Towards a unified vision of classes of language acquisition and change

Abstract

In this study, I analyse Mozambican African Portuguese (MAP), a non-native variety typically acquired as a second language (L2) by children, speakers of Bantu languages, as an instance of language change. Assessed diachronically, MAP/L2 constitutes an interesting object of research since it shows how linguistic and extra-linguistic factors can block the resetting of new/wrong parameter values, making the process of L2 change less gradual than L1 (normal) change and also less dramatic than creolization. To illustrate the argument, I present two case studies on MAP which show why, even assuming that in all cases we are dealing with language acquisition by children who are capable of attaining UG-consistent linguistic systems, these instances of language acquisition and change still look different in kind. These case studies show, more specifically, how the previous knowledge of L1 grammar makes MAP/L2 acquisition more difficult, requiring, for some parametric values, the exposure to extra-data. In post-colonial societies, these extra-triggers are not always available, due to the quantitative preponderance of L2 speakers. As a consequence, the visible effects of some new/wrong parameter settings become more frequent in the input and, instead of showing necessity of resetting, they are taken as cues for new parameter settings.