A Critique on the application of Praat in comparing the intonation of Chinese EFL speakers and native speakers

Abstract
Acoustic features such as intonation are attached great importance by all English learners around the world, so as the Chinese EFL speakers. The study mainly conducts an experiment that collects the intonation data via speech visualization software Praat when both the Chinese EFL speakers and native speakers read the four sentence patterns, namely declarative sentences, interrogative sentences, imperative sentences and exclamatory sentences. In order to control the variables, all participants are female aged from 20 to 30, in which all the Chinese EFL speakers all achieve seven points in IELTS speaking test. The result shows that all these Chinese EFL speakers possess the awareness of intonation variation, but the detailed intonation change is still quite the different from the native speakers. Furthermore, Praat is a reliable and valid software that can benefit people who want to improve intonation, but it could be more perfect if auto-correction function is designed from the author’s perspective.
Keywords: intonation, pitch, Praat, speech visualization software

1. Introduction
Under the unprecedented trend of globalization, English undeniably becomes a must for people, especially the younger generations around the world. Different from the past, people should not only grasp the skill of reading, writing and listening in English, but more importantly, become a master in speaking English. During the pronunciation or intonation practice, English learners are struggling with accent problem due to the negative transfer of the mother tongue. Thus, nowadays, except for the traditional method of correcting pronunciation or intonation such as shadowing, computer-aided pronunciation tools also leap to the non-native speakers’ sight. 
1.1 Research Background
According to Oxford online dictionary, intonation refers to the rise and fall of the voice in speaking, especially as this affects the meaning of what is being said. During a face-to-face communication, people can catch effective message from the content of words, the speakers’ intonation, body language, facial expressions and so forth, in which the sentence’s meaning can be totally reversed in different intonation. Thus, as an intonation language, the pitch, loudness, speech rate and pause of colloquial English can be regarded as the indispensable prerequisite of a successful conversation. English intonation mainly possesses four functions, namely attitudinal function, discourse function, grammatical function and accentual function . Decent and euphonic intonation can help expresser convey information in a more accurate and efficient way.
China is a profound country with splendid history and large population, where numerous people learn English as their second language from an early age. Unlike the exam factory in the past, Chinese people currently pay more attention to spoken English with the mainstream guidance of quality-oriented education. Nevertheless, English and Chinese belong to different language system, namely Indo-European and Sino-Tibetan, which brings Chinese EFL (English as a Foreign Language) learners considerable difficulties in second language intonation acquisition.
1.2 Purpose of the Research
Firstly, the essay aims to find out the intonation difference between Chinese EFL speakers and native speakers with the assistance of speech visualization software. Via visual intonation comparison, second language learners may recognize their problem more intuitively, so that they could improve their phonetic appearance through adjusting their pitch, intensity, stress, pause and so on. In this way, Chinese EFL speakers can imitate the intonation and pronunciation of native speakers consciously. Hence, they might recognize the regular pattern of native intonation and draw inferences about other cases from one instance. 
The second purpose of the study is to critique the application of computer-aided intonation tools in phonetic analysis. The author wants to detect whether the software is reliable or feasible for second language learners, and distinguishes which function of the software is designed reasonably from those needs to be improved. 
1.3 Significance of the Research
According to the British linguist Roger Kingdon, intonation is the spirit of language . To some extent, second language learners may not realize their problems in intonation unless they listen to the same sentence or speech uttered by native speakers. It is utmost important for EFL learners to be aware of the place where the word should be stressed and where should it be paused, instead of listening to the tape over and over again but fail to catch the intonation pattern of the sentence. Only acquiring the general intonation knowledge, such as interrogative sentence should be uttered in a rising tone and declarative sentence should be expressed in a falling tone, is far from enough. In order to be native, second language learners should follow the intonation rules in each segment including pre-head, head, nucleus and tail.
Furthermore, the critique about the computer-aided intonation tools is meaningful for both users and designers. As for software users, they can understand the function of the technical tool in a board picture; while from the perspective of software designers, they can get the feedback from the customers and upgrade the software to meet customer satisfaction. Hence, it is a win-win cooperation for both sides. 
2. Literature Review
The number of research about intonation is not in the minority. Authors who show interests in intonation tend to explore the aspects in terms of phonology teaching, intonation pattern, stressed syllable, tonicity, tone group and so forth. Scholars now not only focus on theoretical research phonetics, but also its application in teaching or studying English like computer assisted language learning.
2.1 Research in western countries
The research of intonation in the western world can be traced back to 16 century. Over hundreds years of discussion and debate, scholars reached a consensus that intonation is a suprasegmental feature which all language prossess and embodied by the pitch variation. Halliday, the representative personage of western linguistics, pointed out that intonation can be divided into three levels, namely tonality, tonicity and tone (the three Ts) . To be specific, tonality refers to division of speech into tone groups or intonation phrases; tonicity is the placement of tonic and nucleus; and tone means different melodic contours. 
2.2 Research in China
In recent years, many Chinese scholars also attach great importance to English intonation, in which intonation teaching account for a large proportion. They tend to conduct empirical studies such as doing case study to test whether learners can improve their intonation appearance via computer assisted language learning. Furthermore, scholars find out that most of the research about Chinese EFL learners still remains on the descriptive level, but lacks of the support and analysis of experimental data and linguistic systematic knowledge. Although Chinese government and educators emphasize the important of intonation repetitively, there are still two negative tendencies of oral English learning: first, teachers mainly impart English knowledge but ignore intonation teaching; second, teachers only focus on the pronunciation of single words instead of the intonation of the whole sentence.
Nevertheless, it is rare to seek out the research on intonation visualization software no matter in China or the western country. Researchers usually conduct empirical study under the assistance of visual software without assessing or criticizing the software itself. In this study, the author will critique the application of a computer assisted intonation software based on a case study.
3. Experimental Study
    Empirical study refers to a kind of research method that the author collects, observes and analyzes data in an attempt to test the correctness of his/her hypothesis. In the narrow sense, the empirical research method is that analyzing the interaction and quantitative relationship between relevant elements through quantitative analysis technology. As for a generalized perspective, the empirical research method regards practice as the research starting point and experience as the basis of science. Thus, people generally classified empirical study into two main categories, namely mathematical empirical research and case-based empirical research . In this chapter, the author aims to give a comprehensive introduction to the experimental study from three aspects, namely research questions, methodological tool and materials as well as participants. 
3.1 Research Questions
This experimental study is implemented to figure out the following questions:
Firstly, what are the differences and similarities between Chinese EFL learners’ and native speakers’ intonation in declarative sentences, interrogative sentences, imperative sentences and exclamatory sentences?
Secondly, how can the Chinese EFL learners improve their intonation based on the experimental result?
Thirdly, what are the merits and demerits of the computer assisted intonation software in this case study?
3.2 Methodological Tool and Materials
With the rapid development of technology, it is not difficult for programming personnel to design sound visualization software, such as Better Accent Tutor, Praat (doing phonetics by computer) and so on, for the purpose of teaching and learning intonation and pronunciation. In this study, the author chooses software Praat as methodological tool. Praat is a speech visualization program designed by professor Paul Boersma and assistant professor David Weenink at the department of Phonetics, the University of Amsterdam, and can be downloaded freely on the internet. Praat is a multi-functional software with easy-to-operate interface. It can be used to analyze the spectrum, pitch, intensity, formant and pulses of the voice in a pictorial and statistic way. In addition, Praat has been widely applied in many emerging fields such as speech or sound synthetic, linguistic data statistics and analysis as well as teaching assisted phonetic test. Since the main focus of this task is intonation analysis, parameter about pitch such as minimum pitch, maximum pitch and average pitch in the pitch contour draws more attention from the author.
In order to compare the Chinese EFL speakers’ and native speakers’ intonation in a multi-dimensional way, the author choose four sentences as the sample this task, in which each sentence belongs to different sentence category: declarative sentences, interrogative sentence, imperative sentences and exclamatory sentences. The author chooses the sentences from the book Teaching Pronunciation──A Reference for Teachers of English to Speakers of Other Languages published by Cambridge University Press and written by Marianne Celce-Murcia, Donna M. Brinton and Janet M. Goodwin. The major reason why the author chooses this book is that Teaching Pronunciation is the core of a comprehensive course in pronunciation pedagogy designed to provide ESL/EFL teachers with a thorough grounding of sound system, including both segmental and suprasegmental aspects, as well as abundant corresponding exercise for students to practice. The author selects four sentences from Teaching Pronunciation, namely “John, we are waiting for you.” (declarative sentence), “Did you cook dinner after you got home?”  (interrogative sentence), “Just drop me off the corner, please.” (imperative sentence), and “How lovely and sweet she is!” (exclamatory sentences). These sentences are selected due to its dailiness and typical.
3.3 Participants
The participants of this study are composed by five Chinese EFL speakers and five native speakers in the United Kingdom. In order to avoid the variables in gender and age, all the participants the author chose are female from 20 to 30. In addition, all participants are required to record the sentences in a quiet room with resonant voice to ensure the software could recognize the voice clearly. 
These five female Chinese EFL speakers all got seven points in IELTS speaking test, which means that their pronunciation of words are articulate and easy to understand. According to IELTS score criteria, Chinese participants’ ability is relatively close to native speaker from the perspective of fluency and coherence, lexical resource, grammar and pronunciation, thus, it is reasonable to believe that apart from intonation, Chinese and native participants’ possess basically the same expression ability. As for native speakers, they are university students in Wales where boasts a profound history of speaking English as its official language. 
4. Findings 
    After collecting data from all participants, the author imports data to the software Praat separately according to the sentence pattern. Apart from Praat, the author also needs to analyze the data under the assistance of EXCEL. The main focus in this chapter is to find out the intonation similarities and differences via the pitch number and pitch contour.
4.1 Findings in declarative sentences
Declarative sentence refers to a kind of sentence pattern that states a fact or declares one’s view. It usually ends with a full stop and should be read in a falling tone. The example the author chose in this sentence pattern is “John, we are waiting for you.”, which is commonly used in daily life. The pitch and intensity parameters and pitch contour are shown as follows:
Participants	Mean Pitch (Hz)	Mean Pitch Range (Hz)	Intensity Average (dB)
Native Speakers	218.79	109.13	71.04
Chinese EFL speakers	212.95	137.20	65.89
Table 4.1 Declarative sentence’s parameters
According to the statistics in table 4.1, the mean pitch and intensity average of native speakers is higher than that of Chinese EFL speakers, while the data in mean pitch range is exactly the opposite. The mean pitch range can be regarded as an epitome of intonation fluctuation, since the intonation variation is caused by the change between maximum pitch and minimum pitch. The statistics shown is mean pitch range represents that Chinese EFL speakers with seven points in IELTS speaking test possess the awareness of intonation variation. Nevertheless, the differences of intonation between Chinese EFL speakers and native speakers will occur when the pitch contour was printed out.
 
Figure 4.1-1 The Pitch Contour of a Native Speaker
 
Figure 4.1-2 The Pitch Contour of a Chinese EFL speaker
    Although the author only makes one pitch contour figure from each side, many typical and representative differences can still be analyzed from the two pictures, since similarities are contained and hided among the two sides. The two figures above reveal the pitch variation of single words when speakers utter the sentence, which demonstrates how the native speakers and Chinese EFL speakers deal with the same word in a different intonation. Starting with the first word “John”, the native speakers chose to utter it in the falling-rising tone, while the Chinese EFL speakers express it in the rising-falling tone. As for “waiting”, the key word in this sentence, the native speakers pronounce the syllable “wai” in the level tone, and improve the pitch at the head of syllable “ting” and then express it in the falling tone. However, the Chinese EFL speakers raise the pitch at the syllable “wai”, and fall down the pitch at the syllable “ting”. Furthermore, the last two words “for you” were expressed relatively in a level tone, but in a falling tone from the Chinese EFL speakers.
4.2 Findings in interrogative sentences
The main social function of interrogative sentences is to raise a question or ask for information. It usually ends with a question mark and should be expressed in a rising tone in most cases. The sample sentence in this part is “Did you cook dinner after you got home?” After data analysis, the pitch and intensity parameters and pitch contour are shown as follows:
Participants	Mean Pitch (Hz)	Mean Pitch Range (Hz)	Intensity Average (dB)
Native Speakers	217.75	141.42	69.60
Chinese EFL speakers	212.97	139.40	65.51
Table 4.2 Interrogative sentence’s parameters
Compare with the data in table 4.1, the number of the mean pitch and the intensity average is relatively the same in table 4.2. With regard to mean pitch range, both native speakers and Chinese EFL speakers utter the interrogative sentences in a more up-and-down way than that of declarative sentences. In this case, the mean pitch range of native speakers is a little higher than that of Chinese EFL speakers, which shows that the Chinese EFL speakers who gain seven points in IELTS speaking test are basically close to native speaker from the perspective of pitch range. However, it is not reliable to infer that Chinese high-level EFL speakers can utter interrogative sentences comparatively the same as native speakers. Specific pitch contour are produced as follows:
 
Figure 4.2-1 The Pitch Contour of a Native Speaker
 
Figure 4.2-2 The Pitch Contour of a Chinese EFL speaker
    As above, the author chooses to pick up two speakers’ voice severally from each side. Intuitively, the variation of pitch wave is much more obvious than that of declarative sentences in this case. To be specific, both the native speakers and Chinese EFL speakers begin with the first three words, “did you cook”, in a rising tone, but differences arise in the rest of the sentence. With regard to the word “dinner”, native speakers read it in a falling-rising tone, while Chinese EFL speakers in a falling tone. The most significant difference implies in the last word of the sentence. Despite both native speakers and Chinese EFL speakers uptrend the intonation of the word “home”, it is evident that the slope of native speakers is much steeper than that of Chinese EFL speakers, which means that the intonation variation in the tail of interrogative sentences is far from adequate for Chinese EFL speakers. 
4.3 Findings in imperative sentences
Imperative sentence is a sentence pattern that can be used to express command, request, advice, warning, prohibition and so forth. In this part, the author selects “just drop me off the corner please.” as sample sentence. Like the two sentence patterns above, the pitch and intensity parameters and pitch contour are shown as follows:
Participants	Mean Pitch (Hz)	Mean Pitch Range (Hz)	Intensity Average (dB)
Native Speakers	215.06	142.43	70.26
Chinese EFL speakers	221.80	137.62	65.37
Table 4.3 Imperative sentence’s parameters
In this part, Chinese EFL speakers’ mean pitch is higher than that of native speakers, while lower in mean pitch range and intensity average. The fluctuation of pitch in imperative sentences is relatively close to interrogative sentences, but still much larger than that in declarative sentences. The pitch range of Chinese EFL speakers who acquire seven points in IELTS speaking test is not considerable, gaping only 4.81Hz with the native speakers. The author also needs to draw the concrete pitch contour to implement a further analysis. 
 
Figure 4.3-1 The Pitch Contour of a Native Speaker

 
Figure 4.3-2 The Pitch Contour of a Chinese EFL speaker
    Based on the two contours above, similarities and differences co-exist between two groups of participants. Both sides read the sentence in a falling-rising tone in the whole picture, but the intonation of each word is disparate. For instance, the word “drop” is uttered in a falling tone with extremely steep slope by native speakers; while for Chinese EFL speakers, the intonation variation of the word “drop” can be rarely seen. In addition, the pitch contour of the word “please” is continuous in figure 4.3-1, but intermittent in figure 4.3-2. The discontinuous pitch contour shows that the frequency of sound is slow and uneven at that point. Thus, according to the native speakers, the word “please” should be expressed in a gently rising tone, instead of raising the sound of tail all of a sudden.
4.4 Findings in exclamatory sentences
Exclamatory sentence is a type of sentence pattern that expresses strong emotional feeling such as pleasure, anger, sorrowful, joy and so on, and normally ends with exclamation point. In this part, the author selects the sentence “how lovely and sweet she is!” as experimental material. As above, the pitch and intensity parameters and pitch contour are necessary to make a detailed analysis.
Participants	Mean Pitch (Hz)	Mean Pitch Range (Hz)	Intensity Average (dB)
Native Speakers	228.18	169.05	70.67
Chinese EFL speakers	231.98	144.95	66.57
Table 4.4 Exclamatory sentence’s parameters
In this section, the mean pitch range and intensity average of the native speakers are higher than that of the Chinese adept EFL speakers, while the result of the mean pitch is reverse. The exclamatory sentence’s mean pitch range is the highest among the four sentence patterns both for the native speakers as well as the Chinese EFL speakers. Additionally, the difference value of the mean pitch range reaches 24.1Hz between both sides, which is also the most considerable gap in this whole experiment. The selected pitch contours are shown as follows:
 
Figure 4.4-1 The Pitch Contour of a Native Speaker
 
Figure 4.4-2 The Pitch Contour of a Chinese EFL speaker
Admittedly, both native speakers and Chinese EFL speakers read the sentence in accordance with the emotional requirement of exclamatory sentences, since the contour fluctuation is quite distinct in the pictures. However, the detailed differences from the separate word level cannot be ignored. At the beginning of the sentence, the native speaker’ tone tend to be flat, but the Chinese EFL speaker rise the tone, especially when comes to the word “lovely”. The second intonation difference emerges in the word “sweet”. To be specific, the native speaker utters it with a sharp falling tone, whereas the Chinese EFL speaker with a slow rising tones. Furthermore, at the tail of the sentence, the native speaker ends it without apparent intonation change, but the Chinese EFL speaker utters it with a falling tone. 
To sum up, in terms of the number of pitch range alone, there is little difference between the Chinese EFL speakers with seven points in IELTS speaking test and the native speakers. However, compared from the perspective of pitch contour, the tone may entirely different, or even opposite when both sides deal with the same word in the same context. Therefore, it is reasonable to conclude that those Chinese EFL speakers who acquire relatively high proficiency in oral English possess the awareness of intonation variation, but still fails to catch the accurate tone change of every single word.
5. Critique about the application of Praat
It is worth mentioning that there is an increasingly number of scholars conduct their experiments with the aid of visible sound software, but the critique about the application of these softwares are rarely seen. In this part, the author will unfold her critique on Praat based on the experiment clarified in the chapter three and four. 
Undoubtedly, Praat is not the only software that specified in phonetics. Software like Speech Analyzer, BetterAccent Tutor and SFS also catch the attention of researchers in this field. According to many researchers’ view in their literature, Praat boasts many outstanding properties such as free to download, strong portability, easy-to-operate interfaces, small memory space and so forth. It can be used to analyze pitch, intensity, formant, spectrum, pulses, etc. As for Speech Analyzer, the main function of it contains converting sound file through international phonetic system, replaying the record in a lower speed, analyzing musical sound and so on . With regard to BetterAccent Tutor, it is speech analysis technology that focuses on intonation, stress and rhythm, and can be used to identify, understand and correct the speakers’ pronunciation errors compared to American English pronunciation . Nevertheless, the major shortcoming of this software is that users can only download a free demo from the official website, and the integrated software should be paid to install. SFS is also a recommended software, since apart from general functions, it can be applied to transcribe speech signals, analyze formant and the like .
Based on the comparative analysis above, Praat can been regard as a convenient sound visualization software for its multi-function and money-consuming. The application of Praat can benefit all the people who are eager to improve their intonation through comparing visualized pitch contours. However, deficiencies cannot be neglected in the process of application. Take the experiment in this paper as an example, the Chinese EFL speakers indeed could find out the intonation difference compared to the native speakers, but the software cannot correct or modify the voice needed to be improved. In the author’s opinion, it is better for the software to provide an auto-correction sound function, so that the learners could listen attentively what the standard intonation should be in his/her own voice, and imitate the sentence in a much easier way.
6. Conclusion
The study is about a critique on the application of Praat in comparing the intonation of relatively high-level Chinese EFL speakers and native speakers. The findings of this experiment show that despite these Chinese EFL speakers pay attention to intonation variation in reading sentences, inaccurate tone still occur frequently in detail. Admittedly, there are some inadequacies in this study, including the limited participant number and sentence examples, which should be improve in the future study. In addition, Praat is the tool to intuitively visualize the sound into numbers, acoustic wave, pitch contours and so forth, which provides significant assistance to sound analysis. Finally, compared to other speech visualization softwares, Praat is by no means inferior. Nevertheless, from the learners’ perspective, customer experiment might be more excellent if auto-correction sound function is designed and added in the interface.

References
(i)  Bahman Gorjian, Abdolmajid Hayati, Parisa Pourkhoni, ‘Using Praat Software In Teaching Prosodic Features To EFL Learners’, Procedia - Social and Behavioral Sciences, 84 ( 2013 ), 34 – 40.
(ii)  Chen Chunhong, ‘A Study of Chinese EFL learners’ use of the Attitudinal Function of English Intonation’, (master’s thesis, Harbin Institute of Technology, 2010). 
(iii)  Cao Ruilan, ‘Improving English Pronunciation Teaching and Learning via Speech Corpora of Learners with Dialectal Backgrounds’, International Journal of Emerging Technologies in Learning, 11(2016),90-94.
(iv)  Cai Ting, ‘An Empirical Stduy on Instructional Strategies of English Intonation in Junior High School’, (master’s thesis, Hangzhou Normal University, 2018).
(v)  Naima Hamlaoui, Nawel Bengrait, ‘Using BetterAccent Tutor and Praat for Learning English Intonation’, Arab World English Journal, 3(2016), 99-112.
(vi)  Zeng Yao, ‘An Analysis of English Intonation Features of Chinese Junior High School English Teachers─A Case Study of the Teacher Training School of Qingpu District in Shanghai’, (master’s thesis, Shanghai Normal University, 2017). (In Chinese)
(vii)  Zhang Yufei, ‘An Experimental Study of Teaching English Intonation to Junior-high Students with Praat’, (master’s thesis, Central China Normal University, 2017). 

(STATEMENT: Sorry, the pictures/figures in this essay cannot be read by the system.)