Namibian teachers’ practices in a multilingual context

ABSTRACT

The scope of the paper is to analyse Namibian mainstream primary teachers’ practices in the multilingual context they work in. As teachers have a central role in implementing language policies, their practices either support or impede learning as well as learners’ multilingual identities. Multilingual practices entailing rich scaffolding and translanguaging strategies are supposed to be beneficial for content and language learning and the learners’ identity development. The methodological approach adopted in this study is ethnographic, and the theoretical framework is linguistically responsive teaching and multilingual education. The data consist of teacher interviews and focus group discussions, a questionnaire and classroom observations. The analysis method is qualitative content analysis. The data were collected in two Namibian government primary schools in the beginning of the 2020 school year. The analysis shows that translanguaging strategies such as peer translation or explaining concepts in home languages are used, but rarely as pre-planned pedagogical strategies. Scaffolding is provided mostly through visual aids. It may be concluded that the translanguaging strategies already in use should be encouraged further, and multilingual pedagogies should be included in both pre-service and in-service teacher education.

KEYWORDS:

• Multilingual education
• translanguaging
• teachers’ practices
• Namibian basic education
• linguistically responsive teaching
• multilingualism

1. Introduction

Namibia is a multilingual country with about 30 different languages spoken on its soil, depending on what variations are counted as languages (Lusakalalu, 2007). English is the only official language (Namibian Constitution, 1990, p. 10). All 14 school languages – Afrikaans, German, English, Ju’hoansi, Khoekhoegowab, Oshikwanyama, Oshindonga, Otjiherero, Rukwangali, Rumanyo, Setswana, Silozi, Thimbukushu and Namibian Sign Language – may be used as media of instruction (MoI) in junior primary (grades 0–3). English is the only MoI from grade 4 onward (NCBE, 2016). However, according to 2011 census, only 3.4% of the population speak English as their first language.

The MoI is central for learning as it influences understanding the subject content. Anxiety caused by a lack of proficiency in the MoI may interfere with learning (Lucas & Villegas, 2011). In a situation, such as that in Namibian schools, where most of the learners are emergent speakers of the MoI, it is crucial to adopt teaching practices that simultaneously help them learn the subject content and develop their linguistic skills. Learners’ multilingual identities have to be acknowledged, and ‘interaction in all the languages of their linguistic repertoire is needed for positive identity development’ (Fielding & Harbon, 2013, p. 529). In multilingual communities, communication is based on speakers negotiating their different linguistic resources (Canagarajah & Wurr, 2011). They ‘shuttle between languages’ (Canagarajah, 2011, p. 401), switching and mixing them. Multilinguals’ fluid language practices have been conceptualised as translanguaging, which ‘changes the perspective from a monolingually oriented approach to one in which the heteroglossic nature of multilingual practice is recognized’ (Jonsson, 2017, p. 20).

The main aim of this paper is to study and document the practices of Namibian primary school teachers in the multilingual context they work in. Teachers have a central role in the implementation of language education policy (Menken & García, 2010) and in determining which languages are used in the classroom (Fielding & Harbon, 2013). Their beliefs and practices thus affect the learning process, its outcomes and the learners’ identity construction. They may act as agents of change and create spaces for a multilingual approach even within a monolingual policy (García & Kleyn, 2016). Multilingual education, entailing rich scaffolding and methods such as translanguaging, enhances learning and supports the learners’ multilingual identities (Erling et al., 2017). Increasing epistemic access and supporting the learners’ multilingual identities is crucial in all postcolonial, multilingual educational settings.

This paper explores the current practices of Namibian mainstream teachers in two government (public) primary schools. The methodological approach is ethnographic, and the theoretical framework is based on linguistically responsive teaching and multilingual education. The data consist of teacher interviews, both individual and focus group discussions, a questionnaire and classroom observations made in Khomas region in the beginning of the 2020 school year. The analysis method is qualitative content analysis. Some typologies of scaffolding strategies used in content and language integrated learning (CLIL) classes and English language learners (ELL) instruction (Mahan, 2020; Pawan, 2008) have been adapted in the analysis of this study as have strategies used in translanguaging pedagogy (e.g. García et al., 2017; Menken & Sánchez, 2019). This perspective may offer insights into how classroom practices relate to the multilingual reality in the society. To begin with, theoretical concepts are discussed in Section 2. The data and method are presented in Section 3 and the findings in Section 4. A discussion of the findings follows in Section 5.

2. Linguistically responsive teaching and teachers’ practices

Although the Namibian language education policy recommends the use of indigenous languages as MoI in junior primary (grades 0–3), 25% of children in grades 1–3 are enrolled in English medium schools (EMIS, 2019). Moreover, according to a survey conducted by the Ministry of Education, Arts and Culture (Towards a New Education Act for Namibia, 2016), 48% of the learners were of the opinion that schools do not provide opportunities to learn in the mother tongue. This despite the fact that learning in mother tongue has been shown to be beneficial for learning and educational quality (Benson, 2019). One of the most important reasons for the high number of learners in monolingual English medium instruction is that parents remain committed to the language that offers social mobility (Qorro, 2009; Shank Lauwo, 2018), a common situation in postcolonial societies. However, when parents are given a multilingual option, they are likely to choose it (Heugh, 2002, 2013). It is important, therefore, to introduce multilingual practices that demonstrate to parents the benefits of multilingual education. Multilingual practices would also contribute to valuing and preserving the indigenous languages and strengthening the multilingual resources in Namibia.

Learning in a foreign language in African classrooms often entails teacher-centred activities, choral responses, repetition and reading aloud (Bunyi, 2005; Guzula et al., 2016), whereas learning in a familiar language is often more learner-centred, as teachers use, for example, more group and pair work (Afitska et al., 2013). It can be inferred that using the learners’ home languages and multilingual methods would offer more opportunities for learners to participate actively, draw on their full linguistic repertoires and thus contribute to the construction of their multilingual identities. Namibian teachers’ practices are documented in this study in order to find out how they relate to the multilingual teaching context. The findings are likely to be relevant not only in the Namibian context, but in other similar ones as well, especially in Sub-Saharan countries.

2.1. Linguistically responsive teaching

The approach adopted in this paper is linguistically responsive teaching, which is grounded in Vygotsky’s theory of learning through social interaction (Lucas et al., 2008). Linguistically responsive teaching requires awareness of the connections between language, identity and education. If the learners’ multilingual identities are ignored, it leads either to their silence and disengagement or to the abandonment of their cultural identities and loyalties towards their linguistic community (Lucas & Villegas, 2011). Instead, the learners’ strong literacy skills in L1 should be seen as an asset for achieving high competencies in L2 literacy (Cummins, 2000).

An essential concept in Vygotsky’s theory is the zone of proximal development (Vygotsky, 1978), which means the space between learners’ current competence and what they are capable of achieving assisted by a peer or an adult. Learning and development take place within this zone (Gibbons, 2015). This means first that classroom language should be just beyond learners’ proficiency: not too difficult but sufficiently challenging (Lucas & Villegas, 2011). Second, scaffolding should be provided in order to make learning efficient.

Scaffolding means temporary support that helps learners extend their performance beyond their present capacities (Kniffka, 2019). It was first introduced by Wood et al. (1976) when they examined parent–child talk. Scaffolding does not mean simplifying the tasks or lowering the demands on the learners, but rather ‘amplification and enrichment of the linguistic and extralinguistic context’ (Walqui, 2008, p. 109). Gibbons (2015) argues that it is actually only when scaffolding is needed that learning will take place, since then learners are likely to operate in their zone of proximal development. Learners should be given multiple opportunities for meaningful social interaction in situations where they have to negotiate meaning (Lucas et al., 2008). Meaningful interaction scaffolds the learners, allowing them to take risks safely and extending their abilities.

Walqui (2008, 2011) distinguishes five components in teachers’ development throughout their career: vision (teachers’ beliefs, objectives and ideologies), knowledge (general and subject-specific pedagogical knowledge, student and self-knowledge), motivation (reasons, emotions and incentives), reflection (planning, recollection) and practice (Walqui, 2008). All development is embedded in a context (classroom, school, society, etc.) The teaching contexts are ever-changing, and teachers need to be supported to meet the increasing demands (Walqui, 2011). Namibian teachers, like all teachers working in a multilingual context, whether postcolonial or other, need to be supported to meet the demands that context sets on them.

The last component mentioned in Walqui’s list is practice, which means the skills and strategies to achieve their goals (Walqui, 2008). Knowledge must be combined with an ability to act on it. ‘At times we meet teachers who can articulate a coherent grasp of what ought to be happening in a class but who demonstrate a discrepancy between their knowledge and the ability to implement it’ (Walqui, 2008, p. 121). The focus of this paper is on teachers’ practices, especially those that are relevant in a multilingual context. Practices that leverage the multilingualism present in a classroom or that simply adapt the teaching strategies to better fit the multilingual context are discussed in sub-section 2.2.

2.2. Multilingual education strategies

In multilingual communities, the theorisation of community is based on shared space, not on shared language, language diversity is the norm, and communication is based on negotiating diverse languages. Multilingual communication practices aim at finding common ground between the codes and resources. Moreover, multilinguals seem to be equipped with language awareness and competence that enable them to act purposefully in different communicative situations (Canagarajah & Wurr, 2011). They draw on all the codes in their linguistic repertoire that forms a continuum rather than separate language competencies (Otheguy et al., 2015). Fluid, flexible and multifaceted language practices characterise communication between multilinguals.

The multiple and flexible multilingual practises have recently been conceptualised as translanguaging (Canagarajah, 2011; García & Wei, 2013). As a concept, it emphasises the fluidity of multilingual practices and contests monolingual language ideologies (Jonsson, 2017). Multilingual speakers use their linguistic repertoire in different situations differently, for example, code-switching, code-mixing, borrowing and using translation. When translanguaging, multilingual speakers ‘shuttle between languages, treating the diverse languages that form their repertoire as an integrated system’ (Canagarajah, 2011, p. 401). Nonetheless, translanguaging has not been accepted in schools until recently (Creese & Blackledge, 2010), largely because of a misconception that children learn a second language better the more they are exposed to it (Cummins & Swain, 2014). Research has shown, however, that there is either no relationship or an inverse relationship between the time-on-task in the target language and its acquisition (e.g. Carey & Cummins, 1983; Egan & Goldsmith, 1981). Using other languages in the classroom does not interfere with learning the official MoI.

Flexible multilingual education that allows using other languages in addition to the medium of instruction and that entails rich scaffolding is likely to enhance learning and student participation (Erling et al., 2017). One such approach that has gained ground in multilingual classes and in additional language pedagogy in America and Europe in recent years is translanguaging pedagogy (Cenoz & Santos, 2020; García et al., 2017; García & Wei, 2013; Paulsrud et al., 2017). Strategies used in translanguaging pedagogy (see e.g. García et al., 2017), but also those used in CLIL and ELL classes to scaffold teaching (Mahan, 2020; Pawan, 2008) are investigated in relation to Namibian teachers’ practices. There is a conceptual difference between these approaches as regards the notion of separate languages. In CLIL and ELL, languages are kept separate, whereas the most commonly cited translanguaging theories reject this notion. Because the purpose of this study is to investigate all teachers’ practices that may be relevant in a multilingual context, both approaches have been included. It is also possible to maintain a translanguaging approach without rejecting the cognitive reality of different languages in a multilingual’s repertoire, as discussed below.

Many translanguaging theories, especially those promoted by Ofelia García and her colleagues, reject the notion of separate, named languages in the cognitive experience of multilinguals (e.g. García & Kleyn, 2016). According to these theories, code-switching or translation would not be considered as translanguaging strategies. Cummins (2021) classifies these theories as Unitary Translanguaging Theories (UTT). He advocates another view of translanguaging that, unlike the unitary theories, affirms the reality of ‘languages’ in the individual’s cognitive and linguistic system. He calls these theories Crosslinguistic Translanguaging Theories (CTT). They are in line with MacSwan’s (2017) ‘multilingual perspective’ on translanguaging that

accepts individual multilingualism as not only psychologically real, in the integrated sense, but also universal. Here, codeswitching may be seen as an instance of translanguaging, alongside other bilingual phenomena such as translation, borrowing, and additional processes, in a range of modalities (MacSwan 2017, 191, my emphasis)

A crosslinguistic or multilingual perspective on translanguaging are adopted in this study, and thus both code-switching and translation are seen as instances of translanguaging. Nevertheless, a distinction is made between spontaneous code-switching and translation versus pre-planned pedagogical use of these strategies.

Translanguaging pedagogy focuses on the flexible instructional strategies used to leverage the students’ linguistic resources fully (Wei & García, 2016). It allows the learners to develop both their home languages and the dominant language they need for social mobility (Shank Lauwo, 2018), and can be used even in multilingual classrooms where the teacher does not speak the students’ home languages (García et al., 2017). Instead of ad hoc code-switching, different languages are used systematically to scaffold learning and enhance student participation (García & Wei, 2013; Otheguy et al., 2015). They may be used in classroom discussions (especially in pair and group discussions), multilingual materials and writing assignments, word walls and cognate walls (Menken & Sánchez, 2019). According to Menken and Sánchez (2019), in addition to the scaffolding mind-set, a translanguaging stance – a transformative social approach – is needed to empower the students and strengthen their multilingual identities.

The possibilities of translanguaging in education in Africa have also been researched recently, especially in South Africa (Charamba, 2019; Guzula et al., 2016; Makalela, 2015; McKinney & Tyler, 2018; Probyn, 2015). Makalela (2016) has theorised translanguaging from an African perspective using the notion of ubuntu translanguaging. It is based on the idea of interdependence of people and of coexisting languages, the boundaries of which are blurred in the meaning-making process.

The use of translanguaging or other multilingual strategies in Namibian basic education has been researched to a minimal extent. Mensah (2015) examined multilingual practices at Windhoek International School, suggesting that dynamic multilingual practices should be encouraged. Van Der Walt (2015) investigated English language proficiency development in a rural primary school in the Kavango region, arguing for the use of translation and preview-view-review strategies. The present study contributes to research by focusing on the practices primary teachers use in the multilingual context in which they work in the Khomas region. The research questions it endeavours to answer are:

1. What self-reported practices do teachers use in the multilingual context they work in?

2. What observed practices do they have?

The data collection and analysis methods and procedures are discussed in Section 3 before presenting the findings in Section 4.

3. Data and method

Ethnographic data collection methods were used to obtain multiple perspectives on teachers’ cognitions and practices. The data consist of semi-structured teacher interviews, focus group discussions, a questionnaire and classroom observations. Teachers’ cognitions and practices shape each other mutually, and the societal, institutional and physical teaching context affect both (Borg, 2006). Research has shown that there may be inconsistency between teachers’ cognitions and practices, as well as between their self-reported and observed practices (Borg, 2006). That is why both self-reporting instruments (questionnaire and interviews) and classroom observations have been used in this study. The data and the participants are presented in sub-section 3.1. and the analysis method in sub-section 3.2.

3.1. Data collection and participants

Permission to conduct research was obtained from the Ministry of Education, Art and Culture of Namibia; the Khomas Regional Council Directorate of Education, Arts and Culture; and the Namibian National Commission on Research, Science and Technology. The ethical preview was also obtained from the affiliated university. The school principals and the participating teachers were informed in a printed letter about the focus of the research, the preservation of anonymity of the participants and the secure storage of the data, according to the data management principles of the affiliated university. The researcher was introduced to the teachers by the school principals, and the teachers were invited to accept the presence of the researcher in their classes and to complete the questionnaire. However, the participation in the questionnaire and the interviews was voluntary, as was allowing the researcher to do the lesson observations. Teachers gave their documented consent for the interviews, the questionnaire and the classroom observations. Information letters were addressed to the pupils and their parents, but not all consent letters for video recording were returned in time. That is why no video recordings of the observed classes were made, and this part of the data consists of observer field notes only. This is one of the limitations of the study, as video recordings would have allowed more accurate observation data on teachers’ practices.

The data collection was made in two government (public) primary schools in the Khomas region in the beginning of the 2020 school year. They are situated in low- to middle-income areas inhabited by Black African populations. In one of the schools, the MoI is English in all grades and in the other one, three local languages are used as MoI in junior primary (grades 0–3). There was no specific reason for including one school with English medium and the other with mother tongue instruction in the data, as the sampling was based on the availability of the schools. Furthermore, the data collection was interrupted due to the COVID-19 lockdown, and the questionnaire and interview data could not be collected according to the initial plan which included eight schools altogether and therefore remain somewhat limited. The number of teachers participating is presented in Table 1.

Table 1. Research participants (interviews and questionnaire).

Research participants (interviews and questionnaire)
Interviews				Questionnaire
	Individual	Focus group	Total
Number of interviews	4	2	6	Junior Primary teachers	10
Participants school 1	1	3		Senior Primary teachers	27
Participants school 2	3	4
Total participants	4	7	11	Total participants	37

As can be seen in Table 1, four teachers were interviewed individually, one in the first and three in the second school. One focus group discussion was conducted in each school, involving three teachers in the first school and four in the second. All interviews and focus group discussions were conducted in English and lasted between 11 and 25 min. One of the interviewed teachers was a junior primary teacher. The others taught grades 4–7, in which one teacher usually teaches two subjects. The subjects represented were ESL, religious and moral education, social studies, home language, natural science and health education, mathematics, Afrikaans, and arts. Eight of the interviewed teachers were female and two were male.

There were 37 questionnaire respondents altogether. Ten of them were junior primary teachers (grades 0–3) and 27 were senior primary teachers (grades 4–7). They represented practically all the subjects. The years in service of the participating teachers are presented in Table 2.

Table 2. Years in service.

Years in service
	Questionnaire	Interviews
Less than 3 yrs	4	0
3–5 yrs	4	2
6–10 yrs	7	5
11–15 yrs	4	1
16–20 yrs	3	1
Over 20 yrs	8	2
Not known	7	0
Total	37	11

As can be seen in Table 2, the distribution of the years in service of the teachers is relatively even and they represent teachers at the beginning of their careers as well as very experienced teachers.

The interviews were semi-structured. Both individual and focus group interviews have been analysed as one dataset and are referred to in the text as ‘interviews’. Qualitative classroom observation (Wragg, 2012) was conducted in 49 lessons altogether, and 30 grade four lessons form the data of this study. Nineteen lessons were excluded altogether. Thirteen lessons were excluded because they were done in other grades or because their contents were not suitable for the purposes of the study (e.g. mere distribution of material). One of the teachers was overrepresented in the original data, and six of her lessons (with parallel groups with the same content) were excluded to obtain a more balanced dataset. Twelve different teachers are represented, with the number of lessons per teacher varying from one to five. The subjects represented are Afrikaans SL, art, ESL, ICT, mathematics, natural science and health education, religious and moral education, and social studies. Observation notes were taken on an observation frame during the lessons. Teachers did not provide any lesson plans, but some reflections and commentaries were made in the interviews afterwards. As the data collection was conducted in the beginning of the school year, the topics of the lessons were mostly new to the pupils.

In the interviews, teachers were asked about their teaching practices and support methods, their beliefs about the language education policy and its implementation, their students’ language proficiency and their experiences and needs for in-service training. Most of the interviewed teachers had been observed in their lessons and the interview provided them the space to explain and comment on their practices. It is worth noting that the notions of ‘multilingual practices’ or ‘translanguaging’ were not familiar to the participants, as they do not figure in curricular documents or in-service trainings. These issues were addressed by formulating the questions without using the terms as such.

The questionnaire contained questions about the respondents’ linguistic and professional backgrounds (Part I), their opinions on language education policy and multilingual education, their teaching practices (Part II), and their experiences with and wishes for in-service training (Part III, see Appendix 1). The questionnaire was distributed to the whole staff in both schools (83 teachers altogether), and resulted in 37 responses (response rate 44.6%). The questions in Part II, Question 2 (a–q) are the focus of this paper. They were presented as statements about teaching practices. The respondents circled ‘often / sometimes / never’ according to their practices. The responses of the junior primary teachers were not included, as some of them teach in learners’ home languages. The questions about multilingual teaching practices would therefore not be relevant. The last question of this section (2.r) was an open-ended question to which the respondents could give examples of their best practices or comment on the previous questions (2a–q). The junior primary teachers’ comments were included. All the questions were in English and the respondents answered them in English.

3.2. Method of analysis

The analysis method is qualitative content analysis. The coding frame was built using the statements of the questionnaire as theory-driven categories and completing it with categories that emerged from the interview and observation data to fit the data and increase the validity of the research (Schreier, 2012). Any occurrences of multilingual practices (a mention in the interviews or an observed instance in the classroom) were coded either into a theory-driven category or a new category or sub-category was founded when necessary. The questionnaire statements were informed by typologies of scaffolding strategies used in multilingual education in CLIL classes and ELL instruction (Mahan, 2020; Pawan, 2008) as well as strategies used in translanguaging pedagogy (e.g. García et al., 2017; Menken & Sánchez, 2019). The coding frame was built using the above-mentioned strategies and adapting them to better fit the research questions and the data. They include language choice, visual and textual support, social strategies (e.g. translation) and linguistic scaffolds. Though these pedagogical practices have been mostly developed in ‘monolingual’ communities, the analysis was completed by practices that arose in the interviews, for example concerning practices outside the classroom that describe the multilingual practices widely used in the Namibian context.

The interview and observation data were transcribed manually and subjected to qualitative content analysis (Schreier, 2012). The main coding was done twice to assess its consistency (Schreier, 2012). The percentage of agreement between the two coding rounds was 92% and the coding of the differing coding units was revised to attain the final coding. The distributions of the questionnaire responses were tabulated using absolute frequencies and percentages to allow triangulation with the two other datasets to increase validity and compare and contrast the results (cf. Creswell & Plano Clark, 2007). The responses to the open-ended question (2.r) were analysed qualitatively.

4. Findings

The findings in interview and questionnaire data (research question 1) are presented in sub-section 4.1, followed by observation data (research question 2) in sub-section 4.2.

4.1. Teachers’ self-reported practices

The practices were divided into four main categories: language choice, material scaffolds, social strategies and linguistic scaffolds that will be explained below. Language choice encompasses the use of languages other than English as a pedagogical tool and spontaneous code-switching. The former included introducing a new topic or explaining in a home language and allowing the use of home languages in class or in assessments. Code-switching occurred related to content teaching, classroom management or outside classroom activities. Not using or accepting the use of other languages was also included in this category. The references made to these practices in the interviews and the distribution of the questionnaire responses are presented in Table 3.

Table 3. Language choice (interviews and questionnaire).

Strategies	Interviews	Questionnaire
	References	Often count / %	Sometimes count / %	Never count / %	N
I Language choice
1. Pedagogical use of other languages than English	5
1.1. Introducing a new topic in home language	0	1/4%	1/4%	22/92%	N = 24
1.2. Explaining core terms in home language	4	3/12%	7/27%	16/62%	N = 26
1.3. Allowing the use of home languages in group work	0	1/4%	9/33%	17/63%	N = 27
1.4. Accepting the use of other languages or mixing codes in class	1	3/12%	7/27%	16/62%	N = 26
1.5. Accepting the use of other languages in assessments	0	0	5 / 20%	20 / 80%	N = 25
2. Code switching
2.1. Relative to content teaching	5	4 / 15%	13 / 48%	10 / 37%	N = 27
2.2. Relative to classroom management	0
2.3. Outside the classroom	5
2.4. Not accepting the use of other languages than English	9

The pedagogical use of languages other than English and code-switching relative to content teaching correspond to questions in the questionnaire, whereas the other code-switching strategies emerged only in the interviews or classroom observations.

As can be seen in Table 3, relatively little pedagogical use of languages other than English is done in class, and the majority do not accept the use of languages other than English in assessments or group work. Group work was actually very little used, as discussed in sub-section 4.2. Home languages are used in classrooms to explain, as was mentioned in the interviews. Some multilingual practices were evident in the findings, namely spontaneous code switching in classroom situations as well as the use of languages other than English outside the classroom. Translanguaging is taking place, though not necessarily recognised as a conscious pedagogical strategy (García & Wei, 2013; Otheguy et al., 2015), and it happens mostly spontaneously (cf. García et al., 2017 for translanguaging shifts).

According to the interviews, home languages are used outside classroom activities among pupils as well as between school staff and pupils. This is in line with other research findings in African schools about English dominating in class and multilingualism occurring more outside the classroom (e.g. Bunyi, 2005; Mensah, 2015). Home languages are also used in after-school learning support sessions and in parents’ meetings.

Not accepting the use of languages other than English in class arose nine times in the interviews and twice in the open-ended questions, which reflects a monoglossic ideology (García et al., 2017).

Ex. 1 I don’t even know their language, so what we do is you just speak English

(School 2, Interview 1)

Ex. 2 It’s a multilinguistic kind of region and if you say I’m going to use Oshiwambo … then you only end up favouring the Oshiwambo speaking (School 2, Interview 3)

The reasons for this practice were either that the teacher did not share a common home language with the pupils, like in excerpt 1 (cf. Haukås, 2016) or, most often, that using home languages in a linguistically diverse context was impossible (see excerpt 2). Teachers felt that if they used their own home language to communicate with those who spoke the same language, they would be discriminating against the other pupils.

The term material scaffolding here means all use of textual or visual material to facilitate learning. It was divided into textual support and visual support. The distribution is presented in Table 4.

Table 4. Material scaffolding (interviews and questionnaire).

Strategies	Interviews	Questionnaire
	References	Often count / %	Sometimes count / %	Never count / %	N
II Material scaffolding
1. Textual support
1.1.Text editing	0	7/27%	10/38%	9/35%	N = 26
1.2. Handouts with some language learners will need	0	6/25%	7/29%	11/46%	N = 24
1.3. Use of vocabularies and other multilingual material	1	15/63%	5/21%	4/17%	N = 24
2. Visual support
2.1. Relative to content teaching	10	19/73%	6/23%	1/4%	N = 26
2.2. Instructions, model, correction	0

Though over half of the questionnaire respondents reported editing texts, giving learners handouts with some language they will need, or using vocabularies and other multilingual material often or sometimes, these did not arise in the interviews. It is also worth noting that over a third of the questionnaire respondents did not edit texts and almost half never provided linguistic help through handouts. Scaffolding learning in a multilingual context by textual support is not a commonly used strategy.

On the contrary, visual support – posters, enlarged pictures, flash cards, and demonstrations – was the most usual strategy to scaffold learning. Visual support was mentioned ten times in the interviews and in three comments to the open-ended question, and almost all questionnaire respondents reported using it.

Ex. 3 I think every lesson when you are planning a lesson you must have the visual materials which we call media (Focus group discussion 1, participant 1)

Visualising instruction seems to be a well-established, commonly used practice. Interestingly, Mahan et al. (2018) observed a similar difference between visual and textual support provided in CLIL classes in Norway.

Social strategies include both affective strategies such as making home languages visible or hearable in class and translation strategies where either another pupil (peer assessment) or a third party is involved in transmitting the message by translation. Their distribution is presented in Table 5.

Table 5. Social strategies (interviews and questionnaire).

Strategies	Interviews	Questionnaire
	References	Often count / %	Sometimes count / %	Never count / %	N
III Social strategies
1. Affective strategies
1.1. Making learners' home languages visible in class	0	2 / 8%	2 / 8%	22 / 85%	N = 26
1.2. Using greetings, etc. in learners' home languages	0	3/11%	9/33%	15/56%	N = 27
2. Translation strategies
2.1. Peer assistance (peer translation, special role assignment)	7	6/25%	15/63%	3/13%	N = 24
2.2. Translation (using a third party to translate)	6

In the questionnaire, the majority reported not making learners’ home languages visible in class. Nine respondents use greetings in home languages sometimes, and only three use them often. Neither of these strategies was mentioned in the interviews.

Peer translation has an important role in the context under study, as it was one of the most evident multilingual practices reported by the teachers. The majority of the questionnaire respondents use it often or sometimes. It arose several times in the interviews, as well as in the comments to the open-ended question.

Ex. 4 You need at least to identify first their mentors so that they can be able to translate the others through those that do not understand very well (School 2, Interview 3)

Teachers ask pupils to translate and explain to their peers in their home language. According to Altinyelken (2010), similar practices were used by Ugandan teachers. One of the interviewed teachers (excerpt 4) called this practice ‘using mentors’.

Ex. 5 Communication was a problem so most of the times I would ask a learner to translate in the mother tongue what I just said, and towards the end of the year the learner opened up and she was much more vocal and could express herself.

(Focus group discussion 2, participant 1)

Another teacher (excerpt 5) told about a situation he had experienced previously, when a pupil coming ‘from the village’ could not speak English at all. His strategy had been to ask another pupil, sharing the same home language with the newcomer, to translate. Using peer translation had been an effective strategy, and he was going to use it again as he had the same situation in his class at the time of the interview.

Teachers also use their colleagues or the school office staff to translate. This may happen during the lessons, when they call in a colleague or send a pupil to ask for translation of a difficult passage. It also happens when communicating with parents if the teacher and the parents do not share a common language.

Linguistic scaffolds include helping the learners formulate their answers by providing prompts or giving a model, practising pronunciation and explicit vocabulary teaching, as well as accepting or using simple or non-standard language. Their distribution is presented in Table 6.

Table 6. Linguistic scaffolds (interviews and questionnaire).

Strategies	Interviews	Questionnaire
	References	Often count / %	Sometimes count / %	Never count / %	N
IV Linguistic scaffolds
1. Teacher helps learners correct their utterances by providing prompts	0	17/65%	9/35%	0	N = 26
2. Teacher reformulates learners' utterances	0	14/54%	8/31%	4/15%	N = 26
3. Practising pronunciation	0
4. Vocabulary teaching	1
5. Accepting non-standard language	0
6. Using simple language	1

All questionnaire respondents reported helping learners correct their utterances by providing prompts and most of them reformulated learners’ utterances. Difficult words were taught in isolation, for example, using flash cards. In one of the interviews and in one comment, the need to use simple language was mentioned. As regards best practices that teachers reported in response to the open-ended question, a learner-centred approach and group work were both mentioned three times.

4.2. Teachers’ observed practices

There was a discrepancy between self-reported best practices (learner-centred pedagogy and group work) and observed practices. Similar discrepancies have often been observed between teachers’ self-reported and observed practices (Borg, 2018) and between their pedagogical knowledge and ability to implement it in practice (Walqui, 2008). Group work was extremely little used, and repetition was a commonly used strategy.

Pedagogical use of languages other than English occurred rarely in the observed lessons. Those that were observed were four instances when the teacher accepted the use of other languages or mixing codes relative to content teaching. Two of them occurred in Afrikaans SL lessons where the teacher allowed the use of Afrikaans and English side by side and one in a natural science lesson where a learner went to the teacher to ask a question, and the conversation took place in their common home language. The fourth event was very interesting as the teacher tried to help a pupil to answer by exhorting him to do it in his home language. He refused and finally answered in English. Later on, the teacher said in the interview that she had been very confused in that situation and supposed that the pupil did not want to answer in his home language for fear that others would think he could not speak English properly. Charalambous, Charalambous, & Zembylas (2016, as cited in Jaspers, 2018, p. 7) report similar resistance by pupils to translanguaging for fear of being taken for Turkish, a problematic identity in Greek Cyprus. Translanguaging has its limitations, too, and it must be implemented considering the contextual factors on all levels, including the language ideologies and attitudes that are prevalent in the context (Jaspers, 2018; Paulsrud et al., 2021). There is a mismatch between what teachers said in the interviews about explaining concepts in home languages (four mentions in the interviews) and what was observed. However, it is possible that the presence of the observer in the classroom had an effect on this.

Spontaneous code-switching occurred more often. Ten of the events were relative to teaching the subject content, seven of which occurred in Afrikaans SL lessons. The rest occurred in natural science lessons where one teacher gave examples in several languages to help learners understand and another teacher gave feedback about homework to one learner. There were even more occurrences (14) of code-switching to maintain discipline or to give instructions. They correspond to what Bunyi (2005) describes as organisational and classroom management tasks. Code-switching also occurred in situations where a pupil came to the classroom for an errand (2). Three events in which the teacher told pupils to speak only English were observed.

Textual support strategies were not observed except for the two ESL lessons where vocabularies were given to support listening comprehension. Instead, visual support (posters, maps, pictures, charts, drawing or writing on the chalkboard) was given dozens of times, mostly to support subject content learning, but sometimes also to give written instructions or corrections on the chalkboard. Altinyelken (2010) similarly observed an extended use of visual aids in Ugandan classrooms.

The classrooms in which the observations took place corroborated the teachers’ reports in the questionnaire about not making the learners’ home languages visible in classrooms. There were many colourful posters on the walls, but most of them were in English. The only exceptions were Afrikaans and those home languages that were taught as subjects in their respective classrooms. Unlike in the interviews and the questionnaire responses, instances of translation strategies occurred only once when a teacher told some pupils to translate her instructions.

Events where a teacher helps pupils correct their utterances by providing prompts and pronunciation practices were observed in Afrikaans SL lessons only. About half of the instances where explicit vocabulary teaching took place were observed in ESL lessons. However, some vocabulary teaching occurred in a social studies lesson, mathematics lesson and an arts lesson where the teachers taught some terminology explicitly. In two lessons, the teachers reprimanded pupils who had been laughing at another pupil’s English pronunciation. They explained that each person’s home language affects the way they pronounce English, thus indicating they accept non-standard varieties.

5. Discussion

Research question 1 was about teachers’ self-reported practices. Especially the practices outside the classroom that arose in the interviews point to multilingual communication practices that belong to multilingual communities. The flexible strategies that the teachers use in classrooms and the practical solutions they provide in different situations reflect the multilingual reality in the society and demonstrate their competence to act purposefully in different communicative situations (Canagarajah & Wurr, 2011). The spontaneous translanguaging practices in the classrooms form a resource as pedagogical practices, though according to the analysis, these practices tend to be spontaneous adaptations to the multilingual context rather than strategically planned pedagogical practices.

Over half of the questionnaire respondents say they never explain core terms in home languages, and as many say they do not accept the use of other languages or mixing codes in class. However, the teachers mention in the interviews using home languages frequently. Moreover, over half of the questionnaire respondents say they switch the language spontaneously if they see the learners do not understand. This may reflect their understanding of the official language education policy that states that the MoI is English from grade 4 onward and the home languages should be used in a supportive role only (NCBE, 2016). The teaching context affects both teachers’ cognitions and their practices (Borg, 2006). Therefore, it is possible that the respondents believe using home languages is not acceptable and for that reason do not report doing so intentionally (cf. Probyn, 2021). It is also possible that they, for any reason, are not aware of code-switching, possibly because they have not reflected on it. Nevertheless, the practical situations in classrooms force them to switch languages anyway, and they admit doing it occasionally.

The context also affects teachers’ willingness to accept the use of other languages or mixing codes in class. Over half of the questionnaire respondents said they do not accept it, and the same practice arose in the interviews several times. The reasons were either that they do not share a common home language with the pupils (cf. Haukås, 2016) or that using their home language in a linguistically diverse context would effectively discriminate against the pupils coming from other language groups. This is probably the reason why teachers do not make learners’ home languages visible or hearable in class either.

As García et al. (2017) point out, transitional early-exit bilingual classrooms, such as in Namibia, suffer from a monoglossic ideology. This means that in multilingual classrooms, the languages are used only in monolingual ways, and code-switching is considered a disruption of monolingual language use and therefore stigmatised (García et al., 2017). It seems that instead of using flexible multilingual strategies, teachers participating in this study align to this monoglossic ideology to a certain extent.

One example of translanguaging practices that teachers report using a lot are translation strategies, especially peer assistance. They also use colleagues as translators, but mainly in an ad hoc way. Instead, teachers could assist one another in a more coordinated way by together preparing multilingual texts, vocabularies, handouts with some language help and other multilingual material. In the questionnaire, over a third of the respondents answered that they never edit texts or provide handouts with language scaffolds. On the other hand, the majority reported using vocabularies and other similar material. Preparing it in collaboration could benefit both teachers and learners.

Visual support was the most commonly used practice. This observation is in line with the National Curriculum (NCBE, 2016), which requires teachers to create and use a stimulating, visually rich learning environment and materials. The curriculum also directs teachers in all subjects to develop learners’ communication skills in English and teach subject-specific terminology (NCBE, 2016). Almost all teachers reported reformulating learners’ utterances or providing prompts for them to auto-correct their answers. This seems to reflect a pursuit of teaching learners correct academic English.

As regards teachers’ observed practices (research question 2), pedagogical use of home languages was very little observed, though there were a few examples of teachers accepting the use of languages other than English. Spontaneous code-switching did occur, both related to content teaching and even more to maintaining discipline, giving instructions and communicating outside classroom activities. This corresponds to Bunyi’s (2005) observations of code-switching taking place in order to perform organisational or classroom management tasks.

Teaching in Namibia is supposed to be learner-centred (NCBE, 2016), and it was the approach mentioned in the responses in the questionnaire about best teaching practices. What was observed in the lessons was that, for example, group work was hardly used, and repetition was a commonly used strategy. This is a typical situation in contexts where the MoI is not the learners’ L1 (cf. Afitska et al., 2013; Bunyi, 2005). The discrepancy between teachers’ knowledge and their observed practices (cf. Walqui, 2008) was obvious, though various working methods (cf. NCBE, 2016) such as teaching discussions, worksheets, pasting pictures, drawing and other activities were used as well.

The instances when teachers forbade learners to speak their home language during the lesson seem to reflect a monoglossic ideology (García et al., 2017), the assumption of maximum exposure leading to maximum learning of the target language (cf. Cummins & Swain, 2014), or both.

Observed use of visual support corroborated teachers’ self-reports of its use. Instead, textual support use was only observed in second-language classes where, for instance, flash cards were used. Explicit vocabulary/terminology teaching took place mostly in ESL lessons, but in some other lessons as well. Lucas et al. state that teachers of emergent speakers of a school language should be aware of basic principles of second-language learning (2008). Familiarising Namibian mainstream teachers with these principles and equipping them with best practices of second-language teaching would help them teach the English language simultaneously with their subject.

6. Conclusion

The analysis produced important findings, the most pertinent of which are summarised here. Importantly, the findings point to multilingual practices both in and out classrooms. However, a monolingual classroom language ideology seems to prevent some teachers from using multilingual practices in teaching, as well as the strategic leveraging of translanguaging practices. First, the translanguaging practices that teachers reported using and that were observed in lessons were mostly used spontaneously, adapting the teaching to the situations that arose in the classroom. They were neither planned beforehand nor intentionally leveraged as pedagogical strategies. Using translanguaging practices should be recognised as a legitimate pedagogical approach to enhance their strategic use when planning and implementing classroom activities. The multilingual practices that teachers use spontaneously and that are common in multilingual communities offer a good starting point for developing multilingual education.

Second, the most widely used strategies, that is, visual support and peer translation, could be encouraged further. Group work offers a natural setting for peer assistance if using home languages were allowed, and using group work more would gear instruction towards a more learner-centred approach. Third, raising mainstream teachers’ awareness of the principles of second-language acquisition and the practices used in second-language teaching would help them plan their instruction in a linguistically responsive way. Fourth, textual multilingual support material was very little used. Teachers from different linguistic backgrounds could prepare it in collaboration if special planning time was allotted.

The limitations of this study, namely the limited amount of interview and questionnaire data and the lack of video recordings of the lessons, do not allow for generalisations of the findings. Furthermore, the data were gathered in a linguistically highly diverse context – the situation is probably quite different in less multilingual regions in Namibia. Further research needs to be done in different schools and areas.

To conclude, implementing linguistically responsive teaching in Namibia necessitates legitimising multilingual teaching practices as well as including translanguaging pedagogies in both pre-service and in-service teacher education.

Disclosure statement

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

Additional information

Funding

This work was supported by Suomen Kulttuurirahasto.