ABSTRACT
Chomsky’s theory of language acquisition shed doubt on the assumption that language can be learned on the basis of experience (see Unsworth, this volume). But currently, input-driven or constructivist theories of language acquisition gain ground quickly, as both corpus and experimental methods allow us to investigate children’s learning processes in more detail. There are different labels for such theories, for instance, usage-based theory of acquisition (Tomasello, 2000; 2003; Goldberg, 2006), or emergentism (Elman et al., 1996; MacWhinney, 1999; O’Grady, 2005). Usage-based theories emphasize that language learning in children is the result of their experience with language use, and consequently, adult linguistic representations are usage-based, not innate. We build linguistic categories through repeated exposure to similar or related structures (e.g. Bybee, 2006; Ellis, 2008). The term “emergentism” emphasizes the progressive nature of the process, namely that qualitatively new and more complex representations can emerge on the basis of knowing its simpler component parts (O’Grady, 2005). These approaches are constructivist because they assume that children, when learning the language, “reconstruct” the system by generalizing over usage-events, that is, utterances they have heard or produced, rather than activating an innate representation of grammar.
