object Marking in the grammar of highly proficient early bilinguals in Catalan and Spanish; and the role language dominance has for the interaction of the languages in early bilingualism. The results show that (a) remaining dominant in the L1 contributes to the maintenance of target-line behavior in the language and (b) that different domains of grammar are affected in different ways. The second study provides an analytical panoramic view of the field of L3/Ln acquisition by reviewing the majority of available L3 morphosyntactic studies published between 2004 and 2018 systematically, examining (and showing) how the methodological practices can explain some of the variability we find in the literature. In the third study, we provide a snapshot of the initial stages of acquisition and we show how confounding proficiency and exposure can introduce potential noise into the study of transfer. In the fourth study we arrive at the culmination of the dissertation by examining the grammars of ab initio L3 learners of English who are Catalan-Spanish bilinguals. The results show that holistic structural similarity is the most deterministic factor for transfer selection in the case of early bilinguals acquiring a third language. More importantly, results of the longitudinal design reveal that developmental sequencing after initial stages transfer is dynamic and nonuniform depending on language dominance in the previous acquired languages. The overall picture of the results of the four studies show that holistic structural similarity plays a role at the initial stages of L3/Ln acquisition, that language dominance plays an important role for L3 development, that certain methodological practices are to be adopted in L3/Ln acquisition and that the study of multilingualism can give answers to the formalization of Negative Concord Items more generally