Bilingual Selection of Syntactic Knowledge

This text motivates a more formal approach in theoretical linguistics by investigating the parameters of syntactic variation and simultaneous acquisition of multiple languages. Taking the behaviour of the Null Subject Parameter (NSP) across languages as an example the book raises important questions concerning the adequacy of standard parameter-setting models in the face of compelling evidence from both monolingual and bilingual child speech data. Teresa Satterfield argues that "universal" premises guiding well-known parametric approaches complicate attempts to construct an economical bilingual analysis. Further, she demonstrates the compatibility of more recent formulations in linguistic theory (i.e. the Minimalist Program) and studies on language learnability (Clark, 1992, 1993; Kapur, 1994) which present the view that while initially convincing, standard parameter models are potentially costly and less effective in terms of monolinguals. Using Clark's application of the Genetic Algorithm as a point of depature, this text describes a number of computational simulations. These simulations not only demonstrate the nature of the GA-as-language-learner, but offer a more detailed account of the parameter-setting task confronting the bilingual child while also making more predictions regarding the process of syntactic knowledge.