Dennis L. Molfese, Ph.D.
Art C. Maerlender, Ph.D.
Date of this Version
Two studies were performed to investigate the temporal structure and organization of the language processing system during the comprehension of coherent contextually ambiguous narrative discourses. In Study 1, participants read short discourses that were contextually ambiguous if read without a descriptive title (Untitled group) or unambiguous with the title (Titled group). Participants identified the title of the discourses after reading 1-3 sentences. Given the unfinished next sentence, they performed a cloze procedure on the sentence-final word. For the Titled Group, cloze probability was greater to the last word of sentence 3 (Critical Word 3) than sentences 2 (Critical Word 2) and 1 (Critical Word 1). For the Untitled group, title identification accuracy was greater after reading the first two sentences than the first sentence and even more so after reading the full three-sentence discourses than the first two sentences alone. This study established 25 discourses in which the contexts were initially ambiguous but became increasingly clearer after reading Critical Word 2 and Critical Word 3.
In Study 2, a different sample of participants read the discourses with (Titled Discourse group) or without (Untitled Discourse group) a descriptive title while undergoing high-density event-related potential (ERP) recording. For the Untitled Discourse group, N400 amplitudes became less negative from Critical Word 1 to Critical Word 2 and again to Critical Word 3, suggesting greater ease of lexical-semantic retrieval as the discourses unfolded. For this group, P600 amplitudes were larger to Critical Word 3 than Critical Word 2. Unexpectedly, a Late Sustained Frontal Positivity (SFP) ERP component occurred to both Critical Word 2 and Critical Word 3 for the Untitled Discourse group only. SFP results corresponded to the title identification findings from Study 1. Thus, it is proposed that the SFP reflects the resolution or revision of contextual ambiguity during discourse comprehension. Alternatively, the P600 is proposed to reflect the updating of the existing context of a discourse when an context is available or after the resolution of contextual ambiguity.
Advisors: Dennis L. Molfese and Art C. Maerlender