TONE LANGUAGE SPEAKERS’ ACQUISITION OF ENGLISH INTONATION英語專業(yè)畢業(yè)論文

TONE LANGUAGE SPEAKERS ACQUISITION OF ENGLISH INTONATION 1. Introduction While the English pronunciation of second language (henceforth L2) learners is a major area of research interest, the acquisition of English intonation has not been considered to a great extent. The previous studies in this area tended to focus on intonation language speakers acquiring other intonation languages or only on production. Those studies mainly found obvious influence from the learners first language (henceforth L1). There are surprisingly few studies concerning tone language speakers acquisition of English intonation. Therefore, this study will aim to present tone language speakers competence in English intonation, based on data from Mandarin English speakers and Cantonese English speakers perception and comprehension of English intonation. Every language has its own intonation system; certain correlations exist between intonation patterns with special meanings. Research tells us that the difficulties in the learning of L2 intonation patterns are often due to the non-equivalence of the intonation structure of the learners L1 and the L2 structure (e.g. Cruz-Ferreira, 1983). Mandarin and Cantonese are tone languages and English is an intonation language, and different use is made of the acoustic components of tone and intonation in these three languages. This paper is organized as follows. Section 2 will provide the theoretical background regarding tone and intonation in Mandarin, Cantonese, and English. Section 3 will present hypotheses, the methodology of this study and results of the tests. Section 4 extends the discussion including the data analysis of the four non-native English groups so as to examine the hypotheses. Section 5 concludes with a summary of the major findings of this study and limitations of this research. 2. Theoretical Background Tone refers to the pitch on the individual syllable and intonation refers to the height of pitch on the sentence level. Trask (1999) and Yip (2002) have suggested explanations for tone and intonation and they agree that tone and intonation are two parts of the same phonetic structure where the different distribution of pitch is the determining factor in distinguishing the two. However, tone only exists in tone languages and intonation exists at the sentence level in all languages. This section will start with an overview of Mandarin tones, particles and intonation and the relationship between tone and intonation. Section 2.2 will introduce the tones and intonation of Cantonese. Section 2.3 will briefly present the intonation system of English. 2.1. Mandarin tones and intonation Mandarin has five tones, including four contour tones and a neutral tone (Chao, 1957). As the different pitch on the word changes its core meaning, the tones are usually called lexical tones. The four contrastive tones in Mandarin are high level , high rising , low-dipping and high-falling which will be represented as T1, T2, T3 and T4, and toneless syllables (Chao, 1968). In linguistics, lexical tones are usually symbolized as well as numbered, which indicates the height of pitch of each tone. Chao (1957; 1968) put forward this numerical system and these measures have been used for decades. The pitch range of a normal speaking voice was divided into five levels, with 1 as the lowest and 5 as the highest. Zero refers to toneless. Most syllables are given two digits, one for the starting pitch and one for the ending pitch. For example: (1) m T1:55 mother; m T2:35 linen m T3:214 horse; m T4:51 curse ma 0 (toneless, one of the neutral tones) particle for question The toneless syllables are another major aspect in the consideration of intonation. Although there is not a large number of toneless syllables in Mandarin, they are used frequently. In studying colloquial speech, it has been found that there is one toneless syllable in each five or seven syllables (Guo, 1993). These are mainly particles and suffixes. In Mandarin particles and suffixes play an important grammatical function as well as an expressive function. Yip (2002) suggested that the sentence-final particles are used to avoid conflicts between the lexical tone of the last syllable and intonation. In other words, the particle takes its pitch partly for realizing the intonation of the sentence. Take two particles of simple questions for example, (2) a. Mary sh xu shng ma? Mary is student (particle)? Is Mary a student? b. Mary sh xu shng ba? Mary is student (particle) ? (I think that) Mary is a student, isnt she?The particles in the above sentences indicate that they will lead to yes/no answers. However, in sentence (a), the particle implies that the speaker slightly doubts about the affirmative answer and the probabilities will be around 50% or less; in sentence (b), it implies that the speaker supposes that Mary is a student and will get an affirmative answer as he/she expected. Comparing the intonation of the two particles, ma usually realizes a bit higher pitch in sentence (a) than ba does in sentence (b), although both of them make the sentence a rising ending. Regarding the classification of intonation patterns in Mandarin, linguists have not reached a common standpoint. The traditional classification suggested that there are three patterns: falling for statements, imperatives, interjectional expressions and wh-questions; rising for interrogative expressions, incomplete sentences, questions with particles and sentences with implications; the mixture of rising and falling to express exaggeration, satire, humor and surprise (Li, 1956; Shi, 1980). Besides these three, there is an extra intonation level - which is used for utterances of thought, recitation and hesitation (Huang, 1956; Jiang and Yin, 1957). On the point of attitudinal function, Guo (1993) analyzed five intonation patterns. Based on all ideas, there are at least three intonation patterns in Mandarin which have been agreed: falling, rising and mixture of falling and rising. Falling and rising are the most common patterns, which mostly occur at the end of sentences. A mixture of falling and rising usually occurs in long sentences or in a sentence with implications. However there are some controversial statements concerning the relationship between lexical tone and intonation. Chao (1968) suggested that tone and intonation have an impact on expressions at the same time. Hu (1987) hypothesized that intonation is an independent phonetic phenomenon beyond the lexical tone， which was shown by the falling intonation at the end of the utterance in Beijing dialects. Xu (1980) suggested that intonation is based on lexical tone unless the pitch appears on the level of extra high. This was investigated through the shouting of statement/question sentences, with the final syllables tone changing into T1:55. It is also suggested that in tone languages, specific tones and overall contour tones are modified to define certain intonation messages and the intonation phenomenon is reported as being superimposed on the lexical tone (Ladd, 1998; Cruttenden, 1997). For instance，the average pitch of each lexical tone in questions is generally discovered to be higher than that of the tones in statement(Chao, 1986; Shen, 1985; Shen, 1989). 2.2. Cantonese tones and intonation Cantonese is one of the Chinese Yue dialects. As Hong Kong Cantonese proceeds from the Cantonese of Canton, its accent is similar to Cantons (Zhang, 1972). Cantonese has 6 basic tones: high-level/falling, middle-rising, middle-level, low-falling, low-rising and low-level (Zeng, 1988). Yet, sometimes high-level/falling is regarded as two tones: high-level and high-falling, in which case Cantonese has 7 basic tones (Yip, 2002; Guo, 1993). In addition to the basic tones, there are another three tones, which have same pitch range of high-level, middle-level and low-level, but end with /p/, /t/ and /k/ (Zeng, 1988). One can conclude that there are 10 tones at most in Cantonese. Taking one syllable si for an example, it refers to different meaning when it receives ten tones respectively. In the example, the number notions used for pitch range follow Mandarins, 1 for the lowest pitch and 5 for the highest. For example: (3) T1: si: 55 poem T6: si: 24 market, cityT2: si: 44 to try, taste T7: si: 53 silkT3: si: 33 affair, undertaking T8: sik: 5 style, typeT4: si: 22/21 time T9: sip: 4 to wedgeT5: si: 35 to cause, make T10: sik: 3 to eat (Yip, 2002) When words come together as compounds or in sentences, most syllables will keep their tones unchanged. Importantly, as one of the Yue dialects, Cantonese has the last tonal phenomenon of interest, which is called changed tone. Yip (2002) suggested that the last syllable usually starts at the starting point of the base tone and ends high. For instance, Cantonese often attaches a prefix a to a surname，turning it into a nickname. For example: (4) tsan: T4:22 surnamea: T3: 33 (prefix) + tsan: T4: 22 (surname) a: T3: 33 tsan:T4: 25 In this example, the basic tone of tsan is T4:22. When it is combined with the prefix a, its tone is changed into T4: 25. Intonation in Cantonese mainly refers to the intonation at the end of sentences where there are three patterns: rising, falling and level (or non-intonation) (Mai, 1989). All three patterns will be realized on the last syllable of the sentence. Rising is usually used in questions, resulting in the pitch of the last syllable rising to a very high level; falling is usually used in confirmation, command, or to express unhappiness and praise. Level will not bring any change to the original tones of syllables, and is used in narration and statement. Regarding the final intonation of declaratives, interrogatives, yes/no questions and echo question, Flynn (2001) found that the falling trend appears in declaratives as well as interrogatives which contain a question word, such as wh- word in English. On the other hand, rising tail is used to express query, doubt or suspicion. Besides echo questions, this can also signal declaratives with attitudes of ironic, confirmative, or impatience. As in Mandarin, particles play a very important role in Cantonese expressions. Cantonese has 35 to 40 particles and almost every Cantonese utterance ends with a particle (Yip, 2002). As Cantonese bears around 10 lexical tones, the toneless carriers for intonation are obviously functional. If the sentence does not need a particle semantically, a speaker tends to place /a/ at the end- if /a/ is high, the utterance invites the hearer to agree to the action proposed; if /a/ is low, the utterance has an air of impatience and boredom (Yip, 2002). Flynn (2001) also hypothesized that the particle is often attached to a grammatical unit to coincide with the end of the intonation pattern. The flexible point is that “a slightly different pronunciation could be more often aligned to another meaning”, which makes Cantonese speech prosodicly unique and complicated. 2.3. English Compared to Mandarin and Cantonese, English makes more use of intonation patterns and the usages of the intonation are more flexible and variable. Halliday (1967) proposed five primary tones for intonation: 1. falling; 2. high rising; 3. low rising; 4. falling-rising; 5.rising-falling and analyzed each intonation with descriptive meaning, such as rising-falling typically for assertions, high rising for seeking information or challenging a statement. Crystal (2003) displayed 9 ways of saying yes and suggested the 9 types of intonation, including level, falling, high-falling, low-falling, rising, high-rising, low-rising, falling-rising and rising-falling. However, to display the diversity of intonation and the wide range of meaning which the intonation can convey, OConnor and Arnold (1973) illustrated ten patterns of intonation with contrastive examples for each pattern. Each pattern is discussed in terms of the attitudinal functions based on grammatical classification. The following examples only cover the different ranges of meanings in statements which two of ten patterns can convey respectively. (5) Low drop with ending of low falling, which conveys the meaning of “weighty,judicial, considered.” For example, A: Have you any news of Malcolm?B: Hes passed his exam. (6) High drop with ending of high falling, which indicates “a sense of involvement, light, airy”. For example, A: Where on earth are my slippers? B: I cant think whats happened to them. 3. The Study 3.1 Hypotheses The present study tested three hypotheses concerning the expected results of speakers of Mandarin and Cantonese learning English: 1. Although there are far fewer intonation patterns in Mandarin and Cantonese, listeners will be able to perceive English intonation. 2. Both Mandarin and Cantonese speakers comprehension of the meanings English intonation patterns represent will depend on the similarities of the intonation patterns between their L1 and English. 3. Participants who have had more exposure to native English input will be more competence than their counterparts in English intonation. 3.2. Methodology There were a total of 59 adults taking part in this study, 5 of which were in the control group and 54 were in 4 non-native speaker groups. The non-native speakers English level was measured by their scores of an international English test which they took before they came to Britain as students. Thirty were from areas of mainland China and all of them are mandarin speakers. They were separated into two groups depending on their exposure to native English: Current Students and New Students. Current Students were postgraduate students and had been in the UK around 11 months when they took part in the tests. New Students were students on a pre-sessional English course who expected to start an MA after finishing the language course. The remaining 24 were from Hong Kong (henceforth HK) and they were studying BA degree in Education in HK. When they took the tests, they were taking summer courses and had been in UK for one month. They were also divided into two groups depending on their exposures to English in HK. Most participants in the HK Group A went to English school or attended English classes after school time while participants of HK Group B went to Chinese school and had no English exposures expect the English class time. Generally, the HK speakers had more exposure to native speakers than those in mainland China. Table 1. Participants by GroupsNative SpeakersCurrent StudentsNew StudentsHK Group AHK Group B TotalNumbers51515121259English ExposureN/A10-year school+ 1year in UK10-year schoolEnglish in daily lifeNo English after classN/AThe participants were asked to take a 40-minute listening test, which consisted of 30 pairs of sentences. The sentences in each pair used same expression but adopted different intonation. The test was designed by Cruz-Ferreira (1983) and was especially used for the research into non-native English speakers comprehension of English intonation patterns. But several changes were made in answer sheets. The original answer sheets are as following: Pair No. xx1. The sentences have the same meaning. ( ) The sentences have different meanings. ( ) 2. The sentences have the same meaning. The meaning of both sentences is:meaning X ( ) meaning Y ( )3. The sentences have different meanings. The meaning of each sentence is:meaning X ( )meaning Y ( )Three parts for one question seem complicated and redundant, which could affect listeners answering question. Thus three parts were reduced to two parts in the present study. Additionally, in the part of matching meaning, writing down Sentence A and B to match Meaning X and Y is not an appropriated format to ask listeners giving answers. Sentence A and B will be played from recorder and the choices X and Y will be in paraphrase sheets. When listeners were doing tests, it requests them to do several things at the same time, listening, reading, recalling what they heard, making decision and do matching. As listening is a kind of temporary resource to the listeners, they would be possible to mix the order of the Sentence A and B or cant remember the difference between A and B after they read the choices sheets. Therefore, listeners can do the same tasks much easier if they are requested to match Sentence A and B to choices X and Y. When they listened to the tape, they were requested to answer two questions for each pair, the first is whether the two sentences (Sentence A and Sentence B) have same meaning and the second question is to match the sentence meaning (X or Y) from the prepared choice sheets to Sentence A and B, writing the answer on the answer sheets. For example: (7) Pair 4. Listening parts: A. Are you sure youve brought with you. B. Are you sure youve brought with you. The prepared choice sheets: X. The speaker is asking a question and is not sure of the answer, which could be either Yes or No. Y. The speaker is asking a question and expects the answer No. The answer sheets: (1) The sentences have the same meaning. ( ) The sentences have different meanings. ( ) (2) Sentence A meaning ( ) Sentence B meaning ( ) The answer of (1) will display whether listeners are able to perceive the different intonation patterns, and the answer of (2) will display whether listeners are able to comprehend the meanings expressed. 3.3. Results As the answers for each question include two parts - perception (Per) answers and comprehension (Com) answers, Table 2 provides the general results for the two parts respectively among the five groups. Table 2. Test results: GroupsAnswersNative Speakers5 participants135 sentencesCurrent Students15 participants405 sentencesNew Students15 participants405 sentencesHK Group A12participants324 sentencesHK Group B 12participants324 sentencesPer answers120353302264227Rate89%87%75%81%70%Comanswers117244170197141Rate87%60%42%61%44%Native Speakers obtained 89% for Per answers and 87% for Com answers. There could be many reasons for that native speakers failed in obtaining 100% in the test. One of the possible reasons could be the different English background of 5 participants: three of them growing up in southern England, one in the north and the other one in the east. Besides this, further investigation will be needed. Anyway, in the latter discussion of the study, the rate of Control group will be regarded as the measure for the non-native speakers groups. In other words, all discussions will be based on Control groups rate. In Per answers, Current Students reached 87%, which was only 2% less than Native Speakers. HK group A was much better than New Students, 81% to75%. HK group B obtained the lowest rate, only 70%. Current Students was 12% higher than New Students and HK group A was 11% higher than HK group B. Thus, Mandarin speakers were better than HK Cantonese speakers in Per answers. The situation changed in the comparisons of Com answers among the five groups. Native Speakers obtained 26% higher than the best group of the non-native speaker groups, 61% of HK group A which was 1% higher than Current Students. HK group B was also 2% higher than New Students. Thus, the HK groups were 3% better than the Mandarin groups in comprehension section. On the other hand, HK Group A obtained 17% higher rate than HK Group B an