1. Introduction

1.1 Research background

Language, an irreplaceable tool in daily lives, plays an intrinsic role in communication. With the increasingly faster pace of the whole society, English is becoming more and more omnipresent no matter where we are and what we are doing, ranging from trawling information online to cross-cultural communication. However, for some EFL learners, how to speak English fluently becomes a conundrum. In China, nowadays there still exist mountains of questions getting students, teachers and researchers anxious and worried. For oral English, empirical studies mainly concern oral abilities, including accuracy, fluency and complexity, and theorical studies largely comprises oral ability developments patterns that form the potent basis for fluency research (Wang Zhou, 2004). Thus, China has seen great strides in fluency investigation. When measuring fluency, researchers have to tackle performance variables or temporal variables (Zhang Wu, 2001). And empirical studies on measuring fluency mainly adopt temporal variables, such as speech rate, mean length of runs, phonation/time ratio, articulation rate, frequencies and lengths of pauses (Towell, 1996; Zhang, 1999b; Zhang Wu, 2001; Yang Mu, 2011b). Pause, an important temporal variable because all the calculations of temporal variables pivot on the definition of pause (Zhang Wu, 2001), is becoming more and more noticeable (Yang Mu, 2011a). For Chinese EFL learners, two reasons of disfluency in language production are alternate occurrence of unfilled pause and filled pause, and superfluous use of unfilled pause and non-lexical fillers (Yang Mu, 2011b). A large wide of students contend that grammar and vocabulary contribute to success in speech accuracy and complexity. However, pause, a problem in their speech production, is often neglected. Under this condition, theorical and empirical studies on pause is essential for research on learning and teaching.

1.2 Research purposes

With the birth of prosody research, intonation has been unearthed for hundreds of years. However, in China, there are not abundant articles or books about the study based on the correlation between Pause and Intonational Phrases (IP). Take Chinese TEM-4 Oral Test as an instance. Intonation is one of the factors listed in the holistic scoring criteria. However, an objective and detailed computerized intonation analysis provides a basis for scoring. As the study depicts, We are left without a clear notion of what linguistic features actually correspond to these general characteristics (Wennerstrom, 2000: 103), and hence more precise and quantitative research needs to be designed and carried out. This paper tries to analyze the pitch levels of intonation before or after speakers pauses, as pause is the criterion in the demarcation of intonation. At the same time, tonality, chunking, or division in IPs (Wells, 2006), helps us signal the structure of IPs and allows the author break the material up into IPs.

1.3 Organization of the study

This paper consists of five chapters.

Chapter One, the present chapter, functions as a guide for the whole paper. Based on respective parts: research background, research purposes and organization of the study, chapter one shows the current situation in English oral tests and unveil the aims of this paper.

Chapter Two offers Pause, definition of Intonation and development of relevant research. After the definition, systems and functions of EI, previous studies are provided to enrich this paper.

Chapter Three is research methodology as a whole, there research questions, participants, tools and measurements will be displayed.

Chapter Four opens the gate of results and discussion.

Chapter Five concludes the preceding analysis. It comprises of major findings, implications, limitations of this study and suggestions for future research.

2. Literature review

2.1Pause

Yang and Mu (2011a) mention that there is no accurate definition about pause and most researchers define it by pause time. By Raupach (1980), pause is the breaks within one sentence or between sentences that span 0.3 seconds or longer. And Zhang Wu (2001) adopt this method as the demarcation of pause. Lennon (1990, 405) groups pause into the filled and unfilled pauses. Cruttenden (1997) also notes that pause, unfilled pause and filled pause included, is the criterion when we distinguish the demarcation of intonation-phrases (IP). This kind of classification of pause routinely draws the attention of Chinese research in language production (Yang Mu, 2011a). At the same time, there exist too many pauses in English majors IP production; incorrect IP production can shift the focus of information, making expression monotonous (Yang, 2014). People need to make tonal phrasing (grouping) and breath, in which circumstance pause is inevitable. Thus, pause is of essence in oral language production.

Cruttenden (1997) concludes that pauses take place at three positions: on the boundary of main utterance constituents, before a high lexical word and after the first word in IP. The first sort can be taken advantage to mark the demarcation of EI and others are considered hesitancy. Yang Mu (2011a) also summarize that the false use of intonation boundaries is one of the reasons of pauses. Thus, due to marking pauses, the author can discover and more easily investigate EI.

2.2 English Intonation and relevant research

2.2.1 Definition of English Intonation

With regard to English Intonation (EI) s definition, there is no sole accurate explanation. Wennerstrom (2000) proposes that intonation is an integral component in judging speech fluency and prefers to put the use of pitch on lexical items in the list of English fluency. Intonation also involves the occurrence of recurring pitch patterns, each of which is used with a set of relatively consistent meanings (Cruttenden, 1997: 7). Wells (2006) believes that intonation is the pitchs ups and downs and the melody in language speaking. EI is also the pitch patterns of oral English (OConnor, 1973: 1) and refers to the use of suprasegmental phonetic features (Ladd, 2008: 4). Though there exist myriads of explanations of EI, the central concerns are pitch patterns and tunes of oral English as well as its musical features (OConnor, 1973: 1). The ups and downs of speech production and musical notes of voice is essential when speaking English. When it comes to prosodic features of intonation, pitch, as the most centrally involved variation (Cruttenden, 1997: 3), needs us to put it into research to a considerable extent. At the same time, the culmination points in pitch enable listeners to gain the most important information (Halliday et al., 2015).

2.2.2 Systems of English Intonation

When it comes to the systems of EI, in the realm of information, The Three Ts Theory cannot be neglected. Halliday (2015) gives a trio system: tonality, tonicity and tone. They are: how to break the material up into chunks, what is to be accented, and what tones are to be used (Wells. 2006). These systems are deemed three Ts theory by Wells (2006). It buttresses this paper and offers the theorical basis.

In the field of theoretical and applied linguistics, when researchers attempt to analyze the discourse and utterance, those tiny sentences need to be grouped into IPs. There is usually one second to two seconds to span each chunk. Apart from IP, there also exist grammatical boundaries to help people analyze syntactically. This process is called tonality (Wells, 2006: 6-7).

After breaking the whole conversation into IPs or chunks, the next step is to operate more detailed and complex analysis in each of its which usually are constituted by pre-head, head, nucleus and tail. Nucleus, lying at the end, is usually the last pitch accent involving tonal movement and always plays the most important role in conveying intonational meanings (Cruttenden, 1997: 49). Routinely there are one or two stressed syllables in each IP and the second one counted from the end is head. Pre-head is before the head and tail is located after the nucleus. This process is called tonicity (Wells, 2006: 7-9).

As the author has mentioned, the most principal correlation with intonation is pitch. Pitch, as a largely used prosodic feature, are acoustically reflected in fundamental frequency (the vibration rate of vocal folds in the larynx) which is measured in Hz. Moreover, to be practical, pitch and fundamental frequency is equated (Cruttenden, 1997: 2-4). Descriptions of fundamental frequency is depicted as its contours which are plots of frequency against time. All of these are more and more being put into great attention with the development of computational linguistic analysis. Pitch movements in the nucleus are able to shape the meanings of IPs syntactically and attitudinally according to various sentence types and pitch patterns. This process is called tone (Wells, 2006: 9-10).

2.2.3 Functions of English Intonation

With regard to the functions, although there are different taxonomies, Chun dissects it and groups into 4 dominant categories: grammatical function, attitudinal function, discourse function and sociolinguistic function (Chun, 2002: 47-67).

Grammatical function means that listeners can make clear of sentence types (statement, exclamation, command, question and so on) by distinguishing the final pitch accent and contours in one IP. Written English can be parted into sentences and clauses by punctuation while oral Englishs demarcations are signaled by intonation. This function is realized by tonality and tone.

Attitudinal function makes people aware of the emotion conveyed or hidden in speech (anger, sadness, exhilaration) by the signals of pitch range and contours. This function is realized by tone.

Discourse function refers to the textual meanings in conversation. Grammatical function helps researchers dissect the materials from the speaker or the listener and Discourse function is used as a tool to analyze the conversation between the speaker and the listener. The latter one tells individuals when to start or end discourse.

Sociolinguistic function aims to help researchers investigate the interactive part of intonation rather than just words and utterances.

2.2.4 Previous empirical studies on English Intonation

Though intonation has been unearthed for a long time, empirical studies on English intonation produced by learners whose mother tongue is Chinese is rare (Chen, 2010). Amid previous studies, as Chen (2010) points out, two groups can be parted: one is to draw a comparison between Chinese students performance and native speakers, aiming to discover problems in intonation systems, and the other is to directly show the characteristics and errors of Chinese learners spoken English.

In the first group, Chen (2006) compares oral English samples differences between 45 learners whose mother tongue is Mandarin Chinese and 8 British college students. On this basis, Chen summarizes some problems of intonational patterns usage. Huang (2008) analyzes Chinese learners intonation characteristics through an English declarative question You mean you are not confident that you can pass them, and compare materials belonging to 30 Chinese English majors and 5 English native speakers.

In the second group, Pan (1994) notes that educators who teach oral English find there exist problems in intonation when students speak English and that students cannot fully comprehend fundamental skills of EI. Hu (2012: 10) also summarizes distinct problems that Chines researchers find on Chinese EFL learners oral production, like the inability to select appropriate tone patterns, destitute of variation in EIs ups and downs and mechanically usage amid sentences. Bu (2016) concludes that the crux of lack of English fluency is intonation acquisition for Chinese learners. And Bu analyzes materials of 30 Chinese college students from tonality, tonicity and tone, i.e. grouping of intonation-phrases, placement of nucleus, stress of syllables and patterns of intonation.

Studies shows students problems in the realm of intonation, ranging from monotonous expression to solely or mistakenly usage of pitch patterns (Chen, 2002; Hong, 2012; Pan, 1994). Reasons can be parted into two dimensions (Pan, 1994). For one, students subjectively harbor that tone and tune do not impair abilities to speak fluently and that written examination is prioritized amid a large pool of examinations. For another, it is an uphill struggle for students to study and practice intonation which is embedded into a deeply subconscious level, thus allowing them opportunities to give up and circumvent those seemingly insurmountable and formidable obstacles.

Research on pause can be divided into characteristics (pause rates, positions, patterns) and functions (physiological, intercourse, intonational and planning function) (Yang Mu, 2011a). There exists a large pool of studies on pause. However, correlation of pause and intonation is scarce. Cruttenden (1997) concludes that pause can signal the demarcation of IP, and Wennerstrom (2000) reveal that inappropriate usage of intonation is one of reasons why pause can be generated. Yang Mu (2011a) summarize that pause in IP is the future of pause research. Thus, correlation of pause and intonation is a must. As pause is the outward criteria of EI, this paper attempts to investigate the intonation around pause produced by English majors in a test, aiming to offer more information to the hall of academia.

References

Chun, D. M. (2002).Discourse intonation in L2: From theory and research to practice(Vol. 1). Amstel dam: John Benjamins Publishing.

Cruttenden, A. (1997).Intonation. New York: Cambridge University Press.

Halliday, M. A. K. (2015).Intonation and grammar in British English. Berlin: De Gruyter Mouton.

Ladd, D. R. (2008).Intonational phonology. New York: Cambridge University Press.

Lennon, P. (1990). Investigating fluency in EFL: A quantitative approach. Language learning, 40(3), 387-417.

OConnor, J. D. (1973).Intonation of colloquial English. London: Longman.

Raupach, M. (1980). Temporal variables in first and second language speech production. Temporal variables in speech, 263270.

Riggenbach, H. (1991). Toward an understanding of fluency: A microanalysis of nonnative speaker conversations. Discourse processes, 14(4), 423-441.

Towell, R., Hawkins, R., Bazergui, N. (1996). The development of fluency in advanced learners of French. Applied linguistics, 17(1), 84-119.

Tench, P. (2015).The intonation systems of English. Manhattan: Bloomsbury Publishing.

Wells, J. C. (2006).English intonation PB and Audio CD: An introduction. Cambridge: Cambridge University Press.

Wennerstrom, A. (2000). The role of intonation in second language fluency. InPerspectives on fluency(pp. 102-127). Michigan: University of Michigan Press.

卜友红（2016），中国英语学习者语调习得问题研究，《外语教学与研究》，（4）：569-582 641.

陈春红（2010），对中国EFL学习者英语语调表态功能使用的研究，硕士学位论文。哈尔滨：哈尔滨工业大学。

陈桦（2006a），中国学生朗读口语中的英语调型特点研究，《现代外语》，（4）： 418-425。

陈文凯（2002），英语语音语调学习主要问题分析及对策，《河南教育学院学报: 哲学社会科学版》，（3）： 126-127。

洪薇（2012），中国学生英语语调格局实验研究，博士学位论文。天津：南开大学。

黄丽娜（2008），中国英语学习者陈述疑问句的语调研究，硕士学位论文。广东：广东外语外贸大学。

胡宇（2012），语境视角下对英语语调能力培养的实证研究，硕士学位论文。南宁：广西大学。

穆凤英、高薇、张云燕（2005），中国学生英语口语实词提取与生成特征分析,《外语教学与研究》，(4)：250-258 321。

潘华慧（1994），中国学生英语语调的错误及其纠正方法，《广西民族大学学报: 哲学社会科学版》，（2）： 29-33。

王立非、周丹丹（2004），我国英语口语研究12年：回顾与现状，《外语界》，（6）：7-14。

杨萌、穆凤英（2011 a），影响英语学习者口语流利性的停顿研究述评，《临沂大学学报》，（6）：60-64。

杨萌、穆凤英（2011 b），影响中国英语学习者口语流利性的语调短语内部停顿研究，《外语与外语教学》，（6）：16-21。

杨晓春（2014），英语专业学生语调短语产出不足及其对信息传递有效性的影响，《宜宾学院学报》，（10）：120-124。

张文忠（1999b），第二语言口语流利性发展的理论模式，《现代外语》，（2）。

张文忠、吴旭东（2001），第二语言口语流利性发展定量研究，《现代外语》，（4）：342-351 341。

3.1 Research questions

This paper tries to investigate the difference between high-fluency speakers and low-fluency speakers in the realm of intonation and pause, in an attempt to offer more enlightenment in oral English teaching.

In the study of Wennerstrom (2000, 102-127), high-fluency speakers are likely to apply different pitch levels appropriately whereas low-fluency speakers tend to use the relatively equal pitch no matter what is the function words or content words. At the same time, pause is the outward criteria of intonation. According to these two facts, the authors research questions are as the following:

1. Is there any significant difference in English Majors pause positions and frequencies in different levels of speakers?

2. Is there any significant difference in English Majors intonation pitch levels in different levels of speakers?

3.2 Participants

Participants are 101 English majors from one university. Due to the coronavirus pandemic, 51 sophomores and 50 juniors take part in an oral test together. Their native languages are all Chinese and there exists homogeneity in English learning. Pause and intonation for students needs to be analyzed.

3.3 Oral test

Materials is taken from a mock test before the TEM-4 Oral Test. This oral test contains a retelling, a talking on a given topic and a role play. Retelling requires the candidates to retell a story within 3 minutes after listening to it twice. Talking offers a topic related to the story in part one, and students have to prepare in 3 minutes and talk in another 3 minutes. Role play provided a topic whereas two candidates are needed. They have to play roles, preparing for 3 minutes and talk in 4 minutes. This paper aims to collect data from part one Retelling as it can fully show students ability of speaking fluently. In this research, the author invites two experts in phonetics to give marks respectively on 101 recording samples. Students are grouped into 3 levels by their test scores: 23 students low-level (0-59 points), 67 students mid-level (60-69 points), 11 students high-level (70-100 points).

3.4 Data collection

With restrictions from the recording device and the circumstance, Zotope RX 8 Audio Editor are used to eliminate noise after date has been collected. Praat (6.1.16), SPPAS (4.2) and SPSS (25) are applied to analyze data samples.

Data is collected in April, 2021 and the test lasts 30 minutes. The recording equipment are headphones with microphones and the place is school classroom. After students finishing the exam, the sound tracks of each one is collected. The next step is to denoise because of the restraints of recording gear and room. In this pre-examination, dozens of students sit together in a room and hence noise is inevitable.

Then materials are transcribed through Automated Speech Recognition (one software platform from the iFLYTEK corporation in China) and exported to texts. Manual modification is inevitable because there are some words and phrases that machine cannot recognize. uh is used to mark filled pauses and signals indistinct utterance.

Praat (6.1.16) provides researchers possibilities to analyze phonetically and for convenience the author applies Praat scripts to annotate. Because of the abundance of data collection, artificial annotation entails great efforts and time. Through automatic computational analysis and tiny corrections, the author is able to give every chunk of utterances a word like adding subtitles on videos. The next step is to annotate pauses using the method of equating or more than 0.3 seconds (Yang Mu, 2011a) and group sentences into IPs. With regard to pause, the first step is to be aware of pause type that include filled pause and unfilled pause. Pause lies at the boundary of main syntactic constituents, before a content word or after the first word in an IP (Cruttenden, 1997). The pause before a content word is the authors concern because it is deemed hesitancy and its proportion is higher than other pause positions (Mu et al., 2005). Pause patterns comprise clusters of dysfluencies, frequent unfilled pause and non-lexical fillers (Riggenbach, 1991; Yang Mu, 2011a), and this study will adopt the last two. SPPAS (4.2) enables the author to annotate unfilled silence automatically. After that, manual annotation of filled pause is requisite.

At the same time, different intonational patterns should be annotated. Intonation of one sentence is decided by nuclear tone that can be grouped into falls, rises, and fall-rises. Falls means a downward pitch movement from the mid or high-pitched nucleus to the low endpoint. Rises is the upward movement from a relatively low-pitched syllable to mid or high endpoint. Fall-rises means that the pitch begins relatively high and move downwards and then afterwards. The endpoint is low and the starting point is mid or high. In these sorts of tones, the syllables before nucleus (prenuclear tone) are disregarded (Wells, 2006).

Drawing and dissecting the picture of 101 students utterances helps the author analyze the occurrences of pause including filled and unfilled pause, and pitch changes of intonation for different levels of students. Finally, the data is put into tables of SPSS (25).