Seeing is knowing? Visual word recognition in non-dyslexic and dyslexic readers: an ERP study

The aim of the current study was to investigate whether phonological/semantic processing of the word takes place simultaneously with, or following, the early processing of its visual features. Event related potentials (ERPs) were recorded in 13 dyslexic (four female) and 14 non-dyslexic (six female) native English speaking young adults in two lexical decision tasks. In Task 1 participants had to make an orthographic lexical decision to distinguish frequently used words (W) from pseudohomophones (PH1)focusing on visual properties of stimuli. In Task 2 they had to make a phonological lexical decision—to pseudohomophones (PH2) and pseudowords (PW) and decide whether stimuli sounded like real words—focusing on non-visual higher order, i.e., phonological and semantic, processing of the stimuli. The behavioural performance was less good and the ERP peaks’ latency longer in dyslexics compared to controls. The reaction times (RTs) and the number of errors (reversed for the controls in Task 2) increased across four conditions for both groups in the following order: W< PH1< PH2< PW. The ERPs were larger in Task 2 compared to Task 1 starting at 100 ms (P1) for the controls and from about 220 ms (P2) for the dyslexics. The latency of N2 peak in left occipito-temporal sites was larger (as was the number of errors) in PH2 compared to PW condition in controls only, which indicates phonological/semantic specific processing at a time latency of 250–260 ms. Thus, the visual task required less effort than the phonological task, dyslexics’ behavioural performance was less good and the brain activation delayed compared to controls. Combined behavioural and ERP results of this study indicated that phonological/semantic processing of the word took place 150 ms after processing of its visual features in controls and possibly later in dyslexics.