Lecturers’ Training Needs in EMI Programmes: Beyond Language Competence1

This paper reports the results of an analysis of English as a Medium Instruction (EMI) lecturers’ training needs at a Spanish state university. In order to carry out this analysis, participants were asked to complete a quantitative questionnaire and participate in a survey with open questions adapted from Pérez-Cañado 2020). Both analysis tools explored the lecturers’ needs in relation to specific areas: linguistic competence, methodology, materials and resources, continuing professional development training and their overall rating of the EMI programme. The results show that, on the one hand, EMI teachers feel that their linguistic competence is sufficient to conduct classes in English. On the other hand, participants seem less confident as far as their methodological competence is concerned and call for more specific ongoing training. These findings are mostly in consonance with the results of other studies carried out in similar contexts (Dafouz 2018; Macaro et al. 2019; Pérez-Cañado 2020; Coelho in preparation) but they also provide some new insights which should be taken into consideration in the design of specific training programmes for lecturers involved in EMI in Higher Education.


Introduction and Literature Review
In recent decades, there has been a rising concern about the internationalisation of European Higher Education Institutions (HEIs). As part of the process of adaptation to the European Higher Education Area (EHEA), internationalisation programmes have become widespread across the continent, resulting in two main lines of development: internationalisation abroad and internationalisation at home (Beelen and Jones 2015). The former refers to "all forms of education across borders" (Beelen and Jones 2015, 61), whereas the latter consists of a series of practices devoted to carrying out activities that help students develop international understanding and intercultural skills while at universities in their home country (Beelen and Jones 2015;Knight, 2005). Beelen and Jones (2015, 69) defined internationalisation at home as "the purposeful integration of international and intercultural dimensions into the formal and informal curriculum for all students within domestic learning environments". This form of internationalisation is being used more and more in HEIs, which, among other practices, are using English as a Medium of Instruction (EMI). European universities are adopting policies to enhance their international profile and to attract students from other nationalities and, as a consequence, they resort to English as the language of communication in the class (Macaro et al. 2019).
The EMI approach advocates the use of English to teach disciplinary subjects, a trend that has had an unprecedented, significant rise over the last 20 years, with the number of EMI programmes implemented at European universities growing by 239% within a period of 7 years (2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014) according to Wächter and Maiworm (2014). However, as these authors pointed out in their report, this increase has not been homogeneous across countries, clearly reflected in the far higher number of EMI programmes in Northern European countries (60.6% and 44.5% of Universities in Nordic and West Central European countries, respectively) than countries in South Western Europe (17%).
Interest in EMI has also been shown in an increasing number of publications devoted to this phenomenon in recent years (Aguilar and Rodríguez 2012;Dafouz 2018;Dafouz and Smit 2020;Escobar Urmeneta and Arnau Sabatés 2018;Fortanet 2013;Macaro 2018;Macaro et al. 2019;Morgado and Coelho 2013; Morell 2020; Pérez-Cañado 2018; Ruiz de Zarobe and Lyster 2018 or Wilkinson 2018, among others). As will be seen in more detail below, all these publications have something in common: they acknowledge the importance of the needs of EMI lecturers beyond their language competence, and it is these needs which is the main focus of this article.
In Higher Education contexts, the teaching of content subjects through a second language has very often been equated to the use of English as the vehicular language for instruction. Pecorari and Malmström (2018) define the main features of EMI programmes emphasising the following: 1) for most participants in this type of setting, English is a second language (L2), 2) English is the language used for instructional purposes, 3) English itself is usually not a subject being taught and 4) language development is not a primary intended outcome. Macaro et al. attribute the prominent role of English, rather than other second languages, at universities to "the status of English as a lingua franca in a more global world, heightened institutional competition, internationalization at home […] and faculty requirements to publish in highly ranked English-medium journals" (2019, 105).
However, it is important to bear in mind that similar educational experiences have also been implemented at other educational stages (predominantly secondary education but also at primary and even pre-primary, to some extent, in some countries, e.g., Spain) with a more multilingual conception, and also a more balanced focus on both content and language. Bilingual educational practices, under the umbrella of Content and Language Integrated Learning (CLIL), have grown exponentially across Europe in the past three decades (Eurydice report 2006; Pérez-Cañado 2012). In many contexts, instructed bilingual education originated earlier in these lower educational levels, and reached HEIs at a later stage. In primary and secondary contexts, though, the movement mostly grew from a great concern for the development of students' foreign language abilities. Although content is vital-CLIL is defined as a dual-focused approach (Coyle et al. 2010;Mehisto et al. 2008)-language is always a fundamental learning outcome, too. In contrast, in Higher Education settings, content seems to have maintained the predominant role.
However, in recent years, there has been a tendency to extrapolate some of the main postulates of CLIL to Higher Education. The term ICLHE (Integrating Content and Language in Higher Education) has been specifically coined for Higher Education (Wilkinson 2018), with its equivalent in Spanish being ICLES (Integración de Contenido y Lengua en la Educación Superior) (Escobar Urmeneta and Arnau Sabatés 2018). Morgado and Coelho (2013) summarise the main similarities and differences between EMI and CLIL at this educational level concluding that, although in EMI contexts, content lecturers may use strategies such as simplifying, classifying or translating to help their learners understand the disciplinary contents, content is always their main focus. Within a CLIL perspective, language is a learning objective in itself, and lecturers are sensitive to synergies between languages (both the L1 and L2) in the construction of new meanings. Within an EMI approach, the L2 is mostly viewed as a communication tool that the subject discipline lecturer neither plans for nor assesses and, although there may be some incidental L2 learning involved, lecturers do not usually make direct reference to it in any way.
Recently, Dafouz andSmit (2016, 2020) have coined the term EME (English-Medium Education) and EMEMUS (English-Medium Education in Multilingual University Settings). They argue that the latter concept is: 1) conceptually wider than previous accounts and more inclusive, comprising various research and pedagogical approaches, 2) it pays the same level of attention to both 'instruction' and 'learning', 3) it establishes a broader sociolinguistic multilingual context which includes languages other than English and 4) it is restricted to Higher Education levels with its specific features.
Some authors have also argued for the 'clil-ization' of EMI programmes at universities both from a theoretical perspective (Alejo-González 2018; Pérez-Cañado 2020) and within a more applied, practical approach (Morgado et al. 2015;Morgado et al. 2020).
Any process is, primarily, promoted by its main stakeholders and, in educational contexts, teachers always have a key role. In the case of EMI lecturers, the research literature has identified two main areas of concern in relation to their needs: their L2 proficiency level (Macaro et al. 2018;O'Dowd 2018) and their methodological training, comprising aspects such as classroom management in an L2, materials development, assessment and even awareness of the theoretical underpinnings of bilingual education (Dafouz 2018;Pérez-Cañado 2018).
In his survey of teacher training for EMI, O' Dowd (2018) notes that the current training for lecturers in European universities is largely focused on language development, and that methodological training, and especially 'bilingual education methodology', is relatively scarce. Dafouz (2018) goes a step forward in the analysis of EMI lecturers' needs and discusses the importance of refocusing their perceptions on identity and ideology. She argues that the process of teaching and learning an L2 is a complex social practise not only restricted to exchanging information but also related to who the participants are and how they relate to the world. In this sense, it is important to value lecturers' (and also students') linguistic background (both English and their L1) and encourage them to use both codes to provide students with disciplinary literacy in both languages so that they become competent professionals in both local and international contexts.
In our view, in order to aid EMI lecturers in the complex task of teaching their subjects in English, the first step would be to analyse their self-perceived needs and that is the main aim of the study reported in this paper. Pérez-Cañado's (2018) set up a framework with the core competences for a CLIL teacher: linguistic competence, pedagogical competence, scientific knowledge, organisational competence, interpersonal and collaborative competence and reflective and developmental competence that she later put into practice for analysis in an EMI context (Pérez-Cañado 2020). Following this framework, we have explored the training needs of EMI lecturers at the University of Extremadura (UEx) in Spain, as part of the regional research project, ICLUEx (Diseño de entornos virtuales de aprendizaje colaborativo para la enseñanza integrada de contenidos y lenguas extranjeras (AICLE) en la Educación Superior adaptados a la Universidad de Extremadura -Developing virtual collaborative learning environments for content and language integrated learning (CLIL) in Higher Education adapted to the University of Extremadura (IB18055), European Regional Development Fund -ERDF). This project aims to contribute to the quality of teaching content subjects through English at the UEx by developing open educational resources in virtual collaborative learning environments, conducting a needs analysis of both the lecturers and students involved in the process as a first step. Apart from being involved in ICLUEx, the two authors of this article are also currently working on an EU-funded project (INCOLLAB) 2 that has also informed many of the reflections included here.
This paper reports our first results in relation to lecturers' needs. In section 2, the specific context of the university of Extremadura in relation to its institutional policies for the implementation of content instruction through English will be presented. Sections 3, 4 and 5 are devoted to the actual study and its results. In order to carry out this analysis, participants were asked to respond to a quantitative questionnaire and to a qualitative, semi-structured survey with open questions, both adapted from Pérez-Cañado (2020). Finally, some conclusions derived from the analysis of the results of the study will be drawn in the final section.

Teaching Through English (EMI/CLIL) at the UEx
As has been pointed out in the previous section, in the last two decades, European integration, and more generally globalisation, has brought important changes to higher education in Europe, with a radical increase in the number of both undergraduate and post-graduate degree programmes taught in English in different European countries (Wächter andMaiworm 2008, 2014). This concern for the internationalisation of HEIs in Europe is a direct consequence of the Bologna process that, over the last two decades, has led to the promotion of international cooperation among HEIs. Important reforms have been undertaken to adapt European higher education systems in order to make them more compatible between countries and to strengthening their quality assurance mechanisms with the aim of increasing staff and student mobility and facilitating employability.
In the region of Extremadura (western Spain), the teaching of content subjects through an L2 (predominantly English, but also French and Portuguese) first started in primary schools, beginning in the academic year 1996/1997 through a programme developed together by the Ministry of Education and Culture (MEC) and the British Council in two state primary schools in the two main towns in the region, Badajoz and Cáceres. CLIL reached secondary schools, first, through pilot programmes that were officially regulated by the educational authorities in the academic year 2004-2005 (Alejo and Piquer-Píriz 2010).
The implementation of EMI experiences at university level started a decade later and has followed several stages. It began with some pilot initiatives promoted by lectures (a 'bottom-up' approach), which were followed by the approval of the official regulations by the university in June 2015 (a 'top-down' approach). These official regulations for 'degree programmes partly taught in other modern languages' established four possible structures: • Option 1: Degrees taught completely in an L2.
• Option 2: Degrees taught in, at least, two parallel groups, one in Spanish and one in the L2. • Option 3: 'Bilingual' degrees (more than 50% of the European Credit Transfer and Accumulation System [ECTS] credits will be taught in English). Two or more groups are also established: one group follows the 'bilingual' strand, and the other/s are offered the modules in Spanish. • Option 4: Degrees in which less than 50% of the ECTs credits will be taught in English. Only some of the subjects are taught in the L2 and, again, there are always two groups to ensure that those students who prefer to follow the subject in Spanish can do so.
In practice, only option 4 has been implemented so far, although option 3 exists in the case of the degree in Primary Education at the Teacher Training College, and only English has been used as the vehicular language.
In the academic year 2015-2016, an official pilot project called 'UEx Bilingual Pilot Project -Learn in English' began in 5 university centres: the faculties of Science, Economics and Veterinary Science, the school of Industrial Engineering and the Polytechnic school. In the same academic year, the degree in Primary Education at the Teacher Training College started a bilingual group, i.e., implemented option 3 (officially approved by the Spanish 'National Agency for Quality Assessment and Accreditation'). A year later (2016-2017), the school of Agriculture joined the programme in which, by the following year (2017-2018), 72 subjects on 25 degrees were being taught in English. During the academic year 2019-2020, a new programme, PALEx -'Programa de Acercamiento a Lenguas Extranjeras' ('Foreign language outreach programme') was designed. Within this new programme, two essential innovations were set up. On the one hand, the option to participate is now open to two foreign languages: English (PALEx-I) and Portuguese (PALEx-P). On the other hand, the offer is divided into three levels: • Basic level: subjects which, taught in Spanish, have at the students' disposal the necessary material to follow the classes, attend to officehour consultations (on demand) and the possibility of carrying out key tasks in English and/or Portuguese. • Intermediate level: subjects that, complying with the requirements of the basic level PALEx, also include some activities in English and/or Portuguese with the students. Some examples of these activities can be: seminars, laboratory practices, field practices, debate on certain topics, etc. For a given subject to have such a condition, at least 5 hours of activities in English and/or Portuguese must be scheduled during the semester in which it is taught. • Advanced level: subjects that are taught entirely in English and/or Portuguese, at least in one group of activity. These would be the type of subjects belonging to option 4 in the established regulations.

Aims
As stated above, the main aim of this study is to explore the training needs of EMI lecturers at UEx in relation to the main five dimensions identified by Pérez-Cañado (2020): (1) linguistic competence, which entails general communication and subject-specific language use; (2) methodology and classroom management, which is related to the teaching method and activities brought to the EMI courses; (3) resources and materials, aimed at identifying the main materials used in EMI lessons; (4) training needs, focused on linguistic and methodological needs, and (5) overall rating, which focused on participants' giving feedback based on their experience in EMI programmes.

Participants
A convenience sample of 27 lecturers involved in teaching through English at UEx took part in the study. Participants' mean age was 50.2, with the largest number being in the 40 to 50 age group, and the fewest in the 20 to 30 group (see table 1 below for more detailed information). As for gender, our sample was quite balanced: 14 males and 13 females. Finally, regarding job category, most respondents were tenured senior lecturers (11), followed by untenured senior lecturers (8) and professors (5). Table 1 below shows more detailed information regarding lecturers' demographic data. Moving on to their area of expertise, most respondents taught subjects on Engineering programmes (41%), and a third taught subjects on Social Sciences degrees. Nearly a fifth of the respondents taught disciplines belonging to Science and the remaining 7% taught subjects related to Health Sciences. In most of these programmes, students are home students, except for the Social Sciences where there is a small percentage of international students coming from Erasmus and other international exchange programmes (mostly to study programmes related to Business Studies).
As for teaching experience, more than half of the respondents had between 11 and 20 years of teaching experience, while 22% had between 21 and 30 years of practice. Very experienced lecturers (with more than 30 years of experience) represented 7% of the whole sample, whereas only 4% were novice teachers (less than one year of teaching practice).
Finally, regarding participants' EMI experience, a large majority (78%) had been teaching a subject through English between one and five academic years, and 11% more than five academic years. For another 11%, this academic course was their first experience in EMI programmes. Regardless of their EMI experience, for most participants, English was the main means of communication (> 95% of the time) in their classes. Table 2 shows the complete information about lecturers' teaching experience. Summing up, the average respondent is a lecturer over 40 years old, with a permanent position at the university and with quite an extensive experience in teaching L1 (more than 10 years) and EMI courses (between one and five years of experience).

Instruments
Data were collected using two online tools: a semi-structured written survey, and a quantitative questionnaire, both adapted from Pérez-Cañado (2020). The questionnaire and the survey were in Spanish in order to avoid any kind of bias that may have been produced by using an L2. The quantitative questionnaire consisted of a 26-item Likert-scale survey in which participants had to indicate their level of agreement with various statements from 1 (totally disagree) to 4 (totally agree). A fifth option -not known, not answered-was also included. Eight items in the questionnaire were aimed at exploring aspects of linguistic competence, three at methodological competence, eight examined aspects related to classroom materials and resources and the remaining seven items explored training needs.

Data Collection and Treatment
Questionnaires were administered using UEx's online Campus (CVUEx) between July 2019 and February 2020. Before the questionnaires were made available, the research team had meetings with some lecturers and coordinators of the programme at the different Schools and Faculties where the objectives of the project were presented.
Due to the nature of the collected data, with open answers in the survey, and the objectives of the study (i.e., to explore the needs of EMI lecturers), a descriptive analysis was carried out.

Linguistic Competence
As explained in the methodology section, eight items in the questionnaire were devoted to exploring linguistic competence. The first three focused on lecturers' perception of their own linguistic capacity to teach subjects in English, whereas the remaining items were more related to their perceptions of classroom interaction.
In terms of their perception of their own abilities, most respondents held a C1 level in English and they considered they had an appropriate knowledge of academic, generic communicative and subject-specific vocabulary. Their selfconception of their subject-specific vocabulary knowledge was particularly high, as can be seen in table 3 below.
Regarding communication in the class, most respondents seemed to agree on the fact that students were able to follow the lessons in English and that the use of the foreign language as the main vehicle of communication in class did not seem to hinder participation. In relation to the attention paid to language in the classroom, there seemed to be no consensus in relation to the explanation of linguistics aspects (grammar and/or vocabulary) in class: on the one hand, nearly 41% of respondents claimed not to have included linguistic explanations in their classes, but, on the other, 37% recognised that they had introduced grammatical or lexical explanations when needed. Finally, some participants acknowledged using Spanish to clarify problems derived from their students' lack of understanding in English, although the majority of respondents avoided the use of Spanish in their classes.

Methodology and classroom management
The great majority of the respondents used lectures as their main teaching method, although more than three-quarters said that they combine lecturing with other more active and participative methodologies and activities. Those most commonly used were students' oral presentations (10 respondents), taskbased learning (9), individual, group and pair work (6), discussions (5), problem resolution (5) and collaborative learning (4). Moreover, some respondents also mentioned other activities and/or methodologies such as flipped classroom (3), gamification (2), and the use of interactive videos with questions inserted in them (1), but these were individual responses rather than actions common to lecturers in general. Finally, two thirds of respondents stated that they made use of different types of classroom arrangements but a third reported to use only one type of arrangement. Table 4 shows participants' degree of agreement with the statements presented in relation to methodological aspects.

Resources and materials
Within this dimension, more than 90% of the respondents showed their satisfaction with the range of materials they had available for teaching their subjects in English, while three-quarters considered the materials they used to be the most appropriate option to help their students to learn subject content in the L2. Most lecturers stated that they created their own materials, although, depending on the language and the content of the materials, a large percentage also adapted materials found on the Internet. Most lecturers also included resources used to teach their subject in English-speaking countries. When asked about the type of materials and resources used, the preferred options were PowerPoint presentations, videos, websites, and some of the respondents also used textbooks in English and materials designed for native speakers of English. Interestingly, one lecturer has even published his own textbook in English. As can be observed, most of these materials are digitally-based, and this fact coincides with the importance that respondents gave to information and communication technology (ICT) and virtual learning in their classes: nearly three-quarters of the respondents made use of online learning resources in their classes, and twothirds considered ICT as a key aspect of their classes.
Finally, when asked about the language of the materials, once again, these EMI lecturers' reluctance to use the mother tongue in their classes was observed, with more than 90% of participants recognising that their materials do not include any instruction in Spanish. Table 5 below shows more detailed information about the lecturers' responses.

Training Needs
The last part of the questionnaire and one of the sections of the survey dealt with lecturers' training needs. This was specifically aimed at examining both their linguistic and their methodological needs. Starting first with linguistics aspects, the first interesting finding is that although most participants had an advanced linguistic level (C1), nearly half of them felt that they need further training on linguistic aspects, although not on subject-specific vocabulary, about which they were more confident.
This latter aspect is linked to another item also explored in this study: lecturers' willingness to participate in international programmes. Most respondents had already taken part in Erasmus+ programmes, international research projects and language immersion programmes offered by the UEx, and, when asked about their readiness to take part in such initiatives, the vast majority responded positively.
As for the second dimension examined, that is, methodological training, most respondents agreed that they needed more training on EMI materials design and that they would also welcome expert advice and training on EMI, as they felt that they lacked training on theoretical aspects of EMI education. In this respect, respondents were specifically asked about their willingness to receive some advice from EMI experts. As can be seen in Table 6, a large proportion of respondents would welcome recommendations about how to conduct their classes in English, and also a large percentage of participants showed that they would appreciate EMI experts coming to their classes to observe their sessions and give personalised feedback, although nearly 15% of the respondents did not provide an answer to that item. Table 6 includes more specific information in relation to lecturers' training needs.

Overall rating of the EMI programme
This final element was only examined in the survey with questions dealing with aspects such as changes in the workload, the value of taking part in EMI programmes, and the main difficulties and advantages of the EMI experience.
Regarding workload, all respondents agreed that teaching subjects in English involved a higher workload. Some respondents even quantified this extra workload, which ranged from one extra hour per week to the assertion that the workload was doubled. However, all respondents agreed that this extra effort was completely worthwhile, both personally and professionally. In relation to the personal advantages, most respondents indicated that participation in this kind of programme made them review their linguistic knowledge, and some of them had even taken up English lessons. Other lecturers pointed out that participation in EMI programmes made them feel more motivated when teaching their lessons. As for professional benefits, some lecturers highlighted that, in the EMI groups, students were highly motivated, the student-lecturer ratio was lower and teaching in such an environment was pleasant. Moreover, some lecturers mentioned that teaching of their subjects in English had made them re-consider their teaching practice in both languages. Finally, other respondents also argued that the implementation of EMI subjects had helped the school to recruit more international and home students.
As for the main disadvantages related to taking part in EMI programmes, respondents highlighted aspects such as the amount of time they devoted to creating materials, their lack of methodological training, and the low number of students taking part in the EMI courses, which in some cases led to the closure of the EMI group due to lack of support and understanding by colleagues and departments, and insufficient institutional support. Table 7 below summarises the lecturers' ideas about the main difficulties and strengths in relation to being part of an EMI programme.

Discussion
The present study was designed to explore five dimensions in relation to EMI lecturers at the University of Extremadura: linguistic competence, methodology and classroom management, materials and resources, training needs and overall rating of the EMI experience. Starting first with the linguistic competence dimension, most participants had an advanced level of English (C1), and this fact was reflected in their answers to the questionnaires. In general, lecturers felt confident with their English level and they considered that their L2 proficiency was enough to teach their subjects in the corresponding foreign language. This is particularly evident in their selfperception of their own knowledge of language specific to their subject, and this was the question which lecturers answered most positively. There may be two possible explanations for this. First, lecturers are also researchers and they use English as a lingua franca in conferences and research publications (Macaro et al. 2019) and, secondly, English is considered the main language of communication in academia and most researchers use it to communicate with others internationally. In academic meetings and at conferences, researchers talk about very specific topics and make use of very technical vocabulary. As such, lecturers may feel that their strongest point is their dominance of subject-specific vocabulary because they come into contact with it in a variety of contexts and they are probably used to using language specific to their subject in a variety of scenarios. Second, there is a substantial body of research identifying lexical knowledge as one of the aspects that benefits most from the implementation of CLIL in compulsory education (Agustín-Llach and Canga Alonso 2016;Alejo-González and Piquer-Píriz 2016;Castellano-Risco 2018;Castellano-Risco et al. 2020;Pladevall-Ballester and Vallbona 2016;Sylvén, 2019). Some lecturers mentioned that one of the main advantages of being involved in EMI programmes was that they felt obliged to start studying English again. It may be that the positive effect of EMI programmes in terms of confidence with and extent of content-specific vocabulary is not only relevant to students, but also for lecturers since delivering their lessons in an L2 places new linguistic demands, both specific vocabulary and more general language to facilitate interactions in the L2 classroom in order to effectively convey non-language content in class. But this would need to be analysed empirically as, to the best of our knowledge, there are no studies that have explored this issue in EMI.
Our results also show that lecturers felt less confident with their knowledge of English for classroom management and interaction with students, and in some cases, greater training in this area was highlighted as a need, in order to be able to deal with unexpected situations in class.
To conclude the analysis of the language competence element, there is a third aspect that drew our attention: these lecturers seem to be quite reluctant to use Spanish, either written or oral, in their classes, which evidences a clear difference with CLIL. Two reasons may explain this finding. First, as Morgado and Coelho (2013) argue, most lecturers at university level basically conceive language as a tool to convey their disciplinary contents (EMI) and they do not follow an ICLHE approach, that is, language objectives are not planned, assessed or integrated with content objectives and this seems to be the case of the participants in our study. Second, it may also be that these lecturers believe, as a consequence of their own language learning experience, that code-switching is not advisable or may even be counter-productive if the aim is to help students to develop their L2 skills, so they avoid the use of their mother tongue in the class.
Moving to the methodological aspect of their sessions, lecturing was employed as the main teaching method, although combined with other more interactive and participative activities such as student presentations, collaborative activities or discussions. This use of a mixture of methods is in line with the findings of Aguilar and Rodríguez (2012), who argue that EMI lecturers should give students opportunities to interact and make use of the L2 in the class, meaning that lecturers need to know methods and activities that help facilitate studentstudent interaction in class. The UEx lecturers who participated in the present study highlighted a need for ongoing training courses devoted to methodological aspects, especially regarding EMI and CLIL. This is, in fact in complete contrast to the situation of the lecturers described by Aguilar and Rodríguez who refused to be trained in CLIL practices. In our case, the lecturers who took part in the study seem to be willing to take CLIL training courses, and, in most cases, they even report to be willing to allow language specialists to go into their classes to observe, record and provide feedback on their lessons. This may be related to the differences in the lecturers' profile, since here, in comparison to Aguilar and Rodríguez's study (2012), the lecturers, generally, have more experience in EMI and, more importantly, were the instigators of the EMI initiative, as courses are not taught in English if the lecturer does make the proposal. The lecturers in Aguilar and Rodríguez (2012) were novice EMI lecturers, as the study was their first experience with EMI, whereas the lecturers in the present study are quite experienced regarding EMI practices. It may be that this greater experience may make them analyse their own teaching practice from a broader perspective, allowing them to be able to recognise their weakest and strongest points and have become more aware of the advantage of using student-centred pedagogies.
In relation to the materials available to lecturers, most respondents show their satisfaction with the materials they use in class. They use a wide variety of resources and most participants consider ICT as a key tool in their classes. However, lecturers state that they create and adapt their own materials, which may be an indicator of the lack of EMI materials available to work with at tertiary level as suggested by Aguilar and Rodríguez 2012. The creation and adaptation of teaching materials is a very time-consuming task, especially if attention must be paid to both content and language.
As regards the fourth dimension explored in our study, i.e., lecturers' specific training needs, despite the fact that the great majority of the lecturers were qualified to C1 level in English, they would welcome the university organising courses or workshops, either as face-to-face or online training, to help them improve their oral skills. Some respondents also highlighted the importance of courses devoted to extending their subject-specific vocabulary and classroom management language. Finally, some lecturers also suggested that the university could organise and facilitate longer international research stays in English speaking countries. To conclude the discussion of training, in terms of the methodological dimension, although the lecturers in our study seem to mostly have an EMI conception about teaching their disciplinary contents in English, they also show significant sensitivity to the linguistic aspects, although in our opinion this could be further developed by making them fully aware of the importance of ideas on teachers' identity (Dafouz 2018), showing them how this has an important impact on the construction of knowledge in an L2. In fact, some lecturers reported having re-considered their teaching practice in both languages as a consequence of their EMI experience.
Finally, when the positive and negative aspects of taking part in EMI programmes were rated, most lecturers defined the experience as a very positive one and showed their satisfaction. However, some wanted greater involvement of and recognition from the institution. They often reported that a negative aspect is that the situation of the EMI group may be threatened because of certain institutional policies: in order to create an EMI course, the L1 course must be also maintained, which, in practice, means a duplication of courses and, in most cases, an unbalanced number of students in the two groups, as L1 groups are usually far more numerous. Moreover, the university requires a minimum of five students for a group to be viable, without taking into account the total amount of students on each degree.
All in all, however, when asked to rate their EMI experience, all respondents gave positive feedback about the experience, highlighting personal enrichment as one of the main advantages, although they also called for greater recognition on the part of the university for both lecturers and students taking part in the EMI programme.

Conclusions
Our analysis of the perceived needs of EMI lecturers at the University of Extremadura allows us to establish a general profile of these particular lecturers. The results of both the quantitative questionnaire and the qualitative survey show that the lecturers involved in the teaching of disciplinary contents through English at the UEx are mostly experienced teachers with a good command of the L2 (in this case, English) who mostly have an EMI conception of the process concentrating their efforts on the contents they teach but, at the same time, show some sensitivity to ICLHE or EMEMUS, that is, to language.
Their reported needs are mostly related to: 1) further training on the specific language needed to efficiently manage the EMI classroom and the different types of interaction that take place in it and 2) developing a greater competence on the theoretical basis of bilingual education. Despite rating their experience as very positive and enriching, they also make clear that it is demanding and time consuming and, thus, they call for more institutional support and recognition. Taking into consideration not only these findings but also the global context of EMI at UEx, we would like to highlight certain aspects that could contribute to further develop these practices at our university.
First of all, there seems to be a need to find a balance between top-down, university imposed, policies and some sort of recognition for the initiatives that emerge from individual lecturers or small groups of lecturers.
Secondly, although many actions are carried out at UEx on the different topics under analysis, they need to be more 'visible' to both lecturers and students (and, also, outside the university). More effective dissemination of these actions is needed.
Finally, in our view, there is a need to give continuity to the subjects taught through English. Currently, subjects can only be taught through English (1) if the lecturer and the Department offer them and (2) if there are more than five students in the EMI group. In some cases, there are EMI subjects which have been cancelled because they did not reach the minimum number of students in a particular academic year. It would be necessary to maintain the offer for a longer period of time even when fewer students than required apply for it at a specific moment.
The findings of this study should, however, be treated with caution, and some of its limitations should be taken into account. First of all, a larger sample, with more lecturers completing the questionnaires, would be welcome. This would increase the significance of the study. Likewise, it would also be beneficial to explore the beliefs of the other stakeholders, such as the students, to compare their respective points of view. Finally, this study should be complemented with class observation to attempt to triangulate the results.