LANGUAGE CIRCLE: Journal of Language and Literature XI/1 (October 2016) p-ISSN 1858-0165 Available online at http://journal.unnes.ac.id e-ISSN 2460-853X __________________________________________________________________________________________________ 47 SAFETALK PRACTICES IN CONTENT AND LANGUAGE INTEGRATED LEARNING (CLIL) CLASSES Sri Wuli Fitriati sriwuli.fitriati@mail.unnes.ac.id English Department Universitas Negeri Semarang Indonesia Received: 15 September 2016. Revised: 20 September 2016. Accepted: 25 October 2016 Abstract This article is drawn from a larger ethnographic case study of a state senior high school in a city in the province of Central Java, Indonesia. The study was classroom discourse analysis, focusing on identifying and examining classroom interactional and pedagogical practices in Content and Language Integrated Learning (CLIL) lessons. CLIL is a type of bilingual education in which some content areas (such as Sciences) are learned through a foreign language. The research participants in this study were Indonesian teachers of Mathematics, Biology, and Geography who were encouraged to teach their subjects in English language because of the political ideology and educational policy at the time the data collected. The data emerged from classroom observations notes, lesson transcripts, and post-lesson interviews with video- stimulated recall. The interpretation and analysis of the data involved cross-checking different sources of evidence. Braun and Clarke‟s thematic analysis (2006) was employed in the analysis of the data. The finding reveals that there was a salient pattern in the interactional and pedagogical practices of the subject teachers which can be called safetalk. The finding also shows that safetalk hinder the students‟ second language acquisition and development. This article offers deeper insights on the interaction and pedagogy in CLIL classes, which can promote critical reflection and contribute to future English bilingual teacher education, particularly in the Indonesian educational contexts. Key words: CLIL classes, interactional practices, safetalk, classroom discourse analysis. How to Cite: Sri Wuli Fitriati. 2016. Safetalk Practices in Content and Language Integrated Learning (ClHIL) Classes Language Circle: Journal of Language and Literature, XI/1. ________________________________________________________________________________________ INTRODUCTION This article focuses on examining classroom practices of three Indonesian senior high school teachers in their attempts to use English alongside Indonesian in the teaching of Biology, Geography, and Mathematics. The context of this study was in the year 2011 to 2013 when the Indonesian central government enacted the educational policy on the use of English alongside Indonesian in some selected Indonesian public schools. The educational policy was eventually withdrawn following the Decree of Mahkamah Konstitusi/the Indonesian Constitutional Court in 2013. This article will not re-open the controversies over the use of English to teach content subjects in some selected state schools, but it will provide deeper insights on what was happening in the classrooms when the subject teachers were expected to teach mailto:sriwuli.fitriati@mail.unnes.ac.id LANGUAGE CIRCLE: Journal of Language and Literature XI/1 (October 2016) __________________________________________________________________________________________________ bilingually. This article is divided into four sections. First, in the introduction, I provide the background to the study, the context of the study, the literature review, and the research objectives. Then, I briefly explained the methodology used. In findings and discussion, I discuss the prevalent interactional and pedagogical practices observed in the three bilingual Science classrooms. Finally, some educational implications of the findings are discussed in the conclusion. The use of English as the medium of instruction alongside Indonesian in some selected schools might be classified as a form of bilingual education. Among the many definitions of bilingual education, Baker (1993), one of the most influential scholars in the field of bilingual education, sometimes used the term bilingual education “to refer to the education of students who are already speakers of two languages, and at other times to the education of those who are studying additional languages” (p. 9). Compared to Baker‟s (1993) definition, Garcia (2009) referred to bilingual education programs as those that use a second or a foreign language “as a medium of instruction; that is, bilingual education programs teach content through an additional language other than the children‟s home language” (p. 6). Bilingual education is different from language education programs that teach a second or a foreign language. Second or foreign language education programs teach the language as a subject, whereas bilingual education programs use the language as a medium of instruction. The use of English in teaching Science and Mathematics as first enforced by the government was categorized as Content and Language Integrated Learning (henceforth, CLIL). In the United States, it is sometimes called content-based second language instruction, while in Europe, it is referred to as Content and Language Integrated Learning CLIL (Baker, 2011). It is not about teaching English language for its own sake as in second or foreign language lessons, but teaching content subjects by using English as the language of instruction. The content subjects are those which are currently studying by students in their school classes, such as history, science and mathematics. In this article, I use the terms bilingual education, bilingual teaching, and CLIL interchangeably as they convey the same meaning, that is, teaching content subjects through an additional language in which the main goal is to prepare students for the integration of language teaching with discipline subject content instruction. CLIL develops access to subject-specific target language terminology and improve overall target language competence, including develop oral communication skills (Daiton- Puffer, 2007). The promotion of the bilingual teaching in the selected schools set the main goal to enable school students and school leavers to communicate with the outside world and to provide the language skills which are marketable, aiding employment and status. The theoretical foundation for CLIL is drawn from second language acquisition theories and research. According to Krashen (1985; 1994), second language acquisition occurs when the learner receives comprehensible input, not when the learner is memorizing vocabulary or completing grammar exercises. Therefore, teachers in bilingual education contexts should be competent in the target language and take account of second language learning pedagogy to be applied in classes (Baker, 2011; Garcia, 2009; Graaff, Koopman&Westhoff, 2007; Snow, 2001). In bilingual education contexts, there Sri Wuli Fitriati. 2016. Safetalk Practices in Content and Language Integrated Learning (ClHIL) Classes Language Circle: Journal of Language and Literature, XI/1. _________________________________________________________________________________________ emerged some prevalent patterns of classroom interactions which Chick (1996) called „safetalk,‟ a recurrent feature in bilingual settings. Safetalk is a very common interactional strategy of classroom language practice. The notion safetalk was termed by Chick (1996) who investigated the characteristics of interactions in schools for black people in South Africa under the former apartheid system. Chick‟s background of study was that he observed classroom practices which were “highly centralized, with teachers adopting authoritarian roles and doing most of the talking, with few pupil initiations, and with most of the pupil responses taking the form of group chorusing” (p. 21). Hornberger then collaborated with Chick (2001) to examine safetalk practices in Peruvian and South African classrooms. They defined safetalk practices as language practices in which teachers and students preserve their dignity by hiding the fact that little or no learning is taking place (Hornberger& Chick, 2001). In research which employed a discourse- analytic study to look at classroom language practices in two classrooms, primary and secondary, in two rural schools, Martin (2005) found that safetalk practices were present. His study was against the backdrop that beginning in the year 2003 Malaysia decided to implement a policy to switch the medium of instruction in Mathematics and Science from Malay to English. His findings revealed that what usually went on in bilingual classrooms were slot-filling, labeling and chorusing. Martin (2005) suggested that “such practices can create an illusion that learning (of both content and English) is taking place” (p. 83). Following Chick (1996) and Hornberger and Chick (2001), Martin (2005) used the term “safe” but he claimed that he gave his own emphasis that safetalk practices allow classroom participants to be seen to accomplish lessons. According to him, safetalk is “a pragmatic discourse strategy (linguistic strategy) that is “safe” in that, potentially, it facilitates comprehension” (Martin, 2005, p. 80). While Martin (2005) asserted that safetalk was used at least for two functions - to be seen to accomplish a lesson and to facilitate comprehension – Chimbutane (2009, 2011) emphasized that “the use of safetalk strategies was prompted by the language barrier and by the teachers‟ limited preparation to using appropriate second- language teaching strategies to minimize this barrier” (p. 87). That is to say, teachers used safetalk strategies because of their limited proficiency in the medium of instruction. Chimbutane (2009, 2011) conducted an ethnographic study in two primary bilingual schools in Mozambique. A language-in- education policy in Mozambique shifted from a Portuguese-only-system of education (which is a second/foreign language for most Mozambican citizens) to a situation in which local African languages were promoted in formal education through the gradual introduction of a bilingual program. In subject classes in which Portuguese was the medium of instruction, Chimbutane (2011) observed that most of the pupils used chorusing, that is they chorused responses and chanted. He concluded that the teachers and their pupils were aware of the lack of communication between them. Therefore, in order to preserve their dignity and give the sense of accomplishment, they colluded in using safetalk strategies such as group chorusing and clued elicitation. To sum up what safetalk is, Chimbutane (2011) explained that safetalk is “a term used to refer to teachers‟ and pupils‟ use of interactional strategies that allow them to preserve their dignity by avoiding opportunities for displays of academic or LANGUAGE CIRCLE: Journal of Language and Literature XI/1 (October 2016) __________________________________________________________________________________________________ linguistic incompetence” (p. 27). Chimbutane (2011) further explains: The key pattern of safetalk is that of teacher prompt and pupils‟ choral responses, that is, teachers routinely provide cues to which pupils respond in chorus. The prompts or cues used by teachers to trigger such pupils‟ chorusing responses include yes/no questions and oral gap filling exercises. In these exercises, teachers provide incomplete words or sentences in which they raise the tone on accented syllable leaving an oral gap for pupils to fill in, for example, with a syllable, word, of phrase (pp. 27-28). Teachers and students may use safetalk to respond to their constraints including language barriers. The use of safetalk is a pervasive discursive strategy used to ensure classroom interaction flow in bilingual and multilingual contexts. Following consideration of the background to the study, research context and literature review I explained above, I developed a key focal area for investigation. This study endeavored to address the following question: How were teachers‟ interactional and pedagogical practices in their bilingual Science classrooms? Therefore, my study aimed to identify and examine the teachers‟ classroom interactional and pedagogical practices as they were expected to teach their subjects bilingually, in Indonesian and English languages. METHODOLOGY The school as the site of the study was a state senior high school in a city in the Central Java Province, Indonesia. During my fieldwork in 2012 to 2013, the school had been implementing English bilingual education since 2009 where Science and Mathematics teachers were expected to teach their subjects bilingually, Indonesian alongside with English. The study was classroom discourse (Chimbutane, 2011). The participants in this study were the teachers of Mathematics, Biology, and Geography. The teachers were qualified with a bachelor degree from local universities. None of them had experienced special training in English bilingual education program during their study at university, but they took English as a compulsory course for one semester. The three teachers undertook short English language training during their in-service at the school. Each lesson ran for 90 minutes. The techniques of data collection were classroom observations, including classroom observation notes and classroom video recordings, and post-lesson interviews with video-stimulated recall, which means that after the classroom observation, I showed the video to the teachers and conducted in-depth interviews with them to seek views on their language behaviors and interactions. I transcribed all the recorded lessons, and analyzed the transcripts to identify the commonalities of classroom interactional and pedagogical practices by using discourse analysis (Palmer, 2011). To analyze and interpret the emerging patterns of the classroom observation notes and interviews, I used thematic analysis (Braun & Clarke, 2006). The three datasets (classroom observations notes, lesson transcripts, and in-depth interviews with video-stimulated recall) functioned to provide “descriptive complementarities” (Palmer, 2011, p. 109). Therefore, the interpretation and analysis of the data of the current study involved triangulation of multiple sources of data and used multiple data-gathering techniques which led to the trustworthiness of this study. Trustworthiness is “validity of the data collection and analysis methods” in Sri Wuli Fitriati. 2016. Safetalk Practices in Content and Language Integrated Learning (ClHIL) Classes Language Circle: Journal of Language and Literature, XI/1. _________________________________________________________________________________________ qualitative research (Mills, 2007, p. 308). I also employed debriefing (Creswell, 2013) for the trustworthiness of the current study. After data collection phase, I showed my participants my preliminary analyses consisting of description of themes. By this way, I tried to seek my participants‟ alternative interpretations to my analysis. In the discussion that follows, all the names of the teachers are pseudonyms, and the transcription conventions used in this study can be seen in the Appendix. FINDINGS AND DISCUSSION The findings revealed that, aware or unaware, teachers together with students appeared to hide their poor command of English which might have given a sense of accomplishing a lesson. The apparent dominant language practices were: students‟ chorused responses to teachers‟ questions and cues, teachers‟ elongating and raising intonation of the last word of phrase of their questions, and response slot. Students’ Chorused Responses Chorusing or answering teachers‟ questions together occurred often during the lessons, as can be seen in Extract 1 which comes from the Mathematics lesson. In this episode, after the orientation stage of the lesson, Ms Tuti attempted to elicit her students‟ responses to define the concept of cyclical permutation. Extract 1 (Math lesson – Cyclical permutation): 1 Ms Tuti : OK, yuk, let we define about cyclical permutation. 2 If we have two volunteers, we get one arrangement. And if we 3 have three volunteers, we have two arrangements. And if we 4 have four volunteers, we get six arrangements. One if you written 5 in one factorial permutation one equal? 6 Ss : One factorial. [chorusing] 7 Ms Tuti : And two? 8 Ss : Two factorial.[chorusing] 9 Ms Tuti : Two factorial. How about six? 10 Ss : Three factorial. [chorusing] It can be seen in Extract 1 that the students provided choral two-word answers, shown in lines 6 to 10. Despite asking her students to voluntarily raise hands to respond to her cues (lines 1 to 5), Ms Tuti let her students answer in chorus (lines 6, 8, 10). This extract suggests that the teacher assumed that all her students must have understood the concept of cyclical permutation by the use of choral responses. Stutz‟s teaching practice shows classroom interactions lacked individual spot checks as a means of checking students‟ understanding of the lesson content. The episode transcribed in Extract 2 was taken from Mr. Cho‟s class where the teacher was eliciting names of parts of the skull in Latin from his students. Extract 2 (Biology lesson – the Skull) 1 Mr. Cho : Apa itu? [pointing on the slide/screen] 2 Ss : [students were mumbling; their voices were indistinct; some were laughing] 3 Mr. Cho : Occipital. [teacher provided a correct answer] 4 Mr. Cho : Number seven? 5 Ss : Mandibula. [chorusing] 6 Mr. Cho : Number six? 7 Ss : {Spenoid. [chorusing] LANGUAGE CIRCLE: Journal of Language and Literature XI/1 (October 2016) __________________________________________________________________________________________________ 8 {Zigomaticum. [chorusing] 9 Mr. Cho : Zigomaticum. [confirming the answer] My classroom observation noted that not all students stated the correct names of parts of the skull that the teacher pointed to on the slide. From the extract 2, it seemed that the teacher was satisfied with the class chorus, but when the teacher selected a student to answer, this student provided a wrong answer. Most students mumbled, saying words quietly and unclearly (lines 7-8) to the teacher‟s question until the teacher himself answered his own question (line 9). These characteristics of interactions demonstrated that the teacher tolerated choral responses and lacked individual spot-checking required to enhance the accomplishment of learning the content. However, Mr. Cho had his own point of view. By allowing the students to chorus their responses to his questions, Mr. Cho justified to himself that he was using a strategy to encourage the students to be active and engaged in the class (Interview, 28 September 2012). By chorusing or answering the teacher‟s questions together in English or Latin, they would not be afraid of making mistakes, since the students and the teacher would not notice which student was answering incorrectly, and this would avoid students‟ embarrassment particularly for the shy students (Mr. Cho, interview, 28 September 2012). This assumption is only partially correct since observation found that there were some students who simply did not answer or respond to the teacher‟s questions. These students were just silent, and it seemed that they relied on other students to answer the questions. There is no harm in a certain amount of chorused responses, but it is not good practice in terms of the learning process because it will never support creative production in students. If this happens all the time during the study in the school, it is possible that some students will depend on their peers and will not achieve maximum progress in the learning process. Response slot Safe talk practices also included teachers‟ particular questioning techniques. My observations suggested that the teachers mostly used a certain type of questioning technique, that is, elongating and raising the last word of phrase of their questions. For example, on several occasions in Ms. Lis‟ lesson, she elongated some last words (see the underlined words in extract 3 line 6). By elongating and raising her intonation for the last words in her questions, she expected her students to complete the slots with correct words or answers. Extract 3 below (line 6) illustrate that the teacher employed a question technique in which she created gapped contexts for which the students provided the deleted element, that is, a one-word or single phrase answer. Extract 3 (Geography lesson - Industry) 1 Ms. Lis : Next number two. Segala seasuatu untuk memenuhi kebutuhan 2 : manusia, to daily needs, untuk memenuhi kebutuhan hidup sehari- 3 : hari. For example? Ilham [pointing a student]. Please mention the 4 : example. Kebutuhan sehari-hari apa? 5 Ilham : Garmen. 6 Ms. Lis : Garmen. Ada kata dihasilkan, berarti ada proses...? 7 Ss : Pembuatan. [chorusing] Another example (Extract 4) was taken Sri Wuli Fitriati. 2016. Safetalk Practices in Content and Language Integrated Learning (ClHIL) Classes Language Circle: Journal of Language and Literature, XI/1. _________________________________________________________________________________________ from the closing stage of Ms. Tuti‟s lesson where the class was encouraged to synthesize or conclude about the concept of permutation. Extract 4 (Math lesson – Cyclical permutation) 1 Ms Tuti : OK... about cyclical permutation. First to make conclusion about 2 : our lesson today. Siapa yang mau membuat kesimpulan mengenai 3 : pelajaran kita hari ini? About cyclical permutation. Who? Who? 4 : Every body? Who? Okay, please you, you can make conclusion. [pointing to a student who is voluntarily raising her hand] 5 S : The conclusion of permutation, we can learn the conclusion is 6 : the number of cyclical permutation is of n object is p equal and 7 : minus one factorial. [the student pointed by the teacher is reading from the textbook] 8 Ms Tuti : OK, good! 9 Ms Tuti : The cyclical permutation, the number of cyclical permutation in 10 : of element is...? 11 Ss : p equals n minus one factorial. [chorusing] 12 Ms Tuti : OK, good! After several prompts from the teacher to make her students say the definition of cyclical permutation (lines 2 to 4), one student voluntarily raised her hand. This student then read from her textbook the definition of cyclical permutation, instead of constructing the definition in her own words. The teacher complimented this student‟s willingness to participate (line 8). Then, the teacher repeated the definition by giving a cue that the whole class had to complete her utterance. The cue was when the teacher paused to cause the students to finish her incomplete utterance (Extract 4 line 10). Elongating words (Extract 3 line 6 and Extract 4 line 10) was a questioning strategy which has been called “response slot” by Martin (2005, p. 80), a term similar to “gap filling” by Swain and Johnson (1997, p. 177). Response slot or gap filling appeared to be a common language practice used by the teachers, regardless of whether they used Indonesian, English, or mixed the two languages. The language practices, students‟ chorused responses to teachers‟ questions and cues, teachers‟ elongating and raising intonation of the last word of phrase of their questions, and response slot, might be considered by the teachers as safe strategies to potentially facilitate comprehension (Martin, 2005). Chick (1996:26) referred to this type of strategy, that is, the “completion chorus phenomenon” as “safe-talk” which enabled teachers and students “to hide their poor command of English,” and “to obscure their inadequate understanding of academic content”. When being questioned about her safetalk practices, Ms. Lis explained: That sort of behavior is very typical of students and commonly found in classes. It happens in Year 12 and also in Year 10. They usually answer without raising hands first but directly answer in chorus and they mumble. I think it is because they do not have self-confidence. So, it seems that when they raise hands and their answer is not correct, they feel embarrassed. I think they are too worried. If I ask them to raise their hands, they do not answer, they are even silent. That is a LANGUAGE CIRCLE: Journal of Language and Literature XI/1 (October 2016) __________________________________________________________________________________________________ weakness. When they are speaking together at the same time, it seems all students are active, and the class sometimes becomes disoriented. I must be able to control and manage the class, so I give them turns and say which student should speak first. (Ms. Lis, interview, 6 September 2012 – translated from the Indonesian language by the researcher) Ms. Lis‟s response suggested that the possible cause of chorus answering was that the students were not confident enough to speak independently. They were afraid of making incorrect answers. It seemed that when there were students who raised their hands, the class‟s attention was directly drawn to a particular student, and if her/his answer was not correct, she/he would feel embarrassed. Ms. Lis reflected on her classroom practices in the following way: I have asked students to put their hands up before answering, but the class became quiet, not what I expected. Actually this has become a concern of mine. I often ask them what makes them afraid to raise their hands or ask questions. I say that they must have guts, but they say that they are afraid if they answer incorrectly. The more I demand them to raise hands, the more the class becomes silent. When the situation was like this, I myself did not feel comfortable. Finally I let it go. (Ms. Lis, interview, 6 September 2012 - translated from the Indonesian language by the researcher) Analysis of classroom observations revealed that students‟ chorus answers, teachers‟ elongating questions, gap-filling or slot response, and teachers‟ lack of individual spot checking were common practices in the three classrooms observed as these practices “allow the classroom participants to be seen to accomplish lessons (Martin, 2005, p. 89). The nature of the teacher and students‟ interaction was the gauge of the quality of teaching and learning process in the classrooms. The illustrations of the data provide evidence that the teachers appeared to hide inadequacy of their English language competence. The teachers and the students might be aware of the lack of communication between them so, in order to give the sense of accomplishment - the impression that the students understood the subject contents - they used the safetalk strategies. The data presented in this section confirmed that the use of English in the classrooms was regarded as a heavy task both for the teachers and the students. Instead of giving the students a language model integrated with teaching content, the teachers‟ language practices in the classroom interactions could hinder the students‟ English language acquisition and development. In a pedagogical viewpoint of bilingual education, teachers‟ language use is their students‟ language input, and this input has a determining function in the acquisition of language (Ellis, 2008). So, if too often teachers gave inaccurate model of English language use in classrooms, it would make the students exposed to what might be unacceptable English use. Based on the features of interactional and pedagogical practices in bilingual classrooms presented above, I can conclude that the teachers used the English language as to be seen complying with the expectation of the government despite their language barrier. This study showed that the teachers were in need of the knowledge of second language pedagogy and guided practices in bilingual education, not to mention improving and developing their own English communicative competence. Taking into account all the evidence found in this study, I believe it was Sri Wuli Fitriati. 2016. Safetalk Practices in Content and Language Integrated Learning (ClHIL) Classes Language Circle: Journal of Language and Literature, XI/1. _________________________________________________________________________________________ such a wise and appropriate decision when the government determined not to continue the enforcement of using English in teaching content subjects. CONCLUSION The teachers in their CLIL classes engaged in interactional and pedagogical practices in what can be called safetalk practices. In the way the teachers performed their talk to interact with their students, they tended to employ the salient features, such as chorusing and response slot with the teachers‟ raising intonation, which were seen as if all the students understood and accomplished the lesson. In my discussion, I have shown that interactional practices in the three classrooms observed lacked adequate language pedagogical practices. The teachers, who were not English language experts, and were not trained as bilingual teachers in teacher education need to learn about second language acquisition theory and language pedagogy so that they would be aware of the importance of providing their students second language comprehensible input. It is essential that teacher education, particularly, those which prepare English bilingual teachers, be informed about the types of classroom practices which I have illustrated in this article and then make critical reflection in order to promote best practices in bilingual lessons. I hope this research might push teacher educators, particularly those who train English bilingual pre-service teachers in teacher education colleges and in-service teachers during professional development, to incorporate values of second language acquisition principles and pedagogy into their teaching. Teacher educators should assist prospective English bilingual teachers to be aware of the tendency of safetalk practices which pedagogically impede the learning of the content subjects and the English language as well. This study gave rich data about the particular context. Findings of the current study can reflect what occurs in other settings in the Indonesian educational context and, therefore, the findings can shed light on understanding of the English bilingual education implementation, not limited in former selected schools which implemented English as the medium of instruction, but also other schools, including private schools which set out to establish bilingual education programs. To sum up, there is an urgent need for teachers to explore classroom interactional and pedagogical practices which might best contribute to English language learning as well as to CLIL classes. . REFERENCES Baker, C. 1993. Bilingual Education in Wales. In H. Baetens Beardsmore (ed.),European Models of Bilingual Education (pp. 7-29). Clevedon: Multilingual Matters. Baker, C. 2011. Foundations of Bilingual Education and Bilingualism (5 th ed.). Bristol, UK: Multilingual Matters. Braun, V.& Clarke, V. 2006.Using Thematic Analysis in Psychology.Qualitative Research in Psychology, 3:2, 77-101. Chick, J. K. 1996. Safe-talk: Collusion in Apartheid Education. In H. Coleman (Ed.), Society and the Language Classroom (pp. 21-39). Cambridge: Cambridge University Press. LANGUAGE CIRCLE: Journal of Language and Literature XI/1 (October 2016) __________________________________________________________________________________________________ Chimbutane, F. S. 2009. The Purpose and Value of Bilingual Education: A Critical, Linguistic Ethnographic Study of Two Rural Primary Schools in Mozambique. (Unpublished doctoral dissertation).The University of Birmingham,United Kingdom. Chimbutane, F. 2011. Rethinking Bilingual Education in PostcolonialContexts.Bristol: Multilingual Matters. Creswell, J. W. 2013. Qualitative Inquiry and Research Design: Choosing among Five Approaches (3rd. ed.). Thousand Oaks: SagePublications Dalton-Puffer, C. 2007.Discourse in CLIL classrooms. Amsterdam:Benjamins. Ellis, R. 2008. The Study of Second Language Acquisition (2 nd ed.). Oxford: Oxford University Press. Fitriati, S.W. 2015. Teachers‟ Language Ideologies and Classroom Practices in English Bilingual Education: An Ethnographic Case Study of a Senior High School in Central Java, Indonesia. (Unpublished doctoral dissertation).The University of Southern Queensland, Australia. Garcia, O. 2009.Bilingual Education in the 21st Century: A GlobalPerspective. United Kingdom: Wiley-Blackwell. Graaff, R. D., Koopman, G. J., &Westhoff, G. 2007. Identifying Effective L2Pedagogy in Content and Language Integrated Learning (CLIL).ViewsVienna English Working Papers, 16(3), 12-19. Hornberger, N. H., & Chick, J. K. 2001. Co-constructing School Safetime: Safetalk practices in Peruvian and South African classrooms. In M. Heller & M. Martin-Jones (Eds.), Voices of Authority: Education andLinguistic Difference (pp. 31-56). London: Ablex. Krashen, S. 1985. The Input Hypothesis. Beverly Hills, CA: Laredo Publishing Company. Krashen, S. 1994. Bilingual Education and Second Language Acquisition Theory. In B. E. Office (Ed.), Schooling and Language MinorityStudents: A Theoretical Framework (2 ed., pp. 47- 75). Los Angeles, CA: California State University, National Evaluation, Dissemination andAssessment Center. Martin, P. 2005. 'Safe' Language Practices in Two Rural Schools in Malaysia: Tensions between Policy and Practice. In A. M. Y. Lin & P. Martin (Eds.), Decolonisation, Globalisation: Language-in-Education Policyand Practice (pp. 74-97). Clevedon: Multilingual Matters. Mills, K. A. 2007. Access to Multiliteracies: A Critical Ethnography.Ethnography and Education, 2(3), 305-325. Palmer, D. 2011. The Discourse of Transition: Teachers' Language Ideologies WithinTransitional Bilingual Education Programs. International Multilingual Research Journal, 5(2), 103-122. Snow, M. A. 2001. Content-Based and Immersion Models for Second andForeign Language Teaching. In M. Celce-Murcia (Ed.), Teaching Englishas a Second or Foreign Language (3 ed., pp. 303-318). USA: Heinle&Heinle. Swain, M., & Johnson, R. K. 1997. Immersion Education: A Category within Bilingual Education. In R. K. S. Johnson, M. (Ed.), ImmersionEducation: International Perspective. UK: Cambridge University Press. Notes: The transcription conventions used in this study: S : Student (unidentified student) Sri Wuli Fitriati. 2016. Safetalk Practices in Content and Language Integrated Learning (ClHIL) Classes Language Circle: Journal of Language and Literature, XI/1. _________________________________________________________________________________________ Ss : Ss Students (several or all students are speaking in chorus) Plain font : Indonesian language Bold font : English language [ ] : Commentary on what is happening in the classroom { : Overlapping speech underlining : Indicates raised intonation from the teacher where the teacher expects students to orally “fill in the blank” LANGUAGE CIRCLE: Journal of Language and Literature XI/1 (October 2016) __________________________________________________________________________________________________