An Eye-Movement Analysis of Ambiguity Resolution: Beyond Meaning Access

Abstract

The research reported here addresses the status of the unselected meaning of a lexically ambiguous word in developing the larger meaning of the text by independently manipulating lexical and discourse-level variables in the text. In a series of 3 eye-movement experiments, participants read passages that contained 2 occurrences of an ambiguous word. The meaning of the ambiguous word either switched from 1 encounter to the next (Experiments 1 and 2) or remained consistent across encounters (Experiments 1 and 3), and the authors manipulated whether the second occurrence of the ambiguous word referred to the same discourse entity or a different entity. Facilitation was observed in initial processing time on ambiguous words when the meaning was consistent from first to second encounter, regardless of referential relations. Switching the meaning between encounters did not influence how accessible the initially unselected meaning was, but it did produce processing difficulty in the post-target region. These results suggest that the lexical status of the (initially) unselected meaning remains unchanged while the availability of this information for the purposes of text integration changes with changes in the configuration of the discourse representation.

Notes

¹To better understand this effect, we deconstructed the post-target region. In addition to the measure we reported, we examined the first fixation following the target work (spillover effects) and the last fixation in the post-target region (wrap-up effects) for each of the three experiments. When we subjected these fixation times to an analysis of variance, we found no significant effects across all three experiments. Thus, it does appear as if the processing difficulty that is found in the experiments is spread across the post-target region.