Indonesian EFL Teacher Educators’ Beliefs and Classroom Practices on The Teaching of Pronunciation: Evidence from A Pilot Study

An exploration into EFL teachers’ teaching beliefs is since such beliefs are capable of their In the effort to fill the gap of the previous research, the current study attempted to unveil teacher educators’ beliefs on teaching English pronunciation in Indonesian EFL context as well as to describe in what ways their classroom practices matched their beliefs. A qualitative research design using a semi-structured interview was employed to collect data from three Indonesian teacher edu- cators who were experienced in teaching EFL. Similar findings to prior were dis-covered about their teaching beliefs, with inconclusive issue remaining about teaching prioritization. Their classroom practices also generally went hand in hand with their under- lying beliefs. noteworthy finding surfaced the class for pronunciation, especially in the context of training prospective EFL teachers.


I. Introduction
With the unavoidable interactions across nations nowadays, the use of English as one of the international languages has been growing considerably in countries beyond what is called the inner circle countries. Increasing number of people whose native language is not English utilize it for a variety of purposes (Fang, 2017;Mukminatien, 2012). Baker & Burri (2016); Setter & Jenkins (2005); Shah et al (2017) claim that, for global interactions to be successful, pronunciation is indispensable to facilitate international speakers in oral communication. Consequently, as Alghazo (2015); Baker (2014); Shah et al (2017) delineate, pronunciation has called for new pedagogical implication whereby the importance of pronunciation instruction needs to be reconsidered since the pedagogy of Second Language Acquisition has fairly neglected it.
According to Baker (2013); Reid (2016), many languages language teaching contexts, including EFL, have relatively ignored pronunciation teaching and belittled its function and importance.
The complexity of English pronunciation including what and how to teach it is the reason for the neglect (Suwartono, 2014). As a result, teachers frequently do not take account of it and tend to demonstrate excessive tolerance. If this is let to happen, pronunciation problems may hinder real communication which takes place outside language classrooms. In addition, it contributes to speakers' low competence (Shah et al., 2017).
In relation to this, pronunciation should be taught in EFL classrooms (Jenkins, 2000) even though Benzies (2013) ;Fraser & Perth (1999) assert that this area is strenuous for both learners and teachers. Teachers should not concern anymore with whether to teach pronunciation, but with how to teach it. Shah et al (2017); Szyszka (2016) exemplify that teacher may acknowledge the values of pronunciation and their pronunciation teaching capabilities, but still, they ignore the teaching of pronunciation. It shows that the underlying beliefs that the teachers have largely influence all their pedagogical decisions. Nevertheless, scant attention is paid on teacher think- ing and its relation to effective teaching (Yero, 2019). For that reason, there is a need to investigate teachers' beliefs and classroom practices in relation to English pronunciation teaching.
Teachers' thoughts and behaviors are based on their unconscious values about their own self, profession, and culture where they were raised. Their experience as students forms these values and they influence their beliefs, not only about their own self and capabilities but also about the character of knowledge as well as the ways to learn knowledge. Kuzborska (2011) asserts that these beliefs affect the professional growth and the pedagogical practices of teachers. Murray et al (2011) avow that their classroom practices also mutually influence their beliefs.
What teachers believe can "powerfully shape both what teachers do and, consequently, the learning opportunities learners receive" (Borg & Al-Busaidi, 2012). They guide their strategic pedagogical actions, such as in determining teaching goals, materials, classroom procedures, and patterns of interactions (Borg, 2015;Harste & Burke, 1977;Kuzborska, 2011). Such beliefs are crucial on teachers' pedagogical actions and hence it is imperative that reflection be made on what they think and do so that they can make informed decisions about what needs changing and how to make improvement on their practices (Kuzborska, 2011).
The beliefs of teachers are affected by social contexts, which include home and school factors (Ernest, 1989). Other than beliefs, these factors also shape their pedagogical practices. Baker (2011); A. Baker & Murphy (2011) claim that they include their previous experiences in learning second language, their education background or training, their experiences in teaching, their sharing with workmates, as well as their own reflection.
There are two principles which underlie L2 pronunciation teaching (Levis, 2005). One principle is the nativist principle, which focuses on developing native like pronunciation because Native Speaker (NS) norms are valued (Curnick, 2012;Sifakis & Sougari, 2005). The other principle is the intelligibility principle, which is viewed as more reasonable and realistic (Derwing, 2010;Roohani, 2013), in that it fosters learners to become comfortably understandable. It is generally understood that achieving native like pronunciation is difficult in EFL/ESL context.
Considering the second principle, World English varieties appear and disregard the inner circle varieties. Jindapitak (2015) argues that there is an urgent need to raise understanding and awareness of the emergence of World Englishers among L2 learners since this strategy will facilitate the effectiveness of nonnative speakers' communication. He advocates the use of listening materials which contain language stimuli produced by nonnative speakers. However, NS varieties should not be eliminated. They are still useful as point of reference for learners to produce sounds which do not deviate too much from the NS norms (Mukminatien, 2012).
There are two general approaches which distinguish the teaching of pronunciation (Alghazo, 2015). 'Bottom-up' approach pays attention to segmental features; meanwhile, 'top-down' approach has more to do with surpassed mental features, which are vital for intelligibility. These two contradictory approaches do not give teachers a clear direction about the best way to approach pronunciation teaching. Consequently, an alternative approach which pays equal attention to both features emerge. Alghazo (2015) claims that teachers have the right to decide teaching prioritization by considering learners' proficiency levels and learning goals. Their teaching priority is usually affected by pronunciation research and coursework (Baker, 2011b). Additionally, the context of teaching and the possibilities of the learners' use of the L2 should be considered. With these in mind, they can design the instruction to meet varied needs and expectations of the learners.
Furthermore, a specific approach can be adopted in the teaching of pronunciation. According to Hismanoglu & Hismanoglu (2010), the options are the intuitive imitative approach and the analytic linguistic approach. The two approaches are suited for youngsters and adults respectively (Roohani, 2013). Another alternative is the integrative approach, which focuses on the use of language for communicative purposes by promoting linguistic, discourse, sociolinguistic, and strategic competence and by practicing pronunciation within meaningful taskbased activities (Hismanoglu & Hismanoglu, 2010).
The teaching of pronunciation is recommended to be included in second language programs (Derwing, 2010). Studies by Baker & Burri (2016;Levis et al (2016;Saito & Shintani (2016) have found that explicit pronunciation teaching is effective to improve segmental features as well as suprasegmental features. However, research conducted by Shah et al (2017) has shown that teachers prefer incorporating pronunciation with other language skills. Additionally, Derwing (2010) asserts that pronunciation should also be evaluated. The provision of corrective feedback will enhance pronunciation Baker & Burri (2016), when it is made available following the teaching of features and reinforced by immediate practice.
Even though pronunciation is suggested to be integrated into second language programs, low confidence as well as reluctance seems to exist among many L2 teachers, be they native or nonnative, in dealing with varied aspects of L2 pronunciation. They are not sufficiently prepared for pronunciation instruction and for providing relevant resources (Baker, 2011b;Baker & Burri, 2016;Derwing, 2010). Accordingly, it is necessary that teachers are equipped with proper coaching (Baker, 2011a(Baker, , 2011bBaker & Burri, 2016;Roohani, 2013). Nonnative teachers also encounter more insecurity problems in pronunciation teaching due to native speaker fallacy. They are not confident as ideal models and teachers of pronunciation (Levis et al., 2016). Luckily, the misconception has been rectified by Alghazo, (2015); Levis et al (2016), who reveal that being nonnative benefits L2 pronunciation teachers in that they have comprehensive knowledge of being the target language learners and of how to cope with the ups and downs. More importantly, teachers' effectiveness in teaching pronunciation is determined by their knowledgeable teaching practices, and not on being native speakers.
Much of the previous studies in the area were conducted predominantly in the context of ESL (Baker, 2011a(Baker, , 2011bBaker & Burri, 2016;Curnick, 2012;Shah et al., 2017) and not many were in the context of EFL (Hismanoglu & Hismanoglu, 2010;Szyszka, 2016). Little is known about teachers' beliefs and practices in the context of teacher education. Besides, there are no conclusive results yet about which pronunciation features should be the focus of teaching, how to deliver them, and what teaching problems may appear. Hence, the present study endeavors to fill the lacuna of the previous research, i.e., by investigating teachers' beliefs and practices in relation to English pronunciation teaching in EFL teacher education. It attempts to specifically unveil the beliefs of EFL teacher educators on teaching pronunciation as well as to describe how their beliefs and their pronunciation teaching practices were related.

II. Method
Three EFL teachers from an EFL teacher training university in Yogyakarta, Indonesia, became the research participant. These EFL teacher educators are female, called Anna, Lina, and Rosy (pseudonyms). Anna and Lina hold Masters' degree in TESOL, while Rosy has Masters' degree in Linguistics. They all had English pronunciation coursework in their undergraduate study. However, no one was formally trained, except Rosy. All the participants were credible informants. They were purposively selected due to their pronunciation teaching experiences. More information about the research participants is detailed in Table 1.
The EFL teacher training university has two separate courses of pronunciation, which are placed in the first and the second semesters of the training. Every session in the two courses lasts 2x50 minutes a week. To ease the discussion, they were respectively termed low-level course, which dealt with segmental features, and high-level course, which concerned suprasegmental features. In the attempt to address the questions of the research, the study employed a qualitative design of which the data gathering tool was a semi structured interview. There were 25 questions, which were adapted from (Baker, 2011a). The reason for choosing the interview format was the possibilities to explore the responses of the participants during the data elicitation. All the three participants were interviewed twice separately. The second interview lasted shorter because the participants were only demanded to clarify some issues from their first interview. Triangulation was conducted by means of multiple sources of data whereby the data collected through interviewing the same people at different time (in follow up interviews) or different people were compared and crosschecked. When all the data were collected, important themes underlying the EFL teacher educators' stated beliefs and practices were extracted from the analyzed data. After comparing the themes across participants, data interpretation was conducted in which the findings of the current study were related to and compared with the findings of previous studies and/or the theoretical background.

A. The Beliefs of EFL Teacher Educators on the Teaching of English Pronunciation
All the participants acknowledged that pronunciation was generally essential to EFL learners. The acknowledgment lent support to (Yates & Zielinski,76  2009). They assert that in accuracy occurring in other parts of utterances will be 'covered up' by pronunciation which is decent. On the contrary, pronunciation which is unintelligible will not be possibly counteracted by remarkable sentence structure as well as vocabulary. The three participants have varied underlying reasons despite their common answer. Two participants put forward the significance of pronunciation to foster one's intelligibility, whereas the other one stated that pronunciation was a critical measure of one's English proficiency (Shah et al., 2017).
Moreover, the three participants emphasized the worth of pronunciation in training future English teachers. Their respective arguments where pronunciation was a skill which a teacher should master, pronunciation was essential for future modelling, and pronunciation was vital to enhance the sensitivity and awareness of prospective teachers to anticipate difficulties of learning encountered during their training to become prospective teachers as well as role models. The last argument confirmed the argument that having nonnative characteristics benefitted teachers in teaching pronunciation (Levis et al., 2016). They would have the ability to assist their students better because they share difficulties in learning and hence, they function as ideal examples of second language learners.
All the participants agreed that the teaching of pronunciation should not be aimed at producing native like speech. Rather, consistent with (Roohani, 2013), the aim of pronunciation teaching should be to develop learners' ability to communicate the messages smoothly without causing any breakdown. The three participants added that, with reference to World Englishers, the teaching of pronunciation should ultimately promote intelligibility. The finding was parallel with (Baker, 2014). Moreover, the three of them admitted that they made use of dictionaries as references and did not impose certain varieties on their students. The finding demonstrated the EFL teacher educators' tendency towards NS varieties as the classroom models despite aiming at intelligibility (Jindapitak, 2015).
All the three participants agreed on the necessity to expose students to World English varieties in the pronunciation classroom (Jindapitak, 2015) and felt that it was urgent to make learners aware of various accents. There were many ways which the EFL teacher educators did. One participant stated that she made use of audios containing a variety of accents. Subsequently, she asked the students to listen to as well as to spot the dissimilarities. Additionally, she told them to predict where the speakers in the audios came from. Her purpose was to show that English language was developing as well as to raise students' phonological awareness. On top of that, she would like to build students' self-assurance in the pronunciation they produced on condition that they could sustain intelligibility. The second participant made use of videos to illustrate accents that she herself could not perfectly exemplify or imitate in her attempt to demonstrate to students that people did not speak in the same ways. What both endeavored was in line with (Jindapitak, 2015) in those listening materials which contain language stimuli produced by nonnative speakers can be utilized to make students understand and aware of World English varieties. Language classrooms were viewed as a beginning to promote the appreciation of the world as well as to nurture a sense of open mindedness towards diversities. Similarly, the third participant claimed that students' language awareness could be raised by exposing them to various accents. Unfortunately, she rarely exercised the claim owing to time limitation. The finding revealed that the belief which one holds may not appear in class practices.
As for the nature of pronunciation class, the three participants had the same opinion that the teaching of pronunciation should be done in a dedicated course in addition to being integrated with other skills. Such a course would be more effective to build proficiency because extra help, energy and time could be spent on shaping students as future models. With a dedicated course, more focused pronunciation exercises could be carried out to avoid distraction, especially in coping with features which were problematical. As they revealed, students usually found difficulties in learning consonants which were different from or absent in their L1, some diphthongs, and some suprasegmental features. Their account was unlike previous studies conducted by Arslan (2013); Kucukoglu (2012), which claim that stress is the biggest problem for second language learners whose first language is syllable timed. Following Kucukoglu (2012), these problematic features were made the center of attention of the intense pronunciation practices, such practices were far more possible when there was more time allotted in a dedicated course. These arguments expressed all the participants' beliefs that improving pronunciation in segmental features as well as suprasegmental features could be done through explicit instruction, as revealed by (A. Baker & Burri, 2016;Levis et al., 2016;Saito & Shintani, 2016).
Various perceptions on their confidence as teachers of pronunciation were demonstrated by the three participants although each of them favorably appraised their own teaching ability (Szyszka, 2016). As (Baker, 2011b) claims, their confidence seemed to stem from the possession of sufficient knowledge of the content as well as the instruction which they obtained from their undergraduate training. However, the finding was different from (Baker, 2011a) investigated was not enhanced by having adequate information from pronunciation related research as well as coursework.
Rosy, being the most experienced of all, was confident due to her seniority, abundant experience in the area, and her profound linguistic knowledge. Nonetheless, she sometimes felt uncertain with her own content knowledge, which pushed her to consult dictionaries or theories. Lina, albeit her mediocre teaching experience, was highly confident. Her confidence may originate from her being an experienced interpreter. Nonetheless, the assertion needed to be empirically validated. By contrast, Anna had the lowest confidence among the three. This may happen because she was the least experienced and she oftentimes encountered uncertainty about a few of her teaching decisions. She was self-assured because she was highly attracted to English pronunciation. The finding demonstrated that all of them were sufficiently confident pronunciation teachers even though they were not native speakers. It further indicated that native speaker fallacy was absent among them. In line with (Baker, 2011b), their confidence may stem from their sufficient understanding of pronunciation concepts as well as the ways to teach them. In addition, it may be affected by their experiences during the process of learning the second language. The finding agreed with (Levis et al., 2016), who argue that knowledgeable teaching practices determine the quality of effective pronunciation teachers.
The significance of formal training in teaching pronunciation effectively was also affirmed by the three participants. Anna and Lina avowed that formal training would guide them and supplied the basics for teaching and learning pronunciation. The former argued that the theories, pronunciation development, and techniques of teaching were knowledge that a pronunciation teacher should have. The latter was consistent with Tudor (2001) in that a better understanding of pronunciation could be fostered through giving explanations phonologically. Rosy opined those linguistic theories which one learned could be chiefly revitalized through formal coaching in which the prospective English teachers would be asked to recollect what they knew of English sound system in comparison with that of the source language. The finding implied that comprehensive training in pronunciation pedagogy would enlighten teachers' knowledge base of pronunciation techniques to be applied in the classroom (Baker, 2014).
Being inadequately trained formally to teach pronunciation, all of them depended largely on their undergraduate coursework experience, which they considered helpful for modelling. Nevertheless, two of them attempted to improve their teaching instead of relying only on their coursework experience. Anna discussed with other colleagues and tried to look for relevant resources. Meanwhile, Rosy, informed by her teaching experience in Phonetics and Phonology, attempted to upgrade her competence by studying varied literature in the pertinent field. Her situation strengthened the assertion that "the teacher's scientific knowledge of articulatory phonetics is shown to be successful in leading students to the correct pronunciation and discrimination of foreign sounds" (Roohani, 2013). The finding partly confirmed (Baker, 2011a). All the participants' cognition and pedagogy were mainly affected by their undergraduate coursework, and in the case of Rosy, by her search of knowledge from various resources.
As for teaching prioritization, all the participants accommodated segmental features as well as suprasegmental features. However, their emphases were not the same. Believing that suprasegmental features were more important than accuracy of sounds (Arslan, 2013), Lina concentrated more on those features, particularly stress. Conversely, Anna opted for segmental features believing that they contributed to intelligible speech. What she believed paralleled previous study which divulged that segmentally deviant pronunciation impedes intelligible speech (Rang & Moran, 2014). Concerning the sequence of teaching, two participants preferred introducing segmental features first then continuing to suprasegmental features.
In a word, they favored 'bottom-up' approach (Alghazo, 2015). Their preference, supporting Shah et al (2017), was to some extent encouraged by their eagerness to enhance intelligible pronunciation among their students besides being advised either by the curriculum or the references. On the contrary, Rosy chose a balanced approach (Alghazo, 2015) whereby both features, believing that they impacted intelligible speech (Baker, 2013), were accommodated in an integrated manner. Despite the integration, she chose to have varied teaching focus per meeting reflecting what she prioritized, which was informed by her capabilities as well as the curriculum. The finding contradicted (Baker, 2011a), in which her research participants opted for suprasegmental features. Their decisions were partially influenced by what they learned and experienced. It concurred with (Baker, 2011b;Baker & Murphy, 2011). Up to this point, teachers' beliefs about prioritized features and teaching sequence remained inconclusive.
In conjunction with techniques of teaching, all the three participants claimed that drilling was the most effective to form mechanical habits. The finding revealed EFL teacher educators' favor for traditional teaching technique because of its merit, as (A. Baker, 2014Baker, , 2013Hismanoglu & Hismanoglu, 2010) discover. Moreover, it unveiled that the three participants were inclined to imitate their past learning (A.  Baker, 2014). Although the technique was criticized as not communicative, the participants still approved of the use of controlled techniques. Furthermore, the finding suggested that these teacher educators partially failed to become good teachers of pronunciation, based on the argument of Shah et al (2017) that to teach pronunciation effectively teachers are required not only to possess a variety of teaching strategies but also to gear the teaching towards communication.
All participants preferred adopting the analytic linguistic approach to train future EFL teachers, which supported (Alghazo, 2015;Roohani, 2013). It was apparent from their acknowledgement that teaching phonetic alphabet was imperative. They referred to International Phonetic Alphabet (IPA) in their teaching. They claimed that students could be made aware of the correct pronunciation through visually representing the words by means of such alphabet. Besides, they stated that pronunciation could be fundamentally developed through the knowledge of sound correspondence.

B. The Relation between What EFL Teacher Educators Believed and Practiced in Teaching Pronunciation
To teach the low-level course, all the three participants used teaching materials which provided learners with rich and clear examples of segmental features in the forms of words and sentences. Besides, the materials also contained enough drilling practices. To teach the high-level course, they utilized learning resources with varied contents on segmental and suprasegmental features. Additionally, they sought resources accompanied by the audio software which exemplified World English varieties for independent learning. These EFL teacher educators' practices were consistent with what they believed about drills as a possible form of practicing intensely and about a pressing need to familiarize learners with World English varieties.
The techniques which the three participants employed in teaching segmental features were reported as follows. One participant exposed student to individual or group practices in which they should pronounce words or sentences containing the target sound. Another participant presented the target sounds, checked how much her students knew about them, provided models, repeated after the model, and practiced in groups. The other participant distinguished the teaching of diphthongs from the teaching of other sounds. Diphthongs were taught by telling students to identify and to make comparison between examples of diphthongs in Indonesian as well as practicing words containing diphthong sounds. In contrast, vowels and consonants were taught by explaining how sounds were produced as well as conducting corresponding practices on the target sound.
Discriminating sounds and doing relevant practices in a more advanced form were the subsequent activities. Overall, traditional teaching techniques obviously dominated the three participants' reported classroom practices, which confirmed their beliefs about the superiority of controlled practice.
The teaching of suprasegmental features adopted varied techniques. To teach stress, the first and the third participants instructed students to locate the word stress in polysyllabic words, and then to confirm their predictions, they should consult dictionaries. Afterwards, they should properly pronounce them. The second participant employed noun verb stress contrasting practices, played games, as well as read poems. Such activities may be carried out independently or collaboratively. To teach rhythm, all the participants used different techniques. The first participant made use of sentence reading to raise students' awareness of beats. The second participant made use of jazz chants, whereas the last participant employed songs, rhythm practicing, as well as creating rap chants. Meanwhile, to deal with intonation teaching, the first participant utilized explanation of theories, listening to examples, and identifying intonation in sentences. Furthermore, the second participant presented modelling in sentences or textual forms, discussed the models, and instructed students to practice. The third participant utilized dialogues and short stories. From this account, controlled techniques evidently still dominated their classroom activities. They may have chosen such techniques to be implemented due to their positive impact on the development of intelligibility and the improvement of phonology (Baker, 2014). In addition, their choice was also aligned with what EFL students needed in pronunciation learning (Alghazo, 2015). Again, this finding indicated that their reported classroom practices corroborated their belief about the strength of controlled practice.
In conjunction with assessment techniques for segmental and suprasegmental levels, the three participants employed the same strategy. Each student was instructed to utter the given words, phrases, and sentences. They all tried to reduce subjectivity (Moskal, 2000) and justify the grading; two of them did it by recording the performance of each student and providing judgement accountability using scoring rubrics, while the other one manually jotted the errors. In conducting pronunciation assessment two participants would always provide corrections to handle any pronunciation problems. It implied that little tolerance was demonstrated towards errors. This finding lent support to (Baker & Burri (2016), who unveil that corrective feedback is vital to improve pronunciation. The feedback modes they employed were immediate and delayed. All participants preferred immediate correction (Alghazo, 2015) when the errors occurred during individual drilling and/or when the errors were encountered on the levels of sound or word. On the other hand, delayed correction was applied when the pronunciation problems were encountered on the level of text. Alternatively, it was also applied when students' practices were done in groups. The pattern of giving feedback agreed with (Baker & Burri (2016), who advise that "effective pronunciation focused feedback should revolve around the following: Explicit form focused instruction of a specific target feature should be followed by corrective feedback and strengthened by immediate practice." The use of dictionary as the grading standard in whichever mode of feedback giving activities reflected what they believed about the supremacy of NS-oriented classroom models

IV. Conclusion
Previous learning, professional experiences, and other contextual factors apparently shaped EFL teacher educators' beliefs. Mean-while, their previous experience as students as well as their own experience while teaching had more influence on their pedagogical practices. The beliefs of the EFL teacher educators which were unveiled in the study were mainly like what previous studies had discovered. The exception was on teachers' beliefs about what they should prioritize in their teaching, which could not be concluded yet. As for the agreement between what the EFL teacher educators believed and practiced in their teaching, no significant discrepancies were found. One thought-provoking issue which emerged from the research was the idea that pronunciation was worth paying attention to and hence should be taught in a dedicated course, specifically in training EFL teachers. Pronunciation will not be intuitively developed over time; it needs explicit training to make students aware of it. The findings of the current study provide initial recommendation for EFL teacher education programs to offer at least a single pronunciation course to equip future EFL teachers adequately with fundamental pedagogical knowledge of pronunciation, which will finally make them confident.