Abstract
A hypothesis was proposed that the presence of appropriate furigana (phonetic signs) would shorten the search time required for reading a kanji in on (Chinese style) or kun (Japanese style). In a factorial design, a series of kanji were presented without furigana, or with on- or kun-furigana, and subjects (Japanese) were instructed to read each kanji in either on or kun. Both the latency of reading and the probability of correct recall were influenced by the furigana, the instruction on reading style, and the customary reading style of each kanji. The results supported the hypothesis, and further reinforced a theory that Japanese subjects attach a meaning to a Chinese character by reading it in Japanese style at the first sight of the character.
