Neural encoding of linguistic features during natural sentence reading

Reading is a complex process that involves translating characters into meaning. Orthographic, phonologic, and semantic features of individual words seem to play a role, but it is not clear how the brain encodes these features during natural reading. To answer this, we analyzed eye tracking and electroencephalography (EEG) signals while proficient adult readers read full sentences of English text. We found that fixation-locked neural responses are modulated by the complete linguistic hierarchy. This included modulation with phoneme and phonotactic information, typically observed during auditory processing of speech. Notably, responses were modulated by the phonotactics of prior words and phonemes of the current word. Our results indicate that natural sentence reading involves, in parallel, comprehensive encoding of the fixated word, phonotactic encoding for previous words, and contextual encoding of words around fixation.