ABSTRACT
Over the last 15 years, there has been considerable psycholinguistic interest in the time course of anaphoric resolution: the process by which a reader or listener determines the antecedent for an anaphoric expression (Clark & Sengul, 1979; Dell, McKoon, & Ratcliffe, 1983; Garrod & Sanford, 1977; Gernsbacher, 1989). This research indicates that resolving anaphora is an integral part of sentence comprehension that can often occur at the earliest possible point in reading or spoken-language comprehension (Dell et al., 1983; Gernsbacher, 1989; Marslen-Wilson, Tyler, & Koster, 1993). However, most of the discussion in psycholinguistics treats anaphoric interpretation as an isolated process only loosely associated with overall resolution of the sentence against its discourse context (Sanford & Garrod, 1989). This is very much in contrast with the treatment in the linguistic literature, where different anaphoric devices have been associated with a range of discourse functions signaling different ways in which a sentence should be resolved (Ariel, 1990; Givon, 1983; see also Marslen-Wilson, Levy, & Tyler, 1982, for a more process-oriented discussion). The present chapter sets out to reconsider some of the processing assumptions about anaphoric resolution in the light of these linguistic analyses.
