漢字, 仮名表記語の情報処理 読みに及ぼすデータ推進型処理と概念推進型処理の効果

野村 幸正

doi:10.4992/jjpsy.51.327

漢字, 仮名表記語の情報処理

読みに及ぼすデータ推進型処理と概念推進型処理の効果

野村幸正

著者情報

ジャーナルフリー

1981 年 51 巻 6 号 p. 327-334

DOI https://doi.org/10.4992/jjpsy.51.327

詳細

発行日: 1981/03/20 受付日: 1980/03/29 J-STAGE公開日: 2010/07/16 受理日: - 早期公開日: - 改訂日: -
訂正情報
訂正日: 2010/07/16 訂正理由: - 訂正箇所: 論文抄録訂正内容: Wrong : The main purpose of this study was to compare the reading processes of Kanji with those of Kana script. The hypotheses upon which these experiments were based are as follows: The meaning of both Kanji and Kana script are extracted by both data-driven and conceptually-driven processing. In the reading of Kanji, however, their pronunciations seem to be generated on the basis of their meaning. On the other hand, while the reading of a sequence of Kana script seems to be generated on the basis of meaning as with Kanji, it is directly constructed from data-driven processing since forty-six syllabics correspond to forty-six phonetic units. In the following three experiments, subjects were presented a random series of Kanji or Kana script one by one and required to read them as fast as they could.
In Exp. I subjects were required to read a random series of Kanji and Kana script under a 2×3 factorial design. One factor was the number of letters forming the semantic unit (i.e. one or two Kanji), the other factor was the types of presentation, i.e. Kanji, Kanji with Furigana (Kana letters printed at the side of Kanji), and Kana script. The results of Exp. I are shown in Fig. 1. The latency data revealed main effects of both of these two factors, and furthermore a significant. interaction between factors. These results seem to support the above hypotheses.
In Exp. II subjects were required to read sequences of Kana script under a 2×2 factorial design. One factor was the number of Kana syllabics (2 or 4), the other factor was meaningfulness or meaninglessness. The results of Exp. II are shown in Fig. 2. The latency data revealed main effects of both these two factors, and furthermore a significant interaction between factors. These results seem to support the possibility that meaningful 4-syllabic sequences are read not only by data-driven processing but also by conceptually-driven processing, while meaningful 2-syllabic sequences or meaninglessness 2- and 4-syllabic sequences are read only by data-driven processing.
In Exp. III subjects were required to read sequences of Kana script under a 2×2×2 factorial design, in which a between-subjects factor was the number Kana syllabics (2 or 4) and two within-subjects factors were meaningfulness or meaninglessness and the presence or absence of prior warning concerning this aspect of the presented sequences. The results of Exp. III are shown in Figs. 3 and 4. The latency data revealed main effects of three factors, and furthermore the significant interaction among these three factors. The results seem to support the above possibility.
Right : The main purpose of this study was to compare the reading processes of Kanji with those of Kana script. The hypotheses upon which these Exps. were based are as follows: the meaning of both Kanji and Kana script are extracted by both data-driven and conceptually-driven processing. In the reading of Kanji, however, their pronunciations seem to be generated on the basis of their meaning. On the other hand, while the reading of a sequence of Kana script seems to be generated on the basis of meaning as with Kanji, it is directly constructed from data-driven processing since forty-six syllabics correspond to forty-six phonetic units. Subjects were required to read Kanji or Kana script as fast as they could. The results of these Exps. seem to support the above hypotheses.

抄録

The main purpose of this study was to compare the reading processes of Kanji with those of Kana script. The hypotheses upon which these Exps. were based are as follows: the meaning of both Kanji and Kana script are extracted by both data-driven and conceptually-driven processing. In the reading of Kanji, however, their pronunciations seem to be generated on the basis of their meaning. On the other hand, while the reading of a sequence of Kana script seems to be generated on the basis of meaning as with Kanji, it is directly constructed from data-driven processing since forty-six syllabics correspond to forty-six phonetic units. Subjects were required to read Kanji or Kana script as fast as they could. The results of these Exps. seem to support the above hypotheses.

責任著者(Corresponding author)

訂正情報

J-STAGEへの登録はこちら（無料）