Overspecification in Early English

Abstract

"After almost 50 years of research in Generative Linguistics, there is little doubt that many properties of human languages are innately determined in our genetic make up. Still, one the most enduring questions is how much is predetermined? Baker (2002) discusses two logical ways in which Universal Grammar might be specified in our genetic endowment. One alternative is to say that Universal Grammar (UG) overdetermines the grammar of particular languages. On this view, all linguistic principles and parameters are known by the child; she just needs to identify the specific parametric values that define the adult language, and discard the rest. On the other view, some grammatical points are left underspecified, with domain-general learning mechanisms filling in the missing information from the environment (Pinker and Bloom, 1990). In this paper, we study three two-year-old children’s acquisition of inflection and negation, and demonstrate that the path of acquisition favours the Overspecified view of Universal Grammar"