Fear and shame: students’ experiences in English-medium secondary classrooms in Tanzania

ABSTRACT

In contexts where schooling is delivered through a language of instruction (LoI) that is unfamiliar to learners, it has repeatedly been observed that students are reluctant to speak, passive, or even silent. This is commonly attributed to lack of understanding of the language in use in the classroom. Although this is, undoubtedly, an important part of the explanation, this paper presents findings from a thematic analysis of lesson observations, interviews, and ethnographic field-notes from two secondary schools in Tanzania to argue that we should also be paying much closer attention to the socio-emotional environment. Specifically, this paper demonstrates the prevalence of feelings of fear and shame in students’ experiences of learning and how these emotions contributed to students’ silence. However, it also shows that the impact of these emotions was not the same for all students. Informal classroom ‘rules’ were found to govern how different groups of students were expected to behave in response to questions from teachers, meaning that the risks of speaking English were greater for some students than others. As such, the prevalence of fear and shame is argued to compound existing inequalities amongst students, thus hindering progress towards equitable, inclusive and safe education for all.

KEYWORDS:

• Fear
• shame
• language of instruction
• secondary education
• Tanzania

Introduction

In contexts where schooling is delivered through an unfamiliar language of instruction (LoI), it has been widely observed that this presents significant challenges for students (and teachers) (Afitska et al. 2013; Alidou et al. 2006; Brock-Utne, Desai, and Qorro 2003; Milligan, Desai, and Benson 2020). Researchers have commonly found that the use of an unfamiliar LoI results in dramatically limiting classroom talk, either to rote, chorus responses, single word answers, or silence. Many of these examples come from countries in Sub-Saharan Africa, including Ghana (Opoku-Amankwa 2009), Botswana (Mokgwathi and Webb 2013), Kenya (Kiramba 2018) and Tanzania (Brock-Utne 2007; Vuzo 2010). But the use of an unfamiliar LoI is not limited to African contexts and similar observations have been made in India (Verma 2017) and Nepal (Sah and Karki 2020).

Most often, lack of understanding of the medium of instruction is pointed to as the key explanation for low student participation (Kamwendo 2016; Rea-Dickens and Yu 2013; Samuelson and Freedman 2010). This paper does not deny the importance of the challenge of understanding when learning in an unfamiliar language. However, it argues that, if we want to develop a fuller picture of students’ learning experiences and reluctance to speak, we must expand the focus of research relating to LoI and classroom participation, in particular to consider the socio-emotional context of learning. Using data from a study of students’ experiences and negotiations of language in two Tanzanian secondary schools, this paper highlights the prevalence and pervasive impact of students’ feelings of fear and shame.

The country context of this research is Tanzania where more than 100 languages are in use in communities (Muzale and Rugemalira 2008). The LoI in all seven years of government primary schooling is the national language and lingua franca, Kiswahili. This shifts to English at the start of secondary schooling, for all subjects except Kiswahili. More broadly, this paper addresses schooling in Sub-Saharan Africa (SSA) and other postcolonial contexts where a dominant, global language is in use as the LoI across multiple subjects and stages of education. In these contexts, exploration of students’ emotions in the learning process has, thus far, been largely absent. I acknowledge that, beyond these contexts, in Europe, North America and parts of Asia, there is a significant and fast-growing body of work about emotions and language learning. Some of this work is acknowledged here in limited ways, but there are significant differences between specific language learning classrooms and contexts where an unfamiliar language is used as the LoI. These differences require caution against assumptions of transferability of this research in its current form. However, the growing international interest in this field only adds strength to the argument that student emotions and the socio-emotional context of learning should also be a focus for research in education in SSA.

Literature review

LoI and classroom participation

It has been estimated that 40% of the world’s population does not have access to schooling in a language that is in regular use in their communities and that they understand (UNESCO 2016; Walter and Benson 2012, 282). In SSA, this may be as high as 80% (World Bank 2021, 9). The use of a dominant, but unfamiliar, LoI is particularly common in countries in the Global South, and is driven by a combination of historical-political and economic forces, alongside the challenges of providing universal education in contexts of high linguistic diversity (Milligan 2020; Tikly 2016). However, it is widely evidenced that learning in an unfamiliar language, particularly in low-income contexts, is a barrier to quality, equitable education and is detrimental to both student outcomes and their experience of schooling (Alidou et al. 2006; Brock-Utne 2012; Milligan, Desai, and Benson 2020; Trudell 2016a).

In Eastern and Southern Africa, many countries have language-in-education policies that recognise the value of learning in the mother tongue, or another familiar language, at least in the early years of schooling (Trudell 2016b). However, even when students begin their education in a familiar language, the vast majority of these countries transition to the use of a European language, most often English, before students have developed sufficient proficiency to use it as an effective medium for learning (Brock-Utne 2014; Heugh 2006). This means that teaching and learning becomes a dual process as students are trying to engage with subject content at the same time as striving to acquire the LoI (Barrett and Bainton 2016). One result of this is that, although official language policies and curricula assume familiarity with English, and unproblematic transition to its use as the monolingual LoI, the reality is frequently multilingual as teachers translate into students’ known languages to help them to understand (Westbrook et al. 2022). However, teachers in Africa are rarely trained to teach multilingually or to strategically support learners who are transitioning between languages (Afitska et al. 2013; Alidou and Brock-Utne 2011). Instead, these practices are often reactive rather than purposefully planned, and are viewed as ‘illicit’ but necessary to compensate for students’ linguistic ‘deficits’ (Kiramba 2019; Li and Martin 2009; Probyn 2015). The use of familiar languages is also often controlled by the teacher, and students are offered little support to express their ideas in the official LoI (Arthur 1996; Mokgwathi and Webb 2013; Simasiku 2016).

Under these circumstances, it is unsurprising that studies based on classroom observation at all levels of education have found that student answers to teachers’ questions are predominantly limited to rote, chorus responses, ‘safe-talk’, single word answers, ‘inaudible mumblings’, or silence (Brock-Utne 2007; Issa Mohamed and Banda 2008, 104; McKinney et al. 2015; Ngwaru 2011; Vuzo 2010; Webb and Mkongo 2013). This is consistently linked to the LoI, which by secondary level in nearly all countries in SSA is a dominant, European language, most commonly English. Commenting on observations from a study of a 4th grade class in Kenya, at the point where the LoI policy shifts to monolingual use of English, Kiramba (2018, 303) notes that ‘the main effect … is silencing’. The most prominent explanation offered by researchers for students’ limited participation and silence is poor understanding of the unfamiliar LoI, an explanation that is strongly supported by studies that compare interaction patterns in English-medium lessons and lessons taught in a familiar language (Afitska et al. 2013; Brock-Utne 2007; Mwinsheikhe 2009; Rea-Dickens and Yu 2013). However, this paper argues that lack of understanding is not a straight-forward explanation. If we do not look more fully at students’ classroom experiences, we risk misrepresenting some students’ levels of understanding and may neglect the important influence of the individual and classroom socio-emotional context.

The socio-emotional context of learning

The socio-emotional context of learning using an unfamiliar LoI is rarely discussed in the language-in-education literature in SSA. Although a number of papers mention fear or anxiety in relation to low classroom participation or the requirement to speak English, it is usually only a passing observation, and the implications are not explored (Early and Norton 2014; Mokgwathi and Webb 2013; Mokibelo 2016; Rea-Dickens and Yu 2013; Sibomana 2022; William and Ndabakurane 2017). Similarly, research considering the use of multiple languages on student learning has sometimes briefly mentioned a positive emotional effect, but without exploring this in more detail (Benson 2002; Gabrieli, Sane, and Alphonce 2018; Shank Lauwo 2020). There is research that highlights the prevalence of fear as part of the broader school environment, often referencing the continued use of corporal punishment or other forms of humiliation in schooling (Joyce-Gibbons et al. 2018; Mligo and Mshana 2018; Ouane and Glanz 2011). Several language commentators point specifically to both historical and contemporary use of fear and shaming as punishments for speaking home languages at school (Kiramba 2018; Ngūgī wa Thiong’o 1986; Rubagumya 2003). Vavrus (2002, 377) recounts observing a student in the Kilimanjaro region of Tanzania being forced to wear a vest with the words, ‘Shame Upon Me’, for speaking Kiswahili in class, in violation of the school’s ‘English Only’ policy. However, apart from identifying strategies for shaming students for the use of unsanctioned languages, students’ feelings and experiences of shame relating to language are largely absent from considerations of language-in-education in this context.

There are a few exceptions to the overall observations made above. For example, Opoku-Amankwa (2009) and Madonsela (2015) draw on their experience of monolingual English instruction in Ghana and South Africa, respectively, and highlight students’ emotions in relation to the unfamiliar LoI. Both authors point to discussions of the concept of language anxiety that have been popular in the global literature about Second Language Acquisition (SLA) since the 1980s (Gkonou, Daubney, and Dewaele 2017; Horwitz, Horwitz, and Cope 1986; Tsui 1996). Writing about his classroom observation-based study in a primary 4 classroom in Ghana, Opoku-Amankwa (2009, 126) identifies language anxiety and self-esteem as ‘the two factors which the pupils consistently alluded to as conditions stalling their participation in classroom communication’. This finding is certainly in line with the assertion of SLA researchers that language anxiety is ‘the affective factor that most persuasively obstructs the learning process’ (Arnold and Brown 1999, 8). Language anxiety is defined as distinct from other forms of anxiety for having features that are unique to language learning situations (Horwitz 2010). Pavlenko writes that language anxiety:

stems from perceived threats to the student’s sense of security or self-esteem, and from fear of failure, fear of negative evaluation, and apprehensions about communicating in a language in which one may appear incompetent or ridiculous. (Pavlenko 2005, 33)

From this definition, it seems that Madonsela’s (2015) argument that experiences of language anxiety are also relevant in the South African context makes sense. Fear of making mistakes and being negatively judged by others are completely understandable in any context where an unfamiliar language is used. However, I cannot agree with Madonsela’s (2015, 448) assertion that ‘the South African school situation is no different from the challenges of language learning in the rest of the world’. Firstly, language anxiety research has focused primarily on the isolated process of language learning, rather than on the use of an unfamiliar language as the medium of instruction across multiple subjects and entire stages of education. Moreover, although some form of language anxiety may be assumed to be a common emotion in language learning around the world, not all language learning is combined with a socio-cultural context where the relationship between teachers and students is strongly hierarchical and where fear and shame are regularly used as forms of classroom control, like in many schools in SSA.

Explorations of shame in relation to language-in-education are also primarily focused in the Global North and Asia, and either in language learning classrooms or in the context of minority language students who experience schooling in the majority language of the country where they are resident (Badwan 2021; Cook 2006; Galmiche 2017; Skutnabb-Kangas and Cummins 1988; Teimouri 2018; Wang 2016). There are a few notable exceptions. Perhaps the most relevant to the contexts discussed in this paper is Speciale’s (2020) consideration of shame and language in a bilingual school in Senegal. She points to the association of shame with local language, highlighting both an active shaming practice that helped to reinforce the school’s official language policies, and the association of English with an educated identity that was positioned as superior to those who used local languages. This delegitimisation of identity and corresponding feelings of shame is also mentioned by Tamim (2013, 2021) in the context of students in Pakistan who have attended government, Urdu-medium schooling, but find themselves at university alongside peers who were educated in a parallel, private, English-medium system. Also of interest are Bartlett’s (2007) and Liyanage and Canagarajah’s (2019) analyses of shame in English language learning in Brazil and Kiribati, respectively. Bartlett also points to the shame attached to local language use, including some compelling examples of learners being frozen by fear and shame in the classroom while struggling to express themselves in English. Liyanage and Canagarajah, however, challenge assumptions that shame in relation to language is always negative, presenting a perspective from the local community where language shaming, this time in relation to those who used English rather than the local language, was viewed as a positive strategy for ensuring community cohesion. Even across these four examples, it is clear that shame plays much more than just a passing role in individuals’ and communities’ experiences of language and learning. Thus, it seems important that language-in-education research in SSA should pay more explicit attention to experiences of fear and shame.

Methodology

The findings presented in this paper are drawn from an ethnographic study that focused on students’ negotiations of language in two secondary schools, one urban and one rural, in the Morogoro Region of Tanzania. The methodological approach was designed to respond to two concerns about existing LoI research. Firstly, there appeared to be a tension between the findings of observational studies that (as discussed above) overwhelmingly highlight the challenges of understanding and low participation in classrooms where an unfamiliar LoI is used, and language attitudes research that reports clear preferences for the continued use of English as LoI, despite these challenges (Kinyaduka and Kiwara 2013; Senkoro 2005). Secondly, I was concerned that students’ voices were largely absent from the existing LoI literature, resulting in limited opportunities for classroom observation and language survey findings to be understood in the broader context of educational and life circumstances and experiences (see Adamson and Milligan In preparation). As such, over the 7 months that were spent in the two schools, I incorporated a wide range of data collection methods, aiming to create space for students to explain both their classroom experiences and how their schooling realities related to their broader education and language-related aspirations. These methods included: classroom and participant observation, informal and semi-structured interviewing, group activity-based workshops and the regular writing of fieldnotes. In addition to activities that I designed as the lead researcher, I worked alongside a group of 10 Student Researchers who designed and conducted their own interviews, helped facilitate workshops and acted as invaluable ‘critical friends’, helping to contextualise findings and commenting on my emerging analyses and interpretations. In the field, I wrote weekly analytical memos that informed both the design of subsequent data generation activities and the later thematic analysis of the complete dataset.

This paper primarily draws from data generated through 51 classroom observations, informal conversations between lessons that were recorded in fieldnotes, and 31 semi-structured group interviews involving 146 students. The relationship between language and feelings of fear and shame was neither an original focus, nor an expected finding of this study. Rather, across the lesson observations, I was struck by the frequency with which teachers’ questions were met with silence and how often I had noted examples of students displaying body-language or behaviours that suggested discomfort and reluctance to speak. My initial assumption was that these simply demonstrated that students were unable to understand the language the teacher had used. However, as I began to follow up on these observed experiences, talking to learners informally between lessons and in interviews, many students challenged this assumption. Instead, they demonstrated a much more nuanced definition of the notion of what it means to ‘know’ and understand language,¹ and described an influential socioemotional context to learning that I would not have accessed through observation alone. In addition to pointing to feelings of fear and shame to explain silences and reluctance to participate, students talked about their experiences of negative emotions in response to interview questions about what they liked and disliked about school, and if they experienced challenges in their learning.

This appreciation of the prevalence of fear and shame in students’ experiences of schooling acted as a turning point in the research. Firstly, it enabled me to explain some of the tensions between students’ expressed language-related aspirations and their observed behaviour as immediate negative emotions in the present prevented students from participating in ways that they knew would support their future goals (Adamson Forthcoming). And secondly, students’ descriptions of their feelings related to language helped me to identify emotions relating to my own experiences as a first-language English speaker learning and using Kiswahili in the context of research (Adamson 2020a).

The study context

The urban school was well-established, with more than 1500 students across both lower and advanced secondary levels (Forms 1–6). There was considerable socio-economic diversity across the student body. Some students had parents in professional jobs and many had siblings who had already completed secondary education. However, other students were living in more disadvantaged situations or were boarding in town to allow them to attend this well-respected school. Students’ levels of English at the point of entry into secondary school varied dramatically as some students had attended private, English-medium primary schools, while the majority had been educated in government, Kiswahili-medium primary schools.² Urban students reported differing levels of access to English-language support outside of school.

The rural school was situated 35 km from town, but only 2 km from a main road. The school was less than 10 years old and had 600 students enrolled in Forms 1–4. The majority of these came from farming families, and many were the first in their family to attend secondary school. Although there were a small number of students who spoke English more confidently, the level of English at the start of Form 1 was more even at the rural school. All students had attended Kiswahili-medium primary schools and very few reported having access to English-language support outside of school.

Ethics

In addition to following official procedures for obtaining research permits with the Tanzanian Commission for Science and Technology (COSTECH) and the Tanzanian Immigration Service, I sought formal permissions from Education officials at the regional and district levels. Once in school, research ethics remained a constant consideration. I placed letters explaining my research on school notice boards in both English and Kiswahili and visited all classes at the start of my time in school to explain and answer questions about my research. However, research ethics and issues of anonymity, confidentiality and informed consent were ongoing concerns and I regularly reiterated to students and staff that participation in research-related activities was completely optional.³

Findings: fear and shame relating to speaking English

The following description is taken from observation notes from a Form Two English lesson at the urban school:

During today’s English lesson Alisha was selected to come up and present the sentences they had been working on in groups (c. every 3rd–4th student was selected so 20 students total). She was probably the 11th or 12th student to come out (so she had seen plenty of examples). She made her reluctance clear as she shuffled her shoes heavily to the front of the classroom. But she absolutely refused to face the class – she stood facing the door. It did not initially look like she had brought any notes with her, although I later saw that she had a piece of paper screwed up under her headscarf.

The teacher kept trying to get her to turn around and speak (I think that she was trying to encourage her, but her tone, as usual, was quite stern and she was quickly getting frustrated). After a few minutes of the teacher trying, but Alisha refusing to turn around and speak, the teacher sent her out of the classroom while she continued with the next few students. When she was called back in, she still refused to face the class. The teacher physically took hold of her arms and tried to turn her around, but Alisha absolutely refused to change the direction of her feet, which stayed pointing out of the door. Eventually, the teacher gave up and Alisha returned to her seat.

At the end of the lesson, after the teacher had left, I asked Alisha why she hadn’t just read her sentences. I knew that she had correct sentences because I had seen her group give their answers to Aiko (a student with good English) to check and she had even added more examples. I asked, ‘Kwa nini hukusoma tu? Ulikuwa na sentensi nzuri’ [Why didn’t you just read them? You had good sentences.]. Alisha said nothing, but she turned away from me and folded her entire body over as if she were trying to fit her head under the desk. I asked, ‘kwa nini unaogopa?’ [Why are you afraid?] and one of her friends replied, ‘anajiona aibu’. [She feels shame]

This example has been chosen to open this findings section because Alisha’s determination to remain silent is a particularly affecting example of student reluctance to speak English in the classroom. But it is by no means the only example. The notes from 51 formal observations that were conducted as part of the study from which this paper is drawn include 5 more examples of students coming into conflict with the teacher over their refusal to speak and 27 examples where a student’s body language visibly demonstrates their discomfort about contributing. In addition, there were countless moments when teachers asked questions and were met with silence.

In this example, Alisha’s friend offered an explanation of her behaviour, saying ‘anajiona aibu’. This Kiswahili phrase could translate either as ‘she feels shy’ or ‘she feels ashamed’, but both the strength of Alisha’s reaction and students’ broader discussions of similar experiences led me to interpret this as an expression of shame. More often, the words that students used to describe emotions that discouraged them from speaking translated as fear. For example, a male student in Form Four at the rural school lamented: ‘So they break a person’s heart … even if you are interested in speaking, you shouldn’t speak … because you are afraid’. In this paper, the emotions of fear and shame are treated together because students repeatedly explained that their fear was predominantly related to the anticipation of shame. A female student in Form Two at the urban school connected these emotions together, stating that, if students make mistakes or say that they don’t know the answers: ‘you will be laughed at, which means we are afraid of the shame … fear, again’. Fear and shame were also frequently connected to feelings of low self-esteem and confidence. One male student in his first year at the rural school explained:

A person is afraid most of the time … A person can have their answer and they go to say it but they fail, they tremble. They are not confident with their answer because they don’t believe in themselves.

The examples in this paper concentrate on verbal participation, partly because this was most easily observed, but also because fear appeared to have a particularly strong influence on tasks that involved speaking English, most likely because this was a public activity that carried the highest risk of exposure, and thus shame and humiliation.

Beyond the understanding/not understanding binary

The focus on fear and shame is not intended to detract from the problem of understanding in classrooms where the LoI is unfamiliar, but rather to develop a richer appreciation of the impact of difficulties with understanding on students’ experiences of learning. Students in this study explained that fear and shame prevented them from even telling the teacher when they had not understood something. For example, a female Form One student at the urban school explained why so many students responded positively to teachers’ questions like ‘do you understand?’ and ‘are we together?’, even when they had not understood:

If the teacher asks if they have understood, there must be those who have not understood, but the majority … many of them say ‘yes’. That voice makes you afraid to say that you haven’t understood … you feel ashamed to say.

There is also a danger that assuming that silence simply indicates lack of understanding may lead to misinterpretation. Although students admitted that there were frequently times when understanding the LoI was a barrier and a struggle, there were also regular examples when students insisted that they had understood the teacher, but still chose not to respond. One male student in Form Two at the urban school explained: ‘It’s not that they don’t know … but if they are asked … that fear … you find it’s difficult’. It also became clear when talking to students that they did not necessarily draw a binary distinction between understanding and not understanding. For example, a female student in Form Four at the rural school explained: ‘There are those questions that you don’t understand completely … you know the language, but you don’t know it completely’. In fact, several students differentiated between the ability to ‘understand’ and the ability to ‘respond to questions’. Understanding what the teacher had said was often enabled through translation into Kiswahili, but student responses, in the vast majority of cases, were restricted to English. An example of this can be seen in this observation of a Form Two English lesson at the rural school:

The boy at the front who has been an active participant in this lesson is the only one to put up his hand. He asks (in English) if he can answer in Kiswahili. The teacher tells him to try English, but the student replies (in English) that he cannot.

At this point the teacher turned his attention to another student, without either allowing the first to answer in Kiswahili, or supporting him to answer in English. These examples demonstrate that the difficulties of understanding an unfamiliar LoI and the subsequent effects on student participation should not be underestimated, but lack of understanding is not a straightforward explanation, nor does it necessarily explain the full picture of students’ experiences.

Speaking English in a classroom atmosphere of fear and shame

The use of English as the LoI was a powerful source of fear. Students were particularly worried about making mistakes in English, the chances of which were high for most students who were quite unfamiliar with the language when starting secondary school. Students often pointed to the fear of making mistakes, and talked about the negative reactions of others. The following discussion between female Form One students at the rural school offers examples of both:

Student 1:

They [students] are afraid to try. If you make a mistake they laugh at you.

Student 2:

(Clicking sound of agreement) You’re right! It’s completely true … 

Student 3:

If a person gets it wrong even just a little they are laughed at.

Student 2:

They laugh … or they hiss … 

Student 1:

Even if you just make a little mistake, they look at you and think, ‘they are retarded … the teacher has already said it, has already explained the spelling and the way of pronouncing it and still they can’t get it right'.

This discussion demonstrates that, in an unfamiliar language, students are not only faced with the possibility that the subject content of their answers will be right or wrong, but that there are a range of different types of mistakes that could be made, including vocabulary, grammar, spelling and pronunciation. The fact that there are so many different elements to get right increases the chances of making a mistake, and thus students’ fear.

The excerpt above also offers a variety of different reactions from other students that constitute negative evaluations, including laughing, hissing, and judgements of intelligence. Although the likelihood of making mistakes may have been particularly high for these students in their first year of secondary school, many older students reported similar fears. One male student in Form Four and approaching the end of his lower secondary education at the rural school reported that this fear of being laughed at persisted: ‘ … because it’s language itself … the words in English. A person is afraid of being laughed at. If they put their hand up and if they ask a question, their classmates will laugh at them’. That these fears are still experienced by some students in their fourth and final year of lower secondary education challenges the assumption that language difficulties will naturally be overcome as students’ progress through schooling.

Students talked about the practice of laughing at mistakes as an inevitable and unchallengeable norm. When I questioned why students laugh at their classmates when they were all learning an unfamiliar language, one female student in Form Two at the urban school explained: ‘It’s … the habit of the person. They have already given themselves the habit that if someone makes a mistake, it’s necessary that I laugh at them’. Students explained that this laughter was also present at primary school, but was more pervasive at secondary level because the requirement to use English greatly increased the risk of making a mistake. However, a female student in Form One at the rural school explained that laughter did not necessarily mean that the other students knew the correct answer:

Even those who laugh, most of them don’t know it. If they see their classmates are laughing, they laugh too … So the person who doesn’t know laughs with gusto until you hear their voice.

Joining in the laughter may itself be motivated by fear of being exposed for not knowing the answer and thus experiencing public humiliation and shame.

The threat of being laughed at contributed to a classroom atmosphere where the fear of shame was a constant spectre that discouraged participation. This was further exacerbated by teaching practices where fear and shaming were seen as methods of control and motivation. In this study, the systematic use of corporal punishment as a response to poor test scores or inability to complete classwork was regularly observed. One particularly memorable example involved a Form One class at the urban school who were all made to individually repeat the sentence, ‘at quarter past six in the morning’ until the teacher was happy with their pronunciation. In the observation notes I wrote:

The teacher got progressively more frustrated with the students, raising her voice … She made some students repeat before moving onto the next. Sometimes students giggled when someone repeated a mistake and some of the teacher’s comments seemed to be targeted at humiliating the student. She did not say anything after each student had read their line, just indicated the next person … One student struggled, despite repeating several times. She told him to come to the front and kneel down in front of the board … The teacher called students to the front, one after the other. She told them that if they ‘shindwa’ [fail] they will kneel by the board like the other student ‘na nitawachapa’ [and I will strike you].

In this example, corporal punishment is combined with the teacher consciously humiliating students in front of their peers. On other occasions, teachers in this study were observed specifically mocking students for difficulties with English, for example in a Form Two English lesson at the rural school when the teacher responded to a student’s silence, by laughing and exclaiming: ‘Shida kwenda Ulaya!’ (You will have a problem going to Europe!) Teachers explained that statements like these were intended to ‘force’ students to participate more or work harder. But students’ descriptions of the impact of fear and shame suggest that they may have had the opposite effect.

Different experiences of fear and shame

Without specific prompting, 21 of the 31 group interviews I conducted as part of this study included discussions about fear and shame, while 17 groups identified lack of confidence as a challenge. Examples of reluctance to speak and silences were also commonly observed across all 51 of the lesson observations. However, it was also noticeable that there were some students who were regular contributors to discussion and appeared to be less affected by negative emotions. Although patterns were not necessarily applicable to all students at all times, there did seem to be different norms and expectations around classroom participation, relating in particular to gender and students’ level of English.

There were plenty of examples in interviews of boys expressing fear of being laughed at for making mistakes. But in some cases there was evidence of a different attitude to laughter. Referring to particular boys in her class who had a reputation for playing the fool, one female student at the urban school explained: ‘There are some who have already got themselves used to it and they speak on purpose to make the class laugh’. This suggestion that boys consciously use comedy and incitement of laughter as a strategy that forms part of their negotiation of the classroom environment was held up by my observations. In the activity that caused Alisha great distress, there were several boys who, when they came to the front of the class, moved their bodies in a way that was designed to induce laughter. Some of them also delivered their sentences through bouts of their own laughter. This gave the impression that the student was in control. If others joined in, it felt more like ‘laughing with’ the student rather than ‘laughing at’ them. Boys still talked about habits and expectations that required that students laugh at mistakes, but they also talked about supporting one another after those knocks. One male, urban student explained:

It’s … let’s say friendship … closeness. … Often you are already good friends … they must laugh at you … because you messed up. After that, because you suffered a lot, they will pick you up, they will correct you. But the first stage is that they must laugh at you because you have both got used to it being that way … it’s every day … So, with that person, you laugh at him, you laugh at one another.

Being laughed at for making mistakes was a common experience for both genders, but some of the data suggests that boys may have been able to cope with laughter more easily, sometimes using humour to assert their own control over the situation. The example above also suggests that boys might be able to move on from being laughed at more quickly.

The most discussed strategy employed by girls to reassert control in situations where they experienced fear and shame was to use cruel words to put down the person that they felt was exposing their shortcomings. This was labelled ‘gossiping’ and students explained it as a form of jealousy. One girl in Form Two at the urban school explained, ‘One thing is gossiping. Like, if I speak fast like you (referring to the way I speak English) they will say, “Oh you are pretending”’. This practice was by no means limited to town. A group of Form One girls at the rural school explained how fellow students would respond if they were to use English when it wasn’t required by the teacher:

Student 1:

If you use English outside of the lessons, they say, ‘Mh! Look at them … they pretend they know English … but they don’t have anything … they are the same as us … quit using English … make way for us'. They are shouting like this … and you don’t even speak again … while all you wanted was to learn … 

Student 2:

Classmates tell you, ‘English … English? Here in this classroom we speak Kiswahili'. So you find that you feel bad.

Researcher:

Why do they speak like that?

Student 1:

Ah! You don’t know people! They don’t like their classmates to see that they are small … they feel jealous and say, ‘Mh! That person is showing off'.

The use of cruel words to criticise students who were seen as breeching expectations and norms around language use was not solely mentioned by girls, but students themselves suggested that ‘gossiping’ was common, and almost expected, amongst girls.

In each class, there were a few students who were regular contributors and appeared not to be hindered by feelings fear and shame. These were almost exclusively the students with the strongest English. In town, these students had often attended English-medium primary schools. In fact, one female student explained that, on the first day, several teachers had asked who had attended English-medium primary schools: ‘We are raising up the hands … and then teacher say, “good, ok” and is starting teaching, and then he just focus on those people’ [Student spoke in English]. The fact that these students had a known background in English meant that both teachers and other students relied on them to answer questions and to support others. Their English ability also meant that the risk of making mistakes and being laughed at was significantly reduced. In fact, a different girl in Form One at the urban school stated, ‘You know, somehow English makes you confident’.

It might be assumed that those who contribute regularly do not experience fear or shame. However, in interviews, these students still talked about these negative emotions playing a powerful role in their schooling experiences, albeit in a slightly different way. Some of these students talked of the pressure they felt to help others, and the criticism they faced when they didn’t, for example one student in town who had been shouted at by her classmates for not passing them the answers in an examination. Several of the students in this group also talked about being criticised for speaking English outside of lessons, to the extent that one urban girl who had attended an English-medium primary school lamented that she could not use English at secondary school, saying: ‘I am heart-broken because I am forgetting English’. The impact of English ability on students’ experiences was most stark in town, due to the presence of a number of students from English-medium primary schools, as well as students whose parents used English regularly and could support them. However, it was not only urban students who spoke about fear and shame relating to speaking English when it wasn’t required by the teacher. In an interview conducted by Student Researchers a group of Form One students from the rural school, students explained what they felt would happen if they spoke English with their peers:

Student 1:

Your classmates don’t know it, so when you speak, they might hate you.

Student 2:

They will hate you … when you walk, you will be alone.

Having a stronger command of English seemed to protect some students from being silenced by negative emotions during lessons, but they still felt that fear of criticism and of being actively shamed by their peers for using English at unsanctioned times resulted in their being silenced in other areas of the school and community.

Discussion

The majority of the existing literature about SSA has associated students’ reluctance to talk with lack of understanding of the unfamiliar LoI (Rea-Dickens and Yu 2013; Samuelson and Freedman 2010; Mokibelo 2016). The findings in the previous section demonstrate that lack of understanding is definitely an important part of the explanation, but it does not fully explain all reluctance to participate. Some students insist that they do understand, at least in part, but are discouraged from talking by the fear of making mistakes and anticipated responses of their teachers and peers. As such, the prevalence of fear and shame acts as an additional layer of learning constraint, preventing students from fully participating in classroom dialogue and constraining their language choices both around school and in their communities.

These constraints on students’ participation and classroom talk are particularly significant in contexts, like Tanzania, where students are simultaneously engaged in two learning processes. They are required both to develop skills in an additional language and learn and demonstrate new subject content using that additional language (Barrett and Bainton 2016). Research into the use of familiar languages to support these parallel processes has highlighted the importance of talk, both informal and formal, for both conceptual learning and language development (McKinney and Tyler 2019; Msimanga and Lelliott 2014; Setati et al. 2002). Students in this study, too, recognised the importance of participating and answering teachers’ questions for their learning and several talked of the ‘heart-break’ they felt when fear and shame prevented them from speaking. However, the findings suggest that, in order to encourage all students to talk more, attention will need to be paid to the nature of the classroom atmosphere, ensuring that students feel safe to experiment with language(s) and ideas without fear of shame.

There may be useful learning that could be drawn from the global literatures looking at classroom pedagogy and language learning, particularly around the role of emotions and related strategies and interventions (Hargreaves 2015; Oxford 2017; Toyama and Yamazaki 2021). Since this paper responds to an observed problem with silence and reluctance to participate, the focus has been on negative emotions, but there is interesting evidence that fostering positive emotions can play an influential role in language learning contexts (Dewaele et al. 2019; MacIntyre, Gregerson, and Mercer 2019). However, in order to evaluate the transferability of this research, more research is needed in SSA classrooms to explore how different emotions are understood locally and the roles that they play in learning. Within SSA, including in Tanzania, multilingual, language supportive approaches to subject teaching are being developed with local teachers and teacher educators, to respond to their particular curriculum demands and local language environments (Erling et al. 2021; Makalela 2015; Opanga and Nsengimana 2021; Probyn 2019; William and Ndabakurane 2017). The data presented in this paper emphasise that changes in pedagogy must include positive classroom management and motivational strategies that decisively move away from the use of fear and shame as methods of control.

Studies in Tanzania have found that corporal punishment and threats and humiliation by teachers are common (Human Rights Watch 2017; Moris 2008; Tangi 2019), but there is not enough evidence to conclusively argue that classroom management strategies based on fear and shame are a direct result of the use of English as LoI. Corporal punishment, and other shame-based punishments, are not only used for language-related misdemeanours and have also been reported in contexts where a familiar language is used as the LoI (Gershoff 2017). However, there is a long history of using shame to punish local language use associated with colonial schooling (Ngūgī wa Thiong’o 1986), and this seems to have been retained in some forms alongside the dominant global LoIs (Kiramba 2018; Mayisela 2018; Vavrus 2002). Research in Tanzania that saw the same teacher teach lessons in English and Kiswahili observed that negative forms of reinforcement, in particular forcing students to stand for failing to answer questions or verbal shaming from the teacher, were observed in English-medium lessons, but not Kiswahili-medium lessons (Mwinsheikhe 2009). It has been noted by Mwinsheikhe (2008), but also elsewhere, that the use of punishment damages the relationship between students and teachers, which undermines both learning and students’ physical and socio-emotional safety (Joyce-Gibbons et al. 2018). More research is needed to better understand the relationship between LoI and negative reinforcement and punishment, and what might be effective to support the transformation of these important teacher-student relationships.

The findings from this study also show that LoI policy and practice is an equity issue, as the silencing effect of fear and shame is not experienced equally. Some students, whether through their level of English proficiency or their ability to take control of classroom laughter, were protected from the full impact of these emotions by classroom norms that made it more acceptable for certain groups to speak. This is another area that would benefit from further research to better understand how these norms are constructed, the similarities and differences across contexts, and the conditions under which they are, or might be ‘transgressed’ (Windle, de Jesus, and Bartlett 2020). From the limited evidence currently available, it appears that the silencing effect of negative emotions might be amplified for girls when combined with gendered behavioural and interactional patterns (Alidou and Brock-Utne 2011; Milligan and Adamson 2022; Rea-Dickens and Yu 2013), however, greater exploration of these patterns is required.

The voices shared here demonstrate that the prevalence of fear and shame in students’ experiences of schooling goes beyond acting as a learning constraint and, in fact, has a much broader impact on students’ self-concept and educational experiences. When talking about fear and shame, students also linked these emotions to lack of confidence, concerns about being perceived as unintelligent and to tension in relationships with their peers. One of the most spoken about and feared responses was laughter from teachers and classmates. This is a clear example of the ‘verbal or symbolic gestures’ that Chase and Walker (2013, 740) argue play a key role in co-construction of shame. Although Chase and Walker make it clear that an individual experiencing shame might make inaccurate assumptions about the thoughts and judgements of others, there were examples in the data from this study that students’ competence in English was used as a measure of students’ overall intelligence and abilities. Similar to the findings of Opoku-Amankwa’s (2009) study in Ghana that showed that students who struggled with English were often labelled as ‘weak’ overall, those who struggled to express themselves and demonstrate learning in English were sometimes labelled as ‘slow’ learners.⁴

The ways in which different dimensions of students’ educational experiences are intermeshed and influence one another is an important, but as yet neglected, aspect of LoI research. Developing greater understanding of students’ holistic experiences, with an emphasis on centring their own voices, is crucial if we want to fully appreciate the impact of LoI policies and practices, both on young people’s learning experiences and outcomes.

Acknowledgements

First and foremost I would like to acknowledge and thank the young people helped to create this research and who shared their experiences with me. I am also grateful to Professors Tristan McCowan and Elaine Unterhalter for their support of my doctoral work, and to Dr Lizzi O. Milligan for her encouragement and comments on drafts of this paper.

Disclosure statement

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

Additional information

Funding

The field research for this study was funded in part by the Tim Morris Award, administered through Education Development Trust. This paper was prepared as part of an ESRC postdoctoral fellowship [grant number ES/W005484/1].

Notes

1 For a fuller discussion of students’ framing of what it meant to understand, see my doctoral thesis: Adamson (2020b).

2 Of 115 students in Forms 1 and 2 at the urban school who completed questionnaires about their home and family background, 13 reported having attended private, English-medium primary schools.

3 This research was also subject to ethical review through the department of Humanities and Social Sciences at the Institute of Education, University of London in 2014.

4 Examples of teachers making this assumption are recorded in fieldnotes from both the urban and the rural school. See also Walker (2020) for discussion of how these labels affected educational inclusion in secondary schools in Morogoro Rural and Handeni districts in Tanzania.