部件覺識之比較研究:以中文為第二語言的漢字圈與非漢字圈學習者為例

Abstract

本研究旨在探討以中文為第二語言的漢字圈與非漢字圈學習者於漢字部件覺識上的比較。主要的研究議題包括透過三項理解能力測驗來測試受試者是否具備漢字部件的隱性知識，並能藉此來判斷出符合結構規則的漢字及提取不熟悉字詞的字義。本研究亦驗證不同操縱變因(如:受試者的語言背景、受試者的中文能力、漢字的透明度、及語境效應)在部件覺識上的影響程度。研究對象包含實驗組的四十位外籍學生，依其語言背景及中文能力分為四組:漢字圈中高級、漢字圈中低級、非漢字圈中高級、非漢字圈中低級，每組各有十位，以及對照組的十位中文母語人士。研究發現如下: 一、所有實驗組均在不符合部件位置的非漢字判別上辨識率最高，此結果顯示出所有的受試者均擁有部件位置規則的覺識。 二、受試者在提取字義的表現取決於漢字的透明度，愈透明的漢字愈容易透過意符來完成字義連結，然而在透明度較低的漢字中，所有受試者皆傾向於使用音符來做字義推斷。 三、語境的加入有助於受試者提取漢字字義，大部分受試者在句子脈絡中的整體表現均優於字詞脈絡，此效應對非漢字圈中低級受試者的表現影響尤為顯著。 四、在相同語言背景、不同中文能力的比較上，中高級受試者的表現較佳。此外，漢字圈受試者的表現並非皆受益於母語正向遷移的影響，其在假字判別及不透明字的字義提取表現上皆呈現負遷移的現象。 綜合上述，本研究的操縱變因均在不同程度上影響了受試者對漢字部件覺識的表現，亦證實受試者對此覺識的理解及使用隨著學習時間的增長而改善。The present study aims to investigate the awareness of radicals by the learners of Chinese as a second language (CSL) from Chinese- and Non-Chinese-Character spheres. Three comprehension tasks (i.e., a character decision task (CD Task), a character meaning task (CM Task) and a semantic cueing function task (SCF Task)) were conducted to examine the participants’ implicit knowledge of Chinese radicals and whether they could make use of such knowledge in identifying legal characters as well as retrieving the meanings of unfamiliar characters. Three factors, including language background, language proficiency, character transparency, and context, were manipulated to examine their influences on the participants’ radical awareness. A total of 20 western and 20 Japanese high- and low-intermediate CSL learners were recruited as experimental groups, together with 10 native adults as controls. The major findings are summarized below: 1. Remarkable performances were found in the identification of ill-formed target characters throughout every experimental group, indicating that all the participants had the awareness of radicals in terms of its positional regularity. 2. The performance in character meaning inference was crucially dependent upon character transparency. The meaning association was triggered more effortlessly with semantic radicals in transparent target characters, whereas in opaque characters the participants resorted to phonetic radicals more frequently. 3. The facilitation of context was attested in meaning retrieval of unfamiliar characters. In most cases the supplement of contextual clues was helpful for the participants, especially for those in the Non-Chinese-Character sphere low-intermediate group. 4. Proficiency effect was evidenced in the comparison between any two experimental groups having similar language backgrounds. What is more, transfer effects were found negative among Chinese-character sphere groups in their identification of pseudo-characters and opaque character meaning retrieval, showing that language transfer from similar L1 to L2 would fail under certain conditions. As mentioned above, the manipulated factors were to some extent influential in the participants’ performance of radical awareness. Furthermore, the results also showed that a gradual mastery of this implicit knowledge was acquired by the participants along with their increasing exposure in Chinese learning.