Tracing the Course of Sentence Comprehension: How Lexical Information is Used

Abstract

There are several reasons why the study of sentence comprehension has become interesting again. Since its earliest days, psycholinguistics was concerned with how people use their knowledge of grammar to construct representations of sentences (Fodor, Bever, & Garrett, 1974; Miller, 1962). This concern has not changed. However, the theoretical and methodological tools for addressing it have improved dramatically. The theories that were entertained two decades ago relied on inappropriate grammatical analyses (as, for example, the derivational theory of complexity relied on transformations) or placed unrealistic demands on the cognitive processor (e.g., analysis by synthesis theories) (cf. Fodor et al., 1974). Most of the available techniques focused on the products of comprehension (Fillenbaum, 1970). The few on-line measures that were used failed to disentangle the results of processes occurring at several different stages of comprehension (Gough, 1965). In the last decade, however, explicit, controversial theories that illuminate the fine-grained details of parsing have been proposed (Abney, 1989; Altmann & Steedman, 1988; Fodor, 1983; Forster, 1979; Frazier, 1978; Kimball, 1973; Marslen-Wilson & Tyler, 1987; Pritchett, 1988; Tanenhaus & Carlson, 1989; cf. Tanenhaus, 1988, for a good overview). Linguistic theory has changed in ways that make it much more consistent with psychological theorizing (Chomsky, 1981; Frazier, 1988).