Over the past three decades the major obstacle for scientists interested in the psychological aspects of human linguistic competence has been generative grammar. Throughout this period, generative grammarians have claimed, and many psychologists have believed, that the only interesting aspect of language is its syntax, and that syntactic structure consists wholly of mathematical algorithms that are independent of meaning, communicative intention, and other psychological processes (e.g., Chomsky, 1986; Pinker, 1994). Recently, however, a new linguistic paradigm has emerged, and it is much more congenial to the traditional concerns of psychologists. The paradigm of cognitive linguistics, and its companion functional linguistics, is explicitly committed to describing and investigating linguistic competence in psychologically meaningful terms (Lakoff, 1990). Cognitive linguists thus describe linguistic structure in terms of such things as symbols, categories, schemas, perspectives, images, communicative functions, and a variety of other fundamentally cognitive and social processes (E. A. Bates & MacWhinney, 1989; Lakoff, 1987; Langacker, 1987, 1991; Talmy, 1988; van Valin, 1993). With the descriptions of linguistic structure emerging in this new paradigm, we have, for the first time, the

276 TOMASELLO possibility of creating a real psychology of language (Tomasello, 1992a, 1992b).