Recycling utterances: A speaker's guide to sentence processing

In recent years, there has been a growing consensus that speakers store large numbers of preconstructed phrases and low-level patterns, even when these can be derived from more abstract constructions, and that ordinary language use relies heavily on such relatively concrete, lexically specific units rather than abstract rules or schemas that apply “across the board”. One of the advantages of such an approach is that it provides a straightforward explanation of how grammar can be learned from the input; and in fact, previous work (e.g. Dąbrowska and Lieven 2005) has demonstrated that the utterances children produce can be derived by superimposing and juxtaposing lexically specific units derived directly from utterances that they had previously experienced. This paper argues that such a “recycling” account can also explain adults' ability to produce complex fluent speech in real time, and explores the implications of such a view for theories of language representation and processing.