A translanguaging perspective on medium of instruction in the CFL classroom

ABSTRACT Although recent years have witnessed increasing interest in learning Chinese as a foreign language (CFL) in China and around the world, there seems to be a paucity of research into the medium of instruction in the CFL classroom. This paper describes a study of the practice of MoI choice in the CFL classroom, situated within the framework of translanguaging. Five focus group interviews and eight classroom observations were conducted across five universities in China. Approximately 43 participants participated in the interview stage of the research, while 149 CFL students across eight CFL classrooms were involved in the observation stage, with six students and four teachers also interviewed after the observations. The findings suggest that translanguaging was commonly used by teachers. However, due to limits on resource availability and to the linguistic complexity of a CFL classroom, these translanguaging practices also posed challenges to the teachers. Interestingly, English was predominantly employed for explanatory and elicitation functions, whereas Chinese was chosen mainly for teaching and instructing purposes. While translanguaging practice was observed in both the beginner CFL class and in the more advanced class, the monolingual approach was still predominantly used and advocated across different levels.


Introduction
Despite English being the world's lingua franca, the number of CFL learners has been increasing significantly around the world. According to the Chinese Ministry of Education (Ministry of Education of the PRC 2018a), approximately 14,000 international students came to China in 1992, growing to 41,000 in 1996. More recently, the total number of international students in China continued to grow by around 10% in 2016 and 2017 (Ministry of Education of the PRC 2018b). Among the 489,200 international students who studied in Chinese higher education institutions in 2017, 49.38% (241,500) enrolled in a degree programme (ibid.). It seems that China has indeed joined the worldwide competition to host international students.
At the same time, learning Chinese seems to be increasingly popular around the world, especially with the support of the Office of Chinese Language Council International (also known as Hanban) and the Confucius Institute programme (Zhao and Huang 2010;Osborne, Zhang, and Zhang 2018). By the end of 2017, 525 Confucius Institutes had been established in 146 countries (Hanban 2018), and it is estimated that about 1.1 million people are registered in Confucius Institute programmes worldwide (Zheng, Luo, and Shang 2014). Additionally, approximately 100 million people may be learning Chinese in schools, universities and private institutions (Wu 2010).
Consistent with the increasing popularity of learning CFL, teaching Chinese as a foreign language (TCFL) has also developed significantly over the past three decades. Initially, the Chinese term for TCFL was 对外汉语教学 (duìwài hànyŭ jiàoxué), literally meaning 'Chinese language teaching towards outside' (Li and Wang 2016). The term was adopted to exhibit that the focus of TCFL was to provide Chinese language teaching to foreign students in Chinato meet 'the internal need' (Gao 2013b). With the expansion of Chinese language study on a global scale, TCFL has also shifted its focus towards learners of Chinese outside China. Against this backdrop, in 2005 the Chinese name for TCFL was changed to 国际汉语教育 (guójì hànyŭ jiàoyù), 'international Chinese language education' (Gao 2013b;He 2018). This alteration of the Chinese term reflects the transition of the main TCFL body from foreign students in China to international learners around the world, and is in line with the global promotion of the Chinese language (Cui 2010;Li and Wang 2016).
Furthermore, TCFL was first proposed as an academic discipline in 1982 and was then developed into an undergraduate programme in 1983 (Gao 2013b), followed by the creation of some MA programmes in 1986 (Lu and Zhao 2011). A complete certificate system for TCFL was not available until 1998, when Beijing Language and Culture University began offering doctorate courses in TCFL (Lu and Zhao 2011). However, a postgraduate programme, International Education of Chinese, was approved in 2007 in China (Gao 2013b).
Because of the worldwide demand for Chinese language study, overseas TCFL training has been made available in recent years. Taking the United Kingdom as an example, the Department for Education announced a joint programme between the Specialist Schools and Academies Trust (SSAT) and Hanban to train a further 1,000 teachers to teach Mandarin in secondary schools in England (Department for Education 2010; Li 2013). A number of schools and teacher associations have also arranged for their staff and members to participate in TCFL training provided by Chinese universities (Shi 2010). For instance, East China Normal University in Shanghai trained about 500 teachers between 2005 and 2007 (Wu 2007).
The development of TCFL courses and training contributed to the review and publication of New Standards for Teachers of Chinese to Speakers of Other Languages (Hanban 2012). Interestingly, the requirement that TCFL teachers have mastered a foreign language (e.g. English) no longer exists in this new 2012 version. It is only briefly mentioned in the 'Chinese Culture and Cross-cultural Communication' section that CFL teachers should be able to 'employ the language of the country which they are teaching in to communicate and teach' (Hou 2016, 103). In comparison, there were specific requirements in relation to foreign language knowledge (e.g. phonology and phonetics, grammar, vocabulary, etc) and skills (e.g. listening, speaking, reading and writing) in the older version published in 2007 (Shao and Shao 2013). As Gao (2013a) indicates, it is important for CFL teachers to be fluent in the language used in the country where they are teaching, such as English in the case of Australia. However, the 2012 version puts more emphasis on Chinese language pedagogy and the dissemination of Chinese culture (Shao and Shao 2013;Hou 2016). The deletion of clear instructions on foreign language requirements for CFL teachers may imply a shift of medium of instruction (henceforth MoI) from the students' first languagewhatever that may be in the relevant countryto the target language, Chinese.
Indeed, there has been a long tradition of adherence to the Chinese-only norm in TCFL in both practice and academic discourse, due either to confusing it with other concepts such as interlanguage, or to the unproven assumption that Chinese is best taught through the target language exclusively (Wang 2019). This assumption may stem from perceiving Chinese as a discrete entity completely separate from other languages. Because of the over-emphasis on linguistically contrasting Chinese with other languages, other aspects (e.g. cross-linguistic pragmatics) have tended to be neglected, which has likely led to grammar and translation methods being used with rote language drills (Singh 2018). Standard language ideology is also likely to play a role in the assumption that using Chinese only to teach Chinese in the CFL classroom is best practice. The standard variety of Chinese (Mandarin) has been promoted across China for the purpose of national unity (ibid). As even some languages of ethnic minorities in China are discouraged from being used as the instructional medium (Zhang and Yang 2018), it is unsurprising that a Chinese-only system has become the norm, in order to enhance loyalty to the nation-state. Although the monolingual principle may maximise exposure to the target language, it also leads to teacher-centred and textbook-focused pedagogies, due to limited, or even lack of, opportunity for students to engage in class interactions (Singh 2018;Wang 2019).
The prevailing monolingual approach is also applicable to the EFL (English as a foreign language) classroom. Interestingly, previous studies (e.g. Van Der Meij and Zhao 2010;Cai and Cook 2015) acknowledge that using a certain amount of Chinese in the EFL classroom is both desirable and functional. In other words, the monolingual principle tends to be confronted by the inevitable practice of multilingualism in real-life classrooms.
Therefore, dynamic engagement of teachers' and students' linguistic repertoires may challenge the target-language-only norm. As Singh (2018, 30) points out, the present language class is in a post-monolingual condition entailing 'tensions in which multilingual practices persist or re-emerge and monolingualism continues to assert itself'.
The current study is situated in the context of the expansion of CFL study and the development of TCFL, in order to scrutinise the current practice of medium of instruction (MoI) in the CFL classroom in China. It first introduces previous research on MoI for TCFL, using translanguaging as the theoretical framework. It then presents the research design, followed by the analysis of the data from interviews and classroom observations. The discussion highlights the main findings regarding the translanguaging practice in CFL classroom, which will contribute to the development of a specific CFL pedagogy, as well as foreign language teaching and training in general.

Medium of instruction from the translanguaging perspective
The term translanguaging was first used to describe pedagogical practices in Welsh bilingual schools where Welsh and English alternate for both input and output (Lewis, Jones, and Baker 2012b;Li 2018). It was initially proposed to refer to a strategy whereby bilingual children would use their stronger language to develop their weaker language, in order to reach a balanced development of two languages (Lewis, Jones, and Baker 2012b). With the development of bilingual and multilingual education, especially the move away from monolingual ideology to pluralingualism, the concept has been evolving into a practical theory to 'interpret hybridity and creativity of language use in foreign language classrooms' (Hou 2016, 2). Translanguaging goes beyond a pedagogy, in that it is 'a powerful mechanism to construct understandings, to include others, and to mediate understandings across language groups' (García 2009, 151). An understanding of a translanguaging act is considered to be highly context-dependent, and involves cognitive and cerebral activity rather than being content-based, or only concerning linguistic codes (Lewis, Jones, and Baker 2012b). Instead of underscoring the learning outcomes, translanguaging emphasises the heteroglossic input and output, to encourage the bilingual process in language learning (Lewis, Jones, and Baker 2012a).
When translanguaging is applied in MoI research, it is often confused with code-switching. However, García and Li (2014) clarify that the perception of code-switching two or more languages as separate entities no longer seems able to capture the fluid and dynamic practices in a multilingual world in which the boundaries between languages are also no longer clear.
Translanguaging, in contrast, refers to a holistic conceptualisation of diverse multilingual and multimodal practices. It is concerned with the entire linguistic repertoire a bi/multilingual person possesses, and draws from real human communication in a globalising world rather than specific and separate structures of languages (Li 2018). The act of translanguaging enables bi/multilingual people to utilise their linguistic repertoire in an integrated way for communication. In this case, multilingual practices through translanguaging no longer mean becoming proficient in different languages, but rather refer to the interweaving of the languages that a person possesses (Li 2018, 15).
Because of the complexity and wide scope that translanguaging implies, Cenoz and Gorter (2017) distinguish pedagogical translanguaging from spontaneous translanguaging. While spontaneous translanguaging refers to fluid bi/multilingual discourse that takes place in naturally occurring contexts including inside the classroom, pedagogical translanguaging is the intentionally planned usage of the whole linguistic repertoire of both teacher and student (Cenoz 2017;Cenoz and Gorter 2017). Building upon this classification, this study examines the data collected from focus group interviews and eight classroom observations across five Chinese universities in relation to the MoI in CFL classrooms, as this area has not yet been explored in the context of the CFL classroom.
In line with the growing interest in learning Chinese as a foreign language, there has been a burgeoning of literature in CFL pedagogies. Previous studies range from creative teaching methods for the acquisition of Chinese language (Xu et al. 2013;Zhang and Lu 2014;Osborne, Zhang, and Zhang 2018) to curriculum design and teaching methodology development (Packard 1990;He and Jiao 2010;Ye 2013). However, there seems to be a scarcity, or perhaps an under-reporting, of this in English-language literature relating to MoI in CFL classrooms. Very few scholars investigate it either in depth (Wang and Kirkpatrick 2012;Hou 2016;Singh 2018) or briefly (Scrimgeour 2009;Lu and Zhao 2011;Zhang 2016) and present the results to international readers in the English language.
Interestingly, monolingual ideologies of using target-language-only or one-language-at-a-time are dominant in the teaching of CFL. In a study conducted by Hou (2016), the attitudes of teachers and students towards the use of translanguaging in the beginner CFL classroom were explored. It was found that for the most part, translanguaging occurred in the classroom; however, teachers were divided on the perceived suitability of employing pedagogical translanguaging. Indeed, Chen and Yeung (2015) highlight that native Chinese CFL teachers' self-efficacy would increase through training of using English in the classroom, which also demonstrates their lack of experience in using English in the instruction of CFL.
Indeed, as Zhang (2016) points out, monolingualism is usually driven by the privilege of native speakers. The monolingual approach tends to emphasise the concept of 'foreignness' in language education, since it sees 'no relevant connections between the students' first language [for example] (English) and the language they are learning (Chinese)' (Singh 2018, 40). Under these circumstances, Wang (2019) says that it is unrealistic to adopt a monolingual approach in a multilingual classroom, and advocates the use of translanguaging strategies. Moreover, she calls for CFL teachers to be able to understand the basics of students' mother tongues (ibid.). This practice of translanguaging has actually been proven to benefit learners for both language acquisition and content instruction in the L2, and is now accepted in many educational contexts (García and Li 2014, 80).
As a result of the somewhat conflicting opinions of CFL teachers surrounding translanguaging, despite the previously reported benefits, it was therefore worthwhile to investigate the opinions of CFL teachers-in-training in order to examine their likely future practices in relation to pedagogical and spontaneous translanguaging, as well as observe some CFL classrooms at varying levels to note any occurrences of pedagogical and/or spontaneous translanguaging.

Research methods
The current study involved interviews and classroom observations in five Chinese universities in southern, central and northern China. All five offer Chinese language courses to international students, which meant the interviewees in the study have ample teaching experience in real-life CFL classrooms. Of the 43 interviewees, eight are full-time TCFL teachers and the rest are full-time TCFL research students.
Because they were divided into five focus groups, their demographic information was collected in advance of the interview, in order to ensure they were all TCFL practitioners. The research students all had at least one year of TCFL experience; four of them had previously been full-time TCFL teachers or tutors in either higher education institutions or international schools, before returning to university for postgraduate study. The TCFL teacher-participants in the interviews were all full-time TCFL staff in the relevant universities. Their experience in TCFL ranged from three years to over thirty years.
The number of students observed in the eight classes ranged from 15 to 22, with an average of 18.625 per class. Levels ranged from beginner level to advanced, and between students and teachers each class included a variety of nationalities, and therefore mother tongues. Students came from a variety of nations across Africa, the Middle East, Australia, the United States and Asia, while all teachers were Chinese. Under these circumstances, the common languages understood in such classrooms are usually Englishthe global lingua francaand Chinese, which is naturally the lingua franca of a CFL class based in China. Six students and four teachers were also interviewed after the classroom observations, in order to gain a clearer picture of the instances of pedagogical and spontaneous translanguaging in the classroom.
A draft of interview questions and protocol was prepared for piloting with two TCFL teachers in a Chinese university involved in the current study. Minor revisions were made after the pilot study, including changing the order of two questions. The class observation form and protocol were also piloted with a class in one of the universities. A few minor changes were made, such as collecting class information (e.g. the main topic or theme of the lesson) before the observation.
All interviews were first transcribed from the recordings by one researcher. After this, the transcriptions were read in detail, manually coded and categorised by two researchers working independently. Using pedagogical and spontaneous translanguaging as two broad categories, any common threads in the entire set of interview data were captured, and are presented in the data analysis section. The classroom observation data was then examined and used to provide real-life examples of translanguaging in current university CFL classrooms under these categories, which is an aspect of the topic not previously explored in the CFL literature.

Data analysis
Firstly, the interview data is presented, in order to gain an overall picture of the stance of some CFL teachers on translanguaging in the classroom. It is presented under two main categories based on Cenoz and Gorter's (2017) definitions of pedagogical translanguaging and spontaneous translanguaging. In this way, both intentionally planned and unplanned instances of translanguaging are detailed according to the findings from the data, and therefore significant information on specific instances of translanguaging in the CFL classroom becomes available.

Pedagogical translanguaging
In relation to the intentionally planned instances of translanguaging in the CFL classroom, the interviewees first said that they would employ English, the mother tongue for some students and a lingua franca for many others, for explanatory purposes. For example, they mentioned specifically that when explaining particular items, the use of translanguaging provided cognitive or metalinguistic scaffolding for meaning-making activities, such as explaining and elaborating on grammar rules and lexical uses, and also for translating new words. Indeed, it is highly likely that students of CFL at any level require detailed explanations of various new items learned, such as grammar, lexical use and translation. As the Chinese language is their learning target, it seems to be most appropriate to explain these items in the language that students are most comfortable with.
One teacher pointed out that in some cases, if a teacher decided not to use English to explain a new item, it is possible that any student who did not understand might 'bother' a classmate by asking him/her to explain it. As a result, this particular teacher believed it was necessary to employ translanguaging not only to help students understand, but also to avoid disrupting the flow of the class. In addition, a student being interviewed after the classroom observations notified the researchers that when their teacher lacked knowledge of a particular language that would aid understanding of the Chinese topic, the class would then rely on dictionaries. However, the student said this method was not as helpful as the teacher being able to explain the topic in English.
A small number of teachers pointed out that they would not simply rely on English to teach a new item, but would try to explain the new item in Chinese first. If any students still did not grasp the meaning, they would then try to use simpler Chinese, before finally using English. In other words, pedagogical translanguaging seemed here to be a last resort.
Additionally, some students noted that they do not like to rely too much on the English language or lingua francain the classroom, as they believe this might hinder or slow their progress of learning Chinese. Also, if a teacher used Chinese only, some students noted that the concentration levels of the class were likely to increase.
The students interviewed also confirmed higher use of translanguaging in the CFL classroom at the beginner level. As they progressed to higher levels, they noted that in their experience, there were fewer instances of translanguaging. One student at HSK 1 4 level also highlighted that in the early stages of their learning, up to 80 percent of the class could be taught using English. In another case, a student detailed that their teacher could speak French and would therefore also use French to explain items to students with knowledge of this language.
A small number of students, however, stated that they preferred to use Chinese as much as possible, so that they could communicate and express themselves to the teacher as accurately as possible, thus hinting at their teacher's lack of English-language skills. Similarly, students communicating with each other in their mother tongues (and therefore avoiding the lingua franca of the entire class) would do so mainly to ensure that they, or a topic, were correctly understood.
The second use of translanguaging that the teachers interviewed mentioned, after explanatory purposes, was for management purposes. For example, they specifically highlighted that when giving students instructions for an activity, giving feedback, praising, disapproving, checking the comprehension of learning content, and planning assignments, it was important to ensure that their students understood these instructionsthis was their justification for using the students' mother tongue or the classroom lingua franca in these instances. The teachers were also of the opinion that the practice of translanguaging in this way was relevant to their students' level of Chinese. The general consensus was that this pedagogical translanguaging usually occurs when beginner learners, who have no previous knowledge of Chinese, are being taught. Depending on their progress, it may take several weeks for them to become familiar with the classroom instructions in Chinese so that a Chinese-only approach can be used. Before that, translanguaging is used as a strategy to manage the classroom.

Spontaneous translanguaging
In terms of more fluid and unplanned instances of translanguaging, participants in this study mentioned that this type of translanguaging occurred in natural in-class interactions. A reason given for this was the contrast between the learners' and the teachers' linguistic repertoire of Chinese and English (or any other mother tongue)the teachers were fluent in Chinese but lacking in the knowledge of the students' mother tongue(s), and the students were not yet fluent in Chinese.
Two main examples were given for when this type of translanguaging occurred. Firstly, learners would translate what the teacher said in order to confirm, or not, their understanding of a certain topic or item. This occurred both in relation to the content of the class and to activity instructions given to students in order to aid clarification. One participant in the study gave an example: One of my students explained to me his understanding of the property of a Chinese word to me using an example in French [first language of the student].
Other examples of this included students clarifying the use or meaning of a word. In addition, teachers sometimes asked students what a certain word referred to, whereby the students would answer using the English translation. Finally, when students were mainly communicating with their teacher in Chinese but wished to use a Chinese word that could not be remembered or had not been previously learned, they employed translanguaging strategies by, for example, using the English translation.
Secondly, students might translate for their classmates what the teacher had said, in order to aid understanding. It was noted by the participants in the interviews that this could occur either on the learners' own initiative or with the encouragement of the teacher. One teacher noted that using English was always the quickest way to clear up any confusion on items addressed in the lesson, but that they used it as a 'last resort' on account of their lack of confidence in using the language. Indeed, students said in their interviews that using their mother tongue to ask a fellow classmate about the meaning of a topic or word was a frequent tactic used to aid clarification. Of course, not all students speak English as their first language, so the teachers in this study found it very difficult to deal with misunderstandings when they had little to no knowledge of the students' mother tongues. In these instances, very simple Chinese could be used to give further examples, or the teachers would need to rely on other students to communicate their understanding to their classmates, and so spontaneous translanguaging could be found.

Guilt of translanguaging
Throughout the course of the interviews, it became apparent to the researchers through comments made by the interviewees that a kind of stigma was attached to the practice of translanguaging in the classroom.
The main source of the teachers' guilt stemmed from the fact that, naturally, an integration of practices from multiple languages in translanguaging results in a deviation from a doctrine of correctness in language. This was an issue for the teachers, as they perceived translanguaging practices as a sacrifice of the standard use of Chinese to support classroom communication. Being teachers of CFL, they did not wish to deviate from speaking Chinese, and in one case a teacher reported being instructed not to use English in the classroom, despite being able to see the benefits of its use. Two participants in the research highlighted: I want to teach them the most accurate usage of the word, so they need to 'feel' [experience and be exposed to] it in Chinese. Using another language cannot help with this.
My English is very poor. If I use it, I may make both their Chinese and English deteriorate. Therefore, despite recognising the benefit and sometimes necessity of employing a practice of planned or unplanned translanguaging in the classroom, as highlighted in previous sections, it seems that some of the teachers are still reluctant to use translanguaging, on account of their dislike of deviation from standard language ideology, and also on account of a lack of confidence in their own foreign language abilities. Secondly, some participants reported a type of guilt surrounding the use of translanguaging in their classroom, as a result of the long-established yet unproven assumptions of the benefits of monolingualism in foreign language teaching (Levine 2003;Hou 2016). Wang (2014) explains that 'Chinese-only' has actually been the predominant language policy across many CFL programmes. Furthermore, CFL teaching borrowed the term 'immersion' from programmes in North America and used it in promoting the monolingual Chinese-only principle in Chinese classrooms. These assumptions of monolingualism are also further cemented in the New Standards for Teachers of Chinese to Speakers of Other Languages (Hanban 2012), which advocates the sole use of Chinese in the CFL classroom. One teacher in this study said: I know that a lot of Chinese scholars advocate the monolingual approach to teach Chinese. I've tried, really, but it really doesn't work for beginner learners. So I have to use some English in my class at the beginning, and then gradually integrate more and more Chinese in the medium of instruction.
This comment demonstrates the internal conflict of the CFL teachers in deviating from the monolingual approach. It seems that the teachers in this study are so accustomed to using only Chinese in the CFL classroom that resisting this practice is not agreeable to them. On the other hand, they are aware that it is necessary to incorporate the mother tongue of beginner CFL students to aid understanding of either CFL-related items or general classroom communication. This indeed highlights a finding from Wang's (2019) research whereby teachers' perceptions of using translanguaging in the classroom also highlighted concerns in relation to practicality and efficiency. Some also described the monolingual approach as an 'ideal', despite claiming in interviews to see value in translanguaging approaches. Therefore, although translanguaging is used as a pedagogical strategy, as shown here, this use seems to be reluctant.
Finally, the notion of using translanguaging as a last resort is also apparent in a final comment by one of the participants: My English is not good, and so I have to use basic English with Chinese to support my students' study of Chinese.
This teacher was referring to a situation also mentioned by other interviewees, whereby they were required to practise translanguaging on account of their low proficiency in English or other mother tongues of students. As a result, they had no choice but to use a combination of Chinese and basic English to further explain an item, in a type of spontaneous translanguaging occurrence. The teacher went on to explain that when a student asked about a particularly challenging grammar point in a low-level class, the teacher could not explain it fully in English, thus causing an issue in their teaching.
The data shows that both pedagogical and spontaneous translanguaging is widespread in CFL practice today and therefore provides significant information surrounding the teaching and learning situation in the CFL classroom. It is also interesting to note the additional finding that the interviews spurred, that is, the internal conflict and guilt reported by teachers when the need to deviate from a monolingual approach to teaching CFL arises.

Observations of translanguaging in the classroom
The following describes some real-life occurrences of translanguaging observed in eight universitylevel CFL classrooms, in terms of both pedagogical and spontaneous translanguaging. After discovering the prevalence of translanguaging strategies in the CFL classroom from the interviews conducted in the current research, the researchers wished to understand more clearly the process of translanguaging in a real-life classroom situation. Eight classes were observed, in order to document instances of pedagogical and/or spontaneous translanguaging that took place during the entire class time among both students and teachers. The researchers noted the incident, the language used, the tactics employed and the outcome for each instance. For this reason, a number of translanguaging instances are highlighted and developed on in the following paragraphs. In each classroom observation, instances of pedagogical and/or spontaneous translanguaging occurred among students and/or teachers, despite many interviewees reporting an affinity for the monolingual approach in the CFL classroom.

Observations of pedagogical translanguaging
Firstly, in all eight classroom observations, teachers employed pedagogical translanguaging for various purposes. These included: providing operational classroom instructions; maintaining discipline; introducing an item; explaining items; checking if information had been understood; and assisting in pronunciation (see details in Table 1).
As a result, it seems that despite a feeling of guilt associated with departing from the monolingual approach, CFL teachers do rely on translanguaging in the classroom for the benefit of students' learning outcomes, with some instances of translanguaging for managerial purposes also noted. In addition, contrary to the findings from the interviews, the use of pedagogical translanguaging did not fluctuate according to class level. In other words, unlike the description in the interviews, the use of pedagogical translanguaging was not necessarily more commonly observed at the beginner level than at the advanced level. Instead, it was used at a similar frequency in all classes.
At the same time, it was noted that students' translanguaging inputs sometimes contributed to the teacher's pedagogy and the lesson in general, as observed in two of the eight classes. For example, students' instances of translanguaging helped to explain vocabulary to their classmates, presumably in instances when these students already understood the new item and could explain it clearly to their classmates. For the most part, however, spontaneous translanguaging among the students was more common, as demonstrated in the following section.

Observations of spontaneous translanguaging
Firstly, in terms of teachers' instances of spontaneous translanguaging, there were significantly fewer instances compared to those of pedagogical translanguaging, as observed in the previous section. In addition, spontaneous translanguaging instances were only observed in five out of the eight classes. These instances can be generally categorised into two types: (a) communicating with students, for example, answering a random question such as when an event will take place ('We will let you know soon') or saying thank you, and (b) confirming students' attempt to understand something, including providing English examples or asking students to repeat something.
Differing from the case of teachers' use of spontaneous translanguaging in the classroom, students' use of spontaneous translanguaging was higher, and occurred in all eight classes observed. In each class, students employed spontaneous translanguaging to ask their teacher or peers a question, or to answer a question from the teacher or their peers. Table 2 shows examples of students' spontaneous translanguaging occurrences in the four aforementioned categories.
The data collected from the classroom observations show a tendency for teachers to employ more translanguaging techniques of the pedagogical kind using English only, while the students tended to engage in more spontaneous translanguaging among their peers using their own mother tongues, a lingua franca and Chinese. The following section discusses these results further, and what they may shed light on in relation to current translanguaging practices.

Discussion
This paper has demonstrated that despite the use of translanguaging in the CFL classroom to aid both pedagogical and administrative content, there is a certain stigma attached to using English, or another foreign language, in the CFL classroom. Indeed, a sense of guilt when employing pedagogical or spontaneous translanguaging is not exclusive to participants in the current research, nor is it exclusive to the CFL classroom. In 2011, Swain, Kirkpatrick, and Cummins compiled a handbook providing practical information on how to use Cantonese in the English as a foreign language (EFL) classroom in Hong Kong without feeling a sense of guilt. Indeed, Swain, Kirkpatrick, and Cummins (2011) note that EFL teachers are pressured to use only English when teaching EFL.
This feeling of guilt for not using the target language can be traced back to standard language ideology. A standard language is socially defined and ideological in nature (Milroy 1999;Campbell-Kibler 2006). Therefore, in the context of a foreign language classroom, the target language that the learners are studying is considered to be the standard language, and any deviation from the target language may be considered incorrect and consequently lead to a sense of guilt. Similarly, in the current study, we see CFL teachers feeling guilty for using languages other than Chinese to aid the comprehension of a given lesson or item. It seems that a longstanding assumption that monolingual practice is ideal for foreign-language learningevident in the New Standards for Teachers of Chinese to Speakers of Other Languages (Hanban 2012), for examplehas indeed prevented CFL teachers from adopting translanguaging practices in their classroom, thus possibly preventing the holistic understanding of new items in the short term. In the long term, it is likely that this also leads to a focus on language learning outcomes rather than processes.
Wang (2019) has found in her research that this guilt even extends to students, and can affect their participation in the classroom. For example, in some cases students would not attempt to speak in the classroom unless they could form a complete Chinese sentence without the use of any English words (ibid.). In this way, it is apparent that students can be so reluctant to use any language other than Chinese that they would rather not speak up in class, resulting in a lack of the participation necessary for foreign language acquisition.
Indeed, in the current research, it was found that when students communicated with each other, it was through a mother tongue or a lingua franca, with some instances of using Chinese. When communicating with teachers, although they still employed translanguaging strategies in some cases, the majority of the interaction was through Chinese. In addition, some students mentioned a preference for communicating in this way as a result of their teacher's inability to speak their language, thus highlighting Wang's (2019) call for teachers to be able to understand at least the basics of their students' languages.
As a result, it seems clear that while translanguaging has been found to be effective in the foreign language classroom, as previously discussed, reservations and even a lack of foreign language skills among teachers have meant that for the most part, a monolingual approach is deemed to be most effective. Despite this, however, each classroom observation of the current study demonstrated a number of pedagogical and spontaneous translanguaging instances. Therefore, it is likely that through training and educating up-and-coming CFL teachers in the use of translanguaging practices, these feelings of guilt will gradually subside, allowing more effective language learning in a multilingual environment, as previously highlighted. As Chen and Yeung (2015) and Wang (2019) advocate, teachers must be trained to use the mother tongues of their students in the CFL classroom, in order to increase their self-efficacy and to be able to teach more effectively in multilingual contexts.
For the moment, it is worthwhile to note that as foreign language teachers are generally reluctant to engage in spontaneous translanguaging techniques in the classroom, it was found that more occurrences of translanguaging actually happen spontaneously among students. This finding from the current study is further supported in studies conducted by the likes of Karlsson, Larsson, and Jakobsson (2018), Makoe (2018) and Rowe (2018), to name a few. In these studies, it is noted that when students collaboratively draw on their linguistic repertoires, more positive learning outcomes occur in the classroom. Indeed, it is seen that these positive learning outcomes are not exclusive to the foreign language classroom, as Karlsson, Larsson, and Jakobsson (2018) actually document occurrences of translanguaging in a science classroom. Although Wang (2019) has reported that in some instances students are reluctant to participate in the class when the only way to do so is to use translanguaging strategies, the aforementioned studies focusing on students and their positive use of translanguaging strategies suggest that students' guilt may possibly have been transferred from their teachers.
In terms of new pedagogical approaches in the multilingual classroom, it is clear that the use of both pedagogical and spontaneous translanguaging is likely to have a positive effect on students' learning outcomes. As previously discussed, it appears that the main issue in terms of incorporating these strategies in the classroom comes from the teachers, who are reluctant to deviate from the monolingual approach that has been so strongly advocated in the foreign language classroom. On the other hand, students appear to embrace their individual linguistic repertoires and work collaboratively with their peers, which appears to lead to more positive learning outcomes. As a result, it is suggested that teachers should be trained in pedagogical translanguaging practices, and at the same time should be encouraged to use spontaneous translanguaging strategies as the opportunity arises.
Wang (2019) highlights the three main principles by which translanguaging has been adopted in various CFL classrooms: for comprehension of the lessons; for efficiency in the classroom; and to motivate learners. These principles are indeed reflected in the translanguaging instances observed in the current study, from both teachers and students. Rowe (2018) has developed six principles that encourage translanguaging in the classroom: (a) explicitly discussing students' language backgrounds; (b) providing dual-language learning materials in students' various languages; (c) promoting collaborative two-way translation; (d) encouraging students to collaborate in creating duallanguage or multi-language texts; (e) providing opportunities for translanguaging through meaningful activities, such as encouraging discussion of curriculum content in relation to individual experience; and (f) designing projects that enable students to connect with multilingual audiences such as family members and the community. In implementing even one of these six principles, it is apparent that students will have more opportunities to draw on their linguistic backgrounds and collaborate with their peers. At the same time, encouragement from the teacher to use their individual linguistic backgrounds should also foster a positive attitude towards translanguaging practices in the classroom.

Conclusion
The current research has contributed significant information in exploring the use of pedagogical and spontaneous translanguaging instances in the CFL classroom by means of interviews and classroom observations. In examining the data collected, this paper has demonstrated that although both pedagogical and spontaneous translanguaging are used frequently by both students and teachers in the classroom at all levels, for the most part the interviewed teachers stated that they try to adhere to the monolingual approach.
The current research has shown that there exists a stigma among teachers towards using translanguaging strategies in the CFL classroom, resulting in this general adherence to the monolingual approach. This stigma mainly stems from a feeling of guilt experienced by CFL teachers when they deviate from the monolingual approach that has been heavily advocated in foreign language learning. However, studies have in fact demonstrated that positive learning outcomes occur when students are exposed to pedagogical or spontaneous translanguaging in the classroom. In the current research, it was found that students are also likely to collaborate and utilise their unique linguistic backgrounds to assist each other in the language learning classroom, and indeed for other managerial aspects. Therefore, the current research also shows that these student-led instances of translanguaging can assist students' overall learning.
As a result, it is suggested by the researchers of this study that incorporating and encouraging translanguaging strategies in foreign-language teacher training is likely to eliminate any guilt associated with translanguaging. Furthermore, training teachers in the basics of their students' languages could also eliminate the guilt of not being proficient in other foreign languages, as advocated by Wang (2019), and could therefore encourage further use of pedagogical or spontaneous translanguaging in the CFL classroom, which appears to have many benefits. It is also advised that Rowe's (2018) principles on encouraging the use of translanguaging in the CFL classroom be examined. Fostering even one of these principles will likely result in a more positive attitude towards translanguaging practices, which should in turn lead to more positive learning outcomes.
Future research may further explore instances of pedagogical or spontaneous translanguaging in other CFL classrooms worldwide, in order to gain a clearer picture on a global level. In addition, as the comments of the expert teachers and of the pre-service teachers are not differentiated in the current research, as it was not the focus of the study, it could also be worthwhile to examine the difference or similarity in opinion of these two groups, for future research.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
This work was supported by Government of Ireland -Academic Mobility Programme.