Tutors’ beliefs about language and roles: practice as language policy in EMI contexts

ABSTRACT

It has been well established that for all students, but particularly second language (L2) English speaking students, academic English speaking skills are key to developing specialist terminology and disciplinary content in an English as a medium of instruction (EMI) context. However, what is less clear in many contexts is the institutional language policy necessary to guide and support both L2 English speaking students and disciplinary tutors. In this paper, we focus on disciplinary tutors’ beliefs of language and their roles with respect to language support to surface implicit and covert language policies. We argue that in the absence of explicit policy, showcasing the range of tutor perspectives and practice around language support can provide a way forward in explicating good practice and highlighting an approach in which all stakeholders take responsibility for supporting students’ academic speaking skills in an EMI context.

KEYWORDS:

• Language policy
• teachers’ beliefs
• EMI
• academic speaking
• disciplinary studies

Introduction

In this paper, we argue that, due to the neglect of explicit language policy in many higher education (HE) institutions, disciplinary tutors’ beliefs about the role of language and their own role in supporting language can reflect their implicit ‘practiced’ language policies. Language is fundamental to students’ ability to access the higher education curriculum (Bond, 2020). Furthermore, academic speaking is central to a student’s academic performance and achievement and is key in developing disciplinary knowledge (Mah 2016). Therefore, a lack of language support for L2 English speaking students can have a detrimental impact on their learning in higher education. Speaking also has a key role for negotiation of meaning in the classroom and beyond (Mauranen 2012), and for demonstrating understanding of the discipline (Arkoudis and Doughney 2014).

However, despite the centrality of language to disciplinary knowledge and understanding, language policy is not always explicit or formalised. The lack of policy in the form of guidance (for example, advice for teachers on managing language use in the classroom) or any other ‘text’ policy (Spolsky 2004) is problematic as support for academic speaking will be idiosyncratic (Yuan, Chen, and Peng 2020), resulting in diverse student experiences. The aim of this study was to focus on a hitherto neglected area of practices in English as a medium of instruction (EMI) contexts, that of academic speaking, to highlight the beliefs and practices of disciplinary tutors (referred to as tutors henceforth) in the absence of any formulated HE policy.

‘Internationalisation’ is not only a buzzword but a daily lived reality for higher education institutions around the world. Since the turn of the millennium, the strive for international student recruitment has led to the exponential rise in the number of programmes across the world that are taught through EMI (Phillipson 2015). Whilst EMI is commonly defined as ‘the use of the English language to teach academic subjects (other than English itself) in countries or jurisdictions where the first language of the majority of the population is not English’ (Macaro et al. 2018, 37), in this paper, we will use EMI to refer to Anglophone higher education, too. In the Anglophone EMI setting, internationalisation is characterised by the influx of overseas students, leading frequently to classrooms that include a large proportion of students for whom English is a second language. Anglophone higher education (HE) institutions share many features of the internationalised HEs in other parts of the world, where both the student and faculty cohort have English as a second or foreign language and English as a Lingua Franca is often practised (Baker and Hüttner 2017). Furthermore, linguistic outcomes are similar for students in Anglophone and non-Anglophone EMI (Humphreys 2017). Despite the internationalisation of campuses, linguistic diversity is not equally welcomed (Jenkins and Mauranen 2019). The infrastructure, for example, teachers’ guidance on managing linguistic diversity in the classroom, is not established (Bond, 2020).

In the absence of official language policy, perspectives and practices of the tutor as ‘key participant’ (Macaro 2018) constitute institutional ‘implicit’ policy through their enactment at the classroom level (Doubleday 2018, 18). To explore tutors’ implicit language policy, we focus on disciplinary tutors since student success is largely dependent on their ability to facilitate and navigate students’ understanding of disciplinary threshold concepts.

Literature review

Language policy

Tutors work in a wider policy environment provided by their institution or other outside agencies. However, a lack of policy concerning language has been noted for both Anglophone and non-Anglophone EMI environments. In a review of the internationalisation policies of UK universities, Dippold (2015) demonstrated that most institutions do not discuss issues of language in learning and teaching explicitly, do not provide support for students for dealing with the speaking demands of the classroom, and do not provide guidance and good practice guides for tutors to support English as a second language (ESL) students (Arkoudis and Doughney 2014). Discussing EMI policy in non-Anglophone countries, Walkinshaw, Fenton-Smith, and Humphreys (2017) argue that countries and institutions have embraced EMI policies without assessing appropriacy and relevance for the sake of internationalisation and profit. In practice, this means that students and staff often lack practical support in dealing with the realities of EMI (Werther et al. 2014; Lasagabaster 2018), including academic speaking, or are left to develop their own, implicit policies.

Spolsky (2004) distinguishes three distinct aspects of a ‘language policy’. Language practices include ‘observable behaviours or practices’, language beliefs are associated with values that individuals attached to language varieties and their features. Finally, language management relates to the efforts of persons of authority to attempt to modify either language practices or beliefs at the document or text level. Given the lack of language management efforts by EMI universities (Lasagabaster 2018), university tutors develop practices based on their own beliefs about English (Yuan, Chen, and Peng 2020). These practices then become the implicit, unstated language policy of the individual teacher, which may potentially become a university-wide unstated policy.

Bonacina-Pugh (2012) refers to policies at the level of practice as ‘practiced language policies’. Using Spolsky’s (2004) conceptual framework, she defines language policy as language choices. Here language choice refers to participants’ opportunities and use of their first language (L1) or second language/foreign language (L2), which implicitly reflects which languages are legitimised (Bonacina-Pugh 2020). Students’ opportunities to participate in academic speaking skills depends on how the tutor perceives the role of different languages and their own role with respect to supporting language development. We, therefore, argue that the tutor’s beliefs and practices implicitly reflect their language policy. Exploring and identifying these unstated language policies are key to ensuring students are supported in their studies and have equitable access to the HE curriculum (Bond 2020).

Academic speaking skills in EMI

Although academic speaking in EMI has been a subject of intense interest in English as a lingua franca (ELF) studies (Jenkins 2011; 2014; Mauranen 2012; Doubleday 2018), and in the wider HE literature (Dippold et al. 2019; Heron 2019), little attention has been paid to tutors’ beliefs and classroom practice with regard to supporting academic speaking, and the link of such practices to policy. This is somewhat surprising given the crucial role of academic speaking to both the curriculum and the academic success of students, who, through the expansion of innovative and active pedagogic approaches, are faced with high demands on their speaking skills (Doherty et al. 2011; Roberts 2017; Heron 2019).

A growing body of literature in both higher education research generally and applied linguistics has identified considerable challenges students face in participating in spoken discourse in their disciplinary studies. Reasons for these difficulties have been reported as lack of linguistic resources, an increased cognitive load, and lack of familiarity with participation structures and routines of interaction (Engin 2017). Part of this relates to students’ awareness and understanding of specialist disciplinary terminology, which is often tacit and assumed by disciplinary tutors (Bond 2020), yet key to learning content (Basturkmen 2018).

In the light of these challenges, some EMI institutions in both Anglophone and non-Anglophone contexts offer students a course prior to their studies in which they can achieve the English language proficiency required by the institution and develop both speaking and writing skills. These courses are often referred to as pre-sessional courses or ‘top-up courses’ (Thorpe et al. 2017). However, the ability of pre-sessional classes to prepare students adequately for the variety of disciplinary practices they may encounter has been questioned. Upon leaving the pre-sessional classes, students experience a variety of disciplinary practices which a generalist EAP (English for Academic Purposes) course may be unable to sufficiently provide for (Dippold 2014). In addition, English language teaching, including EAP, is still firmly rooted in a normative approach favouring national varieties of English (Jenkins 2012; Sifakis 2019). However, when students are subjected to a disciplinary learning space after leaving the pre-sessional, their progression and success are not determined by their investment in their learning and development alone but also by other speakers and the affordances of the environment. In an Anglophone environment, in particular, students will often learn in an ELF environment characterised by mixed nationality groups that may include L1 users of English. This means that effective speaking practices will not be determined by L1 user norms, but by speakers’ ability to use effective strategies for negotiating meaning.

Other research focuses on the role of the tutor in facilitating this transfer and suggests a lack of awareness of students’ linguistic challenges. There is a widely held view that the tutors believe their job is to teach content, not English (Airey 2011; Aguilar 2017) and that responsibility for language proficiency and development lies with the student (Dearden 2014; Dearden and Macaro 2016) and/or the institution.

In the light of this literature, this study aimed to explore tutors’ practiced language policy in four EMI institutions. In the forthcoming sections of this paper, we will explore tutor’s beliefs of the role of language. In addition, we will also explore the role of disciplinary tutors as agents (Dafouz and Smit 2016) at the interface between top-down policy and practice in classrooms ‘on the ground’. Agents refer to those who ‘take part in the planning, implementation, and assessment of language policies in higher education institutions worldwide’ (Dafouz and Smit 2016, 406). With previous research focusing mostly on a deficit view of teachers’ experience and practices in an EMI context, we also aim to surface tutors’ perspectives to build on the range of good practices supporting academic speaking highlighted in this study.

Our research questions were:

• What are tutors’ beliefs about the role of language (academic speaking) in the HE classroom?

• What are tutors’ beliefs about their own role in supporting language (academic speaking) in the HE classroom?

• How are these beliefs reflected in the practiced language policy?

Materials and methods

Research contexts and participants

The two UK HEIs were large universities in the South of England, with University A comprising 6230 international students at the time of the study and University B comprising over 6500. University C was situated in the United Arab Emirates and had a population of 4006 students, 392 of those in the English preparatory programme (academic year 2018/2019). All students shared Arabic as their L1. University D is a foundation EMI university established in Turkey, and as of 2018/2019 academic year, with 13,000 students, 2229 in the English preparatory programme and 864 international students. The majority of students share Turkish as their L1. There are a considerable number of studies focusing on EMI practices in Turkey (Inan, Yuksel, and Gurkan 2012; Dearden 2014; Soruç and Griffiths 2018) but to our knowledge, very little on EMI in UAE, with one notable exception being Belhiah and Elhami (2015).

In terms of explicit language policy, the following observations were made. At the two Anglophone universities in this project, only the relevant admissions policies make explicit reference to language, laying out the expectations (e.g. IELTS scores) for students to be admitted to the university and their relevant programmes. At the two non-Anglophone institutions, explicit language policy is reflected in statements about English being the primary language of instruction. Influences on participants’ practices and beliefs can also occur through opportunities for professional development. In all four institutions occasional workshops and Professional Development (PD) sessions are held for all staff which focus on supporting L2 English users. However, since these are not systematic nor part of a wider PD policy which explicitly seeks to stipulate desired language practices (Yuan, Chen, and Peng 2020), it is doubtful how much impact they can have in terms of their limited length, and how much reach they have on their attended audience. Overall, issues of language are noticeably rare or absent from the HE pedagogic discourse (Bond 2020).

This paper reports on part of a larger study that included students, disciplinary tutors and English for Academic Purposes (EAP) teachers from the four institutions. For the purposes of this paper, we focus on the disciplinary tutors only (Table 1).

Table 1. Summary of participants.

Country	Tutor questionnaire	Tutor interviews
UK	55	5
UK	36	5
UAE	25	5
Turkey	21	5
Total	137	20

Authors’ positioning

Given our insider status at the four research sites and the role of authors 1 and 2 in the analysis and interpretation of the data, it is imperative to explicate our positionality and backgrounds (Baker et al. 2020) in this study. Authors 1 and 3 are disciplinary tutors with previous experience in EAP. Author 1 has also worked at three of the four research sites. Author 2 is a disciplinary tutor with a background in language teaching. Authors 3, 4 and 5 are EAP tutors. All authors were working at one of the four research sites at the time of the study and so were able to gain gatekeeper approval without any difficulty. This insider position offered the study an emic perspective, and we draw on our familiarity with the context and resultant prejudices as ‘resources’ (Holliday 2016). However, mindful of the need for reflexivity, the research assistant on the project conducted the initial data analysis to provide an outsider and non-specialist perspective on the data.

Methods

Questionnaire

A mixed-methods questionnaire was developed to elicit quantitative and qualitative data on tutor perceptions of key academic speaking skills required for the disciplinary study. The data for this paper derives from the qualitative comments organised according to the following questionnaire prompts:

• What informs your practice in planning your teaching and supporting students in using English in spoken academic activities

• Please note any further comments on your students’ use of spoken academic English.

A total of 97 comments were given for question 1, and 66 comments were given in response to question 2.

Interviews

Semi-structured interviews were carried out with disciplinary tutors in the four sites (see Appendix A for interview questions). Tutors were from a range of disciplines, such as engineering, languages and tourism, and volunteered to be interviewed for the project.

Analysis

We utilised thematic analysis (Braun and Clarke 2006) as it offers a flexible approach to analysing the data. We analysed both the questionnaire and interview data iteratively. Firstly, the research assistant coded the data and organised them in NVivo. The first two authors then coded the data separately and discussed their themes with the research assistant. The themes were agreed upon through discussion, explanation, exemplification and reference to the research questions. We acknowledge here that there was considerably more qualitative questionnaire data from the Anglophone contexts, which is reflected in the findings below.

Ethics

Ethical approval was granted by all four institutions. Participants were informed and consent was obtained. Due to the multi-site methodology, all institutions were responsible for the transcription of interviews, and any identifying information was redacted from the transcriptions before sharing as an anonymous document with the research team.

Findings

In this section, we illuminate implicit practiced language policy by presenting tutors’ beliefs of language, their role in supporting language, and how this is reflected in their practice. In terms of beliefs about the role of language, some tutors consciously, with the purpose of supporting academic speaking, legitimised students’ L1 (Bonacina-Pugh 2020) and some did not. Similarly, tutors diverged with respect to how they viewed their own agentic role in supporting language. Whilst the aim is not to compare Anglophone and non-Anglophone contexts, where relevant, we make this distinction. Perhaps the most significant finding is that the key themes were similar across all contexts and part of the experiences of tutors from all linguistic backgrounds (I = interviews, Q = questionnaire, for example, IA = Interview data from institution A).

The role of language

One clear reflection of enacted language policy is the tutor’s belief about the role of language, in particular the use of students’ L1 in the classroom and how to manage linguistic diversity. A range of attitudes towards L1 and L2 (English) was found, highlighting which languages were legitimised. Flexibility with respect to which language was appropriate was reflected in the data, as was a feeling of distrust and potential lack of control. For example, frustration levelled at the use of L1 was expressed by the tutor below, equalling it to ‘bad behaviour’:

(1) Yes and unless you explicitly tell a student, ‘You must now use English’, they will not use English and even if you say to a student, ‘You must now use English’, they will take that as a suggestion. So, for example, if you’ve got a group having a discussion, you sit down to talk with them, they will very quickly, even whilst you’re still sat there trying to converse with them, switch back to [students’ L1]. Actually, in previous years we’ve had some very bad behaviour where they’ve done that even when there have been non-[students’ L1] speakers in the group. (IB)

Here the tutor does not support the use of the students’ L1, but instead legitimizes a monolingual English classroom. The frustration is evident in the phrase ‘even if you tell a student’. The tutor’s argument for a monolingual classroom is the need for a shared language (L2), although there is frustration at students reclaiming their L1 when working in groups.

The importance of using English in the whole class and small group discussions was justified based on the need to check understanding. The tutor in (2) makes this point.

(2) … because they tend to talk in [students’ L1]in smaller groups, so they may be talking about what they did last night or they may be talking about the activity but when it’s about reporting back to us they find, they still find it very hard and there will be one spokes-person who usually will say, will refer to the activity as I have thought this and I don’t know whether it’s linguistically incorrect or whether they’ve just thought about it and they did the talk, in the smaller groups they were talking about what they did last night. So, it’s quite difficult for us to work out whether, what the spokesperson reports is what they talked about in the smaller groups or whether it’s something that they’ve thought. (IA)

The comment above reflects a lack of trust in the students and their task completion (‘they may be talking about what they did last night or they may be talking about the activity’), and there is a reference to a loss of control (‘it’s quite difficult for us to work out’). Interestingly the speaker here uses ‘we’ as an inclusive term suggesting that other colleagues feel the same way, further justifying their perspective.

In contrast to the above beliefs, there was also an attitude which legitimised the use of students’ L1 and which reflected flexibility with the purpose of supporting understanding. In the comment below, the tutor clarifies this:

(3) My feeling is that if they’re talking to people who speak their language and it’s just that group and actually that’s going to help them clarify something, I'm very happy for them to talk in their home language. (IB)

The tutor above also works on the recognition that ‘code-switching’ into L1 can even facilitate understanding of subject content. However, legitimisation of students’ L1 is dependent on others sharing this language, and so when this is not the case, the L1 becomes ‘deviant’. This is reflected in the comments below:

(4) If I’m not part of the audience, for example in a class group discussion, it doesn’t have to be in English and I have absolutely no problem with that, in fact I very much like that. I think that is contributing to the kind of, the kind of multiple language identity, aligned to expert identity that I was talking about just now. So, I don’t think there is a problem with students using other language in the classroom but you do have a kind of a basic point of etiquette really I suppose, that I use the language which is the language that makes my point accessible to all my audience. (IB)

(5) I'm very happy for them to talk in their home language. If that’s going to exclude anybody in their group, then if it’s more than just a quick, ‘blah blah blah’, you know, just trying to help somebody, so two people with one other person, ‘This is what she means’, you know, quick bit of home language, then I'm fine with that. (IB)

Whilst the tutor in (4) recognizes the link between language and identity, and that the possible use of L1 can be beneficial for the development of an expert identity, they also suggest that, in a larger group including multilingual speakers, use of L1 becomes undesirable and implies bad etiquette. The interactional norm therefore is that English is used as the classroom language. As a consequence, this range of tutor expectations and norms can be challenging for students.

However, in the non-Anglophone contexts, and possibly due to the explicit language management policy of English -only instruction, there are tensions between institutional policy expectations and the reality of the classroom. The tutor in (6) expresses this frustration:

(6) But I didn’t know, they're supposed to do things English, but I don’t have all of them conducted in English because the teaching assistant is Turkish. If all the group members are Turkish, they might tend to use it. (ID)

As argued earlier, the social practices and norms as determined by the disciplinary tutor in the classroom reflect the wider issues of language planning and policy. What we aim to reveal subsequently is how tutors supported students with their academic speaking based on how they perceived their own role and agency in the liminal space between policy and practice.

The role of the tutor

Practiced language policy was implicit in how teachers viewed their role with respect to supporting language and how this is then reflected in their classroom practices. Students’ opportunities to use English for the development of terminology and conceptual understanding rests upon several variables, one being that the tutor perceives language support as part of their role. Tutors’ beliefs of their own role in supporting language development were polarized along a disciplinary content–language continuum.

A belief that the tutor’s role is to solely teach the discipline was particularly evident at the two Anglophone institutions. The tutors below, (7) and (8) are very explicit about their responsibility, which is to teach the discipline:

(7) I teach law in English. I am not an English teacher. (QA)

(8) Oh right. I feel the questions are really geared more towards language learning. I don’t teach language, I teach content courses, for which obviously they need to speak language and I get a lot of international students, but I don’t organise speaking activities. (IA)

Some tutors devolved responsibility for language to the students themselves or by ‘outsourcing’ this to other parts of the university, most usually the Language Support teams. This is particularly evident below where, during the course of the interview, the tutor pulls out a referral chart to show the interviewer:

(9) I am not sure, I have this to hand so that I can, this is the flow-chart for various different bits of support. So, I would usually look up on this, there we go, at the bottom, ‘English Language support’, so if I had a student like a personal tutor or somebody else supporting, I know there’s an area somewhere else in the university who can support. I usually pick up a chart like this … (IA)

Outsourcing was seen only in the Anglophone contexts as these institutions provide in-sessional support and a unit dedicated to drop-in support. The two non-Anglophone institutions provided in-sessional support, which was part of the students’ ongoing degree studies, and so was not optional.

Thus far, we have seen perspectives in which tutors’ responsibility is content only. A second view is one in which the disciplinary tutor orients more to a balanced view of responsibility for both content and language. The tutor below explicitly rejects the notion of outsourcing as reflected in the comments above and believes all disciplinary tutors should take some responsibility, whilst acknowledging that this is still the prevailing view in higher education.

(10) The other perception is that pre-sessional eliminates the language challenge for the second language student, people who are studying through English as a second language, and that I think is a mistaken perception. I think it’s quite wrong if you are teaching a course in literature or history or management or computer science that you think, ‘Oh they have either met IELTS or they have done a pre-sessional and that means there is no language issue that I have to deal with in my teaching or in my assessment’. I think that view does prevail in higher education. (IB)

Interestingly, the tutor argues for the need to continue to support students’ language development beyond the pre-sessional course and into their disciplinary studies yet acknowledges the prevailing view in HE of the disciplinary tutor as one who teaches content only.

However, a middle ground whereby the content specialist and the language expert work together collaboratively was reflected in tutors’ comments in both the Anglophone and non-Anglophone contexts. Such a perspective was seen as the optimal working practice and suggests a language policy of shared responsibility. The tutors’ comments below (11) and (12), reflect a belief of shared roles:

(11) I think for us working with pre-sessional people is brilliant because then we really have a chance of talking to each other about how we do things and I think, you know, a conversation is very useful, it’s very useful for us to have. (IB)

(12) Where actual specialists people who know what they are doing with language actually helped the students throughout their projects. (IC)

Tutors’ role in supporting language was further reflected in their awareness of students’ linguistic challenges. Some tutors acknowledged the importance of ensuring their own language was comprehensible to students. This included awareness of how to talk about the content in a way that would be linguistically and culturally accessible. The tutor in (13) below shows a high level of language awareness and considers the grading of their language as part of their role.

(13) The lecturers have to ask, ‘Is my talk too culturally specific? Too linguistically low frequency or specific? And is there a way I can talk about this content in history or management or science or whatever, talk about it in a way that has accessible terminology, accessible real-world events, accessible metaphors? (IB)

Tutors’ account of activities that support academic speaking – their practices – can further provide an insight into their beliefs about language. Some tutors advocated an integration of language-focused activities into key disciplinary practices. Such activities include discussion, group work, scenarios, checking understanding, asking questions, use of visuals, for example, slides and clear rubrics. Tutors talked about ensuring students could discuss topics in small groups due to fear of embarrassment. Formative feedback was mentioned by several tutors as key to supporting students’ development of language.

(14) It’s important for them to practise their speaking and for us faculty to give them feedback. (QC)

There is also some evidence of explicit inclusion of language developmental activities in the classroom in both Anglophone and non-Anglophone EMI environments. For example, the following two accounts in (15) and (16) describe using activities to develop subject-specific vocabulary and input into good oracy practices:

(15) I think it does help if you have something of a genuine query. What I do with a lot of PowerPoints is I have a slide with missing words and if it is kind of summarising an article or summarising a philosophy or distinguishing between two things, you know, I will have a gapped text with words missing and then I will set them the task of, ‘What are the missing words here?’ (IB)

(16) Well, this is a round table discussion where everybody comes from an informed perspective, so students get time to prepare, they come with their notes, they sit in an oval so that everybody is equal, nobody is at the head of the table, they are asked a question and then there are also rules. The rules are, you have to be polite, I give them a few language segments, how to respond, how to encourage others, how to be polite, how to disagree politely with somebody else and then they discuss their particular research paper in this context. (IC)

Finally, a tutor from non-Anglophone EMI also reports on the inclusion of EAP-generated evaluation rubrics into the disciplinary programme, thus providing integration of English support and disciplinary content:

(17) they have to do formal presentations. And, there's an oral presentation evaluation rubric that was actually prepared by faculty academic English program. So, I use that and I share that with my students at the beginning of the semester. (ID)

In summary, the findings reveal how tutors’ practices are intrinsically linked to their understanding of the role of English in the classroom, but also of their own roles of agents at the EMI university. The data highlights that whilst some tutors harbour reservations regarding the use of L1 to support mutual intelligibility and pedagogic effectiveness, many are supportive of codeswitching into L1 to facilitate understanding of subject content. Furthermore, in terms of their own role, tutors’ perspectives range from total denial of responsibility with respect to supporting language to collaborative disciplinary and language support, to recognition and acceptance of a key role in supporting both disciplinary content and language. In addition, questionnaire and interview data show evidence of good practice in terms of supporting L2 English speaking students, which includes adapting their own language, providing opportunities for prior preparation, and the use of activities for developing speaking skills in class.

Discussion and conclusion

The aim of this paper was to explore tutors’ beliefs about the role of language and their own role in supporting language development to explicate implicit language policies. In terms of the research questions posed at the beginning of the paper, our analysis has shown that teachers’ beliefs can be described on a continuum. On one end of this continuum, tutors take no responsibility for language support, instead, viewing support as a duty that is ‘outsourced’ to language specialists. On the other end, some tutors show opposition to siloing language and content in this manner and have developed their own practices, despite a scarcity of support (Aguilar 2017; Yuan, Chen, and Peng 2020).

Our findings provide important insights into tutors’ beliefs and practices in supporting academic speaking. These beliefs reflect tutors’ ‘practiced language policy’ (Bonacina-Pugh 2012) in the absence of explicit and shared institutional policy. One of the key findings, perhaps, is that, in the (almost complete) absence of any official language management efforts in both Anglophone and non-Anglophone institutions tutors are forced to define their own enacted policy, or practiced language policy (Bonacina-Pugh 2012) and which are based on their ideologies or beliefs about language and language varieties, and their agentic role (Dafouz and Smit 2016) in supporting language development. These practices can be described as the micro-level strategies which are carried out by ‘people with expertise’ (i.e. disciplinary and language expertise) (Fenton-Smith and Gurney 2016).

We, therefore, argue that good policy on EMI is not best described in terms of well-worded statements on glossy paper or carefully designed websites. Equally, leaving individual agents to define policy solely through their practices will lead, at best, to a piecemeal and uncoordinated approach. We argue that, instead, that policy best emerges through an effort of identifying good practice on the ground which is then collaboratively documented. We suggest that such a policy should minimally reflect the following two features. Firstly, it should include a definition and description of the institutionally desired role of English in the classroom, which acknowledges the complexities of an ‘English only’ policy in light of issues such as the linguistic provenance of students and tutors. Secondly, there needs to be a definition, and description of the responsibilities of all stakeholders, including management, administration, EAP (language tutors), disciplinary tutors and skills support tutors, in order to decide who takes what role and, crucially, how agents best work together, in ensuring language support.

One way to achieve the above policy suggestions would be to design measurable targets and incentives to encourage programme and module leaders to engage with research in the subject and achieve a ‘realistic, research-based understanding of what can be achieved’ (Fenton-Smith et al. 2017, 471). Similarly, both disciplinary tutors and language tutors can work together on research that focuses on disciplinary interests.

A second initiative could be a framework for improved information transfer between EAP/pre-sessional tutors and disciplinary tutors in order to dispel myths about role, content and realistic aims of the pre-sessional as well as providing EAP staff with an insight into specific disciplinary practices. This framework might include:

• Training and research-based guidelines to develop disciplinary tutors’ understanding of language issues (e.g. English as a lingua franca and negotiation of meaning, language needs for disciplinary genres), and enable them to support development of discipline-specific spoken academic language for all students (Bond 2020).

• Training and research to develop language tutors’ understandings of disciplinary literacies (Airey 2011). As Airey (2020) points out, much of the research to date in EMI stems from a linguistic perspective, it would therefore be timely to focus on research that is disciplinary-driven.

• ‘Embedding’ disciplinary experts into language classes or language experts into content classes to ensure English language support is informed by disciplinary practices and vice versa and better transfer from pre-sessional to disciplinary studies.

Whilst these policy points work primarily on the level of language management; they have the potential for influencing language beliefs and streamline language practices across a programme or module.

We argue that the key policy points outlined above apply equally to Anglophone and non-Anglophone countries as another observation of the study is that, with very few exceptions, the issues and practices discussed by tutors in non-Anglophone countries are very similar, except perhaps the attitudes to L1 in the classroom. Thus, the findings of this paper have reinforced our beliefs and claims – stated in the introduction – that higher education in Anglophone countries can also be subsumed under the EMI terminology. And whilst there are some epistemological differences between the ESL literature which focuses on Anglophone contexts and is rooted in speaker norms, and the EMI/English as a Lingua Franca (ELF) literature which is centred on non-Anglophone settings and uses communicative competence as a norm, we also argue that these two bodies of literature have the potential to cross-reference more than they currently do.

Disclosure statement

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

Additional information

Funding

This work was supported by British Council.

Appendix A:

Interview questions

1. What department do you teach in?

2. In what specific situations do students have to use academic speaking skills? How important are these to be successful in your modules(s)?

3. What oral assessments do you have in your module? Are there any opportunities for formative feedback?

4. Can you describe a speaking activity you organised recently in which students were successful? (prompt why)

5. Can you describe a speaking activity you organised recently in which students were unsuccessful? (prompt why)

6. What is the most crucial skill for students to be successful in academic activities which involve speaking?

7. FOR UK sites: Have you observed some international students being better prepared than others? If so, what are the factors that define success.

8. FOR UK sites: How well prepared do you think home students / native speakers are for academic speaking?

9. FOR non-Anglophone sites: How well prepared do you think students are for academic speaking?

10. How can students be supported to further develop their academic speaking skills? (Prompt pre-sessional, in-sessional, other university support services)

11. Is there anything else you would like to say about students’ readiness for academic speaking during your module this semester?