Combining instruction and immersion to improve pronunciation: The case of Chinese students learning bahasa Indonesia

In pronunciation classes, most teachers over-emphasize the accuracy part and ignore the fluency aspects (Elliot, 1995; Yoshida, 2016). To enhance both accuracy and fluency for Chinese students’ pronunciation in Bahasa Indonesia (BI), this study investigates the effects of combining instruction and immersion. This was classroom action research conducted for one semester at Qujing Normal University (QJNU), Yunnan Sheng, China. Audiolingual Method (ALM) with drilling technique was the instruction procedure while the outdoor project was the technique in the immersion. The research subjects were nine first-semester students in the Indonesian Language Department of QJNU. Their score improvements from the pre-to post-assessments indicate that combining these two methods is effective in enhancing students’ pronunciation in BI. The instruction conducted through explicit phonetic teaching and drilling is appropriate for habituating them with accurate pronunciation. Meanwhile, the fun and enjoyable immersion in contextual settings boosts their fluency and gives them opportunities to explore and practice using the words they have learned in real contexts. Along with pronunciation improvement, enhancement in students’ learning attitude and engagement in the learning activities are the other benefits that may contribute to their success in gaining full fluency in the second language.


INTRODUCTION
The existence of Bahasa Indonesia (BI) in the global world is undeniable as it is an influential language in Southeast Asia (Jegho, 2017) and the third most spoken language in Asia after Chinese-Mandarin and English (Khan, 2018) which is potential to be a regional language in Southeast Asia (Effendi, 2016) or an official language of ASEAN (Afria, 2020;Alfin, 2015).Through various programs, the Indonesian government has been trying to enhance the function of BI as an international language.Teaching Bahasa Indonesia to Foreign Speakers (BIPA) is one of them.BIPA, which has been taught in more than 46 countries all over the world including ones in Asia, Australia, the U.S., Africa, Europe, and the Middle East (Paryono, 2018), plays an important role in implementing government policy in Indonesian Law No. 24 Year 2009 on National Flag, Language, Symbol, and Anthem (Hamid, 2018;Rohimah, 2018) where BI functions as the delivery language in education, a means to develop national culture and a medium to utilize science, art, and technology.
To promote the Indonesian language and culture as well as strengthen the cooperation between Indonesia and other countries through BIPA, the instructors should design the learning well.However, creating a course to make the students able to learn BI effectively while also enjoying the activities which finally lead them to love the language and the culture is not easy, especially when the learners are adults.Language learning in adulthood is complicated and difficult, especially because of age (Birdsong & Molis, 2001;Ghasemi & Hashemi, 2011), particularly when they never use the L2 before coming to the class.
Considering that oral language is the first communication form any foreign people face, mastering oral skills, especially pronunciation, should be the initial stage of foreign language learning.Pronunciation is even the most important skill in spoken language as it supports the learners' communicative power in general (Gilakjani & Ahmadi, 2011).However, one of the problems students often encounter is the different phonological aspects between their first language (L1) and their second language (L2) (Gilakjani & Ahmadi, 2011;Yoshida, 2016).As mispronounced words may cause either semantic confusion or uncertainty that affects the meaning delivery (Odisho, 2014) or leads to communication failure (Yoshida, 2016), effective treatment for beginners is needed in pronunciation class to prevent further errors in the future learning process.
While in the past some language teachers and researchers gave the least attention to pronunciation teaching and research (Gilakjani & Ahmadi, 2011;Gilakjani, 2012), nowadays many innovations are conducted in pronunciation classes (Brinton, 2017).However, these are mostly applied in the English language while the ones related to Chinese students' pronunciation of BI are not many and these limited studies mainly describe the errors made by Chinese students (Wiratsih, 2019;Xu & Setiawan, 2020) without giving any solution on how to overcome the problems.This study aims to fill this gap by implementing the combination of instruction and immersion to improve the Chinese students' pronunciation of BI.
In second language learning, learners' L1 influences the L2 they produce.This effect is broadly known as transfer and is divided into two.Negative transfer or interference is when the influence causes errors in the L2 acquisition or use while positive transfer is when the influence makes the L2 acquisition or use easier (Bardovi-Harlig & Sprouse, 2018).Accordingly, it is crucial to consider the similar and different phonological aspects between L1 and L2 before giving pronunciation practice to L2 learners.
In consonants, most of the inaccurate pronunciations are caused by the absence of Indonesian plosive consonants in Mandarin (Mulyaningsih, 2014).As a result, they tend to change the plosive consonants [b, d, g] into their implosive pairs [p, t, k].Another problem is the alteration of the trill [r] to lateral [l] which also occurs in their English pronunciation (Huang & Radant, 2009).The simple syllable structure where the only permitted final consonants are nasal [n] and [ŋ] (Li & Thompson, 2018) is the reason why they tend to make the final consonants in BI disappear.Consonant cluster, like /ny/, is another difficult thing for Chinese learners as there are no consonant clusters in Standard Mandarin (Li & Thompson, 2018) where consonants are always followed by vowels (Eng, 2012;Zhang & Yin, 2009).
The vowels of BI and Chinese are almost similar, but the consonants diverge in many ways.This leads to the conclusion that negative transfer (interference) of consonants tends to occur in their early time of learning BI and treatment should be given to overcome the problem of pronunciation errors.It is not meant to change their L1 accent to be like BI native speakers but to familiarize them with the differences and to form the habit of pronouncing words in BI accurately to prevent meaning confusion perceived by the hearers.Therefore, ALM whose emphasis is on oral skills and habit formation is worth applying for achieving those goals.
ALM was developed on the principle that language whose main function is for social communication is a system of sounds; writing is the secondary derived system to record the spoken language (Mart, 2014).Within this method, the teacher as the target language model should provide an accurate language form and, by listening to it, the students should be able to imitate the model (Larsen-Freeman & Anderson, 2011).Therefore, listening is the main skill they employ in their first days of learning.By listening, students will get input on the vocabulary to make them able to produce expressions.Besides, listening will make students understand the sound, rhythm, intonation, and stress in L2 (Renukadevi, 2014) which they can adapt to be more perfect in their oral productive skills.
ALM embraces behaviorist learning theory that sees learning as a process of acquiring a set of structure through habit formation (Powel, Honey & Symbaluk, 2013) where stimulus, response, and reinforcement are the main elements that makes the learning take place (Bélanger, 2011).Thus, in ALM, drilling is one of its teaching activities (Mei, 2018).Senel (2006) lists many useful drills for teaching correct and accurate pronunciation.Considering that the students were beginners who were starting to learn the vocabularies, two kinds of drilling adopted in this study were saturation drill and mobility drill.The former is suitable to familiarize students with problematic sounds in all positions.For example, the sound /r/ is drilled in three positions, initial, medial, and final.It is suitable to be employed on the first days to learn many vocabularies with accurate pronunciation.The second drill employed, after the students mastered many vocabularies, was the mobility drill which enables them to mobile their tongue to practice pronouncing the vocabulary they have learned while also producing simple utterances.
By employing such strategies, a claim that "the more often something is repeated, the stronger the habit and the greater the learning" (Larsen-Freeman & Anderson, 2011) is plausible and it becomes one of the strengths of ALM.Mei (2018) mentions the other strengths as to training learners to think in L2 patterns, more easily mastered and applied by learners, and able to bring positive effect on correct habit formation.These many benefits have made some researchers and language teachers, such as Hasanah & Dahniar (2017), Hidayati (2016), andMart (2014), applied the ALM method to improve students' pronunciation and they confirmed that this method is effective.
However, Mei (2018) further provides some shortcomings of this method as ignoring the cognitive process and language creativity, dominating the teacher's roles, ignoring the factors of society as well as the culture of communication, and neglecting the cultivation of reading and writing.Lee & VanPatten (2003, p. 10) adds "what the ALM instructor did not usually provide was the opportunity for students to use the language in a meaningful or communicative way, one involving the exchange of message."Considering these weaknesses, a complementary strategy should be applied.Based on Krashen's theories that "acquisition is more important than learning" (1982, p. 32) and "it requires meaningful interaction in the target language -natural communication-in which speakers are concerned not with the form of their utterances but with the messages they are conveying and understanding," (1981, p. 1) immersion was thought to be a perfect match for the ALM.
That learning L2 would be most efficient in true meaningful communication situations (Krashen, 1981;Nieminen, 2006;Richards, 2006) is the underlying principle of the immersion method.This is a method of language teaching which surrounds (immerses) learners in the language they are trying to learn where the language is used to communicate and to do all activities, such as meals, stories, daily routines, and group activities (First People's Cultural Council, 2016).In this method, the L2 not only becomes the content learnt but also a tool for communication and information gathering.This is a method that emerged since some research agreed that studying abroad is the best way to acquire a foreign language (Lord, 2010) because the students are in L2 real context and they cannot use their mother tongue.
Those theories suggest that immersion is ideally implemented by allowing students to directly experience using the language they are learning in society.It is an effective and comprehensive way to integrate culture into learning the language since the students make direct interactions with the people in a real context (Wirawan, 2018).However, since this study was conducted in Qujing, China, and finding Indonesian native speakers there is hard, the context was created by the teacher by giving much input in BI and students use BI as much as they can.Here the teacher created a situation that enables learners to use and work with the target language and culture, developed learning activities suitable to the learners' context, and provided a variety of language stimuli from various sources.
The immersion method applied in this course was mostly in the form of simple and fun out-ofclass projects intended to encourage the learners to use their language resources in authentic communication.That this method enables learners to develop their linguistic, communicative, and pragmatic competence and improve their levels of accuracy and fluency (Richards, 2014) makes it fits the research objectives.Besides, the opportunity to escape from stale, tedious, and inefficient learning and from being confined in a bubble of language learning materials (Bodnar, 2019) provides a greater chance for learning to take place since language learning will be more successful when it is fun and enjoyable (First People's Cultural Council, 2016).
Pronunciation which is often regarded merely as a linguistic component makes the teachers focus more on accuracy and ignore the fluency aspect (Yoshida, 2016) and thus set pronunciation aside in a communicatively oriented classroom (Elliot, 1995).To place importance on both aspects, this study employed the combination of instruction and immersion in learning pronunciation in 'Basic Indonesian Language and Culture' course.When the problem of pronunciation can be solved in the first semester, this basic skill in speaking will provide students with comprehensible pronunciation which enables them to have effective communication.

METHOD
This was classroom action research aiming at improving Chinese students' pronunciation in BI.
As the nature of the study is to solve problems in the educational field and to make improvements systematically (Pelton, 2010;Sagor, 2000;Tomal, 2010), the method fits the research goal.The study was conducted for nine weeks.The intervention was given in six weeks and each consisted of two meetings: classroom instruction and out-of-class immersion.Since several courses in QJNU had a schedule of regular classrooms and tutorials, immersion was the tutorial program of the course.The classroom instruction lasted for ninety minutes while the immersion duration was more flexible, around 1-3 hours.
Initially, the participants in this study were ten students.They were in the first semester of 'Basic Indonesian Language and Culture' course in QJNU, Yunnan Sheng, China.However, a student moved to the Statistics study program because she thought that learning a language was not her passion.As in most language study programs, the number of female students was dominating.In this class, all nine students were female.They were the first students to enroll in Indonesian Language Education at QJNU.Before the intervention, at the first meeting, their mastery level of BI was zero as they never knew nor used BI before.
The test was the instrument from which the quantitative data were collected.Meanwhile, to gather the qualitative data, the instruments were field notes, students' reflective diaries, and interview protocol.In the pre-assessment, since they have not mastered any vocabulary in Indonesia, reading aloud was considered the best way to measure their pronunciation skill.They were asked to read poetry titled "Pada Suatu Hari Nanti" written by an Indonesian poet, Sapardi Djoko Damono.It was chosen since it contains many sounds thought to be problematic for Chinese students.There are 61 words in the text, but the distinctive words are only 30.Thus, the student's score was the result of the number of correctly pronounced words divided by the number of distinctive words times 100.
The intervention in classroom action research consists of four phases: planning, action, observation, and reflection (Burns, 2010;Ferrance, 2000).In this study, those phases were carried out in two cycles.In the planning phase, it was decided that the aspects each cycle would be focused on are nouns and verbs.The action phase was executed by combining instruction and immersion in the learning process.The action phase was implemented in two cycles within eight meetings.The first cycle focused on pronouncing /r/ and /d/ in nouns.The second cycle focused on pronouncing /n/, /ŋ/, /ny/, and /h/ in verbs.
There were two kinds of data in this study: quantitative data and qualitative data.The quantitative data was in the form of scores resulting from three different tests: pre-assessment, post-assessment-1, and post-assessment-2.Meanwhile, the qualitative data was in the form of participants' behaviors, thoughts, and utterances recorded in the field notes, reflective diaries, and interview transcripts.The quantitative data were analyzed using general statistics and were then interpreted to see students' pronunciation improvement, while the qualitative data were interpreted by the researcher team to evaluate the learning strategies.

Results
In this part, students' scores in the three conducted tests were analyzed.They have been given nicknames.The results of the tests are specified in Table 2. Scores in the pre-test indicate that their ability in pronouncing words in BI was varied.With 33 as the lowest score, this implied that exposing them to Indonesian alphabets and sounds at the first meeting was necessary.One student got an outstanding score of 83 in the first meeting although the [r] she produced was in very strong vibration.It was found that she is good at speaking English and her familiarity with English words has enabled her to pronounce words in BI well.Based on the pre-assessment scores, if the participants were divided into two big groups, where those with very good and excellent qualifications belong to the 'high' category and the rest are in the 'low' category, only 2 of them go to the high one.This means pronunciation remedy was urgent to enhance the effective meaning delivery in their talk and to prevent them from further problems in the learning process, especially in speaking.
After four meetings focusing on improving the sounds /r/, /l/, /t/, and /d/ in nouns, noticeable progress in their average score was presented, from 58 in the pre-assessment to 68.44 in the first postassessment.Because all students performed score improvement, it is reasonable to say that the intervention in the first cycle brought positive effects.
After cycle 1 was conducted, although significant gains had been achieved, their qualifications did not change meaningfully, with three participants (33.3%) in the 'high' category and six of them (66.7%) in the 'low' category.This implies that further intervention needs to be carried out.Then, the next cycle was to be executed with the focus of improving sounds [n], [ŋ], [ny], and [h] in verbs.After the second intervention was given, seven participants had been in the 'high' category and only two were in the 'low' category.Nevertheless, among the two, Rani made great improvement.She performed significant score progress from 47 to 63.With an average score of 80, the conclusion that combining instruction and immersion is beneficial for second-language pronunciation is reasonable.

Discussion
Generally, the success of combining instruction and immersion to improve Chinese students' pronunciation of BI was contributed by three main factors: the learning material and activity, the learners, and the teacher.With the cyclical nature of action research (Denscombe, 2007;Drummond & Themessl-Huber, 2007), the learning design in the study was tentative since it was modified over and over following the evaluation generated from the observation and interview.It helped the researcher to better plan the learning strategies for the next meetings by considering the students' needs and the problems they encountered.Project: talking about food preference Reflection: a. Students were more relaxed and not afraid of making mistakes.They were happily correcting others' mistakes.b.Mistakes in combining words into phrases were found as Chinese language and BI have different ways of forming a noun phrase.8 Activity: reading tongue twisters Reflection: By using tongue twisters, they practiced and laughed when making mistakes.Finally, although at a slow speed, each successfully read several sentences.
Project: describing the surrounding environment in a park Reflection: After vocabulary on the things in the park was mastered, the descriptive text they produced could be easily understood.
Considering that the quality of teaching is an external factor affecting pronunciation learning (Yoshida, 2016) and good language materials influence students' willingness to communicate (Aguskin & Maryani, 2018), selecting interesting teaching techniques is crucial.A song whose benefits have been widely known to improve language skills, especially vocabulary and pronunciation, not only for young learners but also for adults (Javadi-Safa, 2018;Supeno, 2018) were employed twice in this study.However, that students produced accurate pronunciation in drilling and back to inaccurate pronunciation in singing proved that correct pronunciation tends to occur when students are concentrating fully and producing it alone or in a single word (Yoshida, 2016).This suggests that more learning in real-world speaking should be applied to familiarize them with pronouncing words accurately and fluently.It considers that the main goal of learning a language is not to pass an exam but to use the language for authentic communication (MacIntyre & Doucette, 2010).
Besides providing interesting activities, the fact that anxiety is an essential negative emotional factor that affects language learning (Cheng & Chang, 2004;Zheng, 2008) should lead the teacher to choose ones that are innovative and could provide positive beliefs (Loganathan & Meenakshi, 2016).Giving tasks that are low-level pressure is crucial especially at early meetings since this period should be a chance to attract students and make them enjoy the learning.This is a lesson the researcher got in the first immersion from students' reflective diaries.Many of them stated that they were excited to study at the park, but three students said that they were not confident speaking BI in front of their friends.With their low level of BI mastery, they were afraid of making mistakes.Therefore, encouragement was given over and over for them to freely ask questions whenever they found problems.The questions were not necessarily in BI as they could ask the teacher assistant using the Chinese language.It was meant to motivate those who were shy to use BI in front of the teacher to be more confident and active.
Besides authenticity, the interactive nature of the activities should also be considered.In this study, the immersion enables the students to engage in meaningful interactive learning which embraces the principles of Communicative Language Teaching (CLT), especially Task-Based Language Teaching (TBLT) (Klapper, 2003).Cooking Indonesian cuisine was found to be a great pedagogical activity applying the TBLT.That the students were enthusiastic to cook based on the recipes provided by the teacher and eager to ask and answer the name of fruits and vegetables they held by using BI are pieces of evidence that TBLT facilitates 'learning to communicate through interaction in the target language.'(Nunan, 2004, p. 1).Besides, the interactive activities in this study, especially when the students were grouped, were found to be effective to strengthen the bonding among students or between students and teachers.
Not only the out-of-class activities in the immersion that were interactive but the drilling and pronunciation practice in the instruction were also designed to be interesting.Considering that the technique of repeating what the teacher says in ALM is boring and being emotionally engaged with the tasks or the class is an essential motivating factor to maintain students' interest which then leads to better learning and grades (Kahu, Nelson, & Picton, 2017), the drilling activities were sometimes designed to be more like a game.In the form of a game, the repetitive nature of language practice is no longer tedious.
The second factor from which the success of pronunciation improvement was achieved was the learners' efforts.Through observation, the researcher found that the participants of this study had good study habits.They were highly motivated and accustomed to learning independently.Students, whether individually or in a group, read books or memorize words in the campus park or library from morning till night.Through the interview, it was revealed that their hard work was motivated by their dream to get prestigious occupations in the future.This finding is like that of Chidiebere (2018) who investigated China and United States students' cultural learning beliefs and found that Chinese students worked harder, spent more time on academic activities, and had a stronger commitment to educational goals than their American counterparts.
Besides, that the teacher and all students were living in a campus dormitory was another essential factor contributing to their accomplishment in this course.By living in the campus dormitory, it was easy for them to gather to execute the immersion activities.The complete facilities provided in the dormitory, such as canteen, mini market, laundromat, fitness center, swimming pool, and barbershop, enabled them to focus on their education without having to lose much time to mobile to fulfilling their daily needs.The positive impact of student housing on academic performance has been stated in several studies from different places, such as the ones of Fields (1991) in Iowa, USA, Najib, Yusof &Tabassi (2014) in Malaysia, andXulu-Gama (2019) in South Africa.
The last factor contributing to the success of the learning was the teacher.By combining two learning methods, instruction and immersion, he performed two different roles concerning the method being applied.In the instruction, which is teacher-centered, he controlled the class by giving an explanation, exposure, and corrective feedback in the drilling practice and became the role model of language use.Meanwhile, in the student-centered immersion, he became a facilitator guiding the students and managing the activities.Not limited to those roles, in both methods, the teacher tried to always create a positive and supportive learning environment by always paying attention to students' problems and needs.
After the first post-assessment was conducted.it was revealed that the most problematic sound the students faced was [r].They still could not produce the perfect [r] sound as what is usually pronounced by Indonesian people.They either pronounced it as light [r], which almost sounds like [l] or very hard [r] with very strong vibration.Nevertheless, reflecting on the first time listening to their pronunciation, the researcher felt that their pronunciation had noticeably improved.It was not perfect, but their words could be understood more easily than before and there was no meaning confusion perceived by the listener.From the reflective diary, it is known that the students could feel the improvement as well.Two students wrote that the text they read in the test was the same text they read at the first meeting, but they felt like it was easier for them to read it.
Small interviews with five students after they took the first post-assessment also revealed that they were content with the method and strategies implemented in the class.
(This is my first experience.I didn't get bored and I'm always excited. ) Susi : Kegiatannya macam-macam.Saya suka kelas ini.
(In this class, I learn differently.Interesting and enjoyable.) These comments indicated that the way the class was organized was acceptable and it led to good learning attitudes and behaviors shown by students' enthusiasm in participating in every class activity.However, despite those strengths, one student stated that the homework given was too simple.She wanted a more serious task.Reflecting on this, a higher-level task was implemented in the next cycle.
After post-assessment 2 was conducted, the result displayed that students' pronunciation had improved a lot.With a very hard vibration in the /r/ sound, some students tried to show that they could break the barrier.Although in quite a long pause between sentences, the words they pronounced in the test were far easier to be understood.In a small interview, after they took the second post-assessment, it was found that they enjoyed the learning process and the good interaction between the teacher and the students.However, one weakness that they conveyed was that they could not interact with other Indonesian people, except the teacher, from whom they can practice communicating by using BI.
(The many practices trained us to be good speakers of Bahasa Indonesia.)Hana: Kelas ini sangat menarik dan menyenangkan, terutama aktivitas di luar.Tidak seperti belajar.Hanya pergi jalan-jalan dan melakukan hobi.(This course is interesting and enjoyable, especially the outdoor activities.I didn't feel like it is learning.It's just hanging out and doing our hobbies) In this study, the immersion improved students' pronunciation and vocabulary mastery at the same time since they learn the words by directly looking at the objects.For example, when the focus is on pronouncing [r], by learning in a campus park they practiced pronouncing words with [r] sounds in a park, such as rumput (grass), ranting (twigs), rindang (shady), ramai (crowded), and burung (bird).By combining instruction and immersion, students do not learn language merely by theory.They do active learning since they practice using the words they learn in the classroom in the real context.The score improvement proves that combining these two methods is effective to enhance Chinese students' pronunciation in BI.However, this significant improvement does not imply that they were good at speaking BI since they were just beginners in learning the L2 and the text used for both pre and postassessment were the same.However, their positive responses to the learning method applied, their active participation in the activities, their improved self-confidence, and the stronger interpersonal relationship among them were shreds of evidence of their contentedness with the implemented strategies.

CONCLUSION
Many studies confirm that ALM is beneficial in forming a correct habit whereas immersion is effective to boost fluency through meaningful communication.However, ALM which is form-focused tends to neglect communicative competence, while meaning-oriented immersion will not be easy for beginners.Therefore, it is not applying each method in isolation that will contribute to effective language learning, but the combination of the two.The improvement of students' pronunciation scores gives evidence that it brings positive effects on pronunciation accuracy.Besides, the statements that they are more confident, enjoy the learning process and want to learn more are the more essential benefits.The instruction provides the foundation on phonetics and phonology with a lot of pronunciation practices while the immersion creates real language and cultural contexts.This combination provides intensive exposure to phonetic instruction and pronunciation practice with meaningful and communicative tasks.However, some methodological problems, such as the small number of participants, the uncomprehensive pronunciation assessment, and the inability to provide an immersion context with more native speakers might be the flaws of the study.These shall be better addressed by future studies.
stated that they have many differences.Many consonants in BI do not exist in Standard Chinese; they are voiced bilabial stop [b], voiced alveolar stop [d], voiced velar stop [g], palatal nasal [ɲ], velar nasal [ŋ], voiced palatal affricate [j], voiceless palatal affricate [c], voiced palatal fricative [z], voiceless palatal fricative [ʃ], voiceless velar fricative [x], and trill [r].The voiced labiodental fricative [v] considered as one of the differences is not counted since the consonant [v] and [f] in words are pronounced in the same way as [f] in BI.