Keywords

1 Introduction

This small-scale study explores pedagogical issues of EFL (English as a Foreign Language) teachers who shift into teaching content using EMI (English as a Medium of Instruction) in two Japanese tertiary contexts. Despite the relatively rare occurrence of this shift in pedagogical practices, it casts useful micro-level insights into the growth of EMI in its particular Japanese sociocultural context since we as practitioners are primarily language educators with sensitivities about language acquisition. Seen from a wider macro stance, Japanese universities are increasingly adopting EMI to attract students due to government policies to globalize the tertiary sector, meaning that Japanese and expatriate faculty are often pressured to teach their content specialism through the medium of English. Whilst resistance exists to EMI, many engage in EMI in ways which are practically aligned with the realities of student and teacher language proficiencies. Such micro-level and locally contextualized appropriation of EMI manifests itself in diverse forms of delivery. In light of these issues, we critically analyze our lesson plans and co-constructed autoethnographic narratives surrounding our teaching practices in Sociolinguistics and Discourse Analysis courses.

Our study reviews the literature, outlines the methodological process, then presents findings in the form of actual classroom practice and teacher narratives, before moving on to a discussion of these findings. Conclusions and implications for our micro practices are then drawn.

2 Literature Review

We firstly turn to a review of the literature embracing issues of EMI in Japan, Content and Language Integrated Learning (CLIL), teaching methodologies, and translanguaging, all of which impact and shape our pedagogical approaches.

2.1 Japan’s Moves Towards EMI at Tertiary Level

Tertiary education in Japan has experienced a significant shift towards EMI in the last decade. However, as Goodman (2014, p. 130) indicates, although a “global phenomenon”, it is “highly context-dependent” meaning that, as can be seen in the Japanese case, its spread has not been uniform or consistent (Brown 2016). Its impetus stems from the realization at government level of the low international ranking of Japanese universities (Kirkpatrick 2014), the high cost of long-term overseas study (Burgess 2014), and a tendency towards inward-looking Japanese youth (Imoto 2013). In response, government policies have been passed down to elite universities to internationalize, or become ‘global’ in some form, namely, the 2008 Global 30 project, the 2012 Global Human Resources, and most recently the 2014 37 Top Global Universities (Brown 2014). The Ministry of Education, Culture, Sports, Science and Technology (hereafter MEXT 2011) reported that 30% of 800 Japanese universities offered some form of EMI. Such classes started at national universities and were followed by those in the private sector leading to a current 50% increase since 2005 (Brown 2016). Brown and Adamson (2012) noted this trend among non-elite (2nd tier) private universities seeking a competitive edge by increasing enrollment from foreign students and local students, especially due to a decreasing birthrate.

The reality is only 5% of undergrad students (mostly Japanese, not foreign students) actually take EMI classes (Brown and Iyobe 2014). Full EMI programs, as opposed to classes, remain rare (30 at undergraduate and 70 at postgraduate level) and foreign students are often separated from Japanese students in EMI models, for example, in Japanese culture and language degrees. In fact, only 20% of EMI programs actually serve foreign students (Brown 2016) and the perception towards EMI among Japanese faculty persists as primarily for study abroad purposes in Anglophone contexts. Phan’s analysis of the spread of EMI in Asia (2013) reveals that it remains a simplistic notion in Japan, often resisted due to claims of linguistic imperialism and greatly limited by Japanese faculty’s inability to deliver their content in English (Ishikawa 2011; Yonezawa 2011). The drive towards internationalization to create a ‘global’ university is often interpreted as teaching programs to foreigners in Japanese, rather than in English, resulting in unchanged language policy despite the MEXT policy of EMI. Chapple (2014) notes the poor motivation to engage in EMI as universities focus on the all-important requirements of job hunting for Japan-language employment, not employment utilizing English.

As Takagi (2013) summarises, EMI in Japan is only taught by those who can do so, not according to student needs, so that policy is implemented in a dissipated and non-homogeneous manner. This illustrates Tsui and Tollefson’s view (2003) that local policy-makers have deep-rooted political, social and economic beliefs which tend to override “the educational agenda” (p. 2). It is also reflected in Bamond Lozano and Strotmann’s analysis (2015) bemoaning the fact that the rush to EMI in “non-English dominant universities globally often happens with insufficient planning and investment” (p. 848) with the result that language issues in EMI are often “overlooked” (p. 848).

2.2 CLIL

In light of ad hoc shifts towards EMI and lack of language planning in Japan, research into the focus on both content and language on the curriculum plays a pivotal role in informing EMI classroom practice. The integration of content relevant to students’ fields of study and language (CLIL) provides those whose first language is not English with an “authenticity of purpose” (Coyle et al. 2010, p 5; Pinner 2012). With European roots in the mid-1990s and Canadian immersion education in the 1960s (Mehisto et al. 2008), a CLIL approach to instruction is conveniently fluid in methodologies, termed by Ikeda (2012, p. 12) as an “intentional organic” in which language and content teachers collaborate in various forms, from simply information-sharing about lesson content and pedagogy to actual team-teaching. Whether adopted on the language curriculum in English programs or the content curriculum in actual EMI, it can shift strategically during instruction from what Met (2009) terms as “hard” (immersion EMI) to “soft” CLIL (EFL instruction with a touch of content-related materials). This flexibility in delivery carries motivational benefits for students whose language proficiency falls behind their content knowledge (Edsall and Saito 2012; Lasagabaster and Doiz 2016), but risks, if not clearly explained to students or teachers, “disjuncture” (Mehisto 2008, p. 93), where course objectives and pedagogies are misunderstood. This potential confusion is countered by CLIL advocates who outline its cognitive advantages (Lamsfuß-Schenk 2002). Furthermore, Stohler (2006) argues that no significant differences are evident in content knowledge acquisition in L1 or L2 and adds that “...the teaching of non-linguistic topics in an L2 does not impair the acquisition of knowledge” (p. 41). However, Llinares (2015) notes that few CLIL studies measure content improvement, instead focusing solely on development of language proficiency.Specific to the Japanese tertiary context, Taguchi and Naganuma (2006) and Ohmori (2014) identify the potential for university language pedagogy in learning content through English. Further to this, Brown and Adamson (2012) and Brown (2016) argue English for Academic Purposes (EAP) provision in Japanese universities can take on a new purpose, not for preparation for overseas content study, but for localised Japanese forms of EMI involving the integration of academic Japanese and English language materials, a point resonating with Lasagabaster’s (2013) and Merino and Lasagabaster’s (2015) advocacy of the students’ L1 in CLIL instruction. This implies interdisciplinary collaboration between content and language faculty which moves CLIL instruction into a potentially bilingual mode of instruction. Importantly, though, the relationship between content and language faculty should adopt a more “horizontal alignment” (Turner 2012, p. 24) so that their respective professional expertises merge (Lucietto 2008).

2.3 Teaching Methodologies

The premise underpinning the teaching methodologies in this study is that English teachers make the transition to teaching content in English to Japanese undergraduate students. As Fujimoto-Adamson and Adamson (2015) underscore in this shift, it is important to offer not only language support towards students whose first language is not English, but also clarity in EMI content delivery. The “dual focus” (Coyle et al. 2010, p.1) of content requirements for the syllabus and language needs is clearly informed by the experiences and research into CLIL. To simply combine content and language without gauging the level and complexities of language and content input may lead to, as Lasagabaster (2011) warns, lower cognitive engagement by students. This would then necessitate some degree of “sheltered content-based” teaching (Brinton et al. 1989) by focusing strategically on language forms to facilitate effective, or “desirable” engagement with content (Muñoz 2007, p. 23), termed by Lyster and Mori (2008, p. 134) as a “counterbalanced” approach to teaching content. Problems inherent with such attention to language, however, are noted by Ottewill and Drew (2003) who observe that, when transferring skills acquired in language instruction to content classes, students may perceive language and content skills as belonging to “separate spheres” (p. 186). Overcoming this lack of skills transfer would appear at least on the teacher’s part to be less of an issue for language teachers who make the transition into content teaching as language issues are integrated into the flow of the content lesson (Adamson 2005a).

Other strategies recommended are “slowing down one’s pace” to allow students to absorb the content input and engage in peer concept checking (Goodman 2014, p. 139). We argue that it is essential to extend scaffolding of the lesson’s content input from a strategic language focus to an additional sensitivity towards visual representations (grids and matrices etc) of content (Fujimoto-Adamson and Adamson 2015; Adamson 2005b). Such visuals can be effective in clarifying to students complex concepts or processes and reduce time spent on lengthy oral explanations (Kang 2004; Wallace 1980). Pinner (2012) and Er and Kırkgöz (chapter “Introducing Innovation into an ESP Program: Aviation English for Cadets”, this volume) support this use of authentic materials relevant to students’ content fields as they carry more relevance to the instruction and are therefore more likely to engage students in deep learning in EMI contexts, as outlined by Soruç et al. (chapter “Listening Comprehension Strategies of EMI Students in Turkey”, this volume).

3 Translanguaging

The final influence on our practice in this study surrounds the use of the students’ L1 (Japanese) in content classes. As previously argued, our EMI practice is influenced heavily by research into CLIL, the nature of EMI in the Japanese tertiary context, and our strategies concerning the teaching of content by language-sensitive EFL instructors. Embedded within these arguments is the notion that content instruction is porous to the local (L1) academic context and not necessarily intended to prepare students for academic study abroad in Anglophone contexts (Brown and Adamson 2012; Brown 2016). We pursue this argument by addressing how Japanese (the L1) and English (the L2) are combined to enhance the sense of ‘localised’ academic study and fundamentally challenge the predominance of the ‘E’ in EMI, a stance which resonates with Belcher’s (2013 in Paltridge & Starfield eds.) comments about the growth of English for Specific Purposes research in non-center contexts.

The use of two languages in pedagogic contexts is termed as “translanguaging” and defined by Doiz et al. (2013, p. 213) as “the adoption of bilingual supportive scaffolding practices.” It fundamentally opposes the “two solitudes” (Cummins 1994) separation of languages in education, in which L1 use in L2 instruction is viewed as “contamination” rather than “cross-fertilization” (Blackledge and Creese 2010, p. 203). The “integration” of practices from multiple languages is regarded as cognitively beneficial to learners (Garcia and Wei 2014, p. 80), not simply for the acquisition of a new language, but also for content instruction in the L2 (Hornberger 2003; Hult 2007) where collaboration between students on authentic tasks is necessary to mediate understanding (Martin-Beltran et al. 2017). Its use generally signals a shift in language education from monoglossia to hetereoglossia and is now accepted practice in many educational contexts (García and Flores 2013) where there is a growing integration of language, literacy, and content learning through translanguaging. Strategic L1 use is beneficial in “developing an enriched bilingual vocabulary” (Gallagher and Collohan 2014, p.11) and, especially for lower language proficiency students, in developing their sense of autonomy through “code choice” (Levine 2011). Soruç, Dinler and Griffiths (chapter “Listening Comprehension Strategies of EMI Students in Turkey”, this volume) investigate this issue in the Turkish tertiary EMI context, revealing that translation by many students themselves into L1 aids their comprehension of L2 lecture notes. In the Japanese tertiary context, Adamson and Coulson (2014, 2015) reveal that translanguaging through use of Japanese academic readings in English lectures preparation classes enhances lower proficiency students’ comprehension, reduces anxiety, and develops L2 writing confidence. Further to this, Lu and Horner (2013) regard L1 to L2 translation as a kind of “translingual” approach which gives “agency and responsibility” (p. 29) to L2 writers. Also, Lorimer (2013, p. 163) indicates that L1 use in the drafting process of L2 writing develops “rhetorical attunement” and an “ear for difference.” However, despite these reported merits, its use must be also tempered by views by students or instructors of “feelings of guilt, unprofessionalism and subterfuge” (Gallagher and Collohan 2014, p. 2) if their personal views, or institutional policies, on L2 learning or EMI are heavily influenced by monolingual views on learning and teaching. Finally, Cross (2016) stresses that tensions do exist surrounding translanguaging as a pedagogical approach, both in language learning and content learning contexts, and may influence students’ identity formulation and development of proficiency in language and content.

4 Methodology

The methodological approach for this study is a triangulation of descriptive observation of classroom methodologies and collaborative autoethnography (CAE) (Chang et al. 2013, p.17). The two institutions and practitioners are seen as part of a “collected case study” (Stake 1995) in which context and participants are seen as distinct. The first step in this process is to describe the two contexts and participants, noting their particularities. This is followed by a presentation of classroom methodologies which the practitioners see as typifying their focus of content and language(s). It is supported by extracts the CAEs which reveal practitioner perceptions and beliefs about the chosen teaching methodologies not perhaps possible through individual narratives. The formulation of CAEs was undertaken interactively on Google Drive over the duration of the 2016–2017 academic year. Despite being a lengthy process, it creates a co-constructed narrative space which aims for “collective exploration of researcher subjectivity” (Chang et al. 2013, p. 25) to achieve “deeper learning about self and other” (p. 25). This “synergy and harmony” (Chang et al. 2013 p. 24) is argued here as not simply to reveal beliefs but can also serve as a transformative means in teacher development.

5 Findings

Findings are presented firstly concerning the context of the study, followed by selected examples of classroom methodologies and concluded with key extracts from the CAE.

5.1 Context and Participants

The following table represents details of the two contexts and participants (teachers and students) in which we practice EMI.

Table 1 illustrates various factors which help to contextualize the study. As can be noted from the above, the language proficiency of students in the private university is generally below that of those in the prefectural university. Both teachers have experienced studying content in a language other than their own L1 which is potentially a means of sensitising them to their current students’ language difficulties in EMI classes. Interestingly, both teach a progression of EAP, EMI and then have classes in which students write an undergraduate dissertation based on their EMI subjects of Discourse Analysis and/or Sociolinguistics. The two EMI subjects are taught in a similar manner, by means of teacher input scaffolded with various visuals which we present in the next subsection. This input is accompanied by class activities requiring students to work individually or in small groups to complete tasks aimed at raising awareness about the content. These tasks are carried out using either Japanese or English according to student preference. Assessment for the Discourse Analysis class is a mixture of homework, formal test and student group presentations on a theme related to the course content. The Sociolinguistics course is combined with some Discourse Analysis and is assessed through essays on course themes and similarly encourages students to investigate either English or Japanese language themes.

Table 1 Two contexts/participants
Full size table

5.2 Classroom Methodologies

We present here a selection of activities which are in various degrees and forms aided by visuals. As previously argued, this is representative of the scaffolding of content input in order to not simply convey course input for comprehension purposes, but also to act as a means for student activity involving either individual or small group work.

Example 1: L1 Language Awareness/Translanguaging

The first example comes from the Sociolinguistics class and requires students to write the Chinese characters (in Japanese kanji) which are used in everyday life in Japan (Japanese consists of three types of scripts derived originally from Chinese. Kanji are pictorial representations of meaning as opposed to the other two scripts, hiragana and katakana which are phonetic). As students are expected to master a wide range of kanji for daily and academic use, it was considered to be a valuable awareness-raising task to ask them to write three common kanji and analyse them for the inclusion of the character for ‘woman’ in each one. Discussions focused on the potential sexist implications in each character as the meanings of ‘jealous’, ‘hate’ and ‘noisy’ are predominantly negative. Students were frequently surprised in discussions and feedback from groups suggested that most considered the negative connotation to be sexist.

Rubric: Many words or expressions show discrimination towards men or women. In Japanese, what language do you know which is “sexist”? Please write the kanji for 1.netamu (to be jealous), 2. kirau (hate), and 3. kashimashi (noisy).

Why is the kanji for woman (女) in these negative meanings? Is this discrimination? Write your response below and compare with a partner.

figure b

Example 2: Lesson Extract on Speech Events

This next example from the Discourse Analysis class requires students to reflect on class input on speech events by assessing the degree of formality on a speech event continuum. The task is completed firstly individually by marking a number on the continuum. Pair and group comparisons then follow and students are prompted to discuss why they have placed a speech event at a certain point. This task is particularly useful in encouraging student awareness about their own discourse in various speech events which vary in hierarchical nature.

Rubric: Sacks et al. (1974) looked all many types of speech events and made a “continuum” from the most formal type of talk to the most casual.

figure c

In pairs, decide where these speech events go on the continuum of speech events and compare with a partner. How are your answers different or the same?

  1. 1.

    Talking to your mother on the phone

  2. 2.

    Talking to an old friend on the phone

  3. 3.

    Talking to your old teacher on the phone

  4. 4.

    Giving a graduation speech

  5. 5.

    Opening a party at cherry blossom time

  6. 6.

    Speaking to a waiter in a restaurant

  7. 7.

    Speaking to a junior at college

  8. 8.

    Speaking to a senior

Example 3: Dialect

The Sociolinguistics course has a lesson on dialects, part of which focusing on London ‘Cockney.’ The rhyming nature of Cockney with its standard meaning and then producing a sentence proves to be amusing and quite unique for students.

Rubric: In London, some local people use Cockney. The word for an underground train in London is “tube.” In rhyming Cockney English, they say “cube”. Cockney speakers also change the letters of words and add new words, e.g. “Bees and honey” means money.Try to match these Cockney words (left) with the standard English (right) with a line --------------:

Lollipop

home

Gates of Rome

road

Dickey dirt

single

Frog and toad

look

Adam and Eve

shop

Mandy dingle

believe

Butcher’s hook

face

Boat race

shirt

Can you make a sentence?

Example: Let’s take the cube = let’s take the tube

figure d

Example 4: Context and Discourse

The next example comes from both the Discourse Analysis and Sociolinguistics course and is taken from the means to describe the context underlying discourse by Hymes (1974).

Rubric: The following is the SPEAKING model by Hymes (1974). Now use it to analyse an extract from a movie of your choice.

Initial

Word(s)

Explanation

S

Setting and scene

Time and place of a speech act

P

Participants

Speakers and audience

E

Ends

Purposes, goals and outcomes

A

Act sequence

Form and order of the events

K

Key

Tone, manner of spirit of the speech act

I

Instrumentalities

Forms and styles of speech

N

Norm

Social roles governing the event and participants’ action and reaction

G

Genre

Kind of speech act or event

The following is one student’s analysis of a scene from the movie Charlie and the Chocolate Factory and illustrates the compact and visually clear nature of using the SPEAKING model in tabulated form. Student language output is notably short which accommodates linguistically less proficient students in the task.

Initial

Criteria

Analysis

S

Setting& scene

Charlie gets the last ticket that it can go to the chocolate factory therefore his family needed to decide who was going to the factory with Charlie.

P

Participants

Speakers: Charlie and his Family, Audience: Family members

E

Ends

Charlie and family decide go to the Chocolate Factory.

A

Act sequence

Family decided that Grandpa Joe was going with Charlie to the factory. However, Charlie said that he did not go to the factory, because he decided to sell the ticket to someone.

K

Key

Charlie: hiding his true feeling, Father: modest, Grandpa Joe: determined, Mother: honest

I

Instrumentalities

Casual conversation among family members

N

Norm

Charlie did not have appetite of child. Therefore, he could not say his feelings better.

G

Genre

Conversation

Example 5: Cultural Identities

The final example is from the Sociolinguistics course and uses Holliday’s (1994) concepts of multiple affiliations to cultural groups in identity. It acts as a means to raise awareness among students that they belong to diverse ‘cultures’ which impact their relationships. After representing these ‘cultures’ in pyramid form below, students are asked to represent their own cultures in similar form and use the concept as a way to analyse a part of a movie.

figure e

The example below shows one student’s analysis from the movie Notting Hill (1999) in which Anna, a rich American movie star meets William, a struggling English book shop owner in London. Differences and similarities in their respective cultures become visually clear.

figure f

5.3 Collaborative Autoethnographies

The first stage in creating our CAE narratives was to create “frames” (Warwick and Maloch 2003, p. 59), the broad themes around which stories could be told. We negotiated the following seven frames to use for our CAE:

  1. 1.

    Moving from EFL to EMI practice

  2. 2.

    Adaptations in EMI for local Japanese students

  3. 3.

    Scaffolding issues

  4. 4.

    Translanguaging issues

  5. 5.

    Assessing learning outcomes (Language/content)

  6. 6.

    Student perceptions

  7. 7.

    Implications for further practice/investigation

The second stage signalled the start of the actual CAE which took place through 2016 in written mode on Google Drive allowing us to store our narratives and interact without necessarily meeting. For the third stage, data analysis of the discourse was a “crystallization” (Hycner 1985, p. 279) to represent the ‘essence’ of the narratives within each frame. Key parts are presented below (John’s narrative in italics, Naoki’s normal script.)

Extract 1: Moving from EFL to EMI Practice

Themes emerging in this first frame revealed similarities in what John and Naoki teach and highlighted their realization of the importance of linking 1st year EAP with EMI study later.

(Naoki) EMI was very challenging for me because in their final year, my students need to write their BA dissertation in English (5,000 words), oral presentations and oral examinations in English. I realized that it is quite a responsible job to do compared to teaching EFL.

(John) This is a similar situation to me as I also teach 4th students who need to write a dissertation in English doing a discourse analysis followed by a short presentation. In a sense all the academic writing instruction I give to 1st graders has a final objective with the dissertation.

(Naoki) I also started to teach English for Academic Purposes (EAP) for small groups of 1st year students since last year. In this class, I teach basic study skills which can be a preparation for EMI lessons when they become 3rd and 4th years.

Extract 2: Adaptations in EMI for Local Japanese Students

This frame emerged as key in understanding John and Naoki’s adaptations in EMI. Both referred to their ways of scaffolding the classes in terms of language, namely simplification of delivery and modelling of essays. Content simplification was also mentioned, John noting that his prefectural university’s fields of study were not primarily focused on educational or linguistics so many students lacked a background in Applied Linguistics.

(John) As my large EMI lessons are for mixed proficiency students, I need to make some adaptations in terms of language, and also in some sense, content too. Students at my university are not studying linguistics.

(Naoki) Although my classes are smaller, my EMI is similar because my students’ main subjects are international politics, economics and culture. Also, their language proficiency is perhaps lower but still varied, so I believe that simplification of the content delivery and scaffolding the lessons are essential.

Extract 3: Scaffolding Issues

In terms of how both teachers scaffold EMI, John focused more on in-class language issues (pre-teaching vocabulary items) and simplification of content delivery, whereas Naoki’s concern was more on providing final writing models (templates) as a basis for the final year dissertation.

(John) I try to scaffold both English use and content ideas. For language, I write a short list of key language terms at the side of the board either before or concurrent to content input. I tend to slow down in delivery when they take notes and provide teacher copies for this in which terms are defined simply.

(Naoki) As for my 1st year EAP and 4th year seminar classes (Sociolinguistics and Discourse Analysis), I always make sure to show models. Then students can understand what kind of essays or dissertation they need to write. I recommend they use my template with set expressions.

Extract 4: Translanguaging Issues

Both have bilingual stances and advocate L1-L2 use without top-down language policy implementation. This encourages student choice and decision-making about which language to use since the course objective in EMI are more content- than language-focused.

(John) One major difference from English-only EMI practice for me is the use of students’ L1 (Japanese) in various ways. Firstly, I use Japanese at times to manage the lesson or to make strategic explanations of some concepts. My feedback is sometimes in Japanese after a task. Also, students may use either English or Japanese for discussions. They can also analyse Japanese discourse (movies, recorded data etc) with frameworks (from their Discourse Analysis classes) in English. I don’t forbid Japanese because the objective of the lesson is not to develop English skills but content understanding. In this way, I hope that lower proficiency English students will not be reticent in discussions. There is a hybridity of free-flowing L1-L2 use by myself and students simply to ‘get the job done’; each student or group makes their own language choices rather than through ‘policy’ imposed upon them by me.

(Naoki) My stance is also similar as I don’t forbid my students’ Japanese use. I also use both English and Japanese during the lessons. Additionally, I occasionally use a sociolinguistics textbook in Japanese to supplement English handouts and deepen students’ understanding. After all, the purpose of the lessons is to understand the content of Sociolinguistics and Discourse Analysis.

Extract 5: Assessing Learning Outcomes (Language/Content)

In consideration of assessment, John and Naoki indicate that content criteria are prioritized and that only language issues hindering comprehension lead to marking down student grades.

(John) My evaluation criteria are content-focused rather than language-based. Of course, if presentations and reports are written in such poor English that I cannot understand, then they need to be marked down.

(Naoki) My evaluation criteria is similar because I focus on both content and language. The main reason is my relatively small groups of students so I can advise on language use for essays and dissertations.

Extract 6: Student Perceptions

Formal surveying of student perceptions of John’s course was institutionally based revealing generally positive student feedback, although he is concerned about student motivation for his class. And how Japanese faculty who teach content through Japanese may regard his course. Naoki’s course has no institutional survey but feels perceptions are positive.

(John) … official feedback has been positive too although I wonder what motivation some students have for taking the class: An easy option? Curiosity? A content class to practice English? Japanese content teachers may view my content classes as less rigorous than theirs.

(Naoki) Actually, since my EMI lessons are conducted in the seminar classes with small groups of students, there is no official student feedback. However, students seem positive possibly as these are optional so students are genuinely interested.

Extract 7: Implications for Further Practice/Investigation

Both John and Naoki see a need to investigate L1 use more. Naoki particularly wishes to survey student perceptions more formally.

(John) In terms of practice, I think I need to reflect on how I use L1 (Japanese) more – is it effective? too much? too little? Should I learn how to explain technical terms more in their L1?

(Naoki) Yes, I agree with John to investigate more about the use of L1. Also, since there is no formal student feedback in my EMI class, it is difficult to know about students’ perception. Therefore, conducting questionnaire or interview for students would be interesting.

6 Discussion

In light of the findings presented from the contextual details in the two settings (Table 1), methodological practices teaching EMI, and the CAE insights, it emerges that a degree of hybridity in practice and underlying rationale exists. Clearly we both share pivotal experiences in learning content through an L2. We both stress the importance of language sensitivity through models and vocabulary, as in Lyster and Mori’s (2008, p. 134) “counterbalanced approach” to EMI. Additionally, significant to the notion of “hybrid” practice (Fujimoto-Adamson and Adamson 2015), we embrace the students’ L1 for discussions, background reading and within content tasks (see Example 1 on the kanji for woman and Extract 4 on how analysis of Japanese language is encouraged). This stance sees translanguaging for “bilingual supportive scaffolding” (Doiz et al. 2013, p. 213) in language comprehension, and as part of the content syllabus. Language then takes on the role of medium and content accommodating “bilingual language practices” (Garcia and Wei 2014, p. 80) of all participants. In this sense, our own language competencies – Naoki’s as a fluent Japanese speaker and highly competent English speaker, and John’s as fluent English speaker with some Japanese competence – could be contrasted by concluding that the ability to translanguage between English and Japanese is a shift in which Naoki can excel, possibly more so than John when the focus is cast upon subtleties in Japanese. In reverse, when the focus shifts to subtleties in English, John may be seen as possessing deeper knowledge of English; however, our respective competencies are not simply language-bound as content issues of Sociolinguistics and Discourse Analysis in EMI are not solely limited to language proficiency.

Assessment of both EMI courses are content-focused, although John notes extremely poor language leads to downgrading. Important here is the avoidance of confusion as to what assessment is based upon – content or language – as we are recognised in our institutions primarily as language faculty (Lamsfuß-Schenk 2002). The issue of “disjuncture” (Mehisto 2008, p. 93) does not emerge but John mentions (Extract 6) how other faculty may regard his simplified delivery. Naoki appears less concerned with such adaptations and stresses the practicality of language modelling for essay writing, perhaps a consequence of her own EMI experiences in the U.K. and heightened sensitivity to student struggles in academic writing (Extract 3). Interesting, we both note the link in EMI practices to EAP instruction (Extract 1), providing justification and relevance for the “horizontal alignment” (Turner 2012, p. 24) of EAP aims to local EMI norms (Brown and Adamson 2012). As we teach both EAP and EMI, it appears we are well-positioned to transfer skills taught in language classes over to content classes, thereby combining skills from “separate spheres” (Ottewill and Drew 2003, p. 186). This natural familiarity with content and its vocabulary avoids the issues outlined by Celik et al. (chapter “Are We Really Teaching English for Specific Purposes, or Basic English Skills? The Cases of Turkey and Latvia”, this volume) of the language teacher’s struggles with technical terminology.

In terms of government (MEXT) policy advocating the adoption of EMI, there is a clear sense of localization of content and languages which exhibits our own interpretations of “the educational agenda” (Tsui and Tollefson 2003). Our experiences as students ourselves and as language teachers give us sensitivities and insights into EMI practice which are pedagogically pragmatic and reflective. There is a “fluid practice at the institutional level” (Goodman 2014, p. 131) in practice informed by CLIL research and an expansion of the use of language(s) in EMI which challenges the monolingual ‘E’ in EMI. Bilingualism in our local EMI practices is however tempered by awareness that tensions exist concerning translanguaging (Cross 2016) due to potential feelings of monolingual language beliefs among some students who see John’s course merely as a means to practice English (Extract 6). This is in contrast to the acceptance of plurilingual language policies as outlined by Mačianskienė and Bijeikienė in the Lithuanian tertiary context (chapter “Fostering Active Learner Involvement in ESP Classes”, this volume).

7 Conclusions and Implications for Teaching and Research

Our conclusions to this study are limited to the case study boundaries from which the data was gathered. The findings are in this sense intrinsic to our practices alone; however, these practices are shaped by broader trends in Japan and globally to which other EMI practitioners are inevitably exposed. Our first concluding thoughts are that the CLIL literature focusing on how content and language are combined has greatly informed our practice as it gives us, as language and content practitioners, a source of studies which can be contrasted to our own experiences. Although our study has focused on the EMI experiences of teaching Discourse Analysis and Sociolinguistics, our primary teaching activity in EAP course which precede EMI provide a rare but unique informative set of insights into the role of language in content instruction, a sensitivity towards students struggling in language proficiency with which we are engaged daily. The essential message for us as language educators is to encourage students to carry over skills acquired in EAP lessons to EMI learning and not to regard them as “separate spheres” (Ottewill and Drew 2003, p. 186). In our EMI practice itself, we coin the expression ‘“hybrid” (Fujimoto-Adamson and Adamson 2015) to try to encapsulate how teaching and learning take on a “fluid” nature (Goodman 2014, p. 131), noted in CLIL practice as “organic” (Ikeda, p.12). In concrete terms, this manifests itself in the inclusion of the students’ L1 (Japanese), not as a top-down policy but more of a pragmatic and natural means to translanguage into meaning. Language use is seen both to achieve comprehension and as part of the content syllabus. This, as previously argued, fundamentally challenges the ‘E’ in EMI. Also in this hybridity is the scaffolding of content through simplified visuals which is not uncommon in any content instruction, yet takes on greater significance when learning content in one’s L2. Our own experiences studying content abroad in Germany and the U.K. possibly help shape that practice.

The implications for this small-scale study are to firstly collect more data longitudinally, not just to continue the CAE which have provided valuable insights over one academic year for us, but to expand by triangulating with data exploring student perceptions about our hybrid EMI practices. Of particular interest, here would be how students view translanguaging, as a practice which contravenes their beliefs about content instruction or language learning. Fundamentally, this asks whether monolingualism or bilingualism should be integrated into EMI. In the process of adopting EMI in Japanese tertiary institutions, the complaints of a lack of uniformity and clear government direction in implementation due to resistance and teacher competence may actually give rise to the creation of a flexible space for pedagogic experimentation as seen in our study in non-elite universities. Of final note is the methodological approach using CAE which we both feel to have had a synergistic effect (Chang et al. 2013) on how we give voice to our perceptions on our practices.