Delineating the processing of Chinese characters and pictures : an ERP approach

Abstract

The neural processing of reading Chinese scripts has previously sparked a fierce debate, particularly in questioning the uniqueness in processing the logographic Chinese characters compared with pictures. This study was designed to ascertain at which exact time frame, as well as to what extent, the visual, orthographic, and semantic processing of the two domains (Chinese characters and line drawing pictures) resemble and differ, by examining their well-formedness (real, pseudo, and non). By employing an event-related potential (ERP) repetition detection task, 29 participants identified when a stimulus was repeated (fillers) amongst a series of randomly presented target characters or line drawings and varied in well-formedness. ERP results revealed a significantly greater activation in pictures than Chinese characters at the early visual P100 component, and across the subsequent P200-N200 and N400 components, while a greater activation was observed in characters than pictures at the orthographic sensitive N170 component. Furthermore, it was observed that the well-formedness of Chinese characters and pictures was distinguished at only the N170 and P200-N200 components respectively. Later at the semantic N400 component, it was observed that only nonitems were distinguished from real and pseudoitems in both characters and pictures at a frontocentral region. Taken together, the results suggest that Chinese characters differ from pictures at early visual and orthographic processing, as well as their well-formedness, but start to resemble at later semantic processing.