【作者】 杨晋；

【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】 在语言学习中,频次作用(frequency effect)被认为起着无可替代的重要作用。N.Ellis指出,频次作用涉及语言加工和学习的各个方面,随着频次的增加,同时出现的语言单位之间的联系不断被加强和积累,因此易于习得。但是二语习得领域许多专家并不支持频次的核心作用,他们认为二语习得受到许多其他因素影响。N.Ellis认为(2002)显性指导(explicit instruction)的作用不可忽视。以上观点说明了频次作用的争议性,近年来,虽然已有学者开展关于频次作用的实证研究,但是涉及二语语音习得研究相对较少。鉴于此,本研究选取中国学生习得英语重音模式为研究对象,旨在探究频次作用和显性指导在学生聆听模仿本族语者对话过程中重音模式的变化规律。重音是英语最显著的超音段语音特征之一,也是传递话语信息的重要表现形式。重读就是韵律突显；一个语调单位常常包含几个重读音节：调核重音和其他非调核重读音节。所有重读音节和调核一起构成信息链。错误或者不恰当的重音模式将导致交际误解以致失败,重音模式的习得对二语语音习得至关重要。目前为止,二语语音习得研究中超音段特征研究仍然不足。关于中国英语学习者的超音段学习研究,陈桦(2005)比较了学习者朗读口语中容易引起本族语者误解和导致信息传递失败的二语语调特点。毕冉(2009)研究了英语专业学生朗读任务中重音分布的发展规律。但目前为止,大多数研究仍以描述重音模式为主,尚未就语音训练对英语学习者的重音模式所能产生的影响展开深入研究。中国英语学习者的重音模式问题在于学习者话语中重读音节过多,学习者不能恰当地表现重读和非重读音节对比。纵贯研究结果表明,在四年的大学学习中,重音问题不但没有得到改善,反而呈现一定程度的上升趋势,由重音模式所反映的节奏模式,在四年期间也没有明显变化,学习者的话语体现非常明显的音节节拍语言特点。前人研究存在几点不足。首先,各种研究主要限于对于二语学习者重音模式特点的描述。其次,除少数研究外,前人研究较多集中于调核重音,对一个语调单位内其他重读音节的关注不足。同时,仍然缺乏对语调单位层面突显程度不同的音节对比研究。再次,在研究体现重音的声学参数时,前人研究只集中在单个参数上,而没有统一考虑音高、时长和音强三个参数。国外的大多数研究依靠计算机软件或者声学试验仪器对学习者的口语材料进行分析,而国内二语语音习得研究使用软件手段刚刚起步,有待进一步深入。英语重音教学一直在国内语音教学中所占地位不高,传统上教师主要通过学生反复聆听本族语者的范例,再加以模仿来学习英语重音。近年来有人提出采用明示教学法可视化手段等,但是这些教学手段的效果有待进一步验证。本论文对中国英语学习者重音模式的习得过程进行实证研究,探究两种方式对重音模式产出的影响：频次作用和频次作用加上显性指导,具体定义为反复聆听模仿和反复聆听模仿加上教师指导。本研究围绕以下问题：反复聆听模仿和反复聆听模仿加上教师指导对中国英语学习者朗读口语中的重音模式有什么影响?具体来说,(1)对学习者不同句型中的句子重读音节分布有什么影响?(2)对学习者区分句子中不同突显度的音节有何影响?(3)对学习者朗读口语中的节奏特征有何影响?研究对象是12名英语专业一年级学生,这些学习者全都来自南京一所大学英语系,英语口语水平初级。本研究共分两个阶段,在第一阶段,受试者在训练前依次朗读四个对话,对话包含不同句型和不同长度的语调单位,受试者通过反复聆听模仿朗读对话三次。在第二阶段,6名受试者分为两组,一组的训练方式和第一阶段相同。另一组加上教师指导,两组受试者朗读本族语者的对话一次。研究者对收集的录音样本利用语音标注及分析软件Praat进行标注,根据分析结果,主要有以下研究发现：首先,本研究结果表明反复聆听模仿这种练习方式对二语学习者产出英语重音模式有一定的影响,但影响有限。练习前学习者没有考虑词语的信息承载状况,口语中重读音节过多。名词、系动词“be”和功能词的重音分布率高出本族语者,练习后这种状况没有得到明显改善。练习后学习者朗读口语中的调核重音分布的准确率有一定提高,无标记调核的准确率提高快于标记性调核。不同句型的句子的调核重音分布在练习后有所变化,一般陈述句、复合句和特殊疑问句的重音分布正确率要高于其他句型。反复聆听模仿能够在一定程度上帮助学习者从时长特征上区分重读和非重读音节但不能区分调核音节和非调核重读音节。反复聆听模仿不能有效帮助学习者从音高特征上区分调核音节和非调核重读音节以及重读和非重读音节。反复聆听模仿不能对学习者的重音间隔之间的等时性特征产生任何明显的改进作用。反复聆听模仿练习能在有限程度上改善学习者的朗读口语中的节奏特征,比如元音变异度的指数能够得到改善,而对辅音变异度的影响不大。本研究结果同时表明,即使继续增加聆听和模仿的频次,重音模式的各项指数也未能得到有效改善。其次,反复聆听模仿加上教师指导比单独的反复聆听模仿能帮助学习者提高重读音节的分布准确率。各种句型的重音分布准确率比第一阶段都有进步,学习者调核重音和非调核重音的准确率比第一阶段有明显提高,调核重音和非调核重音,重读音节和非重读音节的时长比和音高之比,与第一阶段相比,有接近本族语者的趋势。反复聆听模仿加上教师指导能够帮助学习者学习重读和非重读音节的产出,同样,这种练习方式能帮助学习者在节奏特征上取得进步,学习者的重音间隔之间的等时性特征有接近本族语者的趋势,学习者不仅在元音时长与整个语调单位时长之比以及成对元音的变异指数上取得进步,而且辅音标准误和成对辅音变异度的得分也得到改善。本论文的研究成果在理论上和实践上的意义丰富。理论上揭示了频次作用对二语语音习得的作用和影响,研究结果证明频次作用对二语语音学习具有一定效果,但不能起到决定性作用,显性指导能引导学习者注意目标语特征。本研究结果支持“注意力假设”,即“注意”是第二语言习得的必要因素,学习者反复接触目标语结构能促使他们注意到语言形式,能注意到自身语言产出和目标语的差距。频次作用加上显性指导能够加快学习者学习英语重音模式的进程。在教学实践上,本研究表明反复聆听模仿仍不失为一种有效的学习语音的方式,值得语言教师和学习者的重视,同时表明二语语音习得中,教师指导是不可替代的。更多还原

【Abstract】 Both frequency and explicit instruction are believed to play important roles in the language learning classroom. Although some research has been conducted on their effects on the oral English learning, not enough empirical studies have been conducted on the learning of suprasegmental features. This dissertation reported a study that explored the effects of frequency effects and the effect of explicit instruction on EFL learners’ production of English stress patterns.Stress is a property of syllables that makes them stand out as more prominent than others. In one tone group, there is almost always a peak of stress called "tonic stress" or "nucleus." Nuclear stress is normally found in a content word in utterance-final position. It can also be found in other positions. Stresses other than tonic stress are called non-nuclear stresses. The rest of the syllables are unstressed syllables. The different distributions of stress directly affect the rhythmic features of English.It is believed by many that misuse or inappropriate use of stress patterns can lead to problems in understanding. The learning of appropriate stress patterns, therefore, seems essential before a second language learner can have successful communication with native speakers. Chinese EFL speakers are often described as producing less differentiation between stressed and unstressed syllables and stressing more syllables than do native speakers. Previous studies on the learning of English stress patterns by Chinese EFL learners have involved descriptive analysis of learners’ speech patterns. Chen (2005) has studied the nuclear stress placement by Chinese EFL learners and Bi (2009) has examined the development of stress patterns of some Chinese EFL learners. Yet they have not touched upon the effect of explicit instruction on the learning of stress patterns. More studies are needed to enrich our understanding of the process of learning English suprasegmental features.The present study explored the frequency effects and explicit instruction on12first-year English majors’ command of stress patterns within a tone group. The treatment materials consisted of quasi-spontaneous dialogues between native speakers. The imitations of learners were examined for stress placement, the contrast between syllables of different prominence levels and the rhythmic features. The participants received treatment at two stages. At Stage One, they listened to and imitated the native model dialogues for different times. At Stage Two, in addition to repeated listening and imitation, they also received teacher instruction on the necessary information on English stress patterns. Recorded data were analyzed acoustically. An analysis of the recordings revealed the following findings:1. Repeated listening and imitation is found to have affected the participants’stress patterns in various ways. Before such treatment, there was an overuse of stress on syllables regardless of their information status. The treatment produced some limited effect on the stress patterns, but it had almost no effect on the problem of placing stress on copula "be" and some of the function words. The treatment can only help learners improve the correctness rates of the placement of unmarked nuclei but not those of the placement of marked nuclei. Although there was some improvement made in the non-nuclear placement, the improvement was not seen on all the sentences.With regard to the contrast between nuclear and non-nuclear stresses, the treatment may have helped the learners make negligible progress in the pitch ratio and limited progress in the duration ratios of nuclear to non-nuclear stressed syllables. As for the contrast between the stressed and unstressed syllables, the treatment seemed to help them make improvement in the mean duration ratios but failed to help them make improvement in the mean pitch ratios of stressed to unstressed syllables.Repeated listening and imitation does not have any effect on the duration of inter-stress intervals (ISIs). With continued treatment, no progress could be observed in terms of inter-stress intervals.The treatment produced some improvements in the proportion of vocalic intervals (%V) and variability of vocalic intervals (nPVI). Some limited progress was also made in consonantal variability (AC).2. Repeated listening and imitation accompanied by teacher instruction can improve the correctness rates of the stress placement. It can improve the stress placement for sentences of various lengths. It can enhance the participant’s consciousness of the stressed and unstressed syllables. The instruction pushes learners to pay attention to the way native speakers stress words and reduce stress on those words that are not supposed to be emphasized.Teacher instruction can help learners make noticeable improvement in the mean syllable durations. As a result, the duration ratios of nuclear to non-nuclear stressed syllables and stressed to unstressed syllables can get closer to those of the native speaker.Teacher instruction can lead to greater improvement in pitch ratios than mere listening and imitation. The pitch ratio of the nuclear to non-nuclear stressed syllables and the ratio of stressed to unstressed syllables were both raised and thus closer to those in the speech of native speakers.Repeated listening and imitation accompanied by teacher instruction appears advantageous over mere listening and imitation in helping learners make progress in inter-stress intervals. Before receiving teacher instruction, almost all the durations of ISIs were longer than those of native speakers. After the instruction, the participants often produced shorter ISIs. In contrast, for those who continued to listen and imitate without teacher instruction, there was only limited improvement.Repeated listening and imitation accompanied by teacher instruction seems more useful in helping learners make progress in the four indices of rhythmic features than mere listening and imitation.The findings of the present study have rich theoretical and pedagogical implications. Theoretically, the present study implies that frequency has limited effects on L2speech learning, but it is explicit instruction that directs learners’attention to the target forms and enables them to make progress. Therefore the findings give support to the noticing hypothesis. The repeated experiences with target structure afford learners opportunity to pay attention to language form and notice the gap between one’s language production and the target form. Pedagogically, the practice of repeated listening and imitation can be widely implemented in the EFL classroom. Teacher instruction can better help learners make improvements in the production of English stress patterns. It also implies that much practice cannot replace teacher instruction.更多还原