Accessibility of potential referents following categorical anaphors

In 4 experiments, the authors investigated the accessibility of potential antecedents that were either referents or nonreferents of categorical anaphors using several methodologies. In 2 preliminary experiments, using an immediate probe recognition task, the authors replicated previous findings of inhibition for nonreferents following categorical anaphors. In Experiments 1-3, naming times to nonreferent probes were inhibited, but only at a delay and at the end of the sentence. Eye movements were monitored in Experiment 4, and reading times were found to be slower in end-of-sentence regions following the remention of the nonreferent but not on the nonreferent itself. These results suggest that decreases in nonreferent accessibility are due to postanaphoric processes, such as integrating the nonreferent into the discourse structure, and are not due to more immediate changes in activation occurring as part of the resolution process. Further, some inhibition effects may result from the use of probe tasks that require decisions.