Abstract
Due to the internationalization and marketization of higher education and the increasing prevalence of English as a global language, higher education institutions worldwide have implemented English medium instruction (EMI) courses, highlighting the necessity to examine how languages are chosen and practised in EMI courses. With a mixed-method triangulation approach, this study collected data through classroom observation, questionnaire survey and semi-structured interviews in order to investigate teachers’ and students’ language choices and practices in EMI classrooms at a Chinese university. The questionnaire results showed that although the respondents slightly preferred English-only practice, they found it difficult to implement it and would resort to translanguaging practices, confirming an inconsistency between language ideologies and language practices. The observation findings revealed the adoption of translanguaging for meaning comprehension, terminology explanation and translation of local elements, while interview findings revealed three types of student participants, i.e. believers, sceptics, and inbetweeners, towards translanguaging practice in EMI courses. In contrast, most of the teachers held a favourable stance towards the English-only practice in EMI courses. While the participants all regarded translanguaging practices as facilitators for language and content learning, they believed that hierarchical relationships do not exist among languages. This study carries important implications to bridge the gap between language ideologies and language practices and call for attention to developing language practice from a multilingual perspective.
Keywords
I Introduction
English plays a crucial role not only in English language education itself but also in education through English. The significance of English in education subsequently has resulted in the emergence and growing popularity of English medium instruction (EMI). Propelled by the internationalization and marketization of higher education (Hu, 2008; Hu & Lei, 2014), as well as the increasing prevalence of English as a global language, EMI courses appear to be a default choice for higher education institutions (HEIs) worldwide. For instance, in China, EMI has gained increasing popularity and been implemented in many HEIs due to the great importance attached to internationalization. Nonetheless, the language practice in classroom discourse within many EMI classrooms are still native-speaker-oriented due to the dominance of monolingual language policies (Cenoz & Gorter, 2013; Fang, 2018).
In recent years, the English-only norm in EMI classrooms has been challenged by the emerging paradigm of multilingualism, with increasing attention to how languages are chosen and practised in EMI courses. A growing number of studies have been conducted regarding students’ and teachers’ attitudes towards English-only practices in EMI courses (Fang & Liu, 2020), and the effectiveness of implementation of EMI in various contexts, such as China and Japan (Galloway et al., 2017; Zhang & Pladevall-Ballester, 2022), Iran (Derakhshan et al., 2022), Thailand and Vietnam (Sahan et al., 2022), and Bangladesh (Rahman & Singh, 2022), to name a few. With the need for EMI to divorce itself from the native English model to promote linguistic diversity (Baker & Hüttner, 2018; Jenkins & Mauranen, 2019; Jiang et al., 2022), a new research niche within the multilingualism paradigm, translanguaging, has drawn increasing attention from researchers as a practical theory of language involving code-switching, first language (L1) use, multimodal and further trans-semiotic practices (García, 2009; García & Li, 2014).
This article views translanguaging as a ‘multilingual, multisemiotic, multisensory, and multimodal resource that human beings use for thinking and for communicating thought’ (Li, 2018, p. 26); it is a practical theory of language that encompasses ‘multi-faceted exchanges to convey meanings in communication in situ’ (Fang & Xu, 2022, p. 4). While prior studies have showcased how translanguaging can be used to facilitate content comprehension and communication in EMI courses (An & Macaro, 2022; Fang & Liu, 2020; Muguruza et al., 2023), it remains unclear whether and how teachers and students would respond to it in their EMI-related language practices given the prevalence of the entrenched monolingual ideology in EMI. There is also a need to unpack to what extent the incorporation of translanguaging in EMI courses could address the complex relationship of language ideology, language choice and power issues (Fang & Hu, 2022; McKinney, 2017; Rafi & Morgan, 2022). This study therefore investigates both teachers’ and students’ language choices and practices in EMI courses. As such, the study contributes to the extant EMI literature by further unpacking the complexity and nuances of language policies and translanguaging in the context of higher education in the Chinese context and beyond.
II Literature review
1 Language policy in Chinese EMI settings
Language learning and teaching in EMI settings are influenced by language policy. Investigating language policies for the purpose of increasing the effectiveness of EMI has become increasingly important, as EMI may lead to an extra learning workload for international students (Jenkins, 2014) or even marginalize linguistic resources in the effectiveness of knowledge acquisition (Fang & Hu, 2022). According to Spolsky (2004), language policy involves language belief/ideology, practice and management. Language ideology is the belief about how languages should be practised from one’s perspective; language practice is the habitual pattern when one uses languages, and language management refers to the attempt to influence language practice. Within the language policy theory, the three components interact with each other. For instance, language ideologies may influence people’s language practices, while people’s language practices, to some extent, reflect their language ideologies and are affected by language management (Haidar & Fang, 2019).
Historically, the notion of English-dominated globalization (Bacon & Kim, 2018) has given great impetus to nations and regions in the pursuit of ‘Englishization’. For instance, the demand for a large number of English-proficient experts in China to aid in the modernization process is one major goal the government aimed to achieve since the implementation of the reform and opening-up policy (Hu, 2007). Apart from English language education, more emphasis has been placed on academic education through English in recent decades largely due to the incessant marketization of higher education (Hu & Lei, 2014). Echoing this trend, the Chinese Ministry of Education has attached increasing importance to EMI, announcing that 5%–10% of the courses in Chinese HEIs should be taught in English or other foreign languages (Hu & Lei, 2014). Hu (2007) pointed out that the driving force behind EMI was tied with modernization and led to the persistent top-down endeavour of implementing EMI. In a similar vein, Dearden (2014) also suggested that EMI ‘appears to be a phenomenon which is being introduced ‘top-down’ by policymakers and education managers rather than through consultation with the key stakeholders’ (p. 2).
Given the circumstances, the English-only practice in EMI courses was greatly accentuated in Chinese mainland (Fang & Baker, 2018), similar to what has happened in other places such as Pakistan (Mahboob, 2020) and Taiwan (Chang, 2019). Despite the accentuation of the EMI courses, previous studies show that, paradoxically, there was still a gap between the ideal and the actual language practices in EMI courses (Dafouz & Smit, 2020; Hall & Cook, 2012; Rose et al., 2022). For example, Fang and Liu (2020) noted that while the language ideology and practice in the classroom discourse of Chinese HEIs remained native-speaker-oriented, the actual EMI classroom discourse often involves other languages and linguistic resources. This echoes the need for translanguaging in language education to promote multilingual and multimodal opportunities to maximize both language and content learning (Fang & Liu, 2020; J. E. Liu et al., 2020; Peng, 2019). Considering the mismatch between the ideal and actual language practices, a bottom-up approach addressing language policies and practices in situated contexts is needed and the current study is a response to such a need.
2 Language ideology and practice in EMI courses
To understand the mismatch between the ideal and actual language practices in EMI courses, the construct of language ideology is helpful. Language ideology plays a pivotal role in language choice and practice. It involves one’s beliefs about how languages should be used. For example, Blommaert (1999) stated that the choice of the medium of instruction is not merely a choice regarding a linguistic issue but involves ideological concerns, such as social and political ideas and beliefs pertaining to languages. By investigating key stakeholders’ language ideologies in EMI courses, the gap between language ideologies and actual practices could be uncovered to further improve the overall effectiveness of EMI courses.
Previous research on language ideologies and actual practices in EMI settings has addressed the deep-rooted native-speakerism and monolingual ideology in EMI contexts. For example, in Galloway et al.’s (2020) study, favourable ideologies towards the English-only practice were witnessed. Students regarded the use of their mother tongue as a sign of low English competence and believed that the English-only practice was beneficial for their language learning goals. Similarly, in their investigation of language ideologies in teachers and students from three universities across the UK, Thailand and Austria, Baker and Hüttner (2018) revealed students’ negative views on using other languages in EMI courses. Negative attitudes were also witnessed among teachers in terms of mixing English with other languages in EMI courses. Chang (2019) also discovered a similar monolingual ideological phenomenon among policymakers at a university in Taiwan. Driven by internationalization, the university clearly defined the ‘E’ in ‘EMI’ as ‘English-only’, mandating that English should be used throughout every pedagogical activity in EMI classrooms. Given the strict requirement of EMI, Chang (2019) pointed out that the roles of other languages, semiotic resources and modalities were largely neglected in the process of knowledge-imparting when the English-only policy was reinforced uncritically.
The monolingual ideologies and practices, however, appear to be under scrutiny by other researchers who adopt a more multilingual and translanguaging perspective. Translanguaging as a practical theory of language and a pedagogical tool (Cenoz & Gorter, 2021, 2022; Li, 2018) has viewed communication as a whole by recognizing multilingual speakers’ ‘multiple discursive practices’ (García, 2009, p. 45). More importantly, translanguaging extends beyond the concept of code-switching but emphasizes the nature of multimodal and trans-semiotic practices in communication (García & Li, 2014; Li & Lin, 2019). In this sense, communication or classroom discourse is not viewed through a separate language choice or use but to recognize learners’ linguistic repertoire as a whole and promote inclusiveness of language use and equity through education (Fang et al., 2022; Li & García, 2022). For instance, in M. M. Gu et al. (2021) investigation of content teachers’ language practices in EMI classrooms in elite universities in Shanghai revealed widespread existence of translanguaging practice, with teachers’ acceptance of students’ first language and bodily actions as important resources for learning. Other studies also reported that the limited English proficiency of both teachers and students has indeed made it difficult to sustain the English-only policy in the Chinese EMI classroom practices (Hu & Lei, 2014; Jiang et al., 2019). More recently, Fang and Hu (2022) investigated the nuanced linguistic ideology and practices of multilingual Chinese university students with Teochew as their L1. The participants’ challenged monolingual EMI practices in their knowledge acquisition and recognized the significance of translanguaging practices to promote inclusive education in their academic achievements. Thereby, whether teachers and students would conform to the English-only policy in their language practices within EMI settings remains questionable and warrants further exploration.
We argue that exploring stakeholders’ language ideologies and practices from a multilingual perspective would be beneficial. Although language ideology and language practice are intermingled, the relationship between them is far from straightforward (M. Gu, 2014; Sahan et al., 2022). For instance, people who hold a monolingual language ideology may or may not maintain the English-only ideology in EMI practices. Thus, it is of great significance to investigate language practices in EMI classroom contexts to further unpack the complexity of language ideologies and practices. It is also necessary to understand whether hierarchical relationships exist among languages, especially from the exclusive and native-speaker-centric views of monolithic monolingualism (Li, 2018, 2022; Rose & Galloway, 2019).
3 Monolingual ideologies and power imbalance among languages
The monolingual ideology in language education has been influenced to the power of Anglophone varieties of English as representation of ‘educatedness’ that are chosen as the preferred choice in EMI contexts (Alisaari et al., 2019; Fang, 2020; Jenkins & Mauranen, 2019). English has close ties to colonization, which resulted in its usage in different social contexts; for instance, within inner, outer and expanding circles (Kachru, 1998). Considering language ideology, outer-circle contexts continued with colonial language ideology. In many post-colonial settings throughout Asia and Africa, ‘the use of English as a language of learning and teaching was imposed by imperial powers’ (Galloway, 2020, p. 25). Such powers lead to monolingual language ideology and a pertaining power imbalance among languages.
Such imbalance has also been referred to as linguistic racism (Dovchin, 2020). According to Dovchin (2020), linguistic racism means ‘the ideologies and practices that are utilized to conform, normalize and reformulate an unequal and uneven linguistic power between language users’ (p. 774), which includes L1 use and accent. The spread of English often induces power imbalance among languages by marginalizing other languages. In this sense, the emphasis on EMI in the process of internationalization often implicitly creates a hierarchy of languages, making the role of English in teaching, learning and research indispensable (Chang, 2019).
Previous studies have examined the relationships among language hierarchies and power imbalance in both outer- and expanding-circle contexts (Chang, 2019; M. Gu, 2014; Lee, 2014). The findings of Lee’s (2014) study showed that Korean students in a two-way immersion programme tended to use English predominately because they believed English is the language of power. As such, Korean was treated as subordinating to English, leading to an invisible hierarchy of languages. Similarly, M. Gu (2014) investigated language ideologies and practices of students at a multilingual university in Hong Kong. The students in her study regarded it inappropriate to shuttle between English and Cantonese since they believed that it is not a pure language practice. This notion echoes the monolingual ideology, which underpins the power imbalance of languages.
In recent years, the emerging multilingual turn (May, 2014) advocates a deconstruction of the power differentials among languages through innovative practices such as translanguaging (Cenoz & Gorter, 2013; Fang & Liu, 2020). Previous research shows that with translanguaging practices, it is possible to provide legitimacy to every language in EMI courses. For example, Chang (2019) argued that translanguaging practices provide opportunities to challenge the existing power structures among languages since they empower polyglots to utilize their multilingual repertoire in EMI courses. In a similar vein, Yuan and Yang (2023) revealed translanguaging strategies by a teacher educator including ‘integrating academic discourse with everyday discourse’, ‘linking verbal and other semiotic resources’, and ‘using students’ first language’ to facilitate his EMI content teaching and learning. They argued the need for further research to understand the adoption of planned and generative translanguaging (i.e. pedagogical and spontaneous translanguaging) for effectiveness of EMI learning.
The adoption of translanguaging in EMI courses is argued to promote a de-colonizing pedagogy in that linguistic repertoires of different individuals can not only be recognized but maximized through various discursive practices from teachers and students (Cenoz & Gorter, 2021; Fang & Liu, 2020; Li & García, 2022) that it ‘may contribute to an open attitude towards English variations and language mixing’ (W. Wang & Curdt-Christiansen, 2019, p. 334). However, there is still a paucity of research examining language imbalance and the power differentials underlying the language ideologies and practices in EMI settings in expanding-circle contexts such as China. To address the research gap, this study aimed at revealing the hidden language imbalance, if any, and the underlying power differentials among languages in an EMI setting in China by investigating both teachers’ and students’ language ideologies and practices. The study was guided by the following questions:
Research question 1: What are the language ideologies held by EMI teachers and students?
Research question 2: What are the language choices and practices of teachers and students in EMI courses?
Research question 3: How do the teachers and students perceive the hierarchical relationships between languages in EMI courses?
Contextualized in the Chinese EMI context, this study can help stakeholders ‘to understand the contribution of EMI in China to the global sociolinguistics of English in terms of power balance’ (Y. Wang, 2020, p. 104). The study can also offer a multilingual perspective within EMI contexts and shed light on how legitimate positions to different languages can be provided in EMI classrooms.
III Methodology
1 Setting and participants
This study was conducted at a comprehensive university in southeast China. Resembling many other universities nationwide, it has advocated internationalization through implementing EMI courses in various programmes, such as courses offered by the Centre for Global Studies. Echoing the policy issued by the Ministry of Education in 2001 in the guideline entitled ‘Guidelines for Strengthening Undergraduate Education and Improving the Quality of the Undergraduate Programme’, the university has attached increasing importance to offering EMI courses.
Participants in this study were EMI teachers and students at the focal university. The university has 13 colleges with more than 20,000 students and more than 1,200 academic faculty members. Teacher participants were from five different schools/colleges. They all had study-abroad experiences and offered EMI courses at the university. As per university policy, each EMI class usually has approximately 30–35 students enrolled. During the data collection process in the 2020–21 autumn semester, 161 students enrolled in these five courses. The teachers were willing to open their EMI courses for research purposes and were willing to take part in semi-structured follow-up interviews. All five teachers participated both in the classroom observations and semi-structured interviews. The students who participated in the semi-structured interviews were all recruited through purposive sampling from these five EMI classrooms observed. The student participants who completed the questionnaire were from the five observed EMI classrooms and other EMI courses offered at the focal university. Students from those five EMI classes had all passed the College English Test Band 4, while some had passed the College English Test Band 6. Their English proficiency was at an intermediate level (CEFR B2–C1).
2 Research design and procedure
This study employed a mixed-methods approach, which enables researchers to conduct a multi-level analysis of a complex issue (Creswell, 2009; Dörnyei, 2007). Data were gathered from three main sources: classroom observation, one online bilingual questionnaire survey, and semi-structured interviews.
First, classroom observations were conducted to understand both teachers’ and students’ language practices through classroom discourse (Dörnyei, 2007). Invitation emails were sent to teachers who have the experience of teaching EMI courses. In those emails, the research topic, purpose and procedure were stated. Five teachers were willing to open their classes for observation. Each course was observed at least three times, and each observation lasted for 45 minutes. A total of 855 minutes of classroom observations were audio-recorded, supplemented by field notes (see Tables 1 and 2).
Basic information about teacher participants in class observation and interviews: Qualifications, English medium instruction (EMI) experience and duration of interviews and observation.
Basic information about student participants in interviews: CEFR level, number of English medium instruction (EMI) courses and duration of interview.
Note. CEFR = Common European Framework of Reference for Languages.
Second, an online English–Chinese bilingual questionnaire adapted from Galloway et al. (2020) and Moody et al. (2019) was administered (see Appendix 1). The questionnaire, titled ‘Language ideology and practice of English Medium Instruction (EMI) courses’, included 31 Likert scale questions, with an open question at the end asking students’ overall opinions about EMI courses. The questionnaire is divided into two main themes regarding participants’ language ideologies and language practices respectively; the data were collected via an online platform called Wenjuanxing in the 2020 autumn semester. In total, a number of 163 students completed this online questionnaire by rating on a six-point Likert scale (1 = strongly disagree, 2 = disagree, 3 = partly disagree, 4 = partly agree, 5 = agree, 6 = strongly agree). All student participants were undergraduate students at the focal university. The first section of the questionnaire included statements of Language Belief / Ideology such as ‘I chose EMI because . . .’ and ‘I believe that in EMI . . .’; The second section of the questionnaire included statements of Language Practice such as ‘In EMI courses, I . . .’ and ‘In EMI courses, my teachers . . .’ The Cronbach’s alpha was 0.884 for questions regarding Language Belief / Ideology and 0.765 for Language Practice, which indicated high reliability of the questionnaire (Hair et al., 2010).
Third, interviews (N = 5 for teachers; N = 10 for students) were conducted to gather further information to address the research questions (see Appendix 2). Interviews, as Kvale (2007) argued, are ‘a uniquely sensitive and powerful method for capturing the experiences and lived meanings of the subjects’ everyday world’ (p. 11) to not only ask ‘what’ questions but also to elicit the ‘how’ and ‘why’ from the participants (Lune & Berg, 2017). The third author conducted all 15 semi-structured interviews with the first author as the facilitator ‘to ask a series of regularly structured questions, permitting comparisons across interviews, and to pursue areas spontaneously initiated by the interviewee’ (Lune & Berg, 2017, p. 70). As the first and third authors were working and studying at the focal university, from an emic perspective (an insider perspective to establish close relationship with the participants) of conducting study at the same institute, the researchers were fully aware of subjectivity involved in this study. The researchers also adopted etic strategies as being reflexive and explicit regarding the roles of the participants (Lincoln & Guba, 1985) to establish rapport with the participants (Hennink et al., 2011). The etic perspective allowed the researchers to understand their roles as researchers with the participants, to control the naturalness of the data collection process and not to take the data for granted (Hennink et al., 2011). In terms of ethical issues, anonymity and confidentiality are ensured for all participants as we concealed the names of the participants and addressed them as S1 or T1. In this study specifically, responses related to language beliefs and practices, especially those frequently mentioned ones, were noted to address the three research questions. Five teachers expressed their willingness to take part in the follow-up interviews. Ten students taking EMI courses were purposively selected for interviews (2 from each of the 5 classes observed). The selection criteria were based on how the students used language during class observations, mainly on (1) how they adopted multilingual language practices when answering questions or doing class presentations; and (2) how they expressed willingness to participate in this study (Creswell, 2009). Each interview lasted from 20 to 38 minutes. The interviews were conducted in Mandarin or English according to the participants’ first language, which enabled them to give in-depth responses to interview questions (Mann, 2011).
3 Data analysis
The quantified results from the questionnaire could help obtain data of students’ language ideologies and practices. Qualitative data including classroom observations could help researchers to gain both students’ and teachers’ actual language practices in EMI settings, while interview data could offer reasons of students’ and teachers’ thoughts about their language ideologies, practices, and attitudes about language hierarchy. These triangulated methods could offer more holistic data towards the research questions investigated.
All classes observed were audio-recorded and transcribed. Each recording was listened twice, with the help of field notes, to capture every language practice performed by both teachers and students, especially when translanguaging practices were involved. Further thematic analysis was conducted to code and categorize each possible theme.
The questionnaire data were processed and analysed via SPSS version 20.0. First, each answer on the six-point Likert scale was assigned a numeric value for the degree of agreement on the provided statements. When the data were imported into SPSS 20.0, negative statements in each scale were reverse coded. Reliability analyses were then conducted to examine the internal consistency reliability of different parts of this questionnaire. Descriptive statistics (e.g. means, standard deviation) were calculated and then adopted to analyse the questionnaire data, which were used to address research question 1 about students’ language ideologies. Dependent-samples t-tests were run to examine students’ attitudes towards EMI teachers as well as their preference between the English-only practice or translanguaging practices in EMI courses.
The interview data were audio-recorded and then transcribed verbatim. Repetitive reading of the transcribed data was carried out before qualitative content analysis (Schreier, 2012). To ensure validity and reliability of the data analysis, the transcripts were returned to the interviewees for an accuracy check. NVivo 11 software was utilized to store, organize and analyse the unstructured qualitative data. After reading all the transcripts meticulously, different codes were produced to label relevant excerpts and vignettes. For instance, ‘I prefer English-only teaching because students would be overly dependent on their first language if the course is taught in bilingual form’ was coded as ‘avoidance of over-dependence of L1’ under the label of ‘students’ language ideologies’. The extracted and categorized themes were used to further address research question 1 as well as to answer research questions 2 and 3.
IV Findings
1 Classroom observation findings
The observed translanguaging practices of teachers were classified into three categories, involving meaning comprehension/reinforcement, terminology explanation and translation of local cultural elements. Based on observations, the ratio of translanguaging usage varied among teachers. T2 and T5 abided by the English-only linguistic practice throughout their courses. T3 and T4 adopted translanguaging practices for explanation and translation purposes, but they still utilized English as the predominate teaching language. Although translanguaging practices were observed, in T1’s EMI course, Chinese was the main teaching language mainly because of this teacher’s inadequate proficiency in spoken English.
a Meaning comprehension/reinforcement
One frequent use of translanguaging was meaning comprehension/reinforcement, which refers to the use of translanguaging practices, including L1, pictures and diagrams, as a complement to help students understand subject contents. This practice was observed among both teachers and students throughout all classroom observations.
Extract 1 T2: Suppose a man has faith but does not struggle hard enough [drawing pictures on the whiteboard]. He wanders away from the practice of yoga and fails to reach the profession. What will become of him? He has given up things in this world he could have had, but he did not achieve right this ultimate liberation that life is all about. (16 Nov 2020)
In Extract 1, when T2 was trying to explain a philosophical concept, he utilized visual aids to make the concept more direct for students to grasp. The stick figure refers to the ‘man’, and the vertical line in the middle represents the terminus of his present earthly life; the arrow points towards what may lie beyond, whether in this world (a next earthly incarnation) or another. The arrow likewise signifies the path of the soul in the quest for liberation (Figure 1). By employing visual aids in EMI courses, students could gain a more vivid sense of what the professor was trying to convey.
Translanguaging strategy used by T2: Visual aids for meaning comprehension.
Extract 2 T3: There are several legal terms for this case that I want to remind you to pay attention to. The first one is parens patriae. This is a Latin form, meaning parents of a nation, right? So, what is the Chinese meaning of this phrase? Students: 监护人. [Custodian.] T3: 监护人，可以再准确一点吗？ [Custodian, could you please be more specific?] Students: [no responses.] T3: On the state, right? A state as a parens, 政府监护, the Chinese meaning of this phrase is 政府监护. And some people translate it as 国家亲权. This is another translation, but I prefer 政府监护 (parens patriae). (17 Nov 2020)
Extract 2 illustrates how T3 used the students’ L1 to accentuate a key legal term. When the students’ response did not meet T3’s expectations, T3 resorted to Chinese for clarification, asking 可以再准确一点吗？ (Could you please be more specific?). In such a case, the use of L1 could draw students’ attention and achieve the goal of meaning comprehension. Meanwhile, T3 also repeated the Chinese meaning of parens patriae (政府监护) several times in order to reinforce the meaning of the legal term in the students’ minds. In this case, the facilitation of translanguaging in meaning comprehension eased the difficulty of comprehension if L1 is avoided. When T3 first explained the meaning of parens patriae as ‘parents of a nation’, the students were not able to understand. The use of L1 from the teacher could facilitate students’ understanding of certain concepts, while avoidance of L1 may contribute to some difficulties in meaning comprehension.
b Terminology explanation
The second type of translanguaging practice is terminology explanation. This type of translanguaging practice differs from meaning comprehension/reinforcement in that the teachers would add more explanations to specific terminologies in Chinese. This kind of translanguaging practice was most frequently observed in T1’s courses.
Extract 3 T1: 民主这个词叫做 ‘democracy’. Democracy是由 demos和 kratos这两个词构成. Demos是指所谓的民众, kratos就是统治。所以民主就是人民的统治，民众的统治. Demos里面的民众指的是公民. (The word 民主 is called ‘democracy’. This English word has its roots in the Greek demos and kratos. Demos refers to people, kratos to rule. So, democracy means the regime of the people. Also, demos refers to citizens.) (18 Nov 2020)
Extract 3 is an archetypical example of terminology explanation with translanguaging practice. Although T1 thought all the students knew the word, T1 still explained the term ‘democracy’ in Chinese because he believed that in EMI courses, the main purpose is to let students master the subject contents instead of English ability. Therefore, when encountering important terms in that course, T1 would resort to Chinese to explain to ensure that every student could comprehend them because this course focused on political concepts of the contemporary world. Although most of concepts can be translated to English, the English meaning may not be as straightforward as the original Chinese meaning. Therefore, T1 would resort to Chinese to explain to ensure that every student could comprehend them.
c Translation of local cultural elements
The third type of translanguaging practice is the translation of local cultural elements, which refers to the translation and explanation of locally based content knowledge. In this study, this kind of translanguaging practice was most frequently observed in T2’s course, especially when students were doing PowerPoint presentations.
Extract 4 See Figure 2.
Translanguaging practice used by students: Translation of local culture elements.
In Extract 4, the picture taken from a student’s PowerPoint presentation slide in T2’s course demonstrated the translanguaging practice by translating locally based content knowledge. For instance, although students are familiar with ‘天父堂’, they might still be bemused when it is shown in English – ‘The Taiping Worship Hall’. Thus, a translation of local cultural elements could facilitate students’ understanding of the content knowledge.
2 Questionnaire results
In terms of students’ attitudes towards the English-only practice in EMI courses (see Tables 3 and 5), the questionnaire results showed that, in general, the respondents only held slightly favourable attitudes (M = 3.72, SD = .767) towards the English-only practice in EMI courses. Over half of the respondents (64.42%) believed that the English-only practice in EMI courses is appropriate (M = 3.96, SD = 1.351), and using English-only indicated a high English language proficiency (M = 3.75, SD = 1.362). Around 84% of the respondents believed that the English-only practice could benefit their English learning (M = 4.56, SD = 1.166). In addition, 63.2% of the respondents expressed that the English-only practice indicated a more academic atmosphere in EMI classrooms (M = 3.98, SD = 1.317). Regarding students’ ideas of language practices in EMI courses, 68.7% of the respondents regarded it as necessary to encourage teachers to implement the English-only practice (M = 4.15, SD = 1.327), and 77.3% of them deemed that students should also be encouraged to conform with the English-only practice (M = 4.42, SD = 1.191). However, it is noticeable that even though the respondents held relatively positive attitudes towards the English-only practice, it was also difficult for them to implement it. Only 19.02% of the respondents claimed that it is easy for them to communicate with others using only English (M = 2.49, SD = 1.173), and only 17.8% of them regarded it as easy to learn course contents using only English (M = 2.42, SD = 1.257).
Students’ attitudes towards language practice in English medium instruction (EMI) courses.
The respondents were in favour of using translanguaging in EMI courses (M = 4.03, SD = 0.514) (see Tables 4 and 5). This indicates that the respondents were in favour of translanguaging as a supplementary tool in EMI courses since it could promote both subject learning (M = 3.64, SD = 1.295) and language proficiency (M = 4.19, SD = 1.303). On average, the respondents believed that adopting translanguaging in EMI courses was more appropriate than using English-only. The results showed that although students held slightly favourable attitudes (M = 3.72) towards English-only, as mentioned above, their attitude towards translanguaging was more positive (M = 4.03). This also been confirmed by the t-test result: the difference was significant: t(162) = −3.239, p < .05, r = .247 (see Table 5). In other words, the students were significantly more positive towards the involvement of translanguaging in EMI courses, although the effect size was medium. Therefore, adopting translanguaging in EMI courses was more preferred. The results also showed that although students’ held slightly favourable attitudes towards the English-only ideology, they still resorted to translanguaging practices, confirming that there is an inconsistency between language ideologies and language practices (Deroo & Ponzio, 2019; Fang & Liu, 2020; Qiu et al., 2022; W. Wang & Curdt-Christiansen, 2019).
Students’ attitudes towards translanguaging practices in English medium instruction (EMI) courses.
Results of dependent-samples t-test towards language practices.
As for respondents’ attitudes towards EMI teachers, their preference for native English-speaking teachers (M = 3.51, SD = 1.407) in EMI courses outweighed Chinese teachers (M = 3.34, SD = 1.090). However, this difference was not significant: t(162) = 1.614, p > .05, r = .126 (see Table 6). In terms of students’ and teachers’ language practices in EMI courses, although the English-only practice was valued, the involvement of other languages was also embraced in EMI courses because of its function in promoting communication and subject learning.
Results of dependent-samples t-test towards English medium instruction (EMI) teachers.
3 Interview findings
a Diverse language ideologies
Categorized based on the themes extracted from student respondents, three different language ideologies could be uncovered. These ideologies were held by the believers, the sceptics and the inbetweeners. In general, apart from two students, S4 and S7, who believed the English-only practice should be the norm in EMI courses, the other eight students held slightly negative attitudes towards it, whereas most of the teachers held a favourable stance towards the English-only practice in EMI courses. This is because teachers may believe the use of monolingual English-only practice could establish their power and authority in EMI courses, with some mentioning the invisible university policy, while students tended to adopt their linguistic resources in real-life classroom linguistic practices for knowledge acquisition. Based on teacher participants’ responses, two language ideologies could be unpacked, overlapping with the believers and inbetweeners ideologies in students (Jiang et al., 2022).
The believers’ language ideologies
Among all the responses collected, the students supported the adoption of English-only practice simply because of its efficiency at improving their English proficiency. This was explained in S4’s following response: Extract 5 S4: In fact, I think bilingual courses are not helpful for students to improve their English ability, while English-only teaching is the best way to do it. You have more chances to practise English in an English-only EMI classroom. (23 Nov 2020)
S4 believed that English-only teaching could bring more benefits for enhancing students’ English proficiency than bilingual EMI courses. This notion is frequently seen since the monolingual ideology has long prevailed in not only language learning but also language teaching fields (Alisaari et al., 2019; Hall & Cook, 2012; Y. Liu & Fang, 2022). For instance, T3’s EMI teaching style had long been influenced by her personal language ideology. She believed that an immersive English-only environment could bring the best results for English and subject learning.
Extract 6 T3: I think an English-only practice in EMI courses is helpful for students to learn and think in English. If students are required to use English-only, and are used to it, when they later encounter first-hand information in English, they do not need to rely on translation to understand it. (1 Dec 2020)
There is a possibility that students’ language ideologies towards EMI courses would be affected by the ones that teachers correspondingly hold. For example, S4 mentioned that an English-only environment seems to be the default rule in EMI courses since the EMI courses he took were mostly English-only.
Extract 7 S4: Using English-only seems to be an underlying rule in EMI courses. I think students should participate in class in English rather than in Chinese since the classroom environment is in English and teachers require us to stick to English-only. (23 Nov 2020)
As if to respond to S4’s answer, T4 also mentioned that an English-only immersion classroom was the thing she intended to create.
Extract 8 T4: There is little chance for students to practise their English in China, and language environment is of great importance for learning a language. Thus, I want to create an English-only immersion atmosphere for students in EMI courses. (4 Dec 2020)
Despite the ideological similarities held by students and teachers, there is still one factor – the university’s language policy – that influences the language ideologies of teachers more so than those of students (Jenkins & Mauranen, 2019). For instance, T4 clearly stated that the university’s language policy was the main reason for her to maintain the English-only practice. However, no similar responses were mentioned by student participants.
Extract 9 T4: I stick to English-only practice because it is a requirement from the language policy of the English Language Centre. We are told that all the teaching activities in classes should be in English-only (4 Dec 2020).
Based on the data above, it is noticeable that it is not only the internal thoughts – for instance, students’ yearning to improve their English proficiency and teachers’ personal language ideology – but also the external factors, including the EMI classroom environment and the university’s language policy, that contribute to both stakeholders’ ideas about language practice in EMI courses.
Furthermore, if the goal of immersion is to be achieved, the use of L1 would need to be strictly restrained in EMI courses. For instance, T4 expressed their concerns about the over-dependence on L1 in EMI courses.
Extract 10 T4: I think at present in China, after all, many students’ English ability is not particularly good, especially oral English. If it is a bilingual class, it may turn into a Chinese class in the end. (4 Dec 2020)
Similarly, S7 also deemed that the goal of EMI could not be achieved if Chinese is used in EMI courses.
Extract 11 S7: If I take a bilingual course, I may overly be dependent on Chinese, so I prefer English-only courses. I believe the purpose of choosing EMI courses is to accept knowledge in English. So, if teachers explain the content in Chinese, students would accordingly learn it in Chinese. Thus, the goal of EMI could not be achieved. (27 Nov 2020)
The sceptics language ideologies
Learning subject contents in English seemed to be the primary goal for S7. However, it is another story for those sceptics who believed the English-only practice would only obstruct their learning efficiency. For instance, S1 considered an English-only practice a hypercorrection, as he explained: Extract 12 S1: I think it is a hypercorrection to stick to English-only practice. The purpose of taking EMI courses is to acquire more knowledge, so the language could only be regarded as a tool to help you to learn contents. Thus, it is unnecessary to stick to English if you could be more productive when using your first language. (16 Nov 2020)
Here, S1 expressed his concerns about such an English-only monolingual language ideology, as he saw language as a tool for knowledge acquisition. From the comment, language choice is not regarded as the most important factor compared to the goal of learning, so the student expressed his belief in the ‘hypercorrection’ of monolingual language ideology.
The inbetweeners language ideologies
Aside from the two extremes, some of the students held an in-between position. They believed that language choice and practice should be analysed contextually and flexibly. As such, translanguaging practices in EMI courses appeared to be natural for them, especially when students’ English proficiency was insufficient. For example, S5 believed that teachers should shuttle between languages (e.g. students’ L1s and English) in accordance with students’ English proficiency.
Extract 13 S5: In fact, I think the use of language in EMI courses should be analysed on a case-by-case basis. If teachers use English all the time in classes, it may be unfair to some students because they cannot understand and hardly learn any knowledge. But if teachers use too much Chinese, students could not achieve the goal of learning English. Therefore, I think the use of languages in EMI courses needs to be balanced. (26 Nov 2020)
In line with the inbetweeners, some teachers also believed English-only is not always meaningful. For instance, T1 agreed on the effectiveness that the use of L1 could bring to meaning comprehension and content learning.
Extract 14 T1: The English proficiency of both teachers and students is not good enough to achieve the purpose of communication in English. If we insist on using English all the time, students may not understand the course and teachers may not convey the content clearly in English. I think, in this case, bilingual teaching will be better. (30 Nov 2020)
The interview data extracted above revealed both students’ and teachers’ language ideologies in EMI courses. In these two extracts, neither S5 nor T1 preferred English-only or adopted translanguaging in EMI courses. They expressed more practical reasons for language use and knowledge acquisition. They may prefer a more balanced translanguaging of linguistic performance in EMI courses, depending on the language proficiency of both teachers and students, from the perspective of bottom-up policy and practice. However, as mentioned before, even though one’s ideology is understood, their language practice cannot be easily and directly predicted. Based on the interview data, this study discovered that language practices were mismatched with language ideologies, which is discussed in the following section.
b Mismatched language practices and ideologies
The data also revealed a discrepancy between language ideologies of the students/teachers and their language practices within actual classrooms. For instance, although S4 is an exponent of the English-only practice in EMI courses, S4 still resorted to Chinese when having discussions.
Extract 15 S4: Honestly, I would use Chinese when discussing with my classmates. I think discussion is about constantly expressing and exchanging ideas. In that case, it is more convenient for me to communicate in Chinese. As a matter of fact, due to my inadequate English proficiency, I could not communicate well in English with my classmates. (23 Nov 2020)
As mentioned above, S4 expressed the need to maintain English-only in EMI courses from the believers’ perspective. When asked about his language practice in EMI courses, he resorted to translanguaging practices in real-life discussions with his classmates. The inadequate English capability became one major hindrance for S4 to effectively communicate in EMI courses. Thus, even though he believed that the English-only practice is more beneficial than bilingual practice, he still had to resort to Chinese for more effective communication. Similarly, despite holding the same language ideology as S4, S3 also believed that using Chinese situationally rather than maintaining the English-only practice could make her study more efficient.
Extract 16 S3: I would not stick to English-only practice because I think the main purpose of EMI courses is to impart knowledge to students instead of adhering to the language policy. I believe using various linguistic resources to express my thoughts must be more direct, efficient and accurate than in English-only. (22 Nov 2020)
In line with S4 and S3, based on the observation and interview data, all Chinese teachers would resort to translanguaging practices in EMI courses mainly for enhancing content comprehension purposes. For instance, T4 used Chinese to explain certain jargon and untranslatable expressions in Chinese, as she explained in extract 17.
Extract 17 T4: There are two scenarios that I would utilize Chinese in EMI courses. One is that when it comes to jargons. If there are corresponding expressions in Chinese, I may explain them in Chinese to facilitate students’ understanding. Another is that when it comes to Chinese cultures, if there is no similar expression in English, then I would use Chinese to explain. (4 Dec 2020)
Similarly, T3 also resorted to Chinese to facilitate students’ understanding and content comprehension.
Extract 18 T3: In the classroom, when we come across long and complex legal provisions, I would not only decompose the original text grammatically, but also ask students to translate it. This is the participation of mother tongue. After the translation, students will have a better understanding of the legal provisions. (1 Dec 2020)
As displayed by the findings, both students and teachers embraced translanguaging practices rather than maintaining the English-only practice during the EMI courses regardless of their English-only language ideologies. Thus, the interview data above uncovered a gap between the English-only language ideology and the actual translanguaging practices in EMI courses.
c Uncovering the hierarchical relationships among languages
Language hierarchy, which refers to ‘the accumulated result of language practices and representations of languages carried out by individual and collective actors under specific historical and geographical circumstances’ (Risager, 2012, p. 129), is worth investigating in language use and education. From the perspective of language ideology, with language hierarchy, English is often regarded with a status of privilege alongside the key national languages in terms of academic and social use, while other languages might be marginalized (Fang & Hu, 2022; Risager, 2012). When asked about their perceptions regarding the hierarchical relationships among languages, 13 of the participants refuted the existence of such relationships because they believed that languages were just tools for communication: Extract 19 S1: I think languages are for communication, so there is no hierarchy between English and Chinese. They are both tools for communication. (16 Nov 2020)
In addition to S1’s response, T2 also believed that English is not superior to other languages. However, one thing he added about English was its practicality rather than superiority.
Extract 20 T2: English is not in itself a superior language, but as things stand today, it has tremendous practical value. It opens to the student a whole world of thought and literature. If you are doing business with people in North America and in some parts of Europe, you will have a great advantage when armed with English fluency. (30 Nov 2020)
Some participants, however, did hold different opinions. For instance, S9 thought that hierarchical relationships among languages do exist due to the dominance of western academia in his field.
Extract 21 S9: In my opinion, individuals’ intention to choose EMI courses does imply a hierarchical relationship between languages. Because you feel that English medium education is better, and then you enrol in EMI courses. Besides, nowadays, Western countries hold more rights to speak up, and in terms of academic research, those published in English tend to be more easily to be certified by Western institutions. Naturally, English is considered superior to other languages. (3 Dec 2020)
Additionally, economic strength also contributes to the unbalanced position of languages, as mentioned by S10.
Extract 22 S10: In fact, the hierarchical relationships among languages also relate to economy. University education is a way for many students to achieve self-realization. Naturally, learning one foreign language, especially English, becomes extremely important since it is tied with one’s ability to participate in internationalization. (7 Dec 2020)
In sum, the participants regarded language hierarchy from a more dynamic and inclusive perspective; English is used as a key, dominant language in education, along with some translanguaging practices. The complex power imbalance and hierarchical relationships among the various languages shown above reflect some linguistic ideologies, including language commodification (Heller, 2010) and language as a symbolic power (Kramsch, 2020) in both social and educational settings.
V Discussion and implications
This study investigated both students’ and teachers’ language ideologies and practices in EMI courses. It offers insights into translanguaging practices within the paradigm of multilingualism in the current EMI settings in China. To answer research question 1 regarding language ideologies in EMI, it was found that the student participants showed a complexity of their language ideologies to EMI. The questionnaire results showed that the respondents only held slightly favourable attitudes towards the English-only practice in EMI courses, but more in favour of translanguaging strategies in EMI courses. This indicated that the students preferred translanguaging practices rather than the English-only practice in EMI courses (see also Baker & Hüttner, 2018; Chang, 2019; Fang & Liu, 2020). Compared to the performance in class observation, the questionnaire and interview data revealed an inconsistency between language ideologies and language practices in that while monolingual language ideology has been emphasized in EMI courses from the believers’ perspective (e.g. S4), translanguaging was observed as a de-facto and natural practice in classroom teaching and learning.
The interview findings further revealed some complex language ideologies of the student participants, as three types of the student participants – believers, sceptics, and inbetweeners – were revealed. Regarding the teachers’ ideologies, most held a different position from the students, as they believed that monolingual English-only practice should be implemented in EMI courses, but this was not really revealed from their teaching practices because translanguaging practices through linguistic and multimodal performance were adopted by many of the teachers (e.g. T1, T2, T3 and T4).
In relation to language choices and practices of research question 2, it was revealed that both students and teachers utilize translanguaging in their learning and teaching of EMI courses. For instance, the teacher and student participants utilized translanguaging practices to facilitate content learning and teaching in EMI courses through various strategies from linguistic performance to multimodal and trans-semiotic practices through classroom observation. The three most frequently observed translanguaging practices among the teachers were meaning comprehension/reinforcement, terminology explanation and translation of local cultural elements. This finding is a mismatch from the interview data in that most teachers regarded the importance of English-only practice in EMI courses (e.g. T3 and T4), as they adopted translanguaging practices in their teaching (see Fang & Liu, 2020; W. Wang & Curdt-Christiansen, 2019). For students, they mainly utilized translanguaging practices for effective communication during discussions. The classroom observation findings indicate a practical need from the students and teachers for translanguaging as one teaching/learning strategy. This need was also expressed in the questionnaire data, as students showed a preference for using other languages in EMI courses (see Table 4) and the need for translanguaging to facilitate learning in interviews (e.g. S1, S3, S4 and S5). The findings also echo some previous studies (Fang & Liu, 2020; Ossa-Parra & Proctor, 2021), which also observed translanguaging practices in EMI courses to enable students to perform linguistic analysis for both language and content learning and promote student agency because they are able to use and maximize their full linguistic repertoires (Creese & Blackledge, 2010; Muguruza et al., 2023; Rafi & Morgan, 2022).
In order to answer research question 3 regarding the hierarchical relationships between languages in EMI courses, a surprising finding is that most student participants considered language to be a tool of communication, and they disagreed with the existence of a hierarchical relationships between languages that were involved in their EMI courses. Such a finding indicates the prevalence of a functional orientation towards language use in EMI settings, i.e. adopting a neutral perspective towards languages. A more critical understanding of power imbalance and hierarchical relationships between Chinese and English is also manifested by a few participants, who occupied a minority among the participants though. These participants resorted to academia and economic discourse (e.g. a need to publish in English for international recognition) to support the more powerful and prestigious position of English over other languages such as Chinese. This finding echoes the research results from Bacon and Kim’s (2018) study that English was viewed as being linked with globalization, which was inseparable from Anglo-American economic power. It also reflects Risager’s (2012) study of language hierarchy at a Danish university that language policy and language hierarchy practice are a reflection of ‘multiple complementarity’ (p. 129) with English and local languages with other linguistic resources. More research is warranted to further explore how the belief that ties languages to power and hierarchy can be changed in EMI courses.
The findings yield important implications for the learning and teaching of EMI courses. First, it is helpful for both students and teachers to recognize the effectiveness of translanguaging practices in EMI courses. A monolingual language ideology has long prevailed in EMI fields. However, the effectiveness of the English-only practice in EMI courses may not be guaranteed due to the limited English proficiency of both students and teachers (Hu & Lei, 2014; Jiang et al., 2019). In addition, the English-only practice in EMI courses cannot naturally lead to optimal outcomes in subject learning and English proficiency (Fang & Liu, 2020; Tian & Shepard-Carey, 2020). Thus, stakeholders should advocate translanguaging practices such as using L1s as natural linguistic practices in EMI courses (Creese & Blackledge, 2010; García & Li, 2014; Li, 2018).
Second, raising multilingual awareness among policymakers and teachers when making and implementing language policies is of great significance. Multilingual awareness can enable policymakers to make more conscious decisions when designing language policies and to benefit the implementation of EMI from the macro policymaking level (May, 2014). Additionally, recognition of the legitimate position of a multilingual environment in EMI courses is also important for teachers. As mentioned above, taking an English-only environment as the default rule in EMI courses can largely constrain students’ subject-learning process. We argue that it would be more helpful for subject learning if teachers could embrace a multilingual environment in EMI classrooms and allow students to utilize their diversified linguistic repertoire.
Third, as Hamid and Nguyen (2016) argued, it is often the case that EMI is just ‘dumped’ on the faculty and students with little consultation, preparation and compensation. Thus, it is vital for policymakers to collect suggestions from students and teachers before and after making language policies. Only if all the stakeholders are involved in developing and implementing EMI policies can a systematic and contextualized EMI classroom be developed. By doing this, the goal of building an effective EMI classroom that ‘wholly embraces and builds upon the dynamic cultural and linguistic repertoires of our students across contexts’ (Tian & Shepard-Carey, 2020, p. 11) can be achieved.
The study is not without limitations. First, the sample size was relatively small and thereby the findings should be interpreted with caution in different contexts. Second, this study did not involve policymakers as participants. As mentioned above, the implementation of language policies is a top-down process. The findings can be more representative when more than the bottom layers of stakeholders are involved. Third, due to some concerns mentioned by the teachers, video recordings were not encouraged during the data collection process. Thus, this study only reports certain aspects of translanguaging practices in EMI courses, although some multimodal and trans-semiotic practices were recognized and manifested in both students’ and teachers’ language practices. Notwithstanding its limitations, this study has important implications for EMI language policies to reduce the gap between monolingual English-only language ideologies and translanguaging practices.
VI Conclusions
By employing a mixed-methods triangulation approach, this study showed that while students held slightly negative attitudes towards the English-only practice (due to its ineffectiveness in communication and content learning), most teachers held the opposite view mainly for three reasons: the university’s language policy, personal language immersion ideology, and avoidance of over-dependence on L1. Concerning language practices, despite the monolingual language ideologies students held, they would not stick to the English-only practice due to their inadequate English capability. Similarly, although some teachers expressed some doubts about utilizing translanguaging practices, most of them agreed on its effectiveness for content teaching and learning. Regarding the relationships among languages, both students and teachers believed that hierarchical relationships do not exist among languages themselves. Overall, this study not only challenges the monolingual ideology and practice but also provides evidence for the legitimacy of translanguaging practices within the multilingualism paradigm.
Footnotes
Appendix 1
Questionnaire: Language ideology and practice of English medium instruction (EMI) courses: 英语媒介教学课程中的语言意识形态和实践
Appendix 2
Interview prompts.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Ethics statement
The studies involving human participants were reviewed and approved by Shantou University Research Division. The participants provided their written informed consent to participate in this study.
