The reading brain - Transforming vision into language

Reading is built upon transformations that map representations of written words to their pronunciations, names, and meanings. However, unlike many other visual skills, literacy is unique to humans and has only become widespread in the past few hundred years, therefore neural circuits that underly reading cannot have evolved for reading. Thus, scientists have long debated the nature of the neural re-tuning of visual networks that must occur to support these transformations in a highly skilled manner. A key node of the reading brain, the visual word form area (VWFA), lies where circuits that underpin key aspects of visuolinguistic transformations diverge from earlier visual processing in ventral occipitotemporal cortex. Furthermore, different reading deficits co-occur with parallel deficits in these visuolinguistic transformations. Evidence suggests that as literacy is acquired during development, preexisting visuolinguistic circuits that support representational transformations compatible with reading become tuned for understanding written words. This model implies that the VWFA is where it is because reading requires functional capacities that are parallel to those needed to recognize and name objects on one hand and those needed for multimodal integration during speech/language comprehension on the other. The VWFA is at a key location that gets input from earlier visual regions and has lateral connectivity into the auditory-visual speech and language network and ventral connectivity into visual naming circuits. We review neurophysiological, neuroanatomical, and neurodevelopmental evidence that support a model in which literacy emerges from the specialization of preexisting circuits that have the natural capacities for vision to linguistic transformations needed to acquire and facilitate fluent reading.