Abstract
Reading is a task that engages the vision-language interface, and the efficiency of this process may be related to how much of the text readers use to process visual and/or linguistic information during a fixation (i.e., their reading span size). Compared with typically hearing readers, deaf signers both read more efficiently (i.e., faster with equivalent comprehension) and have wider reading spans, suggesting a critical role for span size in reading efficiency. Using a linear regression analysis, we assessed the effects of two different span sizes (the word identification span, within which readers extract linguistic information, and the perceptual span, within which they extract the visuospatial layout) and reading comprehension ability on reading rate for deaf signers (n = 50) and hearing non-signers (n = 109). For all readers, those with a larger word identification span read faster, but this relationship was weaker for the perceptual span, emphasizing the importance of parafoveal linguistic processing for reading speed. The effect of the word identification span was stronger for the deaf group, and only deaf signers showed a positive relationship between reading comprehension ability and reading rate (and word identification span size). These findings suggest that reading efficiency (exemplified by deaf signers) derives from an ability to extract linguistic information far from fixation in a way that supports both reading speed and comprehension.