Saito H, Masuda H, Kawakami M
Graduate School of Human Informatics, Nagoya University, Japan.
Brain Lang. 1999;68(1-2):75-81. doi: 10.1006/brln.1999.2107.
Kanji are categorized into four types based on the combination of "subword validity" (when the right phonemic radical represents the same On reading as the whole Kanji character) and "radical-neighbor consistency" (when the whole Kanji character represents the same On reading with all its neighbor characters). The study demonstrated that both subword validity and radical-neighbor consistency affect naming latencies and error rates for relatively high-frequency Kanji character words. Moreover, both analytic and holistic decoding processes, which correspond to subword and whole-character activation, are used in the naming of Japanese single Kanji character words. Accounts of the findings by dual-route and connectionist models are discussed.
根据“子词有效性”（右部音素部首与整个汉字的音读相同）和“部首邻接一致性”（整个汉字与其所有邻接汉字的音读相同）的组合，汉字被分为四种类型。该研究表明，子词有效性和部首邻接一致性都会影响相对高频汉字词的命名潜伏期和错误率。此外，与子词和整字激活相对应的分析和解码过程都用于日语单汉字词的命名。文中讨论了双通路模型和联结主义模型对这些发现的解释。
