关于音乐能力对中国大学生英语语音学习影响的实证研究

【作者】 裴正薇；

【导师】 丁言仁；

【作者基本信息】 南京大学 ， 英语语言文学， 2010， 博士

【摘要】 本论文旨在探讨非语言变量对中国大学生英语语音能力的影响,重点研究的非语言变量为音乐能力,即个体音乐学习的潜能和音乐成就的基础。除此以外,十一个与二语语音习得和音乐能力有关的非语言变量(性别、英语学习的起始年龄、目前每周英语的口头使用量、英语语音训练总量、英语语音训练内容、口头模仿能力、其他外语的学习经验、音乐兴趣、音乐训练总量、英语语音学习动机和英语语音学习策略)也被纳入研究的视角。论文突破了现行文献中对音乐能力和英语语音能力狭窄的操作定义,首次从接受性、产出性、音段能力和超音段能力多个层面上较为全面、深入地考察中国大学生音乐能力和英语语音能力两者之间的关系。论文提出了两个研究问题：一、音乐能力和其余十一个非语言变量在多大程度上能共同预测英语语音能力?二、音乐能力能在多大程度上影响英语语音能力?研究工具包括接受性音乐能力测试(听辨乐句在音调和节奏上的异同)、产出性音乐能力测试(模唱音调和节奏模式)、接受性英语语音能力测试(听辨音素最小对、节奏模式、调型模式和调核音调类型的异同及主重音的位置)、产出性英语语音能力测试(朗读和模仿单词、短句及韵律诗)、英语语音学习调查问卷(围绕十一个非语言变量)和访谈提纲(围绕英语语音能力的预测变量)。研究对象为南京农业大学非英语专业一年级的学生,平均年龄19岁。数据收集分为三个阶段。首先,214名学生在语音室随堂参加了接受性音乐能力测试和接受性英语语音能力测试,并填写了调查问卷；其次,155名学生自愿在课后参加了产出性音乐能力测试和产出性英语语音能力测试；最后,根据英语语音测试的总成绩挑选出高分者和低分者各5名进行了访谈。论文的主要研究结果归纳如下：第一、音乐能力和口头模仿能力、英语语音学习策略能共同预测英语语音能力42%的的变异,其中音乐能力的预测力最强(34%)。可见在控制了其他相关变量影响的前提下,音乐能力对英语语音能力仍然具有较强的预测力。对访谈记录的质性分析发现英语语音能力截然不同的学生在这三个预测变量上存在显著的差异。(1)与英语语音能力弱的学生相比,语音能力强的学生具有更强的音乐能力。他们不同的早期音乐经历(家庭音乐环境、个人音乐兴趣和活动、音乐教育)可以在一定程度上解释这一差异。音乐心理学家和教育学家大多认为,音乐能力是天赋和个体早期音乐经历相互作用的产物,九岁是其发展的关键期(Gordon,1987)。英语语音能力强的学生成长于音乐氛围更为浓郁的家庭环境之中。他们更喜欢歌唱活动,接受过为期数年的正规音乐教育,并且音乐教育都开始于九岁之前。(2)英语语音能力强的学生对口头模仿其他方言或外语的兴趣更高,多数有过模仿的经历。(3)英语语音能力强的学生更多地使用语音学习策略,就同一策略而言,他们的使用频率更高。这可能与他们积极的语音学习态度有关。第二、成功地构建了揭示音乐能力和英语语音能力诸变量(接受性音乐能力、产出性音乐能力、接受性英语音段能力、产出性英语音段能力、接受性英语超音段能力、产出性英语超音段能力)之间因果关系的线形结构模型。接受性音乐能力能对两个语音超音段能力的变量产生影响。它对接受性英语超音段能力产生直接影响,解释其12%的变异；对产出性英语超音段能力产生直接和间接影响,解释其约27%的变异。产出性音乐能力能直接影响两个语音产出能力的变量,分别解释产出性英语音段能力和产出性英语超音段能力55%和90%的变异。总的来说,音乐能力对英语语音能力的影响主要体现在超音段和产出能力两个层面。产出性音乐能力对英语语音能力的影响较大,因此将其列入音乐能力的操作定义十分必要。论文的研究结果具有较强的理论价值。(1)重新审视了Flege(1995)提出的言语学习模型。该模型涉及影响二语语音习得的非语言变量,如二语学习的起始年龄、在目标语国家母语和二语的使用量等,研究对象主要是二语语境下的英语学习者。本研究虽然以外语语境下的英语学习者作为研究对象,但研究结果却具有一定的普遍性。换言之,音乐能力、口头模仿能力和英语语音学习策略对二语语音能力也可能产生类似的预测力。鉴于此,笔者思考该模型是否能纳入这三个变量供今后实证研究的不断检验,以期使模型具有更强的适用性。(2)有助于对二语语音个体差异理论框架的构建。目前普遍认为二语学习的起始年龄和语言学能是二语语音习得的个体差异因索。一些国外学者如Dewaele (2009)、Major (2001)等人提出音乐能力也是一个重要的因素。本研究发现音乐能力对英语语音的产出能力和超音段能力有着较强的迁移力,这一发现是对语言学能中音素解码能力很好的补充。笔者认为音乐能力不仅是影响英语语音能力的一个非语言变量,而且还可能是影响二语语音习得的一个个体差异因素。研究结果对二语语音习得的方法论有一定的参考价值。(1)以往有关音乐和语音的研究多使用相关分析,这种单一的研究方法无助于深刻地考察各音乐能力变量对英语语音能力变量的系统影响。本研究采用结构方程建模的方法,依据理论阐释和实证数据,采用验证性因子分析和路径分析软件LISREL建立了音乐和语音潜在变量的线形结构模型,为二语语音研究提供了一种行之有效的研究方法。(2)前人关于非语言变量的研究主要使用多元刚归的定量分析方法,而本研究采用定量和定性相结合的多重研究方法。定量研究的目的是发现英语语音能力的预测变量,定性研究则补充和印证了定量研究的结果,使研究结果更为公正客观。(3)在语言测试领域,语音测试的研究相相对滞后。本研究运用整体法和元素法设计了两份英语语音能力的测试工具,项目分析、信度分析和效度分析的结果表明试卷具有较高的信效度,试卷的研制为我国的英语语音教师和研究者提供了启迪。研究结果的实践意义也不容忽视。中国的家长及学校应充分重视孩子的早期音乐教育,非富他们的音乐经历,通过各种音乐活动提升他们的音乐能力。英语教师可以考虑在语音教学前组织音乐能力的测试,针对学习者不同的能力水平设置相应的语音训练项目,真正实施因材施教。在语音教学中,教师也可以有机地融入音乐元素,帮助学生培养英语语音的产出能力和超音段能力。同时教师应注意提高学生的语音策略意识,加强语音教学中的策略训练和口头模仿训练。后续研究可以继续验证或修正本论文中所呈现的模型,进一步探讨该模型是否适用于不同学段的学生(学前、小学、中学、英语专业学生、研究生),也可以根据中国英语学习者语音学习的具体情况,分析和考察更多相关的非语言变量。更多还原

【Abstract】 L2phonological acquisition is influenced by a complex interplay of extra-linguistic variables. The present study investigated the role of musical aptitude in combination with eleven other extra-linguistic variables identified in the literature as being explanatory of L2phonological acquisition (i.e., gender, age of onset of EFL learning, amount of current oral use of English per week, amount of EFL phonological instruction ever received, content of EFL phonological instruction ever received, aptitude for oral mimicry, experiences with other foreign languages, musical interest, amount of musical training ever received, motivation for EFL phonology learning, and strategies for EFL phonology learning).Two research questions were addressed:(1) To what extent can musical aptitude and eleven other extra-linguistic variables jointly predict EFL phonological skills?(2) To what extent can musical aptitude affect EFL phonological skills?The study yielded two findings. One finding is that musical aptitude, together with aptitude for oral mimicry and strategies for EFL phonology learning, can jointly predict EFL phonological skills to some extent.The three variables combine to account for about42%of the variance in English phonological skills. Among the three predictors, musical aptitude is the most predictive, accounting for a greater amount of variance (34%) than aptitude for oral mimicry (6%) and strategies for EFL phonology learning (3%). Musical aptitude can make a unique contribution to explaining EFL phonological variance.The high and low achievers in EFL phonology exhibit remarkable differences in the three predictors of EFL phonological skills. With respect to musical aptitude, the high achievers on the whole possess stronger musical aptitude than their counterpart. The disparity in musical aptitude can be more or less explained by their different early musical experiences (i.e., home musical environment, personal musical interest and activities, musical education). Compared with the low achievers, the high achievers are brought up in an environment of which music is a part. They tend to show more preference for singing. They have received formal musical training at an early age when their musical aptitude is still growing, whereas the low achievers have not been formally educated in music at all. In terms of aptitude for oral mimicry, the high achievers take more interest in mimicking unfamiliar speech sounds, and the majority of them have the experiences of imitating another L1dialect or foreign language. In matters of strategies for EFL phonology learning, the high achievers tend to employ a greater number of strategies and use the same strategies with a higher frequency than their counterpart. Presumably, the distinction in strategy use is caused by different attitudes the high and low achievers adopt towards EFL phonology learning.The other finding of the study is that a structural model has been constructed to reveal the causal interrelations among two variables of musical aptitude (i.e., receptive musical aptitude and productive musical aptitude) and four variables of EFL phonological skills (i.e., skills of perceiving English segmentals, skills of producing English segmentals, skills of perceiving English suprasegmentals, and skills of producing English suprasegmentals).Receptive musical aptitude has direct and indirect effects on two suprasegmental variables. It can influence skills of perceiving English suprasegmentals directly, accounting for12%of its variance. It can also influence skills of producing English suprasegmentals directly and indirectly, accounting for27%of its variance.Productive musical aptitude has direct effects on two productive variables, viz., skills of producing English segmentals and skills of producing English suprasegmentals. It can account for55%of the variance in the former and90%of the variance in the latter. The strong effects of productive musical aptitude on EFL phonological skills justify the necessity of investigating productive musical aptitude as an integral component of musical aptitude. Overall, the global effects of musical aptitude on EFL phonological skills primarily manifest themselves on suprasegmental level and in productive domain.A number of implications have emerged from the study. Theoretically, the study has two implications.(1) It advances the understanding of Speech Learning Model in terms of its inclusive extra-linguistic variables. Considering that musical aptitude, aptitude for oral mimicry and strategies for EFL phonology learning, which are found to be predicative of EFL phonological skills in this study, may also take effect in ESL context, the study raises the possibility for this model to incorporate the three variables and put them to future empirical tests so that the model is more generalizable and reliable.(2) It strengthens the theoretical building of individual differences in L2phonology. Since musical aptitude is such an undeniable factor in accounting for EFL phonological variance, it may be counted as a candidate of individual differences factors affecting L2phonology.Methodologically, the study is insightful in three ways.(1) Unlike prior studies which mainly describe the music-phonology relationship with correlation analysis, this study makes a successful attempt to use the approach of structural equation modeling for the first time.(2) In their investigations of extra-linguistic variables in the realm of L2phonology, previous researchers make an exclusive use of quantitative approach. In this study, quantitative analysis is supplemented by the interview data to ensure the credibility of the quantitative results.(3) In the existing literature, little attention is paid to the issue of L2phonology testing and evaluation. This study devises two EFL phonological tests with a combination of holistic and atomistic approaches. The tests turn out to enjoy high reliability and validity.Pedagogically, the study points to the need for parents and schools to appreciate the value of early music education. Additionally, the study can facilitate EFL phonological instruction in that it alerts English teachers in China to the importance of incorporating music elements in their classrooms, raising their students’awareness of strategy use in learning English phonology and exercising students’ability to imitate unfamiliar speech sounds through mimicry practice as well.更多还原