In virtually every research paper on this topic we come across we read that, in the particular context that the writer is operating, English as a medium of instruction (EMI) is on the increase. But what exactly is EMI? If we consider every classroom around the world in which learners are exposed to English language as their second language (L2) we are faced with a huge variety: English as a foreign language (EFL); Immersion, English for academic purposes; English for specific purposes, English for examination purposes, Content and language integrated learning (CLIL); content-based teaching; content-based language teaching, and so on. And then we have EMI. Here at the EMI Oxford Centre we define EMI as:
The use of the English language to teach academic subjects (other than English itself) in countries or jurisdictions in which the majority of the population's first language is not English.
This definition was first used in Dearden (2015).Footnote 1
If we accept this definition, at least for the time being, then we need to ask ourselves where on a continuum of objectives, from content-focused to language-focused, would we place EMI? In other words, when a subject teacher (for example a biology teacher) or lecturer enters a classroom, what are the objectives in his/her head? Are they entirely to teach the content or do they also intend to improve the students’ level of proficiency as well – at least to some extent? If EFL is almost completely at the language-end of the ‘objectives’ continuum, is then EMI entirely at the content-end? Presumably we would place CLIL somewhere in the middle of the continuum because the words ‘language and content’ are explicit in its very title. In theory that would be the case although some researchers would dispute that some European CLIL classrooms are integrating both language and content, and would therefore place it more towards the content-end. Moreover, is it the case that at some point, an EMI teacher might want to move gradually and tentatively more towards the middle of the continuum?
The EMI Oxford Centre was established nearly two years ago to act as an observatory of the EMI phenomenon around the world. We are ‘an observatory’ because, by our definition, the UK is not an EMI country. As we are not situated in an EMI country, we propose to take a neutral but nevertheless critical look at the phenomenon as it develops. We try to do this globally, by looking at the big picture, and we try to do this locally, by engaging with individual countries, often through practitioner workshops which sometimes then lead on to research projects.
There is no doubt in my mind that EMI is a controversial issue and these controversies were aired during the course of this symposium. I want to start off with a few ‘global insights’ to set the scene. Our centre carried out a global survey of 54 countries between October 2013 and March 2014 in order to map the size, shape and the future trends of the growing phenomenon of EMI worldwide (Dearden 2015). We have also collected data from an online survey of EMI teachers in secondary and higher education (HE). The key findings from both datasets were that:
1. Apart from a few local exceptions, EMI is expanding rapidly especially in the HE sector as a result of a perceived need to internationalise institutions.
2. The private HE sector is putting pressure on the state HE sector to convert their programmes to EMI as the latter need to compete for students; in turn, the tertiary sector is putting pressure on the secondary sector to also teach subjects through the medium of English.
3. In many countries the introduction of EMI is not supported by pre-service teacher training or teacher professional development.
4. Not all teachers are equipped linguistically to be able to effectively deliver subject content to students.
5. Where there are concerns about the introduction of EMI, these are to do with its potentially divisive nature: the likelihood that it will either create social elites based on English language proficiency or contribute to already existing social inequalities. Other concerns relate to the deleterious impact that EMI may have on the home language.
These five points then outline just some of the themes that were discussed. Here at Oxford we have described the EMI phenomenon as being, in all probability, an unstoppable train which has already left the station. It is the job of researchers to, at the very least, identify the issues surrounding its precarious journey and to communicate these to all the stakeholders. What follows are adaptations of three of the plenary speeches presented to the symposium.
1. The growth of EMI in European HE
A key ambition in the formation of the European Higher Education Area and a European Research Area has been to make Europe a more competitive player in the global knowledge economy (Amaral et al. Reference Amaral, Neave, Musselin and Maassen2009). The underlying political rationale is neoliberalism, defined as ‘a resuscitation of nineteenth century laissez-faire (hence, neoliberal) capitalism based on Adam Smith's competitive equilibrium model in which the unregulated (hence, free) market is assumed to work for the benefit of all if individual competition is given free rein’ (Piller & Cho Reference Piller and Cho2013: 24). Competition itself is premised on measurability and metrics (Styhre Reference Styhre2014). In other words, in order for competition to work, instruments must be in place that can measure how one entity compares vis-à-vis another.
The linguistic consequences of these policies, whether intentional or not, are arguably particularly noticeable in non-English-dominant contexts, where they have led to an increase in English language use in all three areas typically associated with universities: teaching, research and administration (Lillis & Curry Reference Lillis and Curry2010; Haberland, Lønsmann & Preisler Reference Haberland, Lønsmann and Preisler2013; Dearden Reference Dearden2014), effectively reconceptualising English as the ‘lingua academica’ (Vila & Bretxa Reference Vila and Bretxa2014). The growth of EMI is particularly well documented (Dearden Reference Dearden2014; Wächter & Maiworm Reference Wächter and Maiworm2014; Hultgren, Jensen & Dimova Reference Hultgren, Jensen and Dimova2015) with recent figures from Europe suggesting a 239% growth in B.A. and M.A. programmes that are fully taught in English over seven years: from 2,389 in 2007 to 8,089 in 2014 (Wächter & Maiworm Reference Wächter and Maiworm2014). Sociolinguistic practices, however, as explored through linguistic ethnographies, are far more multi- and translingual than these numbers suggest (contributors to Doiz, Lasagabaster & Sierra Reference Doiz, Lasagabaster and Sierra2012; Haberland et al. Reference Haberland, Lønsmann and Preisler2013; Hultgren, Gregersen & Thøgersen Reference Hultgren, Gregersen and Thøgersen2014).
Topics explored as a consequence of these sociolinguistic changes have included the nature and quality of teaching and learning (Hellekjær Reference Hellekjær2010; Björkman Reference Björkman2011; Airey Reference Airey, Dimova, Hultgren and Jensen2015), issues of equity, domain loss, diglossia, language shift and attrition (Phillipson Reference Phillipson2006; Vila & Bretxa Reference Vila and Bretxa2014; Salö Reference Salö2015), language policies and their purposes and effectiveness (Hultgren Reference Hultgren2014b; Airey et al. Reference Airey, Lauridsen, Räsänen, Salö and Schwach2015; Saarinen & Taalas Reference Saarinen and Taalas2016), the varieties and ownership of English (Kuteeva Reference Kuteeva2014; McCambridge & Saarinen Reference McCambridge and Saarinen2015; Karakaş Reference Karakaş2016) and attitudes to the rise of EMI (Hultgren Reference Hultgren, Linn, Bermel and Ferguson2015; Lasagabaster Reference Lasagabaster2015; Soler-Carbonell Reference Soler-Carbonell and Karaoglu2015). But what, more specifically, are the reasons for this drive towards EMI? Why does it happen? And to what extent is it the consequence of rational decisions? These are the questions I consider in this plenary speech.
2. Intentional and non-intentional EMI
When trying to uncover the motivators for introducing EMI, it is often assumed that there is an identifiable agent responsible and that decisions to implement EMI are rational and intentional. Presumably, this is partly an inevitable side-effect of the research methodology, which tends to favour questionnaires and interviews equipped to capture mainly rational intent. Thus one of the most comprehensive European studies to date which has surveyed the motivations of institutions in 28 EU member states that have introduced EMI has identified the move as being a deliberate strategy to facilitate recruitment of international students and improve the international competences of domestic students (Wächter & Maiworm Reference Wächter and Maiworm2014).
However, the current paper argues that it is crucial to recognise that EMI is not always, perhaps not even principally, adopted as a result of deliberate strategies. Rather, it appears to happen as an unintended consequence of other policy decisions, many of which take place at supra-national level by entities such as the OECD and the EU and then trickle down to national and institutional levels (Piller & Cho Reference Piller and Cho2013; Hultgren Reference Hultgren2014b; Ljosland Reference Ljosland2014; Saarinen Reference Saarinen2014). The supra-national policies, centred on neoliberal buzzwords such as knowledge based economy, human capital, competition, innovation and internationalisation (Holborrow Reference Holborrow2013), are contingent on instruments of measurement. These, in turn, result in the prominence of performance indicators such as university rankings, research excellence frameworks and bibliometry. Because such policies are not explicitly concerned with language or with English, they have been referred to as a form of language policy that is ‘covert’ (Piller & Cho Reference Piller and Cho2013), not ‘explicitly’ linguistic (Ljosland Reference Ljosland2014) and in which language is ‘invisible’ (Saarinen Reference Saarinen2014).
One key example of this is the relationship between university rankings and EMI (see Figure 1). The ranking list used is the Academic Ranking of World Universities (ARWU), also known as the Shanghai list. Of the three most influential ranking lists ARWU is the only one which does not have ratio of international staff and students among its indicators, which would arguably confound the results. As can be seen, despite the outlier (Switzerland) and 19 out of 28 countries not having any university listed on ARWU, the relationship comes out as a statistically significant positive correlation (Spearman coefficient = .551; P = .002, N = 28). In other words, the higher the aggregated ranking of a country's universities, the greater the use of EMI. While it should be recognised that this is a correlation and not a causal relationship, it may be analysed as EMI being intrinsically embedded in neoliberal ideologies and time-characteristic battles for excellence (for a more comprehensive and in-depth analysis of this relationship, see Hultgren Reference Hultgren2014a). Hence, EMI, and more generally, language shift cannot be understood in isolation but must be considered as co-constitutive of the political, economic and social sphere.
Correlation between EMI and university rankings
Conclusion: Raising awareness among policy makers
Raising awareness of how university rankings, metrics and neoliberal policies drive EMI and the global expansion of English more generally could lead to better and more informed policies in the education research domain. The aim would be to empower policy makers to make conscious choices about the implementation of EMI and about which support mechanisms should be put into place. It is often the case that EMI is just ‘dumped’ on the faculty and students with little consultation, preparation and compensation (Hamid & Nguyen Reference Hamid and Nguyen2016). It is also worth considering whether the response to EMI should be one of ‘pain relief’ or ‘cure’. Pain relief may involve remedial solutions such as English language support or revising course syllabi to incorporate ‘disciplinary specific language-learning outcomes’ (Airey et al. Reference Airey, Lauridsen, Räsänen, Salö and Schwach2015). Cure, on the other hand, would entail addressing big questions such as ‘What role should rankings and other performance indicators play in our education systems and beyond? How can change be brought about in a system governed by global policies?’
The increase in the number of EMI courses and programmes offered by universities is well documented. This increase is not only seen in Europe (Wächter & Maiworm Reference Wachter and Maiworm2008) but is also evident world-wide, including in non-Anglophone countries (Haberland, Lonsmann & Preisler Reference Haberland, Lonsmann and Preisler2013; Dearden Reference Dearden2014; Kirkpatrick Reference Kirkpatrick2014). Many of these EMI programmes have been established without any real planning or thought for the potential implications. It is therefore important, as noted by Macaro (Reference Macaro2015), to observe these developments with a critical eye and attempt to exert an influence on these developments as far as possible. Questions that researchers need to ask include:
• To what extent does a coherent national and/or institutional EMI policy exist? How aware are staff of these policies? What are the attitudes of staff towards the move to increased EMI?
• Does the move to EMI affect students and staff for whom English is an additional language; and, if so, in what ways? What have been the major challenges for students and staff in the implementation of EMI?
• Do staff receive any training/professional development to prepare them for teaching in EMI and, if so, what is the nature of this professional development?
• What is the attitude of students towards the move to increased use of EMI?
• Does the move to EMI affect the roles accorded to, and the value ascribed to, languages other than English as (a) languages of education and pedagogy and (b) as languages of scholarship and knowledge creation; and, if so, in what ways?
• What variety of English does the ‘E’ in EMI refer to? Is it a native-speaking variety of English, such as British or American, or does the university recognise the use of English as a lingua franca?
From a new study of the increase of EMI programmes across Asia (Fenton-Smith, Walkinshaw & Humphries in press) it is clear that few HE institutions and universities have consulted staff about the issue, nor have they developed coherent language policies. In a case study of one university in Malaysia, Ali (Reference Ali2013) looked at the development of EMI policy at three levels, namely macro (national policy), meso (university documents) and micro (actual stakeholders). Ali found that none of the documents contained any specific reference to the EMI policy, and any reference of classroom language use in the university documents was contradictory. She also found that stakeholders had not been consulted about the implementation of EMI. The situation that Ali describes for one Malaysian university is not atypical.
In Hong Kong, six of the eight government-funded universities are EMI institutions, despite the fact that over 95% of the population has a form of Chinese as their first language. It is important to note that the insistence on EMI at the university level has a washback effect on the school system as parents, not unreasonably, want to have their children taught in English in order to maximise their chances of getting to university. As a result of this parental demand for English in schools, the government has recently allowed schools to teach more subjects in English with a corresponding decrease in subjects taught in Chinese (Kan et al. Reference Kan, Lai, Kirkpatrick and Law2011). Even the Chinese University of Hong Kong (CUHK), which was initially set up to provide a Chinese medium education, has recently increased the number of EMI programmes it offers. A legal challenge brought by students claiming the increase in EMI programmes breached the university charter was defeated in the courts, which ruled that the university had the right to determine the medium of instruction (Li Reference Li, Doiz, Lasagabster and Sierra2013).
An exception is the Hong Kong Institute of Education, where a language education policy has been developed over some years in consultation with stakeholders (Xu 2014) whose major aim is to ensure graduates from the institute are functionally trilingual (in Cantonese, Putonghua and English) and biliterate (in Chinese and English). As such it is the only tertiary institution in Hong Kong which has a policy suited to the government's desire to ensure all its citizens are trilingual and biliterate. The other government universities, by adopting an EMI-only policy (or, in the case of CUHK, increasing the number of EMI programmes on offer) have created a washback effect, which sees more and more classes in schools being taught in English at the expense of Chinese.
In Myanmar, the government policy is that all institutes of HE are to use English as a medium of instruction. This is despite the fact that levels of English proficiency among both staff and students are not adequate enough to allow the successful teaching of content through English. A recent survey of staff across a number of universities in Myanmar (Thant Reference Thant2016) showed that they overwhelmingly favoured a system of bilingual education, reporting that it was impossible for them to teach and for students to learn solely in English and that the use of Burmese was therefore essential. A further study (Khaing Phyu Htut 2016) showed that a great number of English language teachers in Myanmar scored A0 (the lowest possible level on the Common European Framework of Reference for Languages (CEFR) bands) when being tested on their English speaking skills, with the majority scoring only A2. The teachers themselves reported that they could not use EMI, as their own levels of English were so poor. Drinan has summarised the situation with regard to EMI in Myanmar. The three quotes below are excerpts from her report (Reference Drinan2013):
Using English as a Medium of Instruction (MoI): this is fundamentally not working for teaching Maths and Science as few teachers can use English, let alone, teach another subject in English. Students are not learning or understanding important concepts in Maths and Science. They merely remember the technical terms in English for the tests. Most teachers use a mix of Myanmar (for explanation) and English (for technical terms).
The impact (of a weak English Language curriculum) on language proficiency for both teachers and students is very serious. Teachers’ language is fossilised, meaning they seldom develop further than the grade they are teaching. More worrying is that teachers are not even at that level, as many have just memorised the textbooks they are ‘teaching’.
Likewise students’ proficiency levels remain very low – possibly reaching pre-intermediate by Grade 11 but most leaving school with elementary English.
These excerpts show just how serious a challenge it will be to successfully implement EMI in Myanmar. Indeed, the examples of the implementation of EMI given above indicate that there are serious problems with its implementation in many contexts. It is important then that government and institutions take stock before they rush to implement EMI programmes. At the very least, the implementation of EMI at the university level (and indeed at all levels) must be conducted within a framework of multilingualism. By this I mean that:
1. The national language must be seen as a complementary language of education and scholarship. English must not be considered the sole language of education. If English is adopted as the sole language of HE, there will be an inevitable washback effect to the school system where the local and/or national languages will be further undermined as language(s) of education in face of the demand for EMI. The school system will typically see an earlier and earlier introduction of English, often at the expense of local languages (Kirkpatrick Reference Kirkpatrick, Leitner, Hashim and Wolf2016).
2. EMI should not mean English only. On the contrary, the use of the linguistic resources of staff and students needs to be encouraged. We know that wherever an English-only policy is enforced, teachers and students actually use code-mixing on a regular basis (Barnard & McLellan Reference Barnard and McLellan2014) and, as a result, many feel guilty for transgressing policy. Far from feeling so, they should be encouraged to code-mix and use the languages of the classroom in order to convey the content successfully. The benefits of being multilingual need to be exploited.
3. Reading materials and resources must not be limited to those only written or available in English. Students and staff need to be encouraged to use any sources available in the languages they understand.
4. The ‘E’ in EMI is often understood to refer to a native speaker variety of English. This insistence on a native speaker variety as the benchmark needs to be reconsidered, and the use of English as a lingua franca should be supported. After all, a major role of English in today's world is as a lingua franca. The majority of today's speakers of English are multilinguals for whom English is an additional language rather than native speakers. This actual practice should inform language policy.
In conclusion, I would argue that EMI cannot be successfully implemented unless there is a coherent language policy for which all stakeholders have been consulted. This language policy should be constructed within a multilingual framework as outlined above. As language learning is a developmental process, ongoing English language proficiency courses need to be provided as an integral part of any EMI programme. In this context, it is worth remembering that the first languages of Asians will be of a different language family from English, whether this might be a Sinitic, Tibeto-Burman, Mon-Khmer or Austronesian language family. At the same time, teachers and lecturers need to be provided with professional development programmes designed to help them teach content subjects through the medium of English.
Only if coherent language education policies that meet these criteria are met will the implementation of EMI programmes have a chance of lasting success.
EMI programmes are proliferating at great speed in European universities, although their spread varies considerably from country to country. In this presentation I am going to focus on Spain, one of the southern European countries in which EMI is more widespread at pre-university than at university level. In the introduction I will start by briefly making reference to the European context before concentrating on Spain. After the introduction, the following three sections will revolve around the three components that, according to Spolsky (Reference Spolsky2004), make up language policy: language planning, language practices and language beliefs. In these three sections I will pay particular attention to how EMI is affecting the three university bodies: faculty members, students, and administration personnel. In the fifth section I will refer to the research studies undertaken so far with a view to evaluating the impact of EMI on English proficiency, content learning, and L1 academic literacy.
1. Introduction
First, I am going to briefly refer to the study carried out by Wächter & Maiworm (Reference Wächter and Maiworm2014), who analysed the presence of English-taught programmes (ETPs) in Europe, because this will allow me to compare EMI in Spain with the situation in the rest of Europe. In this survey carried out in 28 European countries HE institutions holding an ‘ERASMUS Charter’ were included. Wächter & Maiworm only considered ETPs that were taught fully in English, in other words, this study excluded ‘mixed’ programmes taught both in the local language and English. The authors observed that the number of EMI programmnes had increased by 239% between 2007 and 2014. However, the north-south divide found in their two previous studies still remained.
The study showed Spain to be one of the few countries where non-native-speaker foreign students’ proficiency in English is rated higher than that of domestic students, the others being France, Italy and Turkey. More interestingly, the Spanish respondents were the most critical of their teaching staff's English; nevertheless, 88% rated it positively. In addition, just 28% of the Spanish administrative personnel could operate in English, lower than in any of the participating countries. Similarly, Spain had the lowest proportion of students enrolled in ETPs: only 0.3%. Let's briefly consider the main characteristics of the Spanish university system.
According to the Spanish Ministry of Education (2015), there were 83 universities in Spain in the 2014/2015 academic year, 50 of which were public and the remaining 33 private (but also State funded) and serving a total of 45.6 million inhabitants. The number of Bachelor's and Master's students in the previous academic year was just over 1.5 million. It is worth noting that out of 115.071 faculty members, 96.7% were Spanish and, therefore, it can be surmised that the vast majority of EMI programmes are taught by non-native speakers of English. 4.9% of students were from abroad although this figure was much higher among Master's students: 18%. As far as the Eramus exchange programme is concerned, Spain was the leading country both in the number of outgoing and the number of incoming students.
The following sections are related to the structure of Spolky's (Reference Spolsky2004) tripartite language policy framework.
2. Language planning
Language planning, also known as language intervention or language management, is to do with the formulation of an explicit plan or policy about language use. As Dearden (Reference Dearden2015) points out, the provision of EMI programmes is regarded as a vote winner by Spanish public administrations. In the case of HE institutions, EMI has become a key component of internationalisation and an asset in the labour market, especially at a time when, due to the economic recession, many Spanish graduates are forced to travel abroad in search of work.
However, language policies vary greatly. In some universities the implementation of EMI programmes is the result of bottom-up initiatives (Dafouz, Camacho & Urquia Reference Dafouz, Camacho and Urquia2014); in others it can be top-down (Doiz, Lasagabaster & Sierra Reference Doiz, Lasagabaster and Sierra2013a). An important factor to bear in mind is size; it is much more complex to implement a language policy in an institution such as the University Complutense of Madrid (with over 75,000 students and around 6,000 teachers) than in a private university with just a few thousand students. Consequently, there is great variation in the chosen course of action: some universities have decided to teach 5% of the credits in a foreign language; others deliver whole degrees in English; some others have plans to deliver only some courses depending on qualified teachers’ availability; and other universities are still deciding on how to proceed.
Many faculty members are reluctant to teach in English, and many language policies include no clear language goals (Doiz et al. Reference Doiz, Lasagabaster and Sierra2011; Aguilar & Muñoz Reference Aguilar and Muñoz2014; Halbach & Lázaro Reference Halbach and Lázaro2015). Therefore, it can be concluded that EMI is the right label to use at tertiary level in Spain because there is no dual-focus on language and content, in sharp contrast with pre-university levels where the label CLIL is widely used. In the majority of Spanish HE institutions EMI is optional, in the belief that the best policy is to make headway by implementing EMI gradually rather than imposing it unilaterally. This is driven by the fact that Spanish undergraduates’ command of English is rather limited: whereas 88.3% of international students consider their English to be good or very good, this plummets to just 56.2% for local students (Doiz et al. Reference Doiz, Lasagabaster, Sierra, Doiz, Lasagabaster and Sierra2013b). Consequently, local students regard their command of English as being insufficient for EMI courses, while international students feel more confident.
Halbach & Lázaro (Reference Halbach and Lázaro2015) conducted an interesting survey involving 50 Spanish universities (ten of them private) and compared the results with a study performed five years earlier. Although they observed an increase in the number of EMI courses offered (this is in accordance with the data on Europe from Wächter & Maiworm), the majority of institutions did not implement any quality assurance measure. In fact, only one out of four imposed a C1 level of English as a requisite for the teachers involved in EMI and very few universities provided methodological courses. Similarly, only 31 universities required their students to have reached either B1 or B2 levels of the CEFR. It can therefore be concluded that language planning with regard to EMI still has a long way to go.
3. Language practices
The language practices used are directly related to the language a person decides to use in a given communicative situation. In the university context, language practices refer to the actual use of different languages in different academic situations; in other words, what languages people actually embrace. As mentioned above, research studies demonstrate that EMI teachers in Spain are concerned about content but not language (Dafouz Reference Dafouz, de Zarobe and Catalán2011; Aguilar Reference Aguilar2015) as in their opinion the shift from first language (L1) teaching to EMI simply represents a change in the language of instruction and, in most cases, no adaptation of the teaching methodology is considered (Halbach & Lázaro Reference Halbach and Lázaro2015). The analysis of EMI classes has brought to light that here teachers deploy a more limited stylistic variety and make poorer use of metalinguistic devices, while their language is less subtle in English (Dafouz & Núñez Reference Dafouz, Núñez, Dalton-Puffer, Nikula and Smit2010).
Official documents usually make no reference to the use of students’ and teachers’ L1s in EMI. In addition, little is known about the actual coexistence of students’ L1 and English in EMI classes. In a study undertaken at the University of the Basque Country (an officially bilingual institution using Basque and Spanish), Doiz & Lasagabaster (Reference Doiz, Lasagabaster, Mazak and Carroll2016) observed that in line with the institutional policy, the use of the students’ L1 is allowed on EMI courses, but this was not the preferred option by any of the 13 EMI teachers participating in the study, who tended to stick to English in their classes. However, the authors pointed out that the divide inside/outside class seemed to play a significant role in the participants’ translanguaging practices. Thus, whereas L1 use was restricted in class, and both teachers and students resorted to it only when an explanation or a translation of lexis or whole expressions was required, the use of the L1 was both allowed and indeed habitual outside the classroom context, for example in tutorials, meaning that English did not represent an additional hurdle in teacher/student communication.
It is striking that, despite the fact that all the EMI teachers involved in the study were multilingual speakers (Basque, Spanish and English, and sometimes even a second foreign language) and that ‘studies promote code-switching as a valid and natural learning strategy’ (Costa Reference Costa2012: 35), the participants were attached to a monolingual view of language codes. Doiz & Lasagabaster (Reference Doiz, Lasagabaster, Mazak and Carroll2016) attribute this language compartmentalisation to two main reasons: first, the influence of Canadian immersion programmes on their equivalent Basque programmes, which are the cornerstones of the revitalisation of Basque; and second, to the widespread belief that more exposure to and use of the L2 leads to increased language competence. However, the authors acknowledge that observational studies are needed as triangulation would help to draw more definitive conclusions, given that this study relied on teachers’ self-reported practices which may differ from what is actually going on in EMI classes. In conclusion, language practices in EMI settings deserve further attention.
4. Language beliefs
Language beliefs (also known as language ideology) encompass the beliefs about language, language use and the speakers’ language attitudes. In EMI contexts language beliefs are the way in which English is talked and written about and how it affects language choice, since language choice is influenced by ideological assumptions. In Spain this is by far the most extensively researched component in Spolsky's tripartite language policy theoretical model (Dafouz Reference Dafouz, de Zarobe and Catalán2011; Doiz et al. Reference Doiz, Lasagabaster and Sierra2011; Cots Reference Cots, Doiz, Lasagabaster and Sierra2013; Doiz et al. Reference Doiz, Lasagabaster and Sierra2013a; Fortanet Reference Fortanet2013; Doiz, Lasagabaster & Sierra Reference Doiz, Lasagabaster and Sierra2014; Aguilar Reference Aguilar2015) and two main groups can be distinguished: those dealing with teachers’ beliefs and those tackling students’ beliefs.
In the case of teachers’ beliefs, the general attitude towards EMI is positive. However, there is widespread reluctance to teach language, as EMI teachers consider that they are responsible for content learning but not for language learning. In a recent study Aguilar (Reference Aguilar2015) asked EMI teachers from three universities whether they thought that their focus was EMI (as opposed to CLIL): using a five-point scale the mean was 4.68, 5 being ‘I totally agree’. The participants also pointed out that they had no specific language goals and that they did not assess English in their subjects (mean = 4.17). Studies therefore show that lecturers have no time to reflect on their teaching in English, and language learning plays second or no fiddle at all.
The three main reasons put forward in order to explain teachers’ unwillingness to shift from EMI to CLIL are: (a) their limited proficiency in English; (b) lack of time and the little support drawn from their institution; and (c) involvement in EMI earns them little or no recognition. When asked about the main pros and cons of EMI, they mentioned three positives: (a) teaching in the foreign language does not have a negative effect on subject content learning; (b) it is positive for the institution and all the stakeholders, as the internationalisation process is fostered by EMI courses; and (c) EMI students are usually more motivated (it is optional in the vast majority of cases) and classes are smaller, which makes teaching easier and more rewarding. On the negative side, teachers identified three main issues: (a) they feel more constrained when teaching in English as they have access to fewer linguistic resources than they do when teaching in their L1; (b) adapting materials is a time consuming process; and (c) EMI requires more effort and is a commitment that affects their personal life (reading in English, watching TV in English, etc.). As one of the participants in Doiz & Lasagabaster (Reference Doiz, Lasagabaster, Mazak and Carroll2016) bluntly puts it: ‘The personal cost is extremely high in comparison with the person who is next to you [and teaching in his/her L1] and is being paid the same salary and who presumably does the same chores in the department, right?’
Spanish students’ attitude towards EMI is not as favourable as that of their teachers. Nevertheless, they also believe that EMI has no negative effect on content learning and that it helps to improve their English, particularly their speaking and listening skills. They also emphasise that their English competence is not as high as that of non-native speaker foreign students. Students express concern regarding the negative impact that EMI may have on the presence and use of the minority language in bilingual contexts such as the Basque Country, Catalonia or the Valencian Community (Basque or Catalan/Valencian) at tertiary level (Cots Reference Cots, Doiz, Lasagabaster and Sierra2013; Doiz, Lasagabaster & Sierra Reference Doiz, Lasagabaster, Sierra, Doiz, Lasagabaster and Sierra2013b; Fortanet Reference Fortanet2013). Language strains are ubiquitous, but this is especially so in officially bilingual institutions.
Finally, it must be noted that the number of studies that include the administration personnel is negligible, despite the fact that their support is paramount and that they play a key role in the HE internationalisation process. One of the studies (Doiz, Lasagabaster & Sierra Reference Doiz, Lasagabaster and Sierra2014) that did gather their opinions revealed that they expressed positive attitudes towards EMI.
5. Assessment
In Spain the number of studies on assessment in EMI contexts is rather limited and most of those available so far are based on participants’ impressions. In this section I will focus on the impact of EMI on English competence, content learning and L1 academic literacy by referring to research papers based on actual measurements, and not on impressions.
Aguilar & Muñoz's (Reference Aguilar and Muñoz2014) research was one of the studies in which English language gains were actually measured. Students’ grammar and listening skills were controlled at the beginning of a 60-hour EMI course and the paired-sample t-tests showed significant gains only in their listening skills. The participants were divided into three proficiency levels and only the elementary level students exhibited statistically significant improvement, whereas this was not the case among the intermediate and advanced students. It is worth noting that the proficiency of some of the seven lecturers participating in this study was relatively limited: the Oxford Placement Test completed at the beginning of the term indicated that two were in the lower intermediate band, three in the upper intermediate, one in advanced and one in the very advanced band. Aguilar & Muñoz conclude that in these conditions (a short EMI experience with some teachers not being able to provide rich input (some students had a better command of English than their lecturers)), only those students below the intermediate level benefited from the experience.
As for content learning, Dafouz et al. (Reference Dafouz, Camacho and Urquia2014) compared Spanish university students’ actual academic performance in an EMI Business Administration degree with that of students studying for the same degree through Spanish (non-EMI group). Both groups were matched on their university entrance exam scores at the beginning of the term, and their results in three subjects (Economic History, Financial Accounting I and Principles of Business Financial Management) contrasted at the end of the term. The independent-sample t-tests revealed that both cohorts obtained similar results and, consequently, no statistically significant differences were found. These results allowed the authors to conclude that EMI did not negatively affect the students’ performance although they also included a caveat: ‘the results of this study need to be treated with great caution given the limited size of data, the focus on a single institution and our reduced access to classroom practices’ (Dafouz et al. Reference Dafouz, Camacho and Urquia2014: 233).
To the best of my knowledge, no study has yet examined the impact of EMI on students’ L1 academic literacy. The main reason for the lack of interest in this research area may lie in the fact that EMI has only been implemented on a small scale in Spain, which is why the stakeholders are not preoccupied with this issue, unlike in the Scandinavian countries, where it has become a serious concern.
6. Conclusions
Although the number of EMI programmes is clearly on the increase in Spain (Wächter & Maiworm Reference Wächter and Maiworm2014; Halbach & Lázaro Reference Halbach and Lázaro2015), its presence is still far from substantial. In contrast to small northern European languages, Spanish is a widely spoken language and foreign students already speak it when they participate in exchange programmes and/or are willing to learn it. However, education authorities are well aware that English is nowadays an indispensable asset and are amenable to fostering the spread of EMI, although the lack of sufficient English competence becomes the main stumbling block for further implementation. In this context, quality assurance measures are needed, especially concerning teachers’ and students’ English proficiency. The few studies available indicate that low exposure to EMI does not bring about significant linguistic benefits, although the content learning seems not to be negatively affected.
Spain is a multilingual country (more than 40% of the population live in officially bilingual regions) and linguistic tensions resulting from EMI experiences can be found in both monolingual and bilingual HE institutions, but especially in the latter. Thus, one of the main challenges of bilingual universities is to find a place and value for both local and global languages. There is a compelling need for Spanish universities to establish clear-cut language policies, but very few universities have tackled this issue so far. Probably due to the presence of a co-official language, bilingual universities tend to show more coherent language policies (Halbach & Lázaro Reference Halbach and Lázaro2015). The good news is that the results available so far reveal that the local languages (Spanish and the co-official) are not being upstaged by English, as they are strongly associated with identity and utility (Lasagabaster Reference Lasagabaster2015).
Many more research studies on EMI are needed so that robust conclusions can be reached and EMI programmes underpinned. CLIL is burgeoning at pre-university level and these experiences will undoubtedly have a bearing on the spread of EMI at tertiary level in the short run. Many different concerns are still pending research, but I would like to highlight the following:
(a) There is a dire need to investigate the impact of EMI on both language (L1 and English) development and content.
(b) Team-teaching between English and content lecturers needs to be boosted; EMI teachers refuse to teach English (Airey Reference Airey2012; Aguilar Reference Aguilar2015) and are not prepared to assess language learning, but both they and their students would undoubtedly benefit from language-oriented and pedagogical instruction. Tandem teaching experiences could contribute to improving the quality of EMI, but they are very scarce and, whenever they are put into practice, they become pilot experiences that after their implementation simply vanish into thin air because they have little or no support from the institution (Cots Reference Cots, Doiz, Lasagabaster and Sierra2013). The impact of language and content teachers’ collaboration thus needs to be assessed.
(c) Studies should analyse whether there is any relationship between students’ proficiency and content learning gains and teachers’ proficiency (Aguilar & Muñoz Reference Aguilar and Muñoz2014).
Acknowledgements
This plenary talk falls within the work carried out in the following research projects: FFI2012-34214 (Spanish Ministry of Economy and Competitiveness) and IT311-10 (Department of Education, University and Research of the Basque Government).
Ernesto Macaro is Professor of Applied Linguistics at the University of Oxford where he is Director of the Centre for Research and Development in English Medium Instruction (EMI Oxford). He teaches on the Master's in Applied Linguistics and on the Teacher Education Programme. His research focuses on L2 learning strategies and on the interaction between teachers and learners in L2 classrooms and, more recently, in classrooms where English is the medium of instruction. He has published widely through his books and in applied linguistics journals, and has carried out three systematic reviews. He remains committed to ensuring that his research has practical applications.
Anna Kristina Hultgren (D.Phil., Oxon; M.A. Copenhagen) is a lecturer in English Language and Applied Linguistics at The Open University. Her research focuses on the interconnection between language and the social, political and economic world, and particularly on how globalising processes contribute to making English the world's main international language. She conducts research on institutions in the globalised knowledge economy (call centres and universities), sites characterised by extensive transnational and translingual policies and practices. She has published and spoken widely on EMI in HE and on call centre communication, and is co-editor of the volumes English in Nordic universities: Ideologies and practices (Benjamins 2014) and English as a medium of instruction in European higher education (Mouton 2015). She collaborates with governmental and professional agencies to promote effective language policies and practices in the globalised age.
Andy Kirkpatrick is a Fellow of the Australian Academy of the Humanities and Professor in the Department of Humanities, Languages and Social Sciences at Griffith University, Brisbane, Australia. He has lived and worked in many countries in East and Southeast Asia, including China, Hong Kong, Malaysia, Myanmar and Singapore. His most recent books are English as an Asian language: Implications for language education, co-edited with Roly Sussex and published by Springer, and Chinese rhetoric and writing, co-authored with Xu Zhichang and published by Parlor Press. He is founding and chief editor of the journal and book series Multilingual Education, published by Springer, and is editor-in-chief of the Asia Journal of TEFL. He is Director of the Asian Corpus of English (ACE) project. ACE is freely accessible at http://corpus.ied.edu.hk/ace/.
David Lasagabaster is Professor of Applied Linguistics at the University of the Basque Country UPV/EHU in Spain. He has published on L2/third language acquisition, CLIL, EMI, attitudes and motivation, and multilingualism. Among other publications, he has co-edited Multilingualism in European bilingual contexts: Language use and attitudes (Multilingual Matters, 2007); English-medium instruction at universities: Global challenges (Multilingual Matters, 2013); Motivation and foreign language learning: From theory to practice (John Benjamins, 2014); and CLIL experiences in secondary and tertiary education: In search of good practices (Peter Lang, 2016). He was a member of the executive committee of the international Association for Language Awareness from 2008 to 2016.
