Abstract: Using an on-line window display technique, the present study investigated how tw o features of antecedent information (the relationship and the distance with cur r ent information) affected the access of antecedent information, which was inform ed appropriately or inappropriately in the previous text. In Experiment 1, parti cipants read passages sentence by sentence at their own pace and made judgments about probe words that were either appropriate or inappropriate anaphoric refere nts. The probe words were presented in the normal text through changing the colo r of the words. The data demonstrated that the inappropriate information remained a c andidate antecedent despite the presence of a strong exclusionary statement that indicated it was incorrect. In Experiment 2, we varied the relative lags betwee n different antecedent information and the test anaphor. Participants read passa ges word by word and named the probe words. The results showed that the appropri ate anaphoric referents were accessed faster than the inappropriate ones, while the factor of distance had no effect on the activation level of the anaphoric re ferents.

Key words: discourse comprehension, access of antecedent inf ormation, window display technique, anaphor, antecedent