Introduction

It has been repeatedly documented that the course of L1 acquisition is essentially uniform among children. The observed uniformity has led to the suggestion that all human beings are biologically predisposed to acquire a language (Chomsky, 1981). If the view is taken seriously that linguistic capacity is a common biological endowment, and if it is true that “the structure of the nervous system changes in the course of development” (Kean, 1988: 64), it is plausible that “the structure of linguistic capacity is also changing through the course of development” (Kean, 1988: 65). This view of biological maturation affecting the linguistic capacity was first expressed as the critical period hypothesis (CPH) (Lenneberg, 1967). Since then, Lenneberg's biologically based CPH has been subjected to several variations and revisions: the sensitive period hypothesis (SPH) (Oyama, 1978), the multiple sensitive periods hypothesis (Seliger, 1978), and the windows of opportunity hypothesis (Schachter, 1996).

While evidence for a critical/sensitive period in L1 acquisition is uncontroversial, the question of a critical/sensitive period for L2 acquisition has been more disputed (see Flege and Liu [2001], Flege, Yeni-Komshian, and Liu [1999], and White and Genesee [1996] for arguments against it, and DeKeyser [2000], Johnson and Newport [1989, 1991], E. Kim [1997], Lee and Schachter [1997], and Shim [1994] for strong evidence in its favor). The debate is crucial in resolving one of the central issues of L2 acquisition: whether child L1 acquisition and adult L2 acquisition are essentially similar or different.