Chapter Five Conclusion
5.1 The main findings
The purpose of this study is to investigate Chinese subjects’ performance of liquid pronunciation in the five phonetic environments, namely syllable initial /r, l/, syllable final /r, l/, initial /r, l/ clusters, final /r, l/ clusters and clusters containing /rl/. What follow are the main findings observed from the result of the study.
First of all, syllable initial /l/ is the easiest sound, while /rl/ clusters cause the greatest difficulty to the subjects. Postvocalic /ɫ/, either alone or in clusters, is more problematic than postvocalic /r/. Difficulty of initial /l/ clusters far exceeds that of initial /r/ clusters. Syllable initial /r/ is as easy as initial /l/ when preceding a high front vowel /i/ or /ɪ/, and its difficulty arises when a following mid front vowel /ɛ/ or a back vowel occurs.
Second, backness of the surrounding vowel effect correlates with 62.5% (5/8) of the target sounds while the task size effect is limited to 17.6% (3/17) of them. Source of back vowel effect can be interpreted from the overlap of back vowel resonance found in dark /ɫ/ and postvocalic /r/ especially with their preceding mid back vowel /o/ or /ɔ/.
Third, similarity in phonetic quality affecting liquid performance also includes distribution of Mandarin ㄖ(/ʐ/) and overlap of coronal quality occurring in initial /r/
clusters. The combination of /ʐʊɛ/ ㄖㄨㄟ and /ʐo/ ㄖㄡ are possible syllables in Mandarin, and thus confusion of distinguishing Mandarin /ʐʊɛ/ and /ʐo/ from English /rɛ/ and /ro/ may arise or even interfere the accuracy of prevocalic /r/ as in red and road. On the other hand, overlap of coronal quality occurs in initial /r/ clusters as /tr/
in tree, in which /r/ already owns coronal feature itself and follows another coronal /t/,
and the overlap in place of articulation may cause great difficulty in pronouncing the cluster /tr/.
Fourth, some of the error types are classified as due to L1 interference or developmental factors, or even both. Substitution of /ʐ/and /ʐʊ/ for initial /r/ are considered as interference from Mandarin; /lʊ/ and /l/ for initial /r/ is probably under influence of Taiwanese. On the other hand, vocalization of dark /ɫ/, deletion of /ɫ/ and postvocalic /r/, and gliding of initial /r/ are all regarded as due to developmental process factors because similar phenomena also occur in L1 acquisition. Besides, epenthesis of schwa /ə/ in clusters may result from both sources.
Fifth, the preceding back vowel /o/ or /ɔ/ is observed to be a significant factor in the distribution of the subjects’ main error types for postvocalic /r/ and /l/. The deletion of /ɫ/ usually occurs when it is with a preceding vowel /ɔ/ or /o/, and the substitution of /ɔ/ appears more often when it following a vowel other than /ɔ/ or /o/.
Besides, when a postvocalic /r/ follows an /ɔ/, the deletion of /r/ is usually the only error type, and when it follows a vowel other than /ɔ/, more than one error type occurs.
Finally, other factors such as prominence of stress, assimilation of a neighboring nasal, and some idiosyncrasies of the subjects’ learning all have different degrees of impact on mispronunciation of liquid.
5.2 Pedagogical implications
Based on the findings of the study, some implications for EFL pedagogy are suggested. First, English teachers must be aware of some the factors that impede or facilitate successful acquisition of English phonology. Teachers must exactly know what kind of phonetic input should be provided to optimize the learning of English pronunciation. Therefore, they can direct their students to pay more attention to certain mispronounced sounds, as those found in the present study.
The order of presenting liquid variants and their distribution in a word is considered important in pronunciation teaching. The easier variants of /r/ and /l/ for Chinese students should be introduced earlier than the more difficult ones. Clear /l/ in syllable initial positions (i.e., lead) can be presented first, because it is seen to be the least difficult sound. Prevocalic /r/ (i.e., read, cream) may be taught earlier than postvocalic /r/ (i.e., fear, beard). Dark /ɫ/ either alone or in clusters (i.e., till, twelve) is probably the last one to be practiced. Moreover, all variants of /r/ and /l/ with a non-back vowel are supposed to be presented earlier than those with a back vowel.
That is to say, liquids in the words as lead, read, fear, and till should be acquired earlier than their counterparts load, road, four, and tall.
The amount of time spent on the liquid instruction is also significant in helping students’ performance. A longer duration of time is needed to train learners how to pronounce postvocalic /r, l/ accurately. Teachers should repeatedly demonstrate and explain the different articulatory gestures such as how the tongue tip is curled when articulating a postvocalic /r/ or how the front of tongue is slightly depressed and the back is raised in pronouncing a dark /ɫ/. When students are aware of these distinctive features of /-r, ɫ/, they will be able to pronounce some similar words, such as four &
foe, fall & foe, and poor & pool accurately.
Special efforts should be made to compare the contrastive clusters of /r/ and /l/.
The difference between fry & fly in initial clusters and bears & bells in final clusters should be clearly explained to students. Minimal pairs such as bright & blight, pray &
play, crowd & cloud, and please & police should also be demonstrated by instructors
several times before learners practice them. Besides, two consecutive /rl/ in world, girl, pearl, and Charles are the most difficult sound for Chinese students, and thus a large amount of time and effort is needed before they are successful in pronouncing the troublesome /rl/ correctly.
Particular difficulty in initial /tr/ clusters due to overlap of coronal feature should be interpreted as much as possible to students. Difference between the English true and the Mandarin 出, or the English tray and the Mandarin 脆 are needed to explain clearly and demonstrated by teachers as well.
In addition to liquid pronunciation training through mechanic drills, instructors can also design various activities for students with different levels of English proficiency, such as contests of identifying liquids, games of guessing words containing /r/ or /l/, reading out activities, information gap activities, problem-solving tasks, as well as a role play. These activities not only can increase students’ chances of listening and speaking English, but also may offer more opportunities for teachers to find out students’ pronunciation errors and even correct them at proper time. The aim of these activities is to make students’ pronunciation practice more interesting, communicative and meaningful.
In short, English pronunciation learning and teaching need lots of efforts from both students and teachers. It is important that teachers require themselves to pronounce /r/ and /l/ as accurate as possible so that they can decrease chances of incorrect input to their students. They should also be involved in more input of enhancement researches of language learning and absorb more techniques or knowledge in the relative fields. In addition to teachers’ patient and careful guidance, students’ constant and sufficient practice in pronouncing liquids may help them build confidence in producing more comprehensible English and then they will gradually be capable to face other consonant problems in English pronunciation.
5.3 Limitations and suggestions for further study
Although an equal number of males and females participated in the present experiment, we did not include gender as a possible factor influencing accurate pronunciation of English /r/ and /l/. It would be of interest to regard gender as a
variable in future researches to investigate whether males or females have better performance in liquids.
Moreover, the subjects recruited in our study are all encouraging volunteers.
Therefore, factors of their overall English proficiency and motivation of improving pronunciation skills are not well-controlled. More researches can be conducted to examine if students with better English proficiency produce more accurate /r/ and /l/, or those with high motivation perform better.
Lastly, due to most of the subjects’ inability in expressing themselves orally in English, only reading out task is designed in the present experiment, other methods as an interview or free conversation is not involved. Hence, it is less possible to elicit students’ pronunciation in a more spontaneous and communicative situation. For example, the method of “answering questions” in oral tests applied by GEPT (General English Proficiency Test) may be a good choice to explore students’ erroneous production of liquids elicited naturally and communicatively.
Future researches in this area are expected to adopt a more natural and communicative method to evaluate Chinese students’ mispronunciation in liquids.
Students’ ability in distinguishing /r/ from /l/ in different word positions can also be explored to examine their performance in liquid perception. Effects of special training on liquid perception and pronunciation is also expected to observe degrees of improvement Chinese learners may have in perceiving and producing English /r/ and /l/.
