捷克大學二語學習者華語雙音節聲調偏誤分析

筆者至捷克馬薩里克大學實習期間，發現大三、大四的捷克學生華語雙音節聲調仍有偏誤，興起研究念頭。本研究以對比分析及偏誤分析理論為基，對比捷克語和華語的超音段特徵（聲調，語調，重音）和兩個語言音系。捷克語中的元音有長短區分，音長不同，詞重音固定在第一音節，為了檢視這些差異是不是造成捷克學習者產生聲調偏誤的原因，設計實驗研究。首先以問卷調查訪問七十位在捷克大學學習中文的學習者，初步了解捷克學習者難點。選取捷克大學中文系華語教材中的雙音節詞，作為華語雙音節聲調發音與聽辨測驗材料，徵求六位捷克籍學生參與實驗，以人耳審聽和Praat擷取語圖，描述捷克學習者的發音和聽辨偏誤，推論偏誤形成原因─學習者不易掌握華語音節的音高變化，捷克語的詞重音韻律模式（重輕，長短對比）讓捷克學習者以兩音節間的輕重對比和長短差異模式，產出和感知華語雙音節聲調。接著請捷克學習者與華語母語者聽辨判斷華語雙音節詞的重音位置，綜合討論學習者感知華語雙音節詞重音位置的結果與聲調產出和感知的互動關係，以語音頻譜圖，音高曲線圖，音長、音強圖佐證。參考高成就捷克學習者和捷克教師問卷調查和質化訪談內容，與文獻中捷克語重音聲學特徵的議題，提出華語雙音節聲調教學建議。研究結果發現六位受試者發音偏誤率前三高的聲調組為2+4（40%），2+0（33%），2+3（33%），4+0（28%）。發音偏誤率最低的聲調組合為1+1，1+2，1+3。聽辨偏誤率前三高的聲調組合為1+3（44%），1+0（40%），1+4（38%），聽辨偏誤率最低的聲調組合是1+1（8%），3+0（8%），4+4（8%）。受試者按學習華語時間長短分成中級組、中高級組、高級組，發音偏誤率在中高級組大幅降低（由中級組的47%降為21.5%），聽辨偏誤到了高級組才大幅降低（由中高級組的27.5%降為8%）。捷克學生發音時兩音節的音強對比較強，或前後音節音長差異較大，在發音和聽辨聲調時把降調或低調視為重音音節，非重音音節容易變為平調（第一聲）或低調（第三聲）。針對高頻偏誤教學，並讓學習者認識華語聲調和重音與語調的異同，用相似處如捷克語語調說明華語調形，重音說明華語輕聲。運用雙碼理論和個別化教學理論為基礎進行教學。

關鍵字

華語雙音節詞；偏誤分析；聲調；重音；華語；捷克學習者

並列摘要

During the classroom observations while doing an internship in Masaryk University, the researcher found that many of the Czech students had many tone errors in Chinese disyllabic words, even if they were already junior or senior college year students. Thus, this study focused on disyllabic tone errors produced by Czech college students. This thesis was primarily based on contrastive analysis, error analysis and interlanguage concept . Initially, it discussed the suprasegemental features in Chinese and Czech, making a comparison in tone, intonation and stress for both languages. Difference such as Czech word has fixed stress on the first syllable, etc. In addition, an introduction about Chinese and Czech phonetic system was discussed. After reviewing all the similarities and differences between the two languages, the researcher conducted pronounciation and listening tests of the Czech students’ Chinese disyllabic words. Online Questionnaire was used as a tool to measure, analyze and make a comparison. The questionnaire also included questions to get preliminary knowledge of the Czech students’ learning difficulties in listening and pronunciation. A total of seventy valid questionnaires were gathered, and six out of the seventy respondents were recruited to voluntarily participate in production and perception tests. Three Chinese native speakers reviewed the Czech students’ tone production, writing down their errors in tone combination with the use of phonetic sound software, called “Praat,” that produces wide-band spectrograms. Given the test results, this research hypothesis states that, “Stress system in Czech words influence Czech students’ Chinese disyllabic tone production and perception. To verify the research hypothesis, some Czech students and Chinese native speakers were asked to listen to Chinese disyllabic words and choose which syllables were stressesd. A useful and efficient research method used was Interview Survey. One Czech teacher and chosen learners were interviewed in the test to sum up the research findings. This research found that Czech students’ 2+4 (40%) had higher error rate, while 1+1, 1+2, and 1+3 had the lowest error rate in the test production. 1+3 (44%) had the highest error rate, while 1+1 (8%), 3+0 (8%), and 4+4 (8%) had the lowest error rate in the perception test. The study sample was divided into three groups according to the students’ length of time in learning Chinese language (intermediate group, high-intermediate group, and advanced group). The longer they learned Chinese, the higher the decrease in the students’ tone production and perception error rates. Also, the pronouncing error rate dropped significantly between intermediate group and high-intermediate group (from 47% to 21.5). The Listening error rate dropped significantly between high-intermediate group and advanced group (from 27.8% to 8%). The Czech word stress pattern significantly influenced the Czech students’ disyllabic tone error in Chinese words. When pronouncing Chinese disyllabic words, the Czech students’ syllabic stress was relatively strong or there was a big difference in pronunciation length of the initial or final syllable. For example, the Czech students tend to put one syllable stressed while the other unstressed in Chinese disyllabic words as reflected in syllabic intensity duration and pitch. The Czech students could not prolong or raise the pitch of unstressed syllables and they used Chinese low tone (third tone) or level tone (first tone) to replace the original tone of the unstressed syllable instead. While speaking and listening to Chinese disyllabic words, high falling tone (forth tone) was more likely pronounced in stressed syllables, and it was more often chosen for the stressed syllable in the listening test. Chinese language teachers should help Czech students become aware of the differences between their mother tongue and the other language they are learning , focusing on tone combination with high error rate and inter-learner difficulties. Aside from providing the students knowledge about the differences of the two languages, teachers could use Dual-coding theory to give students sound and image input at the same time. Individualized instruction would help students understand their learning difficulties and enable teachers to provide instant, specialized feedback.