Abstract
In 1999, Caplan and Waters suggested that basic sentence comprehension is an on-line process that does not draw upon working memory resources. While most spoken language comprehension tasks use off-line measures, the present study additionally examined eye-gaze in a speech comprehension task that manipulated syntactic complexity and agency distance. Participants listened to spoken sentences and were asked to click on the “doer” of the action in each sentence from two choices presented on a computer screen in front of them. Among eighteen younger adults, it was found that the times required to click on the correct answers (overt response times or ORTs) were more affected by the manipulations than the times required to fixate on the correct answers (eye-fixation times or EFTs), and that working memory span had a stronger relationship with on-line responses than off-line responses. These results provide partial support for Caplan and Waters’ model; working memory span did not predict EFTs in more syntactically complex sentences, but effects of working memory span were seen on EFTs based on agency distance, suggesting that working memory resources were in fact required for the comprehension task.