Abstract
Although various factors have been considered in syntactic ambiguity research, the influence of formal textual features remains largely unexplored. Poem stimuli were designed to test whether line breaks influence the availability of preceding hosts for RC adjunction. In an online pilot reading study, we considered presentation in 5-line poem format (RC in its own verse) and in prose (RC and hosts on a single line). The prosodic features of rhyme and meter were manipulated to test whether they influence attachment decisions (by making a host more prominent) as well as whether they facilitate or impede the disambiguation process. Results: High Attachment was preferred the most while Low Attachment the least across all +/-Meter & Rhyme conditions. Stimuli were read significantly faster in their +Meter/+Rhyme version relative to the -Meter/-Rhyme condition. The combination of both these rhetorical devices led to processing delay, albeit not significant, when answering an attachment site question.
