Abstract
This paper describes research that investigates how native Japanese speakers read kanji. The controversy concerning whether native speakers of Japanese rely more on visual or phonological coding is of particular interest. Results from recent research question the persistent belief that kanji or Chinese characters are picture‐like symbols representing ideas or meanings of words and that native readers of Chinese or Japanese obtain the meanings of kanji without relying on sounds. An experiment employing eye‐tracking methodology was designed to examine whether a relationship exists between the role of phonology in reading kanji and the spatial ability of the reader. Results suggest that regardless of the reader's spatial orientation, phonological coding is an important factor in the process of understanding kanji. These findings have design implications for computer‐assisted instructional software of non‐alphabetic languages.