ABSTRACT
This chapter describes a research program in language acquisition that has a different emphasis than most. Rather than examining various stages of children's language from the outset and trying to derive implications for the process of acquisition, it began by first considering the empirical fact that the child is ultimately successful at acquiring a language. This led to consider models of language learnability: how, in principle, it may be possible to acquire the rules of an unknown language on the basis of sentences or other relevant inputs to a language learner. A scientist studying the process of language acquisition is, in a sense, a perceiver; he or she attempts to describe the acquisition process using data about adults and children's language as a guide. The chapter shows that a top-down approach, combining studies from theoretical linguistics, language typology, mathematical linguistics, computer science, and empirical studies of children and their parents, deserves a crack at the problem.
