Toward an understanding of the universal word recognition system based on insights from the unique characteristics of Japanese kanji and kana

Hisashi Masuda, Ph. D.
Department of Psychology, Faculty of Humanities and Human Sciences,
Hiroshima Shudo University, Hiroshima, Japan 　

Abstract:
Over the past three decades, psychologists have debated about the nature of the mental representations and recognition systems for words. Through these debates, a number of promising models have been proposed, such as the dual route model, the interactive-activation model, and the parallel distributed processing model. However, most of the visual word recognition research on which these models stand has been concerned primarily with lexical processing in alphabetic writing systems, with only relatively few studies concerned with Chinese, Korean, and Japanese. Generally, Chinese and Japanese kanji are believed to differ from alphabetic writing systems in terms of having a stronger morpheme-meaning relationship than a morpheme-phonology relationship. However, for some kanji characters there is a consistent correspondence between the character elements (i.e., radicals) and pronunciations. The situation is further complicated because Japanese kanji differ from Chinese characters in having two or more pronunciations (typically having On and Kun readings). Another factor adding to the complexity of the Japanese writing system is the fact that it also employs two orthographically shallow kana scripts ( hiragana and katakana ). As most Japanese words are generally written in a particular script in daily life, Japanese readers can experience difficulty in reading a word written in an unfamiliar script, even if the script is kana . In my talk, I will discuss some results from my research concerning subword activation in the reading of kanji , phonological confusions in kanji identification, and script familiarity effects on reading kana words within the frameworks of some models highlighting insights for our understanding of the universal word recognition system.