ABSTRACT

This paper argues for the indispensable value of linguistics for any account of language, symbolic or connectionist. We argue that the kind of data considered by linguists and the theories derived from this data should not be dismissed.