THE EFFECTS OF DISCRIMINATION AND DECODING PROCESS ON THE READING OF KANJI AND KANA SCRIPT IN YOUNG CHILDREN

Abstract

The first purpose of this paper was to reveal the decoding Processes of young children acquiring the reading of semantic units, through an experiment using similaritiebs etweenideographic scriptsa nd real things (Exp. I). The second purpose was to examine the discrimination process and the decoding process of young children by manipulating the number of letters and the meaning of Kana script (Exp. II). The third purpose was to make clear the role of these processes in the learning of Kanji reading (Exp. III).
In Exp. I the Ss were 32 kindergarten children. Each S was required to learn the reading of either four Kanji or four ideographic scripts at each trial. The learning was continued until one perfect criterion had been attained, or else for 10 trials. After these trials, each S immediately received four over leaning trials. The latency of correct reading was recorded.
The results obtained from Exp. I are shown in FIG. 1; (1) Ss under the ideographic condition learned more quickly than Ss under the Kanji condition,(2) there was, however, no difference of latency in any condition. Results indicated that the similari-ties between ideographic scripts and real things facilitated the decoding process: a script was visually decoded and, judging from the subject's discourse, after acquiring its reading, this decoding process was progressed only by repetition of the reading.
In Exp. II the Ss were 32 kindergarten children who had already acquired the reading of Kana scripts. The design of this Exp. was a 2x2 factorial one, in which a between factor was the length of the Kana sequence (2-or4-letters) and a within factor was the meaningfulness, meaninglessness of the Kana sequence. Each S was asked to read eight Kana sequences at one trial and to continue for another 8 trials. And then the latency of reading was recorded.
The results obtained from Exp. II are shown in FIG. 2; (3) 2-letter sequences were asked quickly read than those of 4-letter and meaningful sequences were faster read than meaningless ones.(4) Furthermore, a first order interaction between the degree of meaningfulness and the blocks of trials, and a second order interaction among three main factors were confirmed respectively. These results may be interpreted as supporting the hypothesis that meaningfulness facilitated the decoding processes, and this facilitation was more effective in the 4-letter condition than in the 2-letter.
In Exp. III the Ss were 32 kindergarten children who had not acquired the reading of Kanji scripts used in this Exp. The design of this Exp. was a 2x2 within factorial one, in which one factor was the meaningfulness of the Kanji (On-reading or Kunreading) and the other was the degree of the complexity of Kanji (complex or simple). Each S was asked to learn the reading of four Kanji scripts. The learning was continued until one perfect criterion had been attained, or else for 12 trials. The latency of all correct responses was recorded.
The results obtained from Exp. III are shown in FIG. 3; (5) Kun (Japanese-style reading: high meaningfulness) were more quickly learned and read than On (Chinese-style reading: low meaningfulness), simple scripts were more quickly learned and read than complex ones. Furthermore the latency of reading increased trial by trial. These results indicated that the discrimination and the decoding processes affected the reading processes, but also that these processes were independent of one another.