N400-like effect of stroke-based neighborhood size in Chinese characters

Word-to-letter neighborhood feedback has been proposed to account for the lexical facilitation effect of orthographic neighborhood size (N). In the present study, a new N of Chinese characters was manipulated. Neighbors were created by randomly replacing about one third of the total strokes of characters while keeping the structure and the orientation of its remaining strokes constant. Participants were asked to match probe characters with the preceding target characters while event-related brain potentials elicited by target characters were recorded. In the N400 time window, characters with many neighbors produced more negative amplitudes than characters with few neighbors. The N400-like effect supports the lateral inhibition component of the interactive activation model where the non-target candidates are inhibited; lateral inhibition contributes to the N400-like effect of N more than does character-to-stroke feedback. Likewise, a large neighborhood size yielded higher N250 amplitudes, thus possibly indicating greater summed mismatches between the numerous neighbors and their targets.