| The establishment of coreferential relations between anaphors and the antecedents to which they refer is elemental to understanding discourse. One factor that has been ignored in previous research is the role of "real-world" reference. Anaphors that are used in research rarely refer to actual people, objects or events. In three experiments, I explored the effects of real-world reference on coreferential processing. In Experiment 1, participants responded to recognition targets that appeared immediately before or after the anaphor. I found evidence for enhancement and suppression when the names were "empty". In contrast, enhancement, but not suppression, was found when reference was real. In Experiment 2, event-related potentials were collected as participants read sentences that contained either a correct or an anomalous anaphor (e.g., "she" in reference to a male antecedent). Real-world reference enhanced the magnitude processing difficulty in the anomalous condition. In Experiment 3, eye-tracking data revealed that real reference facilitated coreference during the earliest stages of processing and that its effects continued during later stages. I interpret my results in light of Sanford & Garrod's Scenario Mapping and Focus Theory (1998), a model of text processing offering a detailed account of how entities are represented in a reader's mental "model". |
