an investigation of the acquisition of linking by chinese...

HKBU Papers in Applied Language Studies Vol. 15, 2011

An Investigation of the Acquisition of Linking

by Chinese EFL Learners1

ZHANG Yanyan Wuhan University

Abstract

This paper aims to investigate the acquisition of linking by Chinese EFL learners, with a focus on three major types of linking in English, namely, (1) C-V linking (consonant-vowel linking, e.g. a cup of tea), (2) V-V linking (vowel-vowel linking), including /r/ linking (e.g. there is), /w/ linking (e.g. two eggs), /j/ linking (e.g. three eggs), and (3) C-C linking (consonant-consonant linking, e.g. half finished). 42 Chinese students were asked to read out 20 English sentences as naturally as possible and identify all the cases of linking in 10 different English sentences. Both the reading and identification tasks were administered twice, one in their first semester and the other in their fourth semester in university. It was found that on the whole the learners’ knowledge in linking was rather limited and they had an infrequent use of linking. Moreover, they have not made extensive progress in two years. Among the three major types of linking investigated, the learners could identify C-V linking relatively well, and they used C-C linking more frequently than the other two types. When both identification and production are considered, the participants’ performance in V-V linking was the worst. These results suggest that training of linking in connected speech to our EFL learners needs to be put on the agenda along with other English suprasegmental features.

Zhang: An Investigation of the Acquisition of Linking by Chinese EFL Learners

42

1. Introduction The teaching of English pronunciation and intonation has been receiving increasing attention in China. Its effect has been witnessed in the increased clarity and comprehensibility of university students’ oral English in recent years. However, Chinese teachers and students usually pay more attention to the learning of English segmentals, that is, the pronunciation of discrete sounds. They spend much less time on English suprasegmental features, such as stress, rhythm, intonation, and linking. In fact, suprasegmental features are seldom dealt with either in research or in teaching in the Chinese context (Xiao, 2011). Linking (also called liaison), a suprasegmental feature, is the process of joining the final sound of one word to the initial sound of the following word. It is a common phenomenon in colloquial, connected English speech. A good mastery of English linking is important for both listening and speaking. Failure to comprehend linking and linked words in connected speech may cause misunderstandings or even failed communication. On the other hand, inadequate use of linking in oral communication would make a speaker sound less fluent or natural. English linking is a very complicated phenomenon with various types, and the Chinese language has little use of linking between words. Therefore, it may pose many learning difficulties for Chinese EFL learners. The teaching of linking in China is both infrequent and inadequate. Linking is only sporadically introduced in English textbooks and the types are usually limited to C-V linking. Moreover, research on the acquisition of linking by Chinese EFL learners is rather scarce. To address this lack of research, this paper aims to investigate the acquisition of linking by Chinese EFL learners, hoping to shed some light on the


43

teaching of this particular suprasegmental feature. 2. Literature Review Linking in English has been widely discussed by previous researchers (e.g. Roach, 2008; Wang, 2005a; Windsor, 1975). In a discussion on aspects of connected speech, Roach (2008) concludes that it is valuable to have practice on rhythm and linking. He holds that “an essential part of acquiring fluency in English is learning to produce connected speech without gaps between words, and this is the practical importance of linking” (Roach, 2008: 147). In this section, we shall take a close look at various types of linking and previous studies on the acquisition of linking. 2.1 Types of Linking There are three major types of linking in English, namely, (1) C-V linking (consonant-vowel linking, e.g. a cup of tea), (2) V-V linking (vowel-vowel linking), including /r/ linking (e.g. there is), /w/ linking (e.g. two eggs), /j/ linking (e.g. three eggs), and (3) C-C linking (consonant-consonant linking), such as lengthening (e.g. half finished), unreleased stops (e.g. art classes), /t-j/ (e.g. won’t you), and /d-j/ (e.g. could you). At least, eight specific types of linking have been identified in the literature. In the following is a more detailed descriptiion of each type.

(1) C-V linking: joining the final consonant of a word with the initial vowel of the following word, as in “a cup of tea”.

(2) /r/ linking: linking the final vowel of a word (such as /:/ //) with the initial vowel of the following word, and inserting a /r/ sound, as in “there is” and “the idea of”2.

(3) /w/ linking: linking the final vowel of a word (such as /u:/, //) with the initial vowel of the next word, and adding a slight /w/ sound, as in “two eggs”.


44

(4) /j/ linking: linking the final vowel of a word (such as /i:/, //, /a/, /e/) with the initial vowel of the next word, and adding a slight /j/ sound, as in “three eggs”.

(5) lengthening: blending the final consonant of a word with the same consonant occurring at the beginning of the next word by lengthening, as in “half finished”.

(6) unreleased stops: when the final consonant of a word is a stop sound, and is followed by another consonant of the next word, the stop is unreleased or arrested, as in “art classes” and “best friend”.

(7) /t-j/: joining the final /t/ of one word with the initial /j/ of the following word, and the joined sound is /t/, as in “won’t you”.

(8) /d-j/: joining the final /d/ of one word with the initial /j/ of the following word, and the joined sound is /d/, as in “could you”3. 2.2 Previous Studies on the Acquisition of Linking While there have been plenty of discussions on English linking in various articles and books, research on the acquisition of linking by learners of English as a second/foreign language is much scarcer, especially in the context of China. A search on the study of the acquisition of linking by Chinese EFL learners in the past 20 years has only yielded 3 relevant studies, two general discussions (Liu, 1991; Quan, 2009) and one empirical research (Sun, 2010). Liu (1991) analysed the first two volumes of College English, textbooks for non-English major university students in the 1990s in Mainland China, and found that the English linking phenomenon was only briefly introduced. The types mentioned included C-V linking and V-V linking, but the introduction to V-V linking was so simple that the students were often


45

confused as to when to use it. Moreover, the exercises on linking in the textbooks provided no guidance or help for students to understand and use linking in their speech (Liu, 1991). Quan (2009) claimed that English linking poses learning difficulties for Chinese students because of the prosodic differences between English and Chinese. She said that there is seldom linking between different syllables in Chinese whereas linking occurs frequently in connected speech in English. She proposed a few methods to help learners improve the use of linking, such as contrastive exercises (between English and Chinese), using recorders or mp3, and imitating the speech of native speakers. Sun (2010) investigated 786 university students’ acquisition of four types of English linking - C-V linking, V-V linking (including /w/ and /j/ linking), C-C linking, and /r/ linking - by adopting three instruments, marking out linking, dictation and reading. He took a quasi-experimental method by using a pre-test, experiment, and a post-test. The experiment lasted for 14 weeks, which aimed to improve students’ acquisition of linking by lecturing and practice. Sun (2010) found that the students improved a lot in their marking of linking, particularly in V-V and C-C linking. Their perception of linking in dictation also improved across each type, but their production of linking did not improve by much. Sun (2010) concluded that Chinese university students’ performance of linking is still quite weak. Though insightful, Sun’s study has several limitations. His classification of linking is quite vague. For instance, he did not make distinctions between different types of C-C linking. His paper does not provide detailed information on research method, so we do not know how the data were collected and analyzed. His discussion on the reasons for the inadequate use of linking does not target specific types. As different types of


46

linking often pose different learning difficulties, the reasons for such differences may also vary. In summary, the phenomenon of linking has seldom been dealt with in either teaching or research in Mainland China, and empirical research on the acquisition of linking by Chinese EFL learners is even scarcer.

3. Method 3.1 Research Questions Given the research gaps as discussed above, the current study attempts to analyze the acquisition of linking by Chinese EFL learners. There are three research questions:

(1) How well can Chinese EFL learners identify each type of linking?

(2) What is the frequency of each type of linking produced by Chinese EFL learners?

(3) Is there a significant increase in the identification and production of linking after two years’ study in university? 3.2 Participants The participants are 42 Chinese EFL learners studying at a university in China. They are English majors from two intact classes and their English is at about the intermediate level. They were taught by the researcher in two semesters in 2009 and 2011, respectively. When they first participated in the study, they had just entered university. Their age was between 17 to 19 years old. The second time they participated in the study, they were toward the end of their second year in university. There were 9 male students and 33 female students. The imbalanced number of students of different gender reflects the imbalanced


47

number of male and female students of English majors. 3.3 Instruments Two instruments were used in this study, one for identification and the other for production of English linking. The instrument for identification is a list of 10 sentences containing various types of English linking (see Appendix 1). All the sentences use colloquial English, chosen from a textbook widely used in Mainland China for teaching English pronunciation and intonation for English majors (Wang, 2005a). All the types of linking mentioned above can be found in the 10 sentences. The potential number of each type is indicated in square brackets as follows: (1) C-V [12], (2) /r/ [6], (3) /w/ [3], (4) /j/ [3], (5) lengthening [4], (6) unreleased stops [8], (7) /t-j/ [2], (8) /d-j/ [1]. Each sentence contains at least one linking and at most six linkings, and the average potential number of linkings in each sentence is 3.9. The instrument for production is a list of 20 sentences of colloquial English containing various types of linking (see Appendix 2). The list was taken from a reading exercise on linking in the aforementioned textbook. All the types of linking can be found in the 20 sentences. The potential number of each type is also indicated in square brackets as follows: (1) C-V [29], (2) /r/ [2], (3) /w/ [4], (4) /j/ [10], (5) lengthening [3], (6) unreleased [8], (7) /t-j/ [1], (8) /d-j/ [1]. Each sentence contains at least one linking and at most six linkings, and the average potential number of linkings in each sentence is 2.9.

3.4 Data Collection and Analysis Procedures The study began in 2009 when the participants took the course “English Pronunciation and Intonation” taught by the researcher in their first semester. To investigate the students’


48

acquisition of English segmental and suprasegmental features, the researcher invited them to record their readings of English sounds, sentences and passages. The recording of the 20 English sentences was one of the tasks assigned to the participants at the beginning of the course. They recorded their reading outside of the classroom and saved and submitted their recordings to the researcher the following week after having been given the assignment. The identification task was administered in class a little later, after they had been taught the various types of linking. This arrangement was due to the researcher’s assumption that the students may not be able to identify all types of linking if they had no prior knowledge at all. But production is a different matter, as use of linking is often the natural result of fluent speech. In the fourth semester, the same students attended another course taught by the researcher, “Linguistics”. At the end of the course, the students were required to make another recording of the same 20 sentences in Appendix 2 and identify the same 10 sentences in Appendix 1. To prevent the potential influence of the identification task on their reading, the recording assignment was given one week earlier. When the two batches of recordings were examined, it was found that 4 students’ recordings were not valid (2 were inaudible and 2 were incomplete). All 4 students are females. So the final participants, whose recordings were clear and complete in both times, included 38 students, 9 males and 29 females. Altogether there were 76 recordings for analysis. The researcher listened to each recording carefully and repeatedly, identified every linking used and indicated its specific type. A colleague of the researcher, who also teaches English Pronunciation and Intonation course, analyzed 8 students’ recordings (about 20% of the total number of


49

recordings) independently. The analyses provided by the researcher and her colleague are consistent with each other in over 90% of the cases. The identification task was analyzed by calculating the frequencies of each type of linking identified by the participants. All the data were processed by using SPSS 13.0 in the following steps.

(1) Descriptive statistics of the frequency of each type of linking identified in 2009 and 2011, respectively.

(2) Descriptive statistics of the frequency of each type of linking produced in 2009 and 2011, respectively.

(3) A paired-samples t-test of linking identified by the participants in 2009 and 2011.

(4) A paired-samples t-test of linking produced by the participants in 2009 and 2011. 4. Results 4.1 Descriptive Statistics Results of Identification The frequencies of major types of linking identified by the participants in 2009 and 2011 are summarized in Table 1 and Table 2, respectively. The tables have also provided the identification percentages (%) (against the total number of potential linking) of C-V, V-V, C-C, and all types of linking (SUM).

Table 1: Identification of linking in 2009

Note: Total = total number of potential linking

Linking Total Min Max Mean % C-V 12 8.00 13.00 10.55 87.92 V-V 12 3.00 12.00 9.42 78.5 C-C 15 0.00 15.00 9.74 64.93

SUM 39 17.00 39.00 29.71 76.18


50

Table 2: Identification of linking in 2011

The tables above show that the frequencies of linking identified decreased from 2009 to 2011. The participants could identify most of the linking in 2009 (76.18%), but their performance in 2011 (55.59%) was much worse. Among the three major categories of linking, C-V linking could be identified the most easily, over 70% in both times, followed by V-V linking and C-C linking. The specific frequencies of each sub-type of linking identified in 2009 and 2011 are presented in Table 3 and Table 4, respectively.

Table 3: Identification of each type of linking in 2009

Linking Total Min Max Mean % C-V 12 8.00 13.00 10.55 87.92 /r/ 6 1.00 6.00 5.39 89.83 /w/ 3 0.00 3.00 2.11 70.33 /j/ 3 0.00 3.00 1.92 64 Lengthening 4 0.00 4.00 3.05 76.25 Unreleased 8 0.00 8.00 4.18 52.25 /t-j/ 2 0.00 2.00 1.55 77.5 /d-j/ 1 0.00 1.00 .95 95

Table 4: Identification of each type of linking in 2011

Linking Total Min Max Mean % C-V 12 3.00 12.00 8.79 73.25 /r/ 6 2.00 6.00 4.00 66.67 /w/ 3 0.00 3.00 0.84 28 /j/ 3 0.00 3.00 1.05 35 Lengthening 4 0.00 4.00 2.71 67.75 Unreleased 8 0.00 8.00 1.82 22.75 /t-j/ 2 0.00 2.00 1.53 76.5 /d-j/ 1 0.00 1.00 0.95 95


SUM 39 11.00 38.00 21.68 55.59


51

Table 3 indicates that /d-j/, /r/ and C-V linking each have a high identification percentage (over 85%). /t-j/ and lengthening could also be identified quite frequently, with a percentage above the average in 2009 (76.18%). However, the frequencies of /w/, unreleased stops and /j/ linking are below the average. Table 4 demonstrates that /d-j/ and /t-j/ linking have an identification percentage over 75%. Lengthening and /r/ linking could also be identified relatively frequently, with a percentage above the average in 2011 (55.59%). However, the frequencies of /j/, /w/, and unreleased stops are far below the average, particularly unreleased stops. A comparison of Tables 3 and 4 indicates that except /d-j/, every type of linking was identified more frequently in 2009 than in 2011. The frequencies of unreleased stops, /w/, and /j/ remain the lowest three among all the eight types of linking over time. 4.2 Descriptive Statistics Results of Production The production frequencies of major types of linking are summarized in Table 5 and Table 6, respectively. The tables have also provided the percentages (%) of occurrence (against the total number of potential occurrences).

Table 5: Production of linking in 2009

Note: Total = total number of potential occurrences


SUM 58 9.00 42.00 27.71 47.78


52

Table 6: Production of linking in 2011

The tables above show that on the whole our learners did not display a high frequency in the use of linking (only about 50%). Among the three major categories of linking, C-C linking was used the most frequently (over 70%), followed by C-V linking (around 60%) in both times. V-V linking was seldom used by our students, with the percentages being only about 10%. The frequencies of linking increased from 2009 to 2011. The specific frequencies of each sub-type of linking used in 2009 and 2011 are presented in Table 7 and Table 8, respectively.

Table 7: Production of each type of linking in 2009

Linking Total Min Max Mean % C-V 29 6.00 26.00 17.05 58.79 /r/ 2 0.00 2.00 .34 17.11 /w/ 4 0.00 1.00 .03 0.66 /j/ 10 0.00 2.00 1.00 10.00 Lengthening 3 0.00 3.00 2.66 88.60 Unreleased 8 2.00 8.00 5.16 64.47 /t-j/ 1 0.00 1.00 .61 60.53 /d-j/ 1 0.00 1.00 .87 86.84

Table 7 demonstrates that each of the four subtypes of C-C linking has a higher percentage than any other type of linking. Lengthening and /d-j/ have a percentage over 85%, indicating that these two types of C-C linking have been mastered quite well by our participants. C-V linking is also used relatively frequently, with a percentage of 58.79%. However, the frequencies of V-V linking, including /r/, /w/, and /j/, are extremely low, especially /w/ linking.


SUM 58 19.00 41.00 29.82 51.41


53

Table 8: Production of each type of linking in 2011

Linking Total Min Max Mean % C-V 29 11.00 26.00 18.11 62.43 /r/ 2 0.00 2.00 .47 23.68 /w/ 4 0.00 1.00 .08 1.97 /j/ 10 0.00 2.00 1.26 12.63 Lengthening 3 2.00 3.00 2.87 95.61 Unreleased 8 3.00 8.00 5.21 65.13 /t-j/ 1 0.00 1.00 .84 84.21 /d-j/ 1 0.00 1.00 .97 97.37

Table 8 shows that the four subtypes of C-C linking also have the highest percentages in 2011. Lengthening and /d-j/ have achieved above 95%, indicating that these two subtypes of C-C linking have been well mastered. C-V linking is also used relatively frequently, with a percentage of 62.43%. However, the percentage of each type of V-V linking remains low, especially /w/ linking. A comparison of Tables 7 and 8 indicates that every type of linking was used more frequently in 2011 than in 2009, suggesting that our students made progress in the use of linking. 4.3 T-test Results of Identification To investigate the extent of change in the identification of linking in two years, a paired-samples t-test of each type of linking and the total number of linkings was carried out. The results are presented in Table 9. Table 9 clearly demonstrates that on the whole, the participants have decreased significantly in the identification of linking (p= .000 < .001). More specifically, except /d-j/ linking, /t-j/ linking, and lengthening, the learners have regressed significantly in the identification of all the other five types of


54

linking (p= .000 < .001) from 2009 to 2011.

Table 9: Paired-samples t-test of identification of linking

Linking Mean t df Sig. (2-tailed)

Pair 1 C-V - C-V 1.76 5.03 37 .000 Pair 2 /r/ - /r/ 1.39 5.67 37 .000 Pair 3 /w/ - /w/ 1.26 6.15 37 .000 Pair 4 /j/ - /j/ 0.87 5.73 37 .000 Pair 5 Lengthening - Lengthening 0.34 1.60 37 .119 Pair 6 Unreleased - Unreleased 2.37 4.99 37 .000 Pair 7 /t-j/ - /t-j/ 0.03 0.22 37 .831 Pair 8 /d-j/ - /d-j/ 0.00 0.00 37 1.000 Pair 9 SUM - SUM 8.03 8.51 37 .000

4.4 T-test Results of Production To investigate whether there was a significant increase in the use of linking after two years of university study, a paired-samples t-test of each type of linking and the total number of linkings was carried out. The results are presented in Table 10.

Table 10: Paired-samples t-test of production of linking

Linking Mean t df Sig. (2-tailed)

Pair 1 C-V - C-V -1.05 -1.70 37 .098 Pair 2 /r/ - /r/ -0.13 -1.15 37 .257 Pair 3 /w/ - /w/ -0.05 -1.43 37 .160 Pair 4 /j/ - /j/ -0.26 -2.14 37 .039 Pair 5 Lengthening - Lengthening -0.21 -2.46 37 .019 Pair 6 Unreleased - Unreleased -0.05 -0.26 37 .797 Pair 7 /t-j/ - /t-j/ -0.24 -3.39 37 .002 Pair 8 /d-j/ - /d-j/ -0.11 -2.09 37 .044 Pair 9 SUM - SUM -2.11 -2.43 37 .020

Table 10 clearly demonstrates that on the whole, the participants made significant progress in the use of linking (p= .020 < .05). More specifically, the learners improved in the


55

use of /j/ linking (p= .039 < .05), lengthening (p= .019 < .05), /t-j/ linking (p= .002 < .01), and /d-j/ linking (p= .044 < .05). In contrast, there was no significant increase in the use of C-V linking, /r/ linking, /w/ linking, and unreleased stops.

5. Discussion This study found that the frequencies of linking varied greatly from type to type in both identification and the production tasks. Moreover, the participants have made progress from 2009 to 2011 in the use of linking to different extents according to type. These results suggest that different types of linking may pose different learning difficulties, and the reasons for this differential acquisition of linking may also vary.

5.1 Identification of Linking It has been found that the frequencies of linking identified generally decreased from 2009 to 2011. Whereas the overall identification percentage of 2009 is 76.18%, the overall identification percentage of 2011 is only 55.59%. The paired-samples t-test results further confirm that on the whole, the participants have decreased significantly in the identification of linking (p= .000 < .001). When specific types of linking are considered, it was found that except /d-j/, /t-j/, and lengthening, the participants regressed significantly in the identification of all the other five types of linking (p= .000 < .001) from 2009 to 2011. The participants’ fairly good performance in 2009 may be because the identification test was carried out not long after they had been taught different types of linking, when they still had a fresh memory of each type of linking. However, after one year, their performance worsened a great deal, suggesting that a relapse may have occurred. These findings indicate that the learners had considerable difficulties identifying the various types of linking and their knowledge on


56

linking is far from satisfactory. By comparing the participants’ performance in these two instances more closely, we find that some types of linking were identified more frequently than others. In 2009, the participants could identify most of the cases of linking, and /d-j/, /r/ and C-V linking all had a high identification percentage (over 85%). Only unreleased stops have an identification percentage lower than 60%. In 2011, the participants’ performance was much worse. Except /d-j/, the identification percentages of all the other types of linking are lower than 80%, and unreleased stops, /w/, and /j/ have an identification percentage lower than 60%. In comparison, the frequencies of unreleased stops, /w/, and /j/ remain the lowest three among all the eight types of linking in both instances. These results suggest that our learners had a better knowledge in C-V linking than V-V linking or C-C linking, and their knowledge in unreleased stops, /w/, and /j/ linking is poorer than the other types. One reason for this differential knowledge may be inadequate input. The types of linking mentioned in the English textbooks in China often only include C-V linking, and the introduction to specific types of linking is rather limited (Liu, 1991; Sun, 2010). Another reason is that different types of linking may cause different degrees of difficulty. For instance, lengthening is probably easier than /w/ linking for the students, as the rule of occurrence for the former seems simpler than that for the latter.

5.2 Production of Linking Compared with identification, the participants’ performance in the production of linking was even worse, particularly in V-V linking. This is no surprise, as production has always been found to be more difficult than perception. The good news is that the participants made some progress from 2009 to 2011 in the use of linking. In the following, the results concerning the


57

production of linking will be discussed type by type. 5.2.1 C-V Linking This type of linking was used in around 60% of its potential situations and there was no significant improvement after two years, suggesting that the learners may have had a basic command of C-V linking by the time they entered university and their production did not improve much. This result is consistent with previous research, which also reported a similar percentage (Sun, 2010). C-V linking occurs frequently in connected speech. The condition under which it takes place is relatively easy to remember, that is, when the preceding word ends with a consonant and the following word begins with a vowel. It is often introduced in English textbooks in high schools and universities in China. These might account for the relatively frequent occurrences in the learners’ recordings as well as the higher percentages in the identification task. However, there may be various consonants and vowels involved, so the potential situations of C-V linking usually vary greatly. Moreover, C-V linking often occurs frequently within a sentence, such as “Can I look at it?”, where there are three cases of consecutive linking, making it difficult for learners to produce in oral communication. These two reasons may explain the nearly 40% absence in the learners’ speech. 5.2.2 V-V Linking All the three types of V-V linking were seldom used by the participants. /R/ linking was used the most frequently among the three, though its percentages were also very low (around 20%), and there was no significant improvement in two years. The results of the identification task show that /r/ linking was also the most frequently identified V-V linking and the identification percentages are as high as 89.83% in 2009 and


58

66.67% in 2011. These findings suggest that our learners have a relatively good knowledge in /r/ linking, and they are quite clear about the condition under which it occurs, that is, when the preceding word ends with certain vowels, such as /:/ and //, and the following word also begins with a vowel, the /r/ sound will be inserted (as in “intrusive r”) or realized from spelling (as in “linking r”). The inadequate use of /r/ linking may be attributed to the learners’ lack of practice and the difficulty in producing this type of linking. The learners’ production of /w/ linking was the poorest (near zero) and there was no significant improvement. Almost all the learners have failed in the use of /w/ linking. This result is consistent with Sun (2010), which also reported a low production frequency of /w/ linking. The identification result in 2011 also indicates that learners’ knowledge in /w/ linking was the poorest among all. This type of linking is not often introduced in textbooks in high schools or universities in China and its rule of usage is quite difficult to digest, that is, a slight /w/ should be inserted when the preceding word ends with vowels such as /u:/ and //, and the following word begins with another vowel. These might explain the extremely low frequency of production in the recordings. The learners’ performance in /j/ linking was also very poor. As with /w/ linking, its occurrences require an addition of a slight /j/ sound when the preceding word ends with certain vowels such as /i:/ and // and the following word begins with another vowel. This type of linking is also seldom introduced in textbooks. Although the learners have made significant improvement in their frequency, the percentages were still extremely low, below 13%. The identification percentage of /j/ linking in 2011 was also rather low (35%), showing our learners’ poor knowledge of it.


59

5.2.3 C-C Linking C-C linking was used the most frequently by the participants across each sub-type. The learners have a very good command of lengthening. Its frequency in 2009 was nearly 90% and in 2011 reached 95%. The high percentages and the significant improvement indicate that this type of linking does not pose great learning difficulties for our students. This is one of the easiest types of linking, as the rule of usage - two adjacent identical consonants in two different words resulting in its lengthening - is easy to remember and apply. The result of the identification task confirms the learners’ good mastery of lengthening. “Unreleased stops” is the only type of C-C linking that has not improved significantly. Its percentages remain around 65%, the lowest among all types of C-C linking. It is used when the preceding word ends with a stop consonant and the following word begins with another stop, or fricative, or nasal. So the condition of its usage is quite complicated, as there are 6 oral stops, 9 fricatives and 3 nasals in English. This means that there are many situations where this type of linking may be used. Furthermore, the identification percentages of unreleased stops are the lowest among all, revealing the learners’ limited knowledge in this type of linking. The percentage of /t-j/ linking has increased significantly from 60.53% in 2009 to 84.21% in 2011. This type of linking often occurs when the preceding word is “not”, contracted with auxiliary words, such as “won’t”, “weren’t”, “don’t”, “can’t” and so on, in interrogative sentences. Negative interrogative sentences tend to be avoided by Chinese EFL learners (Milton, 2001), which might explain in part why the percentage in 2009 was not high. Once the learners have mastered its use, as it is relatively simple (with the preceding consonant being /t/ and


60

the following consonant being /j/), its frequency increased a great deal. The /d-j/ linking was used the mostly frequently, with the percentages being 86.84% in 2009 and 97.37% in 2011. It was also the most frequently identified type of linking and the identification percentages remain as high as 95% both in 2009 and in 2011. The condition under which /d-j/ linking occurs is easy to remember and apply, which may explain the nearly full mastery by our participants. 5.3 Acquisition and Teaching of English Linking The discussion about the reasons for the participants’ differential acquisition of different types of linking suggests that input and complexity are two major factors influencing Chinese EFL learners’ identification and production of linking. Implicitly, our discussion also suggests that teaching may help learners improve their acquisition of linking. Although input on linking could be obtained through large amount of listening, the amount of intake is hard to determine. Linking occurs in fast connected speech, so learners may not be able to hear linking clearly every time it occurs. Explicit teaching and training of linking are needed in the curriculum. Students need to be informed of the conditions and situations where each type of linking takes place. One way to fulfill this purpose is that the teachers consciously provide such knowledge to students. A simpler way, and probably a more direct way, is that this knowledge is presented to the students in their English textbooks. This can be done by integrating the research results on second language acquisition into the English teaching materials. As has been recognized, pronunciation is beyond “listen and repeat” (Jones, 1997). Once the EFL learners have been equipped with the knowledge of linking, the next step is to provide them with activities of all kinds, such as imitating


61

fluent connected speech, singing English songs and practicing tongue twists, so that they can actually use linking in communicative contexts. In short, increasing their awareness and creating chances of practice may help students improve their acquisition of linking. Some scholars have pointed out that the learning and teaching objective of English pronunciation and intonation in China should shift from being modeled on the traditional native English speakers to a more practical objective, especially in terms of suprasegmental features (e.g. Wang, 2005b). We also hold the view that our learners need not use linking as if they were native English speakers, but being aware of and being able to use various types of linking in connected speech can help them improve both their listening comprehension and the fluency of their oral English. In this regard, teaching English linking is important and practical for EFL learners.

6. Conclusion This study investigated Chinese EFL learners’ acquisition of various types of linking from two perspectives, identification and production. It was found that on the whole the learners had a poor performance of linking and did not seem to make extensive progress in two years of university study. Among the three major types of linking, the learners identified C-V linking and used C-C linking the most frequently, but their performance in V-V linking, including /r/, /w/ and /j/ linking, was extremely poor. These findings suggest that it is urgent and necessary to train our EFL learners in the use of various types of linking in connected speech. The present study is only a preliminary study with several limitations. First, the number of participants is relatively small and all the students are Chinese EFL learners of English majors.


62

This may limit the generalizability of the findings. Second, the production instrument consists of only 20 sentences. The recordings of the participants’ reading may not reflect the learners’ use of linking in real communication. Third, no native English speaker controls were involved, so the percentages of linking were calculated against the potential occurrences of linking, not the occurrences used by fluent native speakers. These limitations suggest the need for further research with a more careful design and better analyses of the acquisition of linking by Chinese and other EFL learners. Notes 1 This study has been supported by Wuhan University Teaching Research Grants: “A Multi-model Construction of Curriculum System for English Speciality” (No: 20113003) and “A Survey of World Englishes” (No: ts201110), and by “Fundamental Research Funds for the Central Universities” (Post-70s Young Scholar Research Funds at Wuhan University).

2 Some researchers distinguish between “linking r” (the preceding word’s spelling suggests a final r), as in “four eggs”, and “intrusive r” (the preceding word ending with a vowel doesn’t have a final r in the spelling), as in “media event”. 3 Cases of /t-j/ and /d-j/ are often considered as assimilation. References Jones, R. H. (1997). Beyond “Listen and repeat”: Pronunciation teaching materials

and theories of second language acquisition. System, 25(1), 103-112. Liu, Y-P. (1991). English linking and exercises on linking. Journal of Yiyang Teachers’

College. 12 (3), 57-59. Milton, J. (2001). Research Report: Elements of a Written Interlanguage: A Computational

and Corpus-based Study of Institutional Influences on the Acquisition of English by Hong Kong Chinese Students. Hong Kong: HKUST.

Quan, N. (2009). Negative transfer of Chinese prosody on English linking and some

treatments. Guangxi Education Institute Journal. 6, 60-62.


63

Roach, P. (2008). English Phonetics and Phonology: A Practical Course. (3rd ed.). Cambridge: CUP (2000)/Beijing: Foreign Language Teaching and Research Press (reprint).

Sun, Y-C. (2010). An investigation of university students’ knowledge of English

linking and its implications. Shi Ji Qiao. 3, 135-136. Wang, G-Z. (2005a). English Pronunciation and Intonation for Communication. Beijing:

Higher Education Press. Wang, G-Z. (2005b). Revising the goal of EFL pronunciation teaching in the Chinese

context. Foreign Languages in China. 6, 18-23. Windsor, L. (1975). Linking /r/ in the general British pronunciation of English.

Journal of the International Phonetic Association. 5, 37-42. Xiao, J. (2011). A Study of Chinese EFL Learners' Acquisition of English Fricatives.

Unpublished M.A. thesis. Wuhan University. Appendix 1. Identification Task Please mark out the sound linking in the following sentences. Example: I’m afraid he’s out. 1. Can I offer you an orange? 2. Look at the one on the corner of the street. 3. Both my father and mother are coming tonight. 4. Was Susanna in last night? 5. The potential of Solar Energy was still too often neglected. 6. Are they going to take it with them? 7. Can you ask him to come back to the office at eight? 8. Come every once in a while. 9. I wanted to connect you, but you were out. 10. Would you like to try it out? Appendix 2. Reading Task Please read the following twenty sentences as fluently and naturally as possible. Record your reading and save it into a file. Thanks! 1. Take only a little at a time. 2. Come every once in a while. 3. He is the only actor I know. 4. He lives alone. 5. He lives a highly organized life. 6. Who brings it up? 7. She goes to art classes every other day. 8. Who can beat you at tennis?


64

9. Can I look at it? 10. He looked at it in the window. 11. Would you like to try it out? 12. I lose out to my old pal. 13. I’ll go out for a few weeks. 14. I’m just out of sympathy. 15. We’ll enjoy an evening together. 16. You have to stay in shape. 17. You don’t want to end up like this. 18. If only I were there. 19. My back’s out again. 20. Come on in. About the author Zhang Yanyan is currently an associate professor at the College of Foreign Languages and Literature at Wuhan University in P. R. China. Her research interests mainly include second language acquisition, applied linguistics, and world Englishes. E-mail: [email protected]

an investigation of the acquisition of linking by chinese...

Documents