Expertise in reading visual Braille co-opts the reading network

Abstract

AbstractShape processing is crucial for visual object understanding. Visual reading seems to be attuned to it, as most scripts share similar visual features like line junctions and shape cues. Here we investigate the effect of expertise in reading visual Braille, a script developed for touch that does not share the typical explicit shape information of other alphabets. We compared visual Braille readers and a naïve control group and found that individually localized Visual Word Form Area (VWFA) was selectively activated for visual Braille when compared to scrambled Braille only in expert Braille readers. Multivariate analyses showed that linguistic properties can be decoded from Latin script in both groups and from Braille script in expert readers. Yet, cross-scripts generalization failed to reveal common representations across Latin and Braille in experts, suggesting that the VWFA of visual Braille experts contains interdigitated rather than joint representations of visual Braille and Latin scripts. Primary visual cortex, shape-selective areas (Lateral Occipital areas, LO), and linguistic areas (left Posterior Temporal area, l-PosTemp) showed similar multivariate profiles to VWFA, but with aligned representation across scripts in the left Posterior Temporal area only. We conclude that the linguistic status of a visual script, rather than low-level line-junctions properties, play a major role in how the visual system, and VWFA in particular, shows preference for scripts.Significance statementWhat drives the sensitivity in visual cortex for written words? We show that in expert readers, the Visual Word Form Area (VWFA), commonly thought to integrate line-junctions into words, responds to visual Braille, a script without explicit shape cues. Words, regardless the presence or absence of lines, are processed for their linguistic content, which happens in a script-specific manner. Additionally, we observe script-specific linguistic processing similar to VWFA in several key regions involved in vision (V1, LO), and more script-invariant linguistic processing in a language area (l-PosTemp). We conclude that the linguistic properties of a visual script, rather than low-level line-junctions properties, play a major role in how the reading network specialise for processing scripts.