A developmental perspective on the neural code for written words.

Two recent TICS articles [1,2] converge on the idea that in mature reading written words are coded by an abstract neural representation comprising either “open bigrams” [2] or bigrams in combination with larger units [1]. For example, the input TAKE is represented by the units TA TK TE AK AE and KE. However, open bigrams do not map onto phonology. Because phonology is fundamental in beginning and skilled reading [3], we argue that open bigrams cannot be the full story.