Abstract
The purpose of this study was to examine whether semantic processing of presented words occurs, as indicated by ERP measures, even when such processing is not required to perform a given task. ERPs were recorded while 12 students viewed nouns appearing on a CRT every 700 ms and detected the occasional target determined by an orthographic cue. Six word lists were presented in each of three modes created by combinations of two of the three types of writing system : Japanese phonograms, Hiragana and Katakana, and ideograms, Kanji (Chinese characters). In each list of 80 words, 80% of the nouns were written in one system and the remaining 20% (target words) were in the other. In addition, each list contained semantically deviant words, the existence of which was not known to the subject prior to the task. Each stimulus belonged to one of three semantic categories ; animals, plants, and parts of the body. In a list, 80% of the nouns belonged to one category, while the other 20% belonged to a different category.
The semantically deviant words elicited larger N400 than did the frequent category words, though the difference in amplitude was smaller than that evoked in a semantic task. This result suggests that some kind of semantic processing occurred, despite the fact that semantic processing was not necessary for performing the task.