Single and compound logographic Kanji words elicit distinct early neurophysiological responses: ERP evidence from fluent and naive Kanji readers

摘要

While the N170 to printed words in alphabetic scripts is consistently left-lateralized, the available evidence on logographic scripts is much less consistent. Here, we examined the extent to which use of single-character (e.g., 121) versus compound-character (e.g., 2 1) words in a logographic script could account for differences in the laterality of early neural responses in fluent readers. Participants included both fluent Kanji readers (n = 17) as well as naive Kanji readers (n = 19), in order to separate out differences due to low-level physical differences between stimuli. While fluent Kanji readers showed a larger amplitude N170 overall relative to naive readers, this expertise effect was bilateral, rather than left-lateralized as is common with alphabetic scripts. Trend-level differences in laterality were observed at the earlier P1 response, which has not been extensively examined in previous studies of logographic script reading, with naive readers only tending to show right-lateralization across both single-and compound-character Kanji words. Both participant groups differentiated single and compound Kanji words, but contrary to predictions these effects were bilateral and evident during different stages of processing in each group. Among naive readers, compound Kanji elicited a larger amplitude P1, suggesting influence on early perceptual processing, whereas among fluent readers, these differences were not apparent until the later N170. Taken together, findings from the present study indicate that lowlevel visual characteristics and perceptual expertise interact during the specialization of neural circuits for print, with implications for theories of lateralization of neural circuits for print.