Figure 2.

Time courses of decoding accuracy for the key study variables (left panels), peaking at subsequent times (as marked by vertical bars). Horizontal dashed bars in yellow represent the a priori cluster threshold for uncorrected p < 0.05 for decoding accuracy. Data are extracted from left occipitoparietal and posterior temporal sensors areas for object category (top), from left frontal and adjacent anterior temporal sensors for both word length (middle) and phonological neighborhood density (bottom), as shown in the topographic views on the right. The sensor selection was based on a priori hypotheses on the emergence of semantic information in posterior occipitoparietal and temporal regions, and of phonological processes in left frontal regions, as proposed by the Indefrey and Levelt (2004) model.