越南學生一聲和四聲之偏誤探討和矯正對策：華語聲調線上訓練教材之設計

Abstract

對於華語一聲（陰平）和四聲（去聲）的混淆不清，是越南語母語者學習華語時，共通且不易矯正的偏誤。先行研究探討此偏誤成因時，未考量越南語平聲變體對華語一、四聲感知的影響，故尚有可補充之處，而提供有效之聲調教材以協助越南學生矯正聲調偏誤，更是迫切所需。本研究目的即在調查越南學生的華語聲調偏誤模式，分析其偏誤成因，以便據此設計出聲調訓練教材，並驗證此教材之有效性。教材設計為線上教材形式，可作為學生課後自學或教師課堂輔助的教材。 本研究分三個實驗進行。實驗一調查聲調偏誤模式：對25位在台灣的越南籍大學生施以華語聲調「聽辨」、「發音」和「記憶」測驗，以調查其一聲和四聲在單音節詞和雙音節詞的偏誤狀況。結果顯示「聽辨」和「記憶」偏誤模式有其一致性，皆是一聲偏誤率最高，四聲次之，且一、四聲居雙音節詞中之前字時，偏誤率高於居後字。在「發音」方面，四聲偏誤率最高，其次為一聲。四聲居前字時，偏誤率高於居後字，而一聲居前字時，偏誤率低於居後字。 實驗二探討越南學生一、四聲偏誤成因：對35位越南學生施以越南語「雙音節詞發音」調查，對另25位越南學生施以越南語「句子聽辨」調查，再對另17位越南學生施以越南語「雙音節詞聽辨」調查，以考察越南語平聲之聲調調性。結果顯示越南語平聲不僅有高平調之調性，也存在著高降調之變體。由此推論越南學生在學習華語的「語音感知同化」過程中，將華語高平調之一聲和高降調之四聲，皆同化到越南語平聲之同一個音位範疇內，因而造成區分華語一聲和四聲的困難。 實驗三根據偏誤模式及成因，設計華語聲調線上訓練教材，並檢驗其使用成效：由實驗組和對照組各20位越南學生進行實驗。教材為高變異聲調知覺訓練，依據偏誤模式中各聲調組合之難易度，由易而難進行，以大量一、四聲對比之聽辨訓練，協助學習者區分並建立一、四聲各自之聲調範疇。結果顯示，實驗組在一、四聲的聽辨和發音兩方面，進步程度都優於對照組，且發音方面差異更大。而實驗組於擴展測驗和延時後測的顯著進步，也顯示教材中的高變異語音元素，有助於學習者適應不同的語音環境和發音人，且訓練成效可於訓練後繼續保持。

The confusion between Tone 1 (Yin Ping) and Tone 4 (Qu Sheng) in Mandarin is typical and persistent among Vietnamese learners of Mandarin. Existing research, when probing for the causes of this phenomenon, does not take into account the impact of Vietnamese Tone ngang’s variations on the perception of Tone 1 and Tone 4. This paper aims to analyse the patterns and causes of Vietnamese learners’ tone errors and, with the findings as basis, design learning materials, whose effectiveness is to be verified by an experiment. These materials are meant to be available online, serving as self-learning resources after class or supplementary materials for teachers in classroom. This study includes three experiments. Experiment 1 examines the patterns of tone errors, with 25 Vietnamese university students in Taiwan as subjects. Tests of perception, pronunciation and memorization were administered to examine the errors of Tone 1 and Tone 4 in monosyllabic and disyllabic words. The results indicate similar error patterns in perception and memorization, with Tone 1 at the highest error rate followed by Tone 4. Meanwhile, Tone 1 or Tone 4 in the former syllable of disyllabic words reports a higher error rate than in the latter syllable. Tone 4 reports the highest error rate for pronunciation followed by Tone 1: the error rate is higher when Tone 4 is in the former syllable of disyllabic words but higher when Tone 1 is in the latter syllable. Experiment 2 explores the cause of the errors of Tone 1 and Tone 4 among Vietnamese learners. Surveys were conducted to observe the characteristics of Vietnamese Tone ngang, with 35 Vietnamese students tested on pronunciation of Vietnamese disyllabic words, 25 on perception of Vietnamese sentences, and 17 on perception of Vietnamese disyllabic words. The results indicate that the high level Tone ngang is in free variation with a high falling tone. It is thus referred that Vietnamese learners of Mandarin perceptually assimilate both the high level Tone 1 and the high falling Tone 4 to their native Tone ngang category, which induces their difficulties in distinguishing between Tone 1 and Tone 4. Experiment 3 assesses the effectiveness of the online tone training materials, the treatment group and the control group each consisting of 20 Vietnamese students. Specifically designed to address the patterns and cause of Vietnamese learners’ tone errors, the materials provide the treatment group with high-variability perceptual training, starting from easy tasks to more difficult ones. With a large number of contrasts between Tone 1 and Tone 4, the perceptual training helps differentiate and then establish separate tone categories for Tone 1 and Tone 4. The treatment group exceeds the control group in the improvement of perception and pronunciation, particularly in pronunciation. Also, the progress of the treatment group in generalization tests and retention tests highlights the fact that the improvement gained in training was generalized to new stimuli and new talkers and could be retained after training.