Skip to main content


A consistent thread of work in the lab is whether the neural network involved in reading is universal or language specific. Despite differences between Chinese and English in the nature of their linguistic systems, our studies have shown that many brain regions are universally involved, but that there are some interesting cross-language differences that likely depend on the nature of the script. We have shown in a series of studies similar specialization for semantic and phonological processing across languages. However, one important way that Chinese differs from English is the large number of characters that are organized holistically in two-dimensional squares with embedded subcomponents. Chinese reading is different from English in that there are developmental increases in the involvement of bilateral occipito-parietal cortices probably due to the visuo-spatial demands of reading Chinese characters. A series of our studies examining developmental dyslexia in Chinese children shows that they have some of the same neural alterations as alphabetic readers, but with pronounced deficits in right occipital cortex, again probably due to the visuo-spatial demands of learning Chinese characters. This is consistent with our complementary findings of greater developmental increases in right occipital cortex for typically developing Chinese compared to English readers.


Another important characteristic of Chinese characters is that they have a relatively arbitrary mapping to phonology, as compared to English where there is a semi-regular alphabetic mapping. This semi-regular mapping is associated with the robust developmental increases for English children in the engagement of inferior parietal cortex. In contrast, Chinese children show developmental decreases in the involvement of temporal cortex during reading probably due to the relatively arbitrary mapping from orthography to phonology. This arbitrary mapping may also explain the developmental decreases in the involvement of occipital cortex and the lack of orthographic influences on temporal cortex during processing auditory word forms. It may be that phonological information has less of an influence on reading skill in Chinese, but that semantic information plays a more important role. Overall, our studies seem to converge on a universal network for reading, but that unique characteristics of the nature of the orthography result in cross-linguistic differences.