Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Reading Without Lexical Processing: What the Brain Does During Passive Reading
Poster Session D, Thursday, October 1, 4:30 - 6:30 pm, Wangari Maathai
This poster is part of the Sandbox Series.
Jialu Li1, Matthew H. Davis2, Kathleen Rastle3, Jo Taylor1; 1University College London, 2University of Cambridge, 3Royal Holloway University of London
Passive reading paradigms are often used in neuroimaging studies of reading because they are assumed to minimize top-down task demands and reflect automatic language processing. In particular, fMRI studies comparing written words with symbol strings during passive reading have reported greater activation for words in visual word form area. However, it is unclear whether this reflects engagement of lexical representations or lower-level visual or orthographic analysis. The current fMRI study examined whether the lexicality effect is present during passive reading and compared this to active reading. Stimuli consisted of visually presented English words and pseudowords matched on length and orthographic characteristics. Forty-eight adult participants completed one active reading task (read aloud) and two passive reading tasks: fixation cross detection and one-back repetition detection. In the fixation cross detection task, participants pressed a button when the fixation cross in the center of the word/pseudoword changed color. In the one-back repetition detection task, participants pressed a button when a word/pseudoword was repeated on consecutive trials. Stimuli were additionally embedded in varying levels of visual noise to examine how perceptual difficulty modulates lexicality effects across tasks. Items were presented in a randomized order with an event-related design for all tasks. Sparse acquisition was used for the read aloud task (items presented for 1500ms, 2000ms ISI) and continuous acquisition for the passive tasks (items presented for 1700ms, 300ms ISI). In the read aloud task there was greater activity in ventral reading regions (anterior fusiform, middle temporal gyrus) for words relative to pseudowords and greater activity in posterior occipitotemporal cortex as well as dorsal reading regions (inferior parietal and frontal gyri) for pseudowords relative to words. These clusters were used as regions of interest in analyses of the passive reading tasks. In the passive reading tasks, though task engagement was high (>94% accuracy), whole-brain and regions of interest analyses revealed no significant lexicality effects. There was also no interaction between noise level and lexicality. However, there were opposing noise effects in the two passive tasks: in fixation cross detection, activity in posterior occipitotemporal cortex was greater at lower noise levels, whereas in one-back repetition detection, activity in this region was greater at higher noise levels. These findings suggest that passive reading paradigms may not engage lexical representations to the extent often assumed and may instead primarily engage lower-level perceptual or orthographic processing. They also indicate that different passive tasks make different demands on visual reading regions. The results have important implications for interpreting passive paradigms in neuroimaging studies of language and reading.
Topic Areas: Methods, Reading