In standard models, word meanings contribute to reading words aloud and writing them to dictation1,2,3. It is known that categories of knowledge and the associated word meanings can be spared or ...