In search for the referent:

anaphor resolution in linear text and hypertext


The interpretation of natural discourse is a complex dynamic process that evolves in time with amazing speed and efficiency. It crucially depends on context coherence, as well as on cognitive functions such as working memory capacity of the reader. This project focuses on the resolution of anaphoric expressions (i.e. anaphoric noun phrases such as the book, the purchase). Anaphoric expressions can only be fully interpreted if they are linked to entities introduced elsewhere in the discourse.

Therefore they form cohesive links between sentences and sentence fragments, which makes anaphor resolution indispensable for discourse comprehension. In this project the influence of four discourse constructional factors on anaphor resolution will be examined: the presence of multiple possible referents in the discourse, the distance between the anaphoric expression and the referents, the amount of elaboration of possible referents, and the form of the elaboration. I will perform a line of experiments that extends beyond previous research as it addresses the on-line comprehension of anaphor resolution in linear discourse as well as in hypertext.

Although hypertext is a fast growing discourse medium we are more and more exposed to, little is known of the exact influence of discourse constructional factors on the comprehension of hypertext. This project aims to contribute to more detailed insights on this issue by investigating anaphor resolution in hypertext using the sophisticated method of tracking eye movements while reading.

Furthermore, the possible influence of individual differences in working memory capacity on anaphor resolution will be addressed by comparing differences in eye movement patterns of readers with high and low verbal working memory span. Following the eye movements and fixation durations of native speakers of Dutch with high and low working memory will tell us whether different reading strategies interact with individual differences, as well as which factors may facilitate or disrupt the comprehension process. Hence, the outcome of these studies will provide new information on the influence of discourse constructional factors in the comprehension of linear discourse, which in turn may lead to new guidelines for the construction of hypertext.


Principal researcher :Dr. Monique Lamers