中文讀者對於中文字構形規則的知識：一個統計取向之語言學習理論的研究

Chinese Readers’ Knowledge of Orthographic Structure of Chinese Characters: Study from a View of Statistical Approach of Language Learning

羅明(Ming Lo)

指導教授 ： 胡志偉

國立臺灣大學/理學院/心理學研究所/博士(2011年)

https://doi.org/10.6342/NTU.2011.01210

摘要

本研究旨在以統計取向的學習觀探討中文字的習得。統計取向的學習觀主張，任何語言刺激的背後皆有一個統計性的特徵，而一般人天生有能力自發的掌握該特徵，並將其轉化為一種規則。中文字在絕大多數的情況下是由兩個部件所組成，其中又以左右橫排的結構最為常見，而且能夠提供字音線索的部件（簡稱「聲旁」），通常是位於一個字的右半邊的部件。根據統計取向的學習觀，中文讀者在學習中文字的過程中，將會發覺聲旁有很高的機率是一個字的右部件，並據以發展出「聲旁應該是字的右部件」的規則。為了觀察個體在學習發展的過程中可能產生的改變，本研究除了邀請一般的大學生，也邀請了小學低與高年級的學童擔任本研究的受試者。為了瞭解中文讀者學習中文字時所發展出的規則與文字之形音處理歷程的關係，本研究的實驗刺激包含了受試者未曾接觸過的假字，以及日常生活中可能會接觸到的用字，然後請受試者對這些實驗刺激進行兩類涉及語音處理的作業。其中一類作業要求受試者默唸刺激字的發音，然後記錄發音的內容，以及受試者默唸時刺激字誘發的事件相關電位（event-related potentials），另一類作業則要求受試者直接朗讀出刺激字的發音，然後記錄發音的正確性與反應速度。實驗結果發現，受試者會表現出一種以刺激字右部件判斷刺激字發音的傾向，而且這個傾向在大學生與國小高年級學童的反應中尤其明顯，顯示中文讀者對於聲旁在字中的空間位置，在國小的階段就發展出某種抽象的規則。除此之外，不論刺激字屬於真字或假字，當刺激字內的聲旁出現在不符合規則的位置時，腦波成分P200在事件相關電位的波形上的振幅會顯著的增大，但是當受試者是國小低年級學童時，P200的振幅較不會因刺激字的種類而有所變化。而對於朗讀刺激字的行為反應，不符合規則的刺激字會讓大學生與國小高年級學童表現出較慢的反應速度，顯示著處理歷程受到某種抑制的作用。綜合而言，本研究的結果顯示，當中文讀者學習了一定數量的中文字之後，他們對於聲旁的空間位置會發展出一種主觀性的規則。而且一如統計取向學習觀的預期，關於聲旁空間位置的規則反映著語言刺激中最常見的例子，也就是右部件是一個聲旁的字。

關鍵字

統計取向學習 ； 閱讀 ； 中文字學習 ； 事件誘發電位 ； 文字辨識

並列摘要

Learning of Chinese characters is described in terms of learning and regularization of a statistical structure underlying linguistic inputs (Shaffran, 2003; Wonnacott & Newport, 2005). The majority of Chinese characters are LR-characters, in which a character consists of two components that are arraged to form a left-and-right structure. For instance, the character ‘棋’ /qi2 / is composed of the components ‘木’ /mu4/ and ‘其’ /qi2/, and another character ‘期’ /qi2/ consists of the components ‘其’ /qi2/ and ‘月’ /yue4/. Although the component that hints the pronunciation of a LR-character can be the right or the left component of the character, the phonetic component is frequently the right component of a LR-character. If, after learning an amount of characters, a Chinese reader tends to regularize the position of a phonetic component of a LR-character to develop a rule that ‘其’ should be the right part of a LR-character, then ‘棋’ is the one that follows the rule while ‘期’ is the one that violates the rule, and the two types of characters should be treated differently by the reader. To investigate the relationship between the number of characters a reader has learned and the the way that a reader develops rules of the position of phonetic components, a group of college and elementary school students was invited to participate in the present research. In addition, to investigate what is the effect of the rules on the cognitive processing of a LR-character, the participants were asked to read a number of real characters or pseudo-characters. In Exp. 1 and 2, ten college students and twelve first-graders were asked to sub-vocalize pseudo-characters. Event-related potentials were measured, and greater P200 amplitudes were elicited in the condition that the left component of a pseudo-character was a phonetic component. However, the effect was not found for the first-graders. In the third and the fourth experiments, 27 college students and 24 fifth-graders were asked to sub-vocalize regular and irregular characters. Event-related potentials were also measured, and for both groups of the participants, greater P200 amplitudes were elicited when the character is regular and the left component of the character was a phonetic component. In Exp. 5 and 6, twenty college students and thirty sixth-graders were asked to read aloud regular and irregular characters. The regularity effect in Chinese (Hue, 1992; Lee, Tsai, Su, Tzeng, & Hung, 2005) was replicated, and reaction times lengthened in the condition that the left component of a regular character was a phonetic component. The pattern of the results indicates that the joint probability between the structure of characters and the position of a phonetic component is acquired and regularized by readers to form an abstract rule of orthographic structure of LR-characters. And Chinese young readers are able to form the abstract rule of orthographic structure of LR-characters when they are in fifth grade.

並列關鍵字

statistical learning ； reading ； Chinese character learning ； event-related potentials ； visual word recognition

參考文獻

Lo, M., Hue, C. W., & Tsai, F. Z. (2007). Chinese readers’ knowledge of how Chinese orthography represents phonology. Chinese Journal of Psychology, 49, 315-334.

Wang, C. C., Hung, L. Y., Chang, Y., W., & Chen, H. F. (2008). Number of characters school students know from grade 1 to G9. Bulletin of Education Psychology, 39, 555-568.

Lee, C. Y., Tsai, J. L., Su, E. C. I., Tzeng, O. J. L., & Hung, D. L. (2005). Consistency, regularity, and frequency effects in naming Chinese characters. Language and Linguistics, 6, 75-107.

Adrian, E. D., & Matthews, B. H. C. (1934). The Berger rhythm: Potential changes from the occipital lobes in man. Brain, 57, 355-385.

Arciuli, J., & Cupples, L. (2006). The processing of lexical stress during visual word recognition: Typicality effects and orthographic correlates. Quarterly Journal of Experimental Psychology, 59, 920-948.