‘Eventually we’ll all become Anglophones’: a narrative inquiry into language-in-education policy in Rwanda

ABSTRACT This article recounts the lived experience of a Rwandan English teacher’s journey from his village to international opportunities for scholarship and work, enabled by his agency in navigating the country’s changing linguistic landscape. Using the narrative inquiry research method, the study uncovers an insider’s perspective on Rwanda’s switch from French to English-medium education and offers a unique participant-centred contribution to research on language-in-education policy. The respondent’s story is interpreted through the intersecting conceptual themes of linguistic imperialism and cosmopolitanism, which emerge from the narrative. The decolonisation of English-medium education is envisioned through supportive, multilingual pedagogies and a rooted cosmopolitanism.


Introduction
The ascent of the English language in Rwanda in the last twenty-five years has taken place in tandem with powerful political, social and economic forces which have markedly transformed the East African country. Since the tragic genocide of 1994, English has played an increasingly important role in Rwandan society and in its institutions, driven by a rigid post-genocide political paradigm and the interests of the country's elite. Government policy has established English as the medium of instruction (EMI) in nearly all educational contexts, replacing French, the former colonial language, and Kinyarwanda, the national language. However, the imposition of EMI in state schools has presented serious and persistent challenges to Rwanda's education system.
Using the narrative inquiry research method, this study aims to learn from a single educator's lived experience of the English language and the transformation of Rwanda's education system from French to English medium. The single participant whose life story informs this study has been anonymised as Maurice, a Rwandan teacher educator currently working on internationally funded projects which support Rwanda's transition to EMI. Maurice attended French-medium state schools in rural Rwanda in the 1960s and 70s and has been working in English language education since the mid-1980s.
While the single source narrative inquiry method does not aim to be representative or generalisable, this study is augmented by my professional relationship with Maurice and our collective experience in supporting teacher adaptation to EMI in Rwanda, as well as by the work of other scholars who have researched Rwanda's language-in-education policy changes. The data collected in this narrative inquiry is assessed through the lenses of cosmopolitanism and linguistic imperialism, which emerged as intersecting conceptual themes in Maurice's story.
Finally, this study offers a novel, participant-centred contribution to the body of research on language-in-education in Rwanda through the unique and valuable lens of Maurice's lived experience. A life narrative such as Maurice's provides a personal window into the world of educators whose stories often go unheard by policymakers and are seldom the focus of academic research. I, therefore, hope that those who direct language-in-education policy will more closely engage with the personal stories of those who inhabit the classrooms and the communities which most feel their impact.

Rwanda: languages, history and ethnic politics
The legacy of colonialism and ethnic violence informs much of what the world hears about contemporary Rwanda. Indeed, the events which culminated in the genocide of 1994 marked a catastrophic turning point for the country and its people in all aspects of their lives. Policies on language and education have been among the most frequently and radically transformed, with consequential impacts on millions of Rwandans lives.
Rwanda's history of ethnic violence and its relative ethno-linguistic homogeneity presents a distressing paradox. Its population consists of the Hutu (around 84%), the Tutsi (15%) and the Twa (1%), nearly all of whom use Kinyarwanda, the national language (Adekunle 2007). In spite of the widely shared vernacular, French or English have long played prominent roles in Rwanda's governance, business and education, despite only being spoken by a small elite (Borg 2015;Samuelson 2012). French was dominant under Belgian colonial rule up to 1962 and largely maintained its status until 1994, a position politically underpinned by successive French governments (Samuelson 2012). France's support extended to the Hutu-led government's military campaigns in the long-running civil war against Tutsi-led rebel forces, culminating in the genocide of the Tutsi and subsequent reprisals (Borg 2015). Much of the Tutsi rebel military leadership, which seized political power after the genocide, had come to Rwanda from Uganda, where they had grown up with English (Mazrui 2004). They were followed by thousands of other English and Kiswahili-speaking exiles from nearby countries, adding linguistic diversity to communities and establishing a new socio-economic elite who initiated a gradual change from French to English as the principle language of education, government business and high-end commerce (Johnstone 2019;Mazrui 2004). English has thus become the second official language in Rwanda while the status of French has been significantly reduced, the latter partly seen as a snub of France and its political alignment with the Hutus during the genocide (Borg 2015;Mazrui 2004).
Claiming to avert any resurgence of Hutu-Tutsi ethnic violence, the Rwandan government has effectively outlawed all public expression of ethnic identity. However, several scholars of Rwandan language politics (Johnstone 2019;Samuelson 2012;Hintjens 2008) suggest that linguistic identity has become a proxy for ethnic identity in the persistent but unacknowledged strife between Hutu and Tutsi elites. The latter are said to use English as an instrument to retain their stranglehold on positions of power, while relegating Francophone Hutus to less prominent roles (Hintjens 2008).
Rwanda's move to English was bolstered by its accession to membership of the Anglophone East African Community in 2007 and the Commonwealth of Nations in 2009 (Banerji 2010). Along with the government's Singapore-inspired economic development strategy (Knutsson 2012), these dramatic shifts are emblematic of Rwanda's rebranding as an African success story. This narrative lauds Rwanda's impressive development indicators and evokes images of the country's epic journey from the victim of genocidal slaughter to exemplary recipient of international aid (Reyntjens 2013;Knutsson 2012). In this context, English has been used to curry favour with Anglophone governments and international donors (Johnstone 2019;Hintjens 2008;Borg 2015). However, in the eyes of critics, Rwanda's oft-touted successes have been tempered by the government's authoritarian approach to governance, especially the violent suppression of political dissent and the undermining of civil society (Human Rights Watch 2020).

Rwanda: languages-in-education
Under Belgian colonial rule, the provision of education in Rwanda was mostly limited to the Tutsi elite and was offered exclusively in French-medium. Although access to education expanded to the wider community in the post-colonial period, French largely maintained its dominant position. Only early-primary grades were in Kinyarwanda-medium, with English and Kiswahili taught as subjects in secondary school (Pearson 2014). The political shift to English following the 1994 genocide prompted a series of changes to language-in-education policy, culminating in a mandate in 2008 to replace French with EMI from fourth-year primary onward in all state schools. Samuelson and Freedman (2010) provide a dispiriting early assessment of the government's EMI policy: First, claiming that English will contribute to reconciliation by promoting economic growth overlooks the ways that language use is linked to identity amongst Rwanda's elites. Second, claiming that early English education will lead to greater participation in the global economy, and therefore to improved living conditions and prospects for educational advancement, overlooks the ideological implications of the spread of English as a global language. (192) Subsequent classroom-based research uncovered further shortcomings, including teachers lacking adequate training and preparation time to switch to EMI and primary learners struggling to cope with a wholly unfamiliar language (Pearson 2014;Samuelson 2012;Milligan, Clegg, and Tikly 2016).
More than a decade after the policy's enactment, the transition to EMI is still incomplete, as international agencies continue to provide technical assistance to teachers and support staff to adapt to the use of English in the classroom (Education Development Trust 2020). Policy also continues to shift. In late 2019, acting against the advice of international experts, the government announced its intention to switch from Kinyarwanda to EMI from first-year primary (Edwards 2019). This will almost certainly result in further strain on an already overburdened system which has for long struggled with both student literacy in Kinyarwanda and capacity to deliver EMI (Simpson and Muvunyi 2012;Samuelson and Freedman 2010).

English, linguistic imperialism and cosmopolitanism
Based on the data generated with Maurice and supported by the literature, I have positioned the role of English education in Rwanda on the intersection of linguistic imperialism and cosmopolitanism. The former provides a critical assessment of the global spread of English education while the latter framework largely justifies the rise of English as a necessary global lingua franca.

Linguistic imperialism and criticisms of English
Critical applied linguists such as Canagarajah (1999), Pennycook (1994) and Phillipson (1992) advance the notion of linguistic imperialism, which highlights the hegemonic impact of the global spread of English, especially the consolidation of political and economic power of the West in postcolonial contexts. They further point out that imposing English in the global south tends to benefit a privileged elite who possess the cultural and socio-economic capital (Bourdieu 2018) to access high quality language education and resources, thereby maintaining their hold on political and economic power connected with the language. Moreover, there is evidence in much of the global south that English programmes deplete a significant portion of public education budgets and donor funds, thereby limiting the resources dedicated to more locally relevant learning objectives such as first language literacy and numeracy (Bruthiaux 2002).
The growth of EMI in developing countries has been under much scrutiny by researchers, whose findings largely concur that the switch from local languages to English has been imposed too early and too suddenly, with most teachers and students lacking the necessary language skills and resources to achieve successful educational outcomes (Brock-Utne 2015). Indeed there is a broad consensus among scholars and development agencies working in sub-Saharan Africa that supports mother tongue-based multilingual education until at least the end of primary level and a transition to an international language only after sufficient academic proficiency in that language has been achieved (Chumbow 2013;Simpson 2019). Transcending the argument for multilingual education, some scholars evoke the concept of translanguaging within 'the African value system of ubuntu where languages are interwoven in a system of infinite dependent relations that recognise no boundaries between them' (Makalela 2019, 238).
Despite significant empirical evidence supporting education in local languages, the drive toward promoting English and implementing EMI is accelerating rapidly in Africa and worldwide (Macaro 2018). Support for EMI is widespread among policymakers, teachers and parents in much of Africa (Rassool, Canvin, and Heugh 2007) as well as in Rwanda (Samuelson 2012;Milligan, Clegg, and Tikly 2016;Pearson 2014). Scholars attribute this to the widely held perception that English and other colonial languages 'are the only means for upward economic mobility' (Ouane and Glanz 2010, 4-5), resulting in policies which are in fact contrary to the public interest and which reinforce political and social inequalities (Rassool, Canvin, and Heugh 2007).

Cosmopolitanism
Cosmopolitanism is one of the common threads in many of the articulations which promote global English and defend its use as a medium of instruction. The cosmopolitan argument suggests that access to English, as the world's predominant lingua franca, facilitates socio-economic mobility, fosters world citizenship and enables greater participation in political discourse across linguistic and transnational boundaries (Van Parijs 2011). To cosmopolitan theorists, as well as to many policymakers, teachers and parents, the purported hegemonic cultural and political impacts of English are peripheral to the need for a lingua franca in a globalising world and to provide greater access to the benefits that English can bring (May 2015;Ives 2010). Such an aspirational positioning of English likely contributes to the lack of political will in many African contexts to develop local languages for academic purposes (Simpson 2019, 12;Ouane and Glanz 2010, 4-5).
Diverging from the linguistic imperialism narrative, some scholars suggest that English has already been de-centred from its Anglo-American roots, as non-native speakers far outnumber native ones and numerous localised varieties of English have widened ownership of the world's primary lingua franca (Jenkins 2007). Furthermore, sociolinguists such as Canagarajah (1999), Guilherme (2007) and Imperiale (2017) outline how English can be appropriated and decolonised, enabling 'local communities to bring in their own values, discourses, and interests into the English language' (Canagarajah and Said 2013, 393-394). Several scholars refute the argument that cosmopolitanism represents a universal rootlessness with English as its hegemonic force, upholding instead the notion of a rooted cosmopolitanism (Appiah 1997), reflecting hybrid local and global identities (Nederveen Pieterse 2011), using 'many languages, tongues and grammars' (Beck and Sznaider 2006, 14).
The seemingly incongruous critical and cosmopolitan perspectives on global English frequently intersect in Maurice's story. Likewise, the two frameworks disclose my own conflicting positions, first as an educator who assists teacher transition to EMI, and second, as a researcher attuned to the decolonial agenda of critical scholars. The two positions are reconciled in Maurice's lived experience and in the ethos by which this study has been conducted, which is also persuasively articulated by Pennycook's (1994) anti-deterministic stance: For many who have learned English, the experience has opened up new possibilities of personal gain and communal interaction, and to dismiss their learning and using of English as a colonization is to position them within a new academic imperialism. (69)

Research question
It is in the spirit of decolonising both English and the academic research around it that I have chosen to conduct a narrative inquiry with Maurice and to address the following research question.
How has a Rwandan teacher educator experienced, understood and coped with the country's changing language policy and linguistic landscape as a student, teacher and teacher trainer?

Data collection and analysis
The principle source of the primary data collected for this study is the narrative inquiry, a qualitative research method defined by Clandinin and Caine (2008, 542) as 'an intimate study of individuals' experiences over time and in context', articulated through stories, 'situated and understood within larger cultural, social, and institutional narratives' (542). Unlike data collection methods in which multiple respondents follow a researcher's agenda, the narrative inquiry has a bottomup, collaborative ethos which aims to give the participant an autonomous voice and integrate them into the research process (Atkinson 1998). Narrative research is, therefore, 'more concerned with individual truths than identifying generalisable and repeatable events' (Webster and Mertova 2007, 89).
The single participant in this study is Maurice, whose story begins in the 1960s in Rwanda's rural north, in French-medium primary and secondary schools, continuing on to English-medium university in Tanzania, a scholarship in the UK, and a return to Rwanda as a teacher of English. Maurice and I have worked together since 2018 on two residential teacher development courses in Rwanda, focused on upskilling primary teachers to support their transition to EMI. While doing this fieldwork we developed a trustful personal and professional relationship, a factor which would become crucial for our subsequent research collaboration (Atkins and Duckworth 2019). In the course of our many informal chats in the field, I noted Maurice's flair for telling compelling stories of his remarkable life experiences. Thinking back to these stories inspired me to invite Maurice to collaborate on this narrative research study.
As a result of the participant-centred spirit of the narrative inquiry method, effective collaboration between researcher and informant is especially important, as both are actively involved in creating meaning (Atkinson 1998). In practice, this means deferring 'theoretical assumptions until after the interview and to see then if and what theory emerges from the story' (Atkinson 1998, 68) and inviting the participant to amend transcripts and other research texts to ensure an agreed understanding of the narrative (Clandinin and Caine 2008;Atkinson 1998). I have adhered to these principles throughout my collaboration with Maurice.
The main part of this collaboration took place in two 90-minute online interviews conducted in English, using the Zoom platform. To establish a space in which data is generated rather than extracted, I was guided by Atkinson's (1998) suggestion that 'the less structure a life story interview has, the more effective it will be in achieving the goal of getting the person's own story in the way, form, and style that the individual wants to tell it in' (41). I, therefore, prepared a very loose interview guide with prompts to elicit stories rather than with specific questions, allowing these to emerge from the interviews when I sought clarification or more detail. I produced a 16,000-word transcription of the two interviews which I invited Maurice to review and which he affirmed as true to his life story.
I followed several principles of data analysis which are specific to narrative inquiry research. According to Webster and Mertova (2007, 83), the starting point of analysing a life story is to identify critical events; these are unplanned and unanticipated but life-changing events which have a significant impact on the storyteller. The critical events in Maurice's life story were evident because he signalled their profound impact on his life trajectory and on his worldview. Connelly and Clandinin (1990) note that storytellers highlight critical events in their narrative by reflecting on their significance to their current and future lives.
Additionally, the narrative inquiry method compels the researcher to avoid judging or analysing the storyteller, but instead to identify connections, meanings and patterns in his story (Atkinson 1998, 65). Likewise, it is imperative to first establish the point of view of the storyteller and only thereafter to explore the theoretical perspectives of the researcher (Atkinson 1998). Thus, after several readings of the interview transcriptions, focusing on the critical events, patterns, and connections within Maurice's narrative, I identified linguistic imperialism and cosmopolitanism as the most compelling theoretical perspectives in which to frame the study. It is at the intersection of these two theories that I discuss the tension within Maurice's story between how he negotiates the hegemonic power structures that have bolstered English and EMI in Rwanda and the opportunities that global engagement through English has granted him.

Decolonising the research
An especially appealing principle of narrative inquiry is that research ethics and the decolonisation of the research space are at the heart of the methodology (Clandinin and Caine 2008; Wright 2019). I have therefore pursued an approach to ethics beyond the formalised assurance of confidentiality and informed consent to which Maurice and I agreed at the outset of the research process. Ongoing correspondence with Maurice and maintaining 'thoughtful sensitivity and wide-awakeness' (Clandinin and Caine 2008, 544) in representing his lived experience in my research text have been central to my approach.
I am reflexively aligned with scholars who are 'mindful of contemporary research methodologies' historical impact in the establishment of Colonization' (Wright 2019, 177) and which 'life history methodology has the potential to offset' (178). Enacting the rich tradition of storytelling in African and other colonised lands, narrative inquiry is positioned as an instrument for resisting dominant Eurocentric research paradigms (Brock-Utne 2018; Wright 2019). Furthermore, narrative inquiry methodology asks researchers to seek ways in which to be 'helpful to the participant(s) both during and after the research' (Clandinin and Caine 2008, 543), a request I will continue to honour on Maurice's terms in our ongoing professional relationship.
Decolonising the research context also requires reflection on the part of the researcher. Positioned largely as an outsider who has done occasional work in Rwanda, I am mindful of my own subjective biases, as well as my privileged position as a PhD scholar from the global north and the resulting power imbalance embedded in the collaboration with my Rwandan colleague (Atkins and Duckworth 2019). Additionally, as other international researchers have noted, I am aware that 'in the case of current-day Rwanda, outsiders or expatriate Rwandans have the luxury of speaking critically about Rwanda's current language policy, whereas Rwandans living within the borders of their country do not' (Samuelson and Freedman 2010, 211). As Webster and Mertova (2007, 31) suggest, narrative inquiry enables tacit knowledge to be articulated through stories and creates a more socially acceptable way to voice criticism.
This narrative inquiry, therefore, contributes Maurice's unique personal perspective to the existing body of research on Rwanda's language-in-education policy changes. Following the enactment of the EMI policy in 2008, a number of ethnographic studies with Rwandan teachers in the field were conducted, with data largely collected through participant observation, semi-structured interviews and focus group discussions (Samuelson and Freedman 2010;Pearson 2014;Kagwesage 2013). Such multi-participant research methods tend to take on a normative epistemological position, in which a phenomenon is studied from an outsider's perspective (Wright 2019). In contrast, the narrative inquiry aims at an interpretive epistemology, in which the starting point is the participant's insider perspective.
While single participant narrative research has been conducted by several language education scholars (Benson 2014), none has, to my knowledge, been published about the Rwandan context. This narrative inquiry, therefore, responds to the appeal for the inclusion of local educators like Maurice in the dialogue about English language teaching reform in Rwanda (Simpson 2012).

Maurice's story
The village and beyond: Francophone colonial education Maurice spent his childhood in a remote part of northern Rwanda just as the country was emerging from Belgian colonial rule. Despite the formal end of colonialism in 1962, Maurice's education remained heavily influenced by the surviving colonial structures. Apart from early years in Kinyarwanda-medium, Maurice's primary and secondary education was all in French.
You had to study … in French, but at the same time you had to even speak French outside of class. That was obligatory. So we carried some sort of padlocks, you see what we used to lock doors, so once you heard someone speaking Kinyarwanda you would say 'OK take your padlock' so that was like a punishment now. That's if you spoke Kinyarwanda.
While contemporary scholars discredit such mandated monolingualism (Makalela 2019), Maurice reported an enthusiasm among his peer group for the obligation to use French, 'because we knew by knowing French that would definitely take us quite far'. He elaborates further: People used it to identify themselves as civilised people, speaking very good French. You can imagine that's how things were. Yes, even today when somebody is eloquent in French or in English, definitely, you will consider this person as a learned person. And yeah, people like to appear speaking good languages.
The enduring perception that using a colonial language signals being 'civilised' and 'educated' underscores the persistent hegemonic weight of privileged languages in post-colonial contexts (Pennycook 1998) and the 'unilateral flow of power from centre to periphery' (Canagarajah and Said 2013, 394). Maurice observed this unidirectional flow in both linguistic and cultural assimilation: People in the post-colonial period, especially teachers, used to actually almost try to lookwe even used to comb our hairlike the colonisers, like white people. … So people actually used to identify themselves like people who are enlightened, learned.
Maurice describes the social stratification in access to languages of prestige, a situation which persists in the modern day.
The broader community was purely Kinyarwanda but sometimes of course there was radio and … they used to air some news in French but at that time only rich people could afford a radio, so it wasn't easy to hear French outside of the classroom.
Maurice further reflects on the teaching he experienced and the privilege of accessing education at the time.
Yeah, the teachers were very good at that time. The teachers had gone through colonial education, … had very good French, actually better than the ones … finishing today. So, the education was not really tough but it was also elitist, so … the only people who were selected were the best ones.
Since Maurice's student days access to both primary and secondary education has increased significantly in Rwanda (UNESCO Institute for Statistics 2020). The surge in student numbers and the change from French to English-medium teaching resulted in an enormous demand for English proficient teachers which has been difficult to fulfil (Pearson 2014; UNICEF 2020). The teachers and students who struggle with English today stand in marked contrast to Maurice's depiction of his successful studies in French-medium. It is perhaps this experience that shapes Maurice's current view that EMI can be effective.
Because Maurice held his teachers in such high esteem, he chose to pursue a career in education.
The only people that were educated were teachers so this was my reference and I couldn't think of becoming an engineer, I couldn't think of becoming a doctor, I didn't know what a doctor meant, so I ended up becoming a teacher because this is the people that I saw very early as the enlightened people in my village.
Even more than wanting to become a teacher, Maurice wished to leave village life behind.
I wanted not to become really a rural person, because the rural life, I would see, was quite tough. … I remember when I was told that I had succeeded P6 leaving examination I was very happy. When I got the news, we were cultivating in the field with my parents, and I took my hoe and I threw it away … and said 'bye bye hoe, I will never cultivate again, so I am going to study and definitely have a better life than cultivating.' So that's how people felt about rural life.
Cosmopolitan scholars here might note Maurice's self-responsibility and agency which led him to pursue teacher education with the goal of becoming 'enlightened', 'civilised' and emancipated from 'traditional habits' (Popkewitz 2012, 3). Maurice completed secondary school as a trained primary school teacher. However, he had a different career path in mind: I could see the teachers who taught me in primary school had not actually gone beyond the village life … I knew that if I went to [teach at] primary school in the villages, that would be my end, so I started … looking for ways to … to escape.
Maurice's 'escape' would be by means of earning scholarships to study in Tanzania and the UK.

Scholarships in Tanzania and the UK: from French to Kiswahili to English
At secondary school, Maurice studied English and Kiswahili as subjects, the latter offered as part of a government policy which aimed at closer integration with Rwanda's East African trading partners. As part of this drive for pan-African cooperation Maurice was offered a scholarship to attend university in Tanzania, with the understanding that he would return to Rwanda as a teacher of Kiswahili. Although he was studying to become a Kiswahili teacher, the courses delivered in Englishmedium proved to be a significant challenge: When I went to register for my course, I went to the office of one lecturer who was British. … I said, 'Good morning teacher' so that was a translation from 'Bonjour Monsieur le professeur' and he went very angry. He said, 'I'm not your teacher.' … I didn't know what was the mistake because that was the first time I was now confronting using [English] in my ordinary interaction. … I called him a teacher, that was too low for him, it was too diminishing, so he said, 'I cannot register you in my course.' Although he had formally studied English and Kiswahili in secondary school, Maurice suddenly discovered his lack of proficiency and its potential to jeopardise his academic success in Tanzania. And though he needed to learn both languages he focused on the formal study of English, conscious of its global value, 'because I knew English … had some more potential than Kiswahili for my future'. However, he still felt than his spoken fluency in English fell short: Whatever [the lecturer] said I knew how to transcribe it phonetically, but I could not speak it, that was my biggest issue. … We were studying English usage, linguistics, stylistics, … literature, but I never really learned how to speak English. While many sociolinguists who study African contexts favour the teaching of international languages as subjects, rather than as medium-of-instruction (Brock-Utne 2015, 180), Maurice's experience studying the mechanics of English as opposed to communicating in it have likely pointed him to make the opposite conclusion. His high level of proficiency in French, the result of his French-medium education in Rwanda, likely consolidate his current view that EMI is an overall benefit for Rwanda. I could read English and I could write a very good piece of work. … And to exemplify … how my language was very poor … when I got a scholarship to the UK … in one of the shops I remember I was unable to refer to 'the bag.' So after I bought the thing I said I want 'a bag' but I didn't have the word to mean 'bag' and so when the shop assistant said 'do you want a bag?', I said 'Yes, yes, yes.' I have never met the word 'bag' in my life. … I've got a Masters degree in linguistics in English so that was … a shock for me. So that's how I realised that I never learned how to speak English.
Although his six years of formal study focussed on English, Maurice expected to return to Kigali as a lecturer of Kiswahili at the University of Rwanda. However, the anticipated academic post did not materialise, so he reluctantly took a position teaching English in a rural secondary school. 'It was opposite the Mount Karisimbi, opposite the volcano. In that place it was raining almost every day. I was climbing the mountain on foot. My first work … was a very big challenge for me.' After two years teaching secondary English, Maurice was offered a scholarship to participate in a TESOL Diploma course in the UK, thanks to a senior education officer who saw his potential.
Because that guy was in charge of the professional development in the Ministry of Education, so that was how I got my luck. For me, although it was very hard getting on top of that mountain, for teaching in that school, … today I survive because of that scholarship to the UK. … Everywhere they forget all my Masters and my background. What I did in the UK is the most helpful.
The next pivotal point in Maurice's career as an educator would come later; events in Rwanda would first take a tragic turn.

Returns to Rwanda: careers with English
When he returned to Rwanda in 1991 Maurice was again unable to secure work teaching in Kigali. Instead, he joined a friend in a fruit export business for which his English was a significant asset. However, everything in Rwanda came to a halt in 1994 when the genocide reached its peak, and like millions of others, Maurice fled the country, returning only in 1997. He describes this period as the 'dark years of our lives. … We wasted a lot of time. It was very dark'.
In 1999 Maurice achieved his goal of working in higher education, becoming Dean of Students at a newly formed teacher training college in Kigali. He would stay there until 2010, alternating between leadership and academic roles. His fluency in French and English were instrumental, as the institute was providing language training for student teachers and other public servant trainees to support Rwanda's transition from French to English. Maurice describes the complex and highly politicised sociolinguistic context: So those who had come from Uganda, Tanzania and Kenya … who had English background had to study French and those who had the French background had to study English, at least one year, to support their proficiency. So at this time students were, especially from the Anglophone side, very unhappy to study French. And so I remember even they had a riot … almost in every university in Rwanda.
Maurice perceives the violent rejection of French 'was a kind of reaction against France, against their policies in Africa and it was also political because, actually, those people who are Anglophone are identifying themselves with the power now in place'. There was no similar backlash against English. 'On the side of Francophone people … actually English was very much welcome because we know it's the language of the world, the language of science, technology, so nobody rejected the language.' Research conducted with teachers at the time confirms their 'genuine investment in the policy' (Pearson 2014, 54) to switch to EMI, despite the significant challenges of the transition. While English might have been welcomed for its currency as a cosmopolitan lingua franca, a domestic political force supporting the language was exerting its power. External observers at the time recognised Rwanda's language policy 'as a critical factor in continuing tensions' (Samuelson 2012, 213).
The language change in Rwanda also touched Maurice in a personal way.
I liked [English] as a subject. I studied it, but when this time it was declared as a medium of instruction it's as if something really opened my mind. … I saw English differently now, I found … I loved it, actually. I said, 'how did I know I was being prepared for this change, which I didn't know, which was going to take place?' Because when I was studying English really nobody could ever think that French would ever lose its position. … I felt I was really lucky and fortunate compared to many people, compared to my colleagues at the Institute. 2010 would be a pivotal year for Maurice. He was awarded a grant to attend a prestigious language teaching conference in the UK. Also, overburdened by administrative work, he resigned his position at the training college despite having no other immediate work prospects. Much to his surprise, an offer came to become a teacher with a newly opened international language centre. I saw this email and I felt weepy really. I remember when I answered … the country director … my letter was emotional. … That was something very very positive in my life and as I started loving this language, I started identifying myself with … world users of English or even felt that I was beyond Rwandan. … If I could speak English then I am an English person, I am British, I am an American, I'm an Australian, I'm a person. So, wherever I would go in those places I would adjust, I could easily fit in that society. So that came with a very strong feeling about English and it really changed my life actually.
Maurice's self-identification as an Anglophone global citizen was never expressed in opposition to his identity as Rwandan. While the dichotomy between cosmopolitan and local identity is a frequent topic of debate among scholars (Nussbaum 1994;Levy 2012), Maurice's cosmopolitan identity is clearly 'rooted' (Appiah 1997) in a Rwandan and pan-African context. He is proud that Kinyarwanda 'enjoys the privilege of being spoken throughout the country' as the sole national language and expresses his appreciation of Kiswahili as a regional lingua franca, which shares Bantu roots with Kinyarwanda and other African languages.
Above all, however, Maurice repeatedly affirms his love of English and embraces its role as the principal lingua franca of a globalising world. His affection for the language extends beyond its instrumental role, to its register and emotive pragmatics: 'You feel English is so sweet compared with some languages. … When people talk to you in English you feel better than in some languages where you feel like there is a confrontation.' Indeed, researchers of English pragmatics note the tendency to avoid imperatives and personal criticisms in the 'Anglo script of "pleasant interaction"' (Wierzbicka 2010, 58), though critical scholars maintain that justifying the global spread of English on the basis of its perceived virtues reinforces an Anglocentric neo-colonial world view (Pennycook 1998).
However, given the opportunities English has brought him, it is no surprise that Maurice sees little wrong with how English has come to dominate Rwandan education and the global linguistic landscape: It's going all over the world and eventually we'll all become Anglophones, I think. Some languages will lose their status and it doesn't matter because what we need is something useful as a tool which we can use everywhere and which we can use to earn our life, to earn our bread. So personally, I don't see any bad thing.
Maurice's sentiments echo the cosmopolitan position that extol the instrumental value of English as a lingua franca while absolving it of its hegemonic socio-political impacts (House 2003;Ives 2010). Reflecting on the contentious recent decision to extend EMI to early primary classes (Edwards 2019), Maurice suggests: The policy for me is not that bad, but the ability to implement this policy is what caused the problem, because in primary school the classes are overloaded; [for] the teachers it's very very difficult to have … to [teach] in this language at this level.
Maurice expresses optimism about these challenges being overcome with teacher development projects organised through international partnerships. He reflects on his own recent contribution to this work: We were teaching … global citizenship, … preparing people to adjust meeting others … without being shocked, without shocking others, intercultural practices. … But later on it developed into core skills, which we're supporting teachers to improve the performance of their children, through these twenty-first century skills which are needed out … in the labour market, like critical thinking and problem solving, communication and collaboration, digital literacy, student leadership and personal development.
In addition to the cosmopolitan cultural hybridisation of the local and the global (Nederveen Pieterse 2011), Maurice highlights the link between English and the needs of the market. Critical applied linguists have frequently noted the convergence of the global spread of English, free market ideology and digital technology as a neo-colonial project (Pennycook 1994).
Maurice finishes the interview with a reflection on UK support for Rwanda's language transition: No single country would really go through this kind of challenge without support, especially from … those who know how to teach even this language, those who have developed materials. So, I was very glad from the support of Britain to this change in Rwanda.
Despite advocating for multilingual primary education and supporting a principled transition to EMI (Simpson 2019), such UK bilateral aid which bolsters English is typically depicted as an investment in Britain's soft power (MacDonald 2018) thereby consolidating the argument of critical scholars who decry such aid as linguistic imperialism (Phillipson 1992).

Mitigating linguistic imperialism
Maurice is keenly aware of the power of English but does not problematise the potential neo-colonial and hegemonic attributes which critics attach to it. Given that educators in Rwanda lack the agency to critique EMI policy, Maurice focuses on overcoming challenges rather than lamenting government decisions. Importantly, Maurice's enthusiasm and advocacy for English in Rwandan education are deeply embedded in his lived experience as a student and an educator. His studies in French-medium demonstrate to him the potential for a foreign language medium education to effectively lead to both language acquisition and subject content learning, even when that language is not used in the local community. Maurice's later success in gaining international scholarships and rewarding employment provides him evidence that English can offer social mobility for someone from a modest rural background. However, Maurice acknowledges that he 'was different from other kids' and understands that he was especially fortunate with scholarship and work opportunities. He also sees that providing universal education in an international language in the modern day is far more cumbersome than doing so for a privileged group, as in the past. Indeed the lack of suitably trained English-proficient teachers in rural areas has coincided with poor learning outcomes for many rural students (Milligan, Clegg, and Tikly 2016;Simpson and Muvunyi 2012). There is thus little evidence that the vast divide between the English-speaking urban elites and the rural population is waning (Reyntjens 2013). It is also unclear whether extending quality English education beyond Rwanda's urban elite is a challenge which can be met and thereby whether English is a relevant priority for educating the rural poor. Critical scholars like Bruthiaux (2002), call 'the global spread of English … a sideshow compared with the issue of basic economic development and poverty reduction' (290) and investment in English education for the rural poor 'an outlandish irrelevance' (292).
Nevertheless, Maurice is optimistic that internationally funded technical assistance projects are helping to overcome some of the obstacles that EMI policy has put up. Many initiatives are undoubtedly doing good work, including those in which Maurice is involved, such as teacher development, educational leadership and building systemic capacity (Education Development Trust 2020). Much of this support has focused on mentoring and supporting educators in remote and economically marginalised regions. One notable pilot project trained teachers in language supportive, learner-centred pedagogy, underpinned by bilingual teachers' guides; Anglocentric reading texts were replaced with linguistically graded, bilingual and culturally relevant alternatives, while Kinyarwanda was integrated into scaffolding learning activities (Milligan, Clegg, and Tikly 2016, 337-338). Maurice and others on the ground favour scaling up initiatives like this which mitigate some of the deficiencies associated with EMI (Simpson 2019), especially if the mother tongue-based multilingual education model favoured by scholars is off the table.
However, support for such work is threatened by cuts to international aid budgets due to economic fallouts from the COVID-19 pandemic (Worely 2020) as well as from heightening concerns about Rwanda's human rights practices (Human Rights Watch 2020). Consequently, it is imperative for Rwandan education authorities to nurture the expertise of local educators like Maurice to sustain the reform taking place in Rwanda's schools independently of donor funding. Together with a policy shift toward officially endorsing multilingualism in classrooms, such reforms would enable the Rwandan teachers and students to more successfully navigate and appropriate EMI, thereby mitigating the hazards attributed to linguistic imperialism.

Rooted cosmopolitanism
In his stories, Maurice describes a common perception which throughout his lived experience has correlated the knowledge of English (and French) with cosmopolitan notions such as being enlightened, educated and civilised. Through his language skills Maurice acquired the cultural and social capital (Bourdieu 2018) which enabled his escape from the toil of the village and his opportunities to interact with the world through international scholarships and employment. Though such a trajectory might be interpreted through binary codes such as rural-urban, local-global or tribal-transnational, Levy (2012) instead points out that, … the analytic approach to cosmopolitanization refers to a process in which universalism and particularism are no longer exclusive 'either-or' categories but instead a coexisting pair. This is premised on the notion that meaningful identities are predicated on particular attachments, as one's identity is always embedded in the story of the communities from which identities are constructed. (300) Although Maurice places a high value on global citizenship, his cosmopolitan identity is symbiotic with his Rwandan one. Indeed, as Guilherme (2007, 81) suggests, 'being an active cosmopolitan citizen does not start only beyond national borders'. In this light, Maurice takes extra pride in his work on international projects which aim to embed citizenship and cross-cultural communication skills in the Rwandan education system. Though it is a sentiment rarely spoken, Rwandan educators must surely hope that success in developing such cosmopolitan skills reduces the likelihood of sliding back into what Maurice calls 'the dark times'.
Building on Maurice's lived experience, education for peace in Rwanda would most effectively take the shape of a rooted cosmopolitanism (Appiah 1997), 'which eschews abstract concepts for daily practices of tolerance at whatever scale is meaningful at that particular time and place' (Mitchell 2007, 710). According to Guilherme (2007), this type of localised education for cosmopolitan citizenship 'carries the responsibility of simultaneously promoting a shared identity, the appreciation of diversity, the respect for difference, the pride in one's own identifications and the commitment to taking action in the interest of the weaker members of our communities' (82).
Whatever direction language-in-education policy takes, English must be seen as only part of the process of developing cosmopolitan citizenship. Some scholars contend that to be cosmopolitan is to be multilingual (Skutnabb-Kangas 2000); others suggest notions of individual agency, solidarity and reciprocity (Guilherme 2007, 82). These qualities are all embedded in Maurice's life story and must be cultivated among Rwanda's teachers and students if they are to build a socially just and peaceful society.

Conclusion
The cosmopolitan worldview which Maurice projects is inextricably connected to his multilingual education and his resolve to leave his village, gain international scholarships and build a career as an educator and teacher of English. Maurice's story demonstrates that international languages like English need not be the sole domain of the urban globalised elite and that a cosmopolitan identity can transcend the assumed binaries of global and local, rich and poor. Likewise, Maurice's evident acumen as a global and a local citizen support notions that transnational cultures are hybridised (Nederveen Pieterse 2011), and local ones resilient to colonial language policies (Vaish 2005).
While Maurice does not explicitly seek to 'decolonise English', he instinctively appropriates it for his own local use, like when he effortlessly blends English phrases into the translingual mix in chats with colleagues, thereby practicing ubuntu translanguaging and thus celebrating the interdependence of languages and the people who speak them (Brock-Utne 2018, 731;Makalela 2019, 240). The spirit of appropriating English into a multilingual education more relevant to the Rwandan context uncovers a potential alternative to the government's drive for more EMI. Approaching English as part of a translingual mix in schools could accelerate the process of decentring the language from its colonial baggage and the politicised elite in Rwanda.
By recounting the story of Maurice's successes through English, this study seeks neither to defend EMI policy nor to challenge the well-researched merits of educational provision in local languages. It aims instead to reflect on the lived experience of an educator who has acquired the agency to effectively navigate imposed language-in-education policies. Stories of how the decisions made by Rwanda's political elite impact those who occupy classrooms are largely absent from policy deliberation. Indeed, scholars on the ground have called for the inclusion of 'pedagogues, and practitioners in dialogue which helps to better inform and enhance the quality of English language teaching reform' (Simpson 2012, 127). To this end, there are countless stories waiting to be heard.