关键词: 盲文阅读, 盲人, 脑可塑性, 视觉词形区

Abstract: Both the deprivation of sensory experience and the acquisition of cultural skills such as reading can induce plastic changes in the brain. On the one hand, sensory experience deprivation, such as lifelong loss of visual input in congenital blindness leads to large-scale reorganization of cortical function, with occipital “visual” regions becoming responsive to non-visual modalities and even higher-order cognition such as language, memory or numerical reasoning. On the other hand, the acquisition of cultural inventions such as reading sculpts the cortex in domain-specific ways, with the ventral occipito-temporal cortex (vOTC) developing a specialized “visual word form area” (VWFA) in literate individuals. Braille reading in blind individuals provides a unique natural model in which sensory modality (tactile rather than visual) and lifelong visual deprivation jointly shape the reading network. In this review, we summarize recent neuroimaging evidence with a focus on three central questions: (1) whether the early visual cortex in blind individuals contains reading-specific representations; (2) whether the visual word form area within the ventral occipito-temporal cortex retains cross-modal orthographic processing functions; and (3) whether the parietal cortex could host an analogue of the VWFA, a putative “tactile word form area”.
First, converging evidence from fMRI, TMS and lesion studies indicates that early visual cortex contributes to Braille reading. However, what information these regions encode remains unclear. Some evidence indicates that early visual cortex is involved in higher-level tactile discrimination, but whether this activity is specific to Braille reading remains controversial. Other evidence suggests that early visual cortex participates in higher-level linguistic processing and may be integrated into the language network in blindness, implying that its activation during Braille reading could reflect lexical or semantic processing rather than tactile analysis. Paradigms that confound tactile shape processing with naming or phonological tasks further complicate interpretation. A key next step is to determine whether occipital activation in Braille reading reflects a reading-selective mechanism or instead stems from broader linguistic or executive processing, and, if selectivity exists, to identify the nature of the representations involved.
The second question concerns whether the VWFA in blind individuals preserves its canonical role in orthographic analysis via cross-modal plasticity, or whether it is repurposed entirely for higher-order linguistic computation, as proposed by the cognitively pluripotent cortex hypothesis. Although some studies report Braille-selective activation patterns resembling those in sighted readers, their reliance on contrasts between Braille words and low-level tactile controls makes it difficult to determine whether such effects reflect true orthographic processing or simply the presence of linguistic content. Other findings instead reveal sensitivity of the VWFA to spoken language, syntactic structure and broader linguistic demands—properties not typically seen in the sighted brain. Moreover, blindness appears to alter the internal representational hierarchy of ventral occipito-temporal cortex, blurring the traditional gradient from low-level letter features to higher-order word forms. Disentangling genuinely orthographic computations from more general linguistic processing therefore remains a central challenge for future work.
The third question concerns whether tactile orthographic processing may be supported by the parietal lobe rather than the ventral stream, a hypothesis that remains relatively underexplored but potentially important. The posterior parietal cortex is a plausible candidate for such a role, given its documented involvement in tactile shape discrimination and its anatomical and functional connectivity with language-related frontal regions. Converging evidence shows that parietal activity is sensitive to Braille letter length and sublexical structure, while other findings suggest a hierarchical gradient from somatosensory encoding anteriorly to word-level preference posteriorly. These observations raise the possibility of a parietal “tactile word form area.”
Taken together, current research provides compelling evidence that Braille reading engages a distributed network spanning somatosensory, parietal, occipital and language regions, but the division of labor between parietal and occipital cortices remains unresolved. Progress will require more cautiously controlled and methodologically diverse experimental designs to disentangle tactile, orthographic and linguistic computations. Moreover, the neural basis of individual differences in Braille reading proficiency remains largely unexplored, despite its clear relevance for educational interventions. Addressing these open questions will not only refine theories of cortical recycling and pluripotency but also provide principled guidance for designing targeted Braille literacy training programs.

Key words: Braille reading, blindness, neuroplasticity, visual word form area