Teachers’ Perceptions of the Use of Translanguaging within English Education in Grades 4-6

KULTUR–SPRÅK–MEDIER

Examensarbete i engelska och lärande

Teachers’ Perceptions of the Use of

Translanguaging within English

Education in Grades 4-6

Lärares förhållningssätt till användandet av translanguaging i

engelskundervisning i klass 4–6

Cajsa Grenner

Niri Hagelin Jönsson

Grundlärarexamen med inriktning mot arbete i årskurs 4–6, 240 högskolepoäng

Examensarbete i engelska och lärande, 15 hp Slutseminarium: 2020-03-24

Examinator: Chrysogonus Siddha Malilang

There were two authors of this research study, Cajsa Grenner and Niri Hagelin Jönsson. We have both contributed with an equal share of work throughout the stages of this project.

The stages being:

- Formulating the research question - The search for research and literature - Gathering data

- Transcribing data

- Analysing and selecting the gathered material - Authoring the degree project

We hereby state that we together have planned for, conducted and completed every part of this study in agreement.

Abstract

This degree project aims to explore, in the context of translanguaging, teachers' perceptions of the use of pupils’ first language within English as a second language education in Sweden. Following a review of the concept of translanguaging from a historic and pedagogical perspective, teachers’ views on the roles of their pupils’ first and second languages as reported in international research, recent research pertaining to teachers' perceptions and pedagogical methods within translanguaging is highlighted. Results from a methodological triangulation examination of data emanating from a survey, interviews and classroom observations show that: even though the term translanguaging is relatively unknown, teacher respondents use translanguaging as a method within ESL education in Sweden; a majority of teachers value their English-use higher than their use of the pupils’ first language, but are not averse to using translanguaging when deemed appropriate; and a more positive attitude is displayed towards the pupils use of their first language, but the use of English is preferred.

Keywords: code-switching, ESL, L1, L2, perceptions, plurilingualism, multilingualism translanguaging

1. Introduction ... 1

2. Purpose and research questions ... 3

3. Theoretical Background ... 4

3.1. Translanguaging ... 4

3.1.1. Code-switching ... 7

3.1.2. Summary ... 9

3.2. Teachers’ view on the roles of first and second languages in language education.. 9

3.3. Aspects of translanguaging within the steering documents ... 10

3.4. Recent research in code-switching and translanguaging ... 11

3.4.1. Teachers attitudes/perceptions ... 11 3.4.2. Pedagogical Methods ... 14 4. Methods ... 15 4.1. Interviews ... 15 4.2. Observations ... 16 4.3. Survey ... 16

4.4. Ethical and Legal Considerations ... 17

4.5. The participants of the interviews and observations ... 18

5. Results ... 19

5.1. Result of the survey ... 19

5.1.1. Teachers’ perceptions of their own language use ... 19

5.1.2. Teachers’ perceptions of their pupils’ language use ... 21

5.2. Results of five interviews ... 23

5.2.1. Teachers’ perceptions of their own language use ... 24

5.3. Results of four observations ... 29

6. Discussion ... 31

6.1. Teachers’ practices and perceptions of their own use of L1 ... 31

6.2. Teachers practices and perceptions of their pupils’ use of L1 ... 33

7. Conclusion ... 36 7.1. Limitations ... 36 7.2. Further Research ... 37 8. References ... 39 9. Appendices ... 44 9.1. Consent form ... 44 9.2. Survey ... 48

1

1. Introduction

Over the past decades, Sweden has become more globalised. In May 2018, Studieförbundet Näringsliv och Samhälle - a non-profit organisation with many of Sweden’s biggest companies as members - issued a report regarding the increased amount of globalised companies in Sweden. This development results in a growing need for professions demanding personal interaction (Heyman and Sjöholm, 2018). In this new society, strong and flexible linguistic knowledge with English as one of the primary languages is crucial. It falls to us as teachers to prepare our pupils to be able to thrive and take part in this future.

One large part within this emerging multilinguistic society is translanguaging, which could be defined as the dynamic use of different languages as linguistic resources when thinking and communicating (Svensson, 2018). The use of translanguaging as a pedagogical method, with its’ origin in Cen Williams’ classrooms in Wales (Lewis, Jones & Baker, 2012) has now started to expand and has become a component in the aim for language education targeted in the Common European Framework of Reference for Languages: learning, teaching, assessment (hereafter CEFR).

Apart from an ongoing research project at Lund University, conducted by Marie Källkvist and Henrik Gyllstad, not much research has targeted the use of translanguaging within the field English as a Second Language (ESL) in Sweden. Therefore, our study targets this area. While the research within this field has been limited in this country, researchers from other countries have conducted such studies concerning English as a Second Language (ESL), translanguaging and teachers’ attitudes in diverse countries such as the U.S.A., Japan and Turkey. Research points to similar results, where teachers are positive towards using translanguaging (Nambisan, 2014, and Yuvayapan, 2019). However, the implementation of translanguaging as a teaching method is still not widespread. The lack of proliferation of this method is due to various reasons, such as a perceived non-correlation with the policy documents (Yuvayapan, 2019) or an unfeasibility to view translanguaging as an asset (Nambisan, 2014). Further support for this notion can be found in other research, which has pointed towards a discord within the teacher community (McMillan & Rivers, 2011).

In the Curriculum for the compulsory school, preschool class and school-age educare 2011/2018 (hereafter LGR11) it is stated that “Language is the primary tool human beings use

2

for thinking, communicating and learning. Having a knowledge of several languages can provide new perspectives on the surrounding world, enhanced opportunities to create contacts and greater understanding of different ways of living” (p. 34). The foundation for LGR11 is the CEFR, in which the Council of Europe (2001) explicitly states that the aim with language education is to “develop a linguistic repertory, in which all linguistic abilities have a place” (p. 5) without achieving mastery of languages in isolation. It is therefore not unrealistic to assume that LGR11 favours multilingualism within English education. However, in the commentary to the English curriculum (LGR11), Skolverket (2017) has stated that mediation - which according to CEFR (2001) is an extension of translanguaging - should not be prioritised, since it should be possible to conduct the entire lesson using the target language. This statement can - and has - been interpreted as promoting a monolinguistic view where English is the primary language used while teaching English.

In the CEFR (2001), it is stated that one of the aims of language education is to develop “specific fields of action, such as strategies for diversifying and intensifying language learning in order to promote plurilingualism in a pan-European context” (p. 4). In the Companion Volume to the CEFR, the Council of Europe (2018) refers to the paper Plurilingual and Pluricultural Competence by Coste, Moore and Zarate (2009), commissioned by the European Union’s Language Policy Division, and explicitly links the two concepts of translanguaging and code-switching into flexible language strategies. These strategies are to be used to further possibilities to express oneself and encourage communication between peoples.

In line with the previous statement, our job as future teachers are to give our pupils “all-round communicative skills” (Skolverket, 2018, p. 34).While a teacher’s perception of their subject does not influence the proficiency of the teaching, the outlook of the teacher might impact the methodology the teacher chooses to utilize (Kagan, 1992) - and by extension the pupils’ learning. The perception of a concept is therefore a valid consideration when evaluating pedagogical methods. The purpose of this project is not to prescribe the use of translanguaging within English education in Sweden, but to investigate the emerging phenomenon and its current application in teaching. Therefore, the aim of our study is to examine teachers’ perceptions of their own and their pupils' use of their first language in the English language classroom.

3

2. Purpose and research questions

The purpose of this work is to empirically examine teachers’ perceptions of translanguaging as a pedagogical method in the English classroom in years 4-6. The tension within the Swedish policy documents mirrors the perceived diverse opinions about a monolinguistic view in the English classroom within the teacher community. The emerging pedagogical method of translanguaging creates polarised beliefs on which focus on how English is taught, which might influence the English education in Sweden. Our research questions subsequently are as follows:

Which information have English teachers received about translanguaging?

What are English teachers' perceptions towards their own usage of translanguaging in the English language classroom?

What are English teachers' perceptions towards their pupils' usage of translanguaging in the English classroom?

The choice of posing three questions mirrors our view on the complexity of the classroom context. As a teacher, in most teaching situations, there is a need to be aware of our own knowledge, our perceptions of our own actions and, finally, our perception of our pupils’ actions. This choice is an attempt at addressing this complexity.

4

3. Theoretical Background

This section aims to provide relevant theory and background to the research questions. Firstly, we aim to define and explain the concept of translanguaging, while highlighting the connection to the linguistic feature of code-switching. Secondly, teachers' views on the role of first and second language (hereafter L1 and L2) within language education in international research will be briefly discussed. Thirdly, the role of translanguaging within the steering documents, which includes Curriculum for the compulsory school, preschool class and school-age educare 2011/2018 (LGR11), the proposal for the new curriculum and the Common European Framework of Reference for Languages: learning, teaching, assessment (CEFR), will be accounted for. Finally, we review recent research within the field of translanguaging pertaining to the research questions.

3.1. Translanguaging

In a publication from Skolverket, Svensson (2018) refers to translanguaging as the dynamic use of multiple linguistic resources when thinking and communicating. This is only one among many definitions of the concept. However, apart from the use of different definitions, at the moment much research is consistent and built on the work of peers in agreement. The term “translanguaging” was first used in Welsh as “trawsiethu” by researcher Cen Williams in the middle of the 1990´s (Williams cited in Wei, 2016). Williams explored and adopted the term translanguaging as a way to explain how pupils can develop their language proficiency and gain deeper knowledge of the subject at hand through the parallel usage of two languages in their education. Initially, the term originated as a way to characterize pedagogical methods in bilingual classrooms (Williams, cited in Baker, 2001). Examining the use of translanguaging, Williams discovered that when the pupils systematically had to alternate between their two languages in terms of receptive input and productive output, their skills in both languages improved. Moreover, Williams concluded that it does not matter if the pupils spend more time using one language over the other, but rather that successful language development is about which activities are being used (Williams, cited in Baker, 2001).

Baker (2001), a pioneer within the field, maintains that the use of translanguaging is beneficial in many ways. For instance, one of the advantages with translanguaging is that the

5

pupils using it can grow stronger in their less advanced languages. Baker (2001) states this is possible because, when given the option of translanguaging, pupils may be less inclined to utilize their weaker language solely when performing easier work. In the same fashion, allowing the pupils to work with material via translanguaging is beneficial because it lets them absorb the information in one language first and then in another. As such, this can deepen their comprehension. However, for this to be successful, it is important that teachers take time to reflect on their teaching practices and how the two languages can be strategically used to stimulate the pupils’ language growth and minds (Baker, 2001). Furthermore, Baker (2001) notes that the parallel usage of two languages works best with older learners (e.g. high school) but should be more carefully used with younger pupils who have not yet fully developed their language repertoires. This is because younger children have less separated and more unstable language skills. Additionally, regardless of its benefits, translanguaging may not always be the appropriate approach for all pupils; there are times when it is best used as a way to create a differentiated learning environment (Baker, 2001).

Building on Williams' definition of the term, Garcia (2011) defines the concept of translanguaging as something that exceeds the general description of what code-switching is. Additionally, translanguaging can be viewed as a form of mixing languages in a strategic and systematic way in order to create meaning (Gutiérrez, Baquedano-López & Alvarez cited in Garcia, 2011). In essence, Garcia (2011) believes that translanguaging promotes a limitless use of different languages and clearer communication between people in a bilingual society. Furthermore, Garcia (2011) adds that a pupil’s success in second language development begins with the teacher’s attitude towards the usage of the family language in the classroom. Garcia (2011) also suggests that it may be best if teachers started to think about their learners as “emergent bilinguals” (p. 141).

Discussing the subject, Garcia and Wei (2014) mention that translanguaging goes beyond the belief of separation and treating languages as independent structures. In fact, translanguaging instead refers to individuals with two or more languages as having “one linguistic repertoire” (Garcia & Wei, 2014, p 2). Garcia and Wei (2014), also explain that in contrast to code-switching, translanguaging does not force bilinguals to alternate between their languages when speaking, but instead allows them to pull from their whole repertoire when communicating. This means that teachers can use translanguaging as a way to connect with their pupils, give intricate instructions, manage their classrooms, accelerate the pupils learning process, and develop the grammatical knowledge of the learners (Macaro, cited in Garcia &

6

Wei, 2014). However, within the translanguaging approach there are typically two ways of using it; naturally and officially. The key difference between the two is what drives each approach and if its usage is planned or not. Not planned and natural, translanguaging can be used by the pupils themselves in order to learn, or by the teachers as a way to check what the pupils have understood. Planned and official, translanguaging may be used by the teachers to expand the information that they are giving, or by the learners themselves in order to demonstrate their full knowledge of the subject at hand (Williams cited in Garcia & Wei, 2014).

Elaborating on the subject, Lewis, Jones and Baker (2012) emphasize that teachers who translate are not necessarily teachers who translanguate. This is because translation in its purest form is about separating the languages from each other, thus causing pupils to only process content in their more advanced language. Nevertheless, when used strategically and in combination with other methods, translations are definitely a part of translanguaging. For example, Lewis, Jones and Baker (2012) state when teachers intentionally choose to translate certain parts of the subject matter in order to support their learners' understanding, they are translanguaging. Another example of when translations are a part of translanguaging is when the teachers use it with the purpose of teaching the pupils specific terms and to aid them finishing assignments. Additionally, translations can be specifically used to help individual second language learners with the interpretation and understanding of teachings (Garcia & Kleifgen cited in, Lewis, Jones & Baker, 2012). However, this is only true as long as the usage of the stronger language is not favoured (Williams cited in Lewis, Jones & Baker, 2012).

Translanguaging as a concept is still in transition. Gaining more ground within the field of education, the definition has started to move closer towards the field of code-switching (Nagy, 2018). This is also mirrored in the connection between code-switching and translanguaging made in the CEFR and further publications from the EU (Council of Europe 2019; Language Policy Division, Strasbourg, 2009). It also has foundation in independent research. Lewis, Jones and Baker (2012) finds support for a close connection in the classroom context between the concepts in the works of Chitera (2009) Kamwangamalu (2010) and Wei (2011). Furthermore, Lewis Jones and Baker (2012) states that there is “clearly much overlap between code-switching and translanguaging” (p. 659), while the concepts target different fields of research. Garcia (2009) supports these statements stating that translanguaging encompasses

7

code-switching and other bilingual features, but eventually goes beyond the sole concept of code-switching.

3.1.1. Code-switching

Today, many researchers adhere to or to a variation of Poplack's (1980) definition where code-switching is described as the "alternation of two languages within a single discourse, sentence or constituent" (p. 583). However historically, the term code-switching has had various definitions since the 1950s when the term switching code was minted. According to Alvarez-Caccamo (1998) there has been a transformation from a view where code-switching is researched as a linguistic action, where languages are alternated in grammar and lexicon, to a more varied research which contains not only linguistic interest, but also conversional and anthropological work on the subject. It subsequently follows that the definition of the term code-switching has gone through a metamorphosis since its origin.

According to Lewis, Jones and Baker (2012) there is a distinction between the denotations of code-switching and translanguaging, as code-switching is considered a linguistic term that describes the analysis of bilingual speech patterns, whereas translanguaging is situated, sociolinguistic and ecological. Lewis, Jones and Baker (2012) also highlight a difference in two diverse ideologies, where code-switching has an affiliation with a separation between different languages while translanguaging favours adaptability in language use and the learning through two or more languages. Nevertheless, there is much overlap between the denotations for code-switching and translanguaging, which makes the connotations of the two terms almost identical and the two terms are used interchangeably by many researchers. One example of this is Wei (2011) who states that “code-switching is not simply a combination and mixture of two languages but creative strategies by the language user” (p. 374) and should be considered a multicultural approach.

Code-switching is not only a substitution of one lemma for another, the proficient speaker also displays a recognition of the shift between the different languages, indicated by seamless transitions without hesitations, false starts and repetitions. According to Garcia (2009), code-switching has two different varieties. One variation is intrasentential, where the switch between languages transpire within a clause or a sentence, and one is intersentential where the switch is outside a clause or sentence. Furthermore, the underlying purpose of switching between languages should be the provision of greater meaning than conveying meaning of

8

untranslatable glossary (Poplack, 1980). Poplack and Garcia's definitions target naturalistic code-switching where the code-switching is not a construct, but a natural occurrence. This kind of code-switching has traditionally been researched within the field of sociolinguistics with the focus point on structural issues, where the research often examines language choice as a reflection of power structures. The other field that concerns itself with code-switching focuses on grammar and examines syntactic restrictions within a chosen theory (Auer, 1998). A third field has also emerged where the research examines a virtual construct, such as the classroom context. This field targets interactional code-switching, where the switching of languages correlates with the preferences of the participants to facilitate the meaning of a pronouncement (Auer, 1998).

Research focusing on learning situations show that teachers use code-switching for various reasons, but foremost when expecting that learners will have problems understanding an utterance in the target language, thus prompting the teacher to fall back on the L1 (Nussbaum, 1990, quoted in Liebscher & Dailey-O'Cain, 2005). Moore’s (2002) research corroborates these findings and highlights more uses for codeswitching, such as a corrective function from the L1 to the second language by repeating the same glossary in L1 and L2. Furthermore, Moore (2002) has documented evidence that the teacher encourages code-switching to override communication issues where the learner’s insufficient knowledge of the second language is displayed. When used in the same way by the teacher, the code-switching takes a coaching turn where the teacher tries to supply support to coax out the correct sentence from the learner. Moore (2002) believes that using code-switching this way could be compared with different methods and modalities to teach a second language. Additionally, Macaro’s (2001) research on student teachers shows a frequent use of the learners L1 by the teacher for admonishing pupils and keeping control in the classroom.

The scientific discourse on educational code-switching contains many diverse opinions on whether code-switching benefits the learners (McMillan & Rivers, 2011). Ellis and Shintani (2014) concur, stating that a positive outlook among the teacher community on the use of L1 might not be due to a perceived cognitive benefit, but a reaction against a native-speaker ideal or a response to an urgent demand in the classroom. In addition, Ellis and Shintani (2014) are of the opinion that there is a lack of research on the effects of the use of L1, which makes it difficult to draw conclusions on the general benefits of the use of L1 within teaching. However, Garcia (2009) concludes that if properly applied code-switching can enhance

9

learners’ cognitive skills, for instance Jacobsen’s Concurrent Approach which uses intersentential switches as a method to provide support and reinforcement.

3.1.2. Summary

Since the view on translanguaging emanates from an interpretation of a context-based occurrence, the definition of the concept varies with the perspective of the researcher. What is indisputable, is that the term contains many multilingual phenomena. Code-switching is one of the most prominent features within this concept, due to the great overlap between the denotations and connotations of the two terms (Lewis, Jones & Baker, 2012). However, translanguaging encompasses more than linguistic terms (Garcia, 2009). The definition of the term might therefore be that translanguaging is a social phenomenon pertaining to the subject's identity, group affiliation and culture which is displayed with linguistic features, such as code-switching. These multi-linguistic elements are used for mediation in communication, to further understanding and create meaning.

3.2. Teachers’ view on the roles of first and second languages

in language education

Teachers’ beliefs influence decisions and views in the classroom and pedagogical methods. In an extensive review of research Kagan (1992) has concluded that pre-existing beliefs and prior experiences play a substantial role in the teacher’s practice. What is most influential in changing these above-mentioned beliefs is not theory, but the influence of practical classroom variables, such as the pupils' parents and the principal’s and co-worker’s attitudes toward teaching. Grenfall’s (1998) research corroborates this, stating that the dominating view is that teachers believe that their own and their colleagues’ craft knowledge supersedes theory. Elliott (1991) agrees with this expressing the opinion that in reality educational beliefs among teachers often are derived from practice and professional discourse instead of application of theories learned during the teacher education.

Macaro (1997, 2001, 2009) has through extensive research drawn the conclusion that teachers can be split up in three groups where each group primarily represents one personal theory about the usage of first and second languages in the language classroom. One group believes that knowledge of the second language only can be obtained through that language and that the classroom should echo the environment that migrants and children learning the language

10

are subjected to in real life. The second group believes that the target language only can be fully obtained by the target language, but that the classroom could never mirror the reality needed for these perfect learning conditions. This leads to the conclusion that the teacher should utilise the target language as much as possible. The third group believes that there are values in using the L1, and that in certain occasions usage of the L1 enhances the learning. Macaro has labelled these three positions the virtual position, the maximal position and the optional position (Macaro, 2009).

3.3. Aspects of translanguaging within the steering documents

Skolverket (2019a) states that the Swedish steering documents, Lgr11, are based on the CEFR. In the commentary to the English Curriculum, Skolverket (2017) proposes that the levels and categorisation in the curriculum mirrors the categories - or what the Council of Europe (2001) denotes activities - and levels in CEFR. In the original version of CEFR, Council of Europe (2001) labels these Reception, Production and Interaction, which in Lgr11 could be translated into Listening and reading (reception), Speaking, writing and conversing (production and interaction).

As previously noted, the quote “Language is the primary tool human beings use for thinking, communicating and learning. Having a knowledge of several languages can provide new perspectives on the surrounding world, enhanced opportunities to create contacts and greater understanding of different ways of living” (Skolverket, 2018, p. 34) could mildly be interpreted to favour plurilingualism, which could be defined as making use of two or more languages for communicative purposes (Crystal, 2008) and by extension translanguaging, which is a part of this concept. However, Skolverket (2018) does not put much emphasis on this aspect of language learning. In contrast, plurilingualism has become more important in the Council of Europe’s view on language learning. In the CEFR it is stated that language education should aim to give learners an opportunity to develop competence within this field (Council of Europe, 2001).

Furthermore, in the CEFR, the definition of plurilingualism encompasses both definitions of translanguaging and code-switching stating that the plurilingual approach accentuates that when an individual’s experience with an language in its context expands, the borders become fluid between different languages and the person “builds up a communicative competence to which all knowledge and experience of language contributes and in which languages

11

interrelate and interact” (Council of Europe, 2001 pp. 3-4). The methods of communication can vary, from code-switching to taking advantage of root-knowledge of one language to understanding another language (Council of Europe, 2001). The Council of Europe (2001) further expands on this stating that mediation which consists of translation and interpretation, is a large part of the language and communication process.

The focus on plurilingualism and mediation can in retrospect be said to foreshadow the addition made in the companion volume, where mediation has been added to the three previous activities, reception, production and interaction (Council of Europe, 2018). In the upcoming revision of the curriculum for the elementary school (Skolverket, 2019b) mediation has not been added, which could be due to the expectations of language levels the pupils are expected to reach. In year 4-6 you are still considered a basic user and in years 7-9 reaching the beginning of the intermediate user-scale (Skolverket, 2017). In the proposal for the curriculum for upper secondary school (Skolverket, 2019c), however, it is now stated that the pupils should be given an opportunity to develop plurilingualism, which could be considered a large step from the previous statement in the commentary where it is written that the English subject in essence should be taught using the target language (Skolverket, 2011)

3.4. Recent research in code-switching and translanguaging

3.4.1. Teachers attitudes/perceptions

Further research has been done regarding teachers' beliefs concerning the use of the pupils’ L1 within English education. With a focus on Communicative Language Teaching (hereafter abbreviated CLT) McMillan and Rivers (2011) conducted a project with 29 foreign language teachers in Japan as participants. Using an anonymous survey, McMillan and Rivers (2011) examined the teachers’ feelings toward their own and the pupils’ use of the pupils’ L1. The result showed mixed feelings toward both teachers’ and pupils’ use of the L1. However, a significant number believed that the L1 could play a positive role in both teaching and learning. Surprisingly, the teachers’ positive and negative perceptions did not mirror their own competence within the pupils’ L1. The findings showed that many of the teachers that were positive towards the pupils’ use of Japanese were not themselves proficient Japanese speakers, albeit all teachers agreed that overuse of the L1 is detrimental for the pupils. McMillan and Rivers (2011) also noticed that the teachers in favour of sagacious use of the L1 could provide more theoretical background for their reasoning than the arguments against

12

the use of the L1. This prompted the conclusion that many teachers' beliefs and pedagogical methods are influenced by external actors and that the teachers that believe that the L1 might play a positive role may feel a pressure to exclude it from the methodology (McMillan & Rivers, 2011).

Conducted in the United States, Nambisan’s (2014) research study investigated nineteen English language teachers’ opinions on and methods of using the pupils’ primary languages during class. The researcher made connections to the study conducted by McMillan and Rivers and used a survey to collect data. During this, Nambisan (2014) discovered that the majority of the teachers from the survey pool used English as the main language of instruction but considered translanguaging as helpful for developing a pupils’ language. However, Nambisan (2014) highlights the fact that one of the teachers felt translanguaging was not as beneficial as sole use of the targeted language. Nambisan (2014) also declares that the teachers rarely saw or promoted the use of the learners’ primary languages when discussing subject matter during the lessons. Nevertheless, most of the teachers also reported using translanguaging as a way to increase pupils’ engagement, and to support less skilled learners. Nambisan (2014) concludes that although the teachers generally thought of translanguaging as beneficial, they did not see it as an effective way to connect with their pupils.

Burton’s and Rajendram’s (2019) research study, which was conducted at a Canadian University, also focuses on the use of translanguaging in schools and explores teachers' views on it in an English language learning context. During this, Burton and Rajendram (2019) discovered that most of the teachers thought the use of translanguaging was regressive rather than progressive in terms of developing the pupils’ skills in English. One of the teachers, however, felt translanguaging could be used as a resource to develop language and the pupils should decide if and how they wanted to use it in the classroom. On the other hand, most of them considered translanguaging as useful when teaching or instructing the pupils about things that involved their real lives, although some felt it was best used with lower levels and beginner pupils. Additionally, some of the teachers mentioned that, due to time constraints, they determined that the limited opportunities the pupils had to speak English were not to be wasted on using other languages (Burton & Rajendram, 2019).

Yuvayapan’s (2019) research study conducted in Turkey and included 50 EFL teachers from different schools in the country. Exploring English language teachers’ perceptions of translanguaging, Yuvayapan (2019) used a questionnaire, observations and semi structured

13

interviews to find answers for her research questions. These set of questions asked were, “What are the perceptions and practises of English language teachers towards translanguaging? (Yuvayapan, 2019, p. 682), and “For what purpose do EFL teachers use Turkish in their classrooms?” (Yuvayapan, 2019, p. 682). During these interviews Yuvayapan (2019) discovered that the teachers recognized the use of translanguaging as beneficial for helping lower-proficiency pupils with their learning. They also reported themselves as actively promoting the use of translanguaging within their classrooms. However, Yuvayapan (2019) also notes that the teachers seemed to avoid using the pupils’ L1s in certain instances, and that this went against what they stated earlier about their own perceptions. Yuvayapan (2019) states a possible explanation for this is due to a predominant belief amongst Turkish teachers, as well as those who enact policy changes, that developing skills similar to primary English speakers requires less application and prioritizing of Turkish. Thus, the use of translanguaging is not promoted as it is seen as being a hindrance for the development of the pupils’ second language. Yuvayapan (2019) concludes that the teachers did not use translanguaging in a systematic way, and therefore it did not help the pupils with achieving set goals.

Menken and Sánchez (2019) examined how the introduction of translanguaging in schools changed “educational practices, language ideologies, and/or language education policies” (p. 3). Using data from eight New York public schools to find answers for their research, Menken and Sánchez (2019) found that, over time, the teachers had become more positive towards the use of translanguaging. For example, some teachers went from holding monolinguistic views, forbidding their pupils to speak any other language than English, to taking a more translanguaging stance. Menken and Sánchez (2019) mention that this means, with the introduction of translanguaging, that the teachers had started to think of the pupils’ home languages as a resource for developing their English language. Moreover, several teachers noted that some pupils’ engagement had increased with the use of translanguaging, and these pupils had also started to take more responsibility over their own learning. Furthermore, the teachers stated that some pupils seemed to feel comfortable because they had implemented translanguaging in their education. Additionally, Menken and Sánchez (2019) noted that other teachers also began using different translanguaging methods during their lessons after seeing the success of teachers who used translanguaging.

14

3.4.2. Pedagogical Methods

Makonye’s (2019) research study, conducted in Zimbabwe, points out the positive effects of using translanguaging in learning mathematical concepts. Examining the use of translanguaging, Makonye (2019) had sixth-grade pupils divided into two experimental translanguaging groups and two non-translanguaging control groups. The pupils in the experimental groups received their lessons partially in their home language, whereas the pupils in the control group had their lessons conducted only in English. At the conclusion of the study, Makonye (2019) found that there was significant improvement in the understanding of perimeter and area amongst pupils that had been in the experimental group. Makonye (2019) states a possible explanation for this is that translanguaging helps pupils with learning mathematics when it is used as a strategy to provide scaffolding. Furthermore, Makonye (2019) states that “One of the most important roles of education is to cater for learner diversity in all its forms” (p. 224) and, therefore, different languages should be used in the classroom. Makonye (2019) adds, from a Vygotskian perspective, that a teacher “cannot ignore the contexts of its learners such as their home language” (Makonye, 2019, p. 222) if they want their pupils to successfully learn new things.

Duarte’s (2019) research corroborates Makonye’s conclusion. The aforementioned study draws on observations of task-based lessons with an aim to find answers concerning pupils' interaction and the role of translanguaging in peer-to-peer-interaction. The analysis of the result yielded key segments that showed that translanguaging and code-switching could be used to scaffold meaning, advance understanding and “reinforce the creative process of knowledge building, by mediating the emergence of high-order thinking” (Duarte, 2019, p. 162). Duarte (2019) further concludes that translanguaging is beneficial, if used systematically and in a transparent fashion, suggesting that allowing translanguaging in phases of collaborative talk might be a tool to further the quality of pupils’ language and work.

15

4. Methods

This study aims to partly replicate three studies conducted in three different countries. Two of the studies carried out in Japan (McMillan and Rivers, 2011) and in the USA (Nambisan, 2014) use surveys which are analysed with both quantitative and qualitative methods. The third study carried out in Turkey (Yuvayapan, 2019) uses qualitative interviews combined with a quantitative survey.

The purpose is to investigate teachers’ perceptions of the use of translanguaging in English language teaching. Since we believe that perceptions are a subjective matter, qualitative methods were chosen to find answers to the research questions. As a research design, methodological triangulation was chosen in order to increase the validity of the study and to gain a deeper understanding of the research questions (Alvehus, 2013). Interviews, observations and a survey were specifically selected as tools to collect the data.

The language primarily used to collect the data has been Swedish. The choice of language was based on the belief that the use of a language that is most recognized to both the interviewers and interviewees is preferable to avoid linguistic misunderstandings and enhance nuances (Filep, 2009), thus gaining more richness in our material.

4.1. Interviews

Part of the data was collected by means of interviews, since this is the preferred method to gain a better understanding of research subjects’ perceptions and experiences (Alvehus, 2013). More specifically, semi-structured interviews where the respondents had to answer open ended questions were utilized. The rationale for choosing this approach was that this form of interview has the advantage of giving the conversation a flexible aspect, letting the interviewer and interviewee follow points of interest that arise during the dialogue (Denscombe, 2014). The interviews were conducted face-to-face. The possibility that the personal effect could incur, where the interviewees are influenced by the identity of the interviewers (Denscombe, 2014), has been taken into account. The teachers that were interviewed might be influenced by the fact that we are students examining their attitudes within a field where we soon might be colleagues. A counter to this effect has been to correlate the answers with observation data.

16

The interviews were conducted at the workplaces and classrooms of the interviewees. This method, suggested by Ryen (2004), was to counter the critique posed by Kvale and Brinkmann (2014) that many qualitative interviews are immovable and removed from the environment pertaining to the interview. The conversations were, in all but one case, recorded by equipment supplied by Malmö University. In one of the cases, the interviewee declined being recorded and field notes were used instead. The interviews were later transcribed. No sensitive data were stored in a cloud service. Furthermore, the interview questions were sent to the teachers in advance, thus making the interviewees more comfortable, and give everyone a chance to consider beforehand. Qualitative interviews are criticized for being verbalised (Ryen, 2011), in other words focusing on the interviewees' thoughts and experiences instead of actual actions. This has been countered by observations of the teachers in the classroom context.

4.2. Observations

The observations have been conducted within the classroom environment and focused on the teacher. The pupils’ actions and reactions have thus not been recorded. This method is often used to gain insight into naturally occurring and context-specific behaviours (Alvehus, 2013, pp. 94-95) and complements the conducted interviews. The mediation mode used has been observation charts and written comments. The observation charts are used as a means to counter the fact that observations and perceptions will always be selective (Denscombe, 2014) and increase objectivity. To further add clarification, context and background information field notes have been used (Denscombe, 2014). When analysing the results, the fact that observations only records what happens, not why (Denscombe, 2014) has been taken into consideration. The interviews have been used to clarify and explain the information that was gathered.

4.3. Survey

Paired with the other methods, a survey has also been used to collect data. While in some cases, it is preferable to do a pilot survey, this survey builds on previously validated and published surveys by Nambisan in 2014 and Yuvayapan in 2019. Thus, it was decided that no pilot was needed, since the survey has been previously tested and the validity has been tried. The survey consisted of ten questions in total (see appendix) and the planned survey length for the participants was three to five minutes, thus making it more attractive to participate.

17

The participants have been chosen with the help of Grundskoleförvaltningen in Malmö. The distribution to teachers for grades 4-6 has been conducted by email from Grundskoleförvaltningen to all directors of studies and principals within the municipality, thus gaining objectivity and randomization. Furthermore, an additional reminder was emailed one week later to all principals and vice principals of the schools that received the first letter. The answers from the recipients of the emails have varied, but many have answered stating that the lack of time only affords them to help the students at Malmö University that are doing their vocational training with them.

The survey received 17 responses. The response rate was previously indicated by the e-mails received from the principals. The data were gathered and compiled using Sunet Survey, which ensured anonymity and the correct handling of data according to referenced laws and ethical guidelines. A quantitative analysis was conducted on the data. The survey in its entire form can be found in the appendix.

4.4. Ethical and Legal Considerations

When conducting this research Vetenskapsrådet’s ethical guidelines have been followed. The participants have all been informed about the research and consented in writing to participate in accordance to Vetenskapsrådet’s ethical guidelines (Vetenskapsrådet, 2017; Art. 6:1: a,

Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016

General Data Protection Regulation on the protection of natural persons with regard to the processing of personal data and on the free movement of such data (GDPR)). An anonymisation has been achieved by eliminating the connection between the answers and the participants (Vetenskapsrådet, 2017). Furthermore, the participants were informed of the fact that they can withdraw their consent and have their contribution erased during the entire process (Vetenskapsrådet, 2017; GDPR Art 17:1:a). A copy of the consent form can be found in the appendix.

Additionally, in the name of transparency, the authors of this project sought to maintain as an objective approach as possible to the subject material.

18

4.5. The participants of the interviews and observations

Out of twelve elementary schools that were initially contacted, seven schools showed an interest in participating in our study. The schools that were contacted were all situated in southern Sweden, in three different municipalities - one larger city, one small town and one municipality that could be defined as a suburb. Furthermore, the schools were selected on the basis of contextual reasons and aimed to give a diverse foundation for the interviews and observations. The selection targeted four different classroom contexts pertaining to the cultural and linguistic backgrounds of the pupils, as well as the pedagogical methods the schools adhered to. However, due to some time limitations and unforeseen events, two schools had to decline.

Five teachers eventually participated in the study. The teachers were, in all but one case, invited by the school’s principal to take part in our research and five interviews and four observations were organised. Due to special needs of the pupils, an observation could not be conducted in one of the classes, but an interview with the teacher was organised in the after-school-hours. In the result section our interviewees are named Teacher 1, 2, 3, 4 and 5.

The teachers all work in year 4-6 with a work experience that ranges from one to twenty-five years. Three of the teachers work in schools with a high quota of multilingual pupils. Two teachers teach mainly monolingual pupils, or pupils where Swedish is spoken at home. One of the schools adhere to the method of immersive English, whereas the rest of the schools have Swedish as the main school language.

19

5. Results

In this section we will report and analyse the results gathered through the three above mentioned methods. The results are organised in sections according to the method through which they were obtained - a survey, interviews and observations. Finally, we analyse and discuss the results in relation to the referenced framework in the theoretical section.

5.1. Result of the survey

The majority of the respondents (14 out of 17) work in schools where Swedish is the main language. On the question of which language that is the main language used in the English lessons, a clear majority (14) have answered Swedish and English, 2 have reported English and 1 Swedish. Furthermore, the result of the survey showed that a majority of the respondents (15 out of 16) believed that the use of the main language is favourable.

5.1.1. Teachers’ perceptions of their own language use

Two of the survey’s questions addressed the teachers’ perceptions and practices towards their use of the pupils’ first language. The questions were formulated to show in which situations the pupils’ L1 was used and the teacher’s perceptions of how beneficial they deemed the usage to be. The first question (see fig. 1) suggested certain situations where the pupils’ L1 could be used. The situations were helping pupils with a lower language proficiency, making quick clarifications during activities, bonding with pupils, praising pupils, giving feedback to pupils, situations demanding specific pedagogical leadership, giving instructions, explaining vocabulary and explaining concepts. A majority of the respondents answered never/not often to almost all the proposed situations in the first question.

20

Figure 1 Percentages of the teachers’ estimations of their use of the pupils’ L1

The only situation where a majority of the teachers perceived themselves often using the pupils’ L1 was when explaining concepts. In this situation, 8 out of 17 respondents answered never/not often. The situation concerning support to pupils with lower language proficiency generated the same reply, where 8 out of 17 teachers responded that they use the pupils L1 in this situation. However, in situations concerning instructions a larger number (11 out of 17) perceived themselves using English. Similar results were reported when making quick clarifications.

The second question asked in which situations the teachers believed it to be important to use the pupils’ L1 within the English education. Overall, the teachers’ answer yielded a different result than the question about actual use (see fig. 2). Here the majority of the respondents replied important or very important in all situations.

21

Figure 2 Percentages of the teachers’ belief about the importance of the teachers use of the pupils’ L1

All of the teachers believed the use of the pupils’ L1 to be important or very important when explaining concepts and when supporting pupils with lesser language proficiency. The use of the pupils L1 was also seen as beneficial when giving instructions, but not as high on the list. 13 teachers believed the use of L1 to be important in this instance. Almost the same number (12) believed the use of L1 to be important for making quick clarifications, as well as when explaining vocabulary.

5.1.2. Teachers’ perceptions of their pupils’ language use

In the survey, two of the questions were formulated to target the examination of the teachers’ perception and opinion of their pupils’ use of their L1 in the English classroom. Also, here the first question within this target area proposed certain situations where the pupils’ L1 could be used. The situations addressed were to discuss content and activities in small groups, to give aid to classmates during activities, to brainstorm during activities, to explain problems not related to lesson content, to enable participation for pupils with lower language proficiency, to answer teachers questions and to ask for permission.

22

Figure 3 Percentages about the estimation of allowance or encouragement of pupils’ use of L1

When asked about the pupils' use of the L1 (see fig. 3), a majority of the respondents answered never/not often in the majority in four situations. There was a great agreement that English was used when asking permission, answering the teacher’s questions, brainstorming during activities, discussing content or activities in small groups. discussing content and activities in small groups, brainstorming during activities, answering the teachers’ questions and asking permission.

More participants replied that pupils used their L1 when aiding classmates during activities (11) and enabling participation for pupils with lower proficiency (12). Comparable results were reported in the situation pertaining to explaining problems not related to lesson content, where 10 answered in a positive fashion.

The final question (see fig. 4) asked how important the responding teachers believed the pupils’ usage of L1 was in the above-mentioned situations.

23

Figure 4 Percentages about the teachers’ belief about the importance of the pupils use of L1 in English classes

All of the respondents believed the pupils’ use of L1 to be important when giving aid to classmates during activities. Another situation high on the list was enabling participants with lower language proficiency, where only 1 participant deemed the use of L1 to be unimportant.

The two situations concerning pupils’ usage of L1 while discussing content and activities in small groups and brainstorming yielded similar results. In the two instances 13 and 14 of the respondents respectively believed a use of L1 to be important or very important.

The situations concerning social care issues also yielded results displaying a positive view of the use of L1. A clear majority believed that the pupils’ use of L1 was important when explaining problems not related to lesson content (15). However, when asking permission more (6) respondents than in the previous situation believed that the use of English was beneficial.

5.2. Results of five interviews

The results of the interviews have been divided into two themes. The first theme targets the perception of the teachers’ own language use. The second theme focuses on the teachers' perception of their pupils’ language use.

24

5.2.1. Teachers’ perceptions of their own language use

All of the interviewed teachers were in agreement about the languages utilized in the classroom during the English lessons. The predominant languages are Swedish and English. However, the share of the estimated usage between these languages varies between the teachers. Two of the teachers stated a high preference for using English in the classroom - one in a fully immersive way and another with high percentage of English (70%):

“Jag vill helst prata engelska hela tiden men sedan ibland så känner jag att jag har försökt säga på engelska först, sedan ibland säger jag på svenska. Men jag behöver inte göra det lika ofta längre” [I prefer to speak English all the time, but sometimes I feel that I have tried to say something in English, but then afterwards in Swedish. But I don’t have to do that as often anymore.].

Two of the other teachers estimated their English-use to be within the spectrum of 40 percent (English) and 60 percent (Swedish) to a 50-50-percent-usage. However, one teacher stressed the importance of using Swedish in the classroom:

“Många säger att man ska använda mycket engelska i klassrummet, men jag tycker att man tappar de svaga eleverna om man använder för mycket engelska.” [Many people say that one should use much English in the classroom, but I believe that one loses the weaker pupils if one uses too much English.].

Even though teacher 2 emphasized the importance of the use of Swedish, she added that the share of English could be increased with the pupils’ age and proficiency: “Uppe i sexan pratar jag mer engelska och ber någon elev översätta vad jag sagt” [In sixth grade, I speak more English and ask a pupil to translate what I have said.]. The increase of percentage of the teachers’ English use in combination with the pupils’ age and proficiency is corroborated by all the interviewees. However, the focus is more on proficiency than age, meaning a fourth-grade class is sometimes more developed in their understanding of English than a sixth-fourth-grade class. Subsequently, in certain cases more English could be used in the fourth grade than the sixth.

25

The systematic use of English and Swedish is by the teachers’ estimations varying in different degrees and modality. Four of the teachers stated that they did have a systematic plan for their use of the two languages. Teacher 1 did not consider herself having a pre-decided system but amended this statement by stating that instructions were usually spoken in English and Swedish.

The value of the use of Swedish while giving instructions seems to be an emergent theme. However, the teachers’ use of Swedish differs within this context. Two of the teachers accounted for a system using English - Swedish - English, or as teacher 3 expresses it: “[...]with more beginners I give the instruction in English, translation in Swedish and then repeat it in English”. Teacher 2 expressed a parallel use of both languages with instruction, but with an emphasis on Swedish:

“Vid instruktioner används alltid både engelska och svenska, främst svenska. Först säger jag sakerna på engelska sedan översätter jag till svenska. Ibland används bara svenska.” [When giving instructions, both English and Swedish are utilized, for and foremost Swedish. First, I’ll say the things in English, then I translate into Swedish. Sometimes only Swedish is used.].

Teacher 1 agrees with this system, stating the reason behind this approach:

“Jag kan ta dem på engelska innan också, men främst på svenska... så att jag är säker på att de förstår vad det är de ska göra [...]” [I can give them in English before as well, but primarily in Swedish...so I am sure that they understand what they should do].

Another method a majority of the teachers also utilized was to say the instructions in English, then let a pupil translate it into Swedish. Teacher 5 explains the process:

“Den kör vi alltid på engelska, eller alltid på engelska först. Sen brukar jag [...] fråga om det är någon elev som kan översätta [...]. Så är det någon som kan översätta den så att den sammansättningen alltid kommer två gånger på, en gång på engelska och en gång på svenska.” [We always do it in English, or always English first. Then I usually [...] ask if there is a pupil who can translate [...] Then there is somebody who

26

can translate so the composition will always come twice, once in English and once in Swedish.].

Teacher 4 gives almost an identical explanation of a similar process.

Furthermore, the use of Swedish has been deemed to be beneficial when teaching grammar both by Teacher 2 and 4. Teacher 3 added an extra value in translating glossary from English to Swedish, when the word is abstract and hard to explain by other means. Two teachers mentioned translation as a means to save time and being able to help as many pupils as possible. Other uses of Swedish that are mentioned are when handling social care issues and other problems, where the pupils’ grasp of English is insufficient, or as Teacher 3 explains it:

“I feel that it is more important that they can communicate their needs and say what they need rather than force them and add that extra level of stress of a language that they can't necessarily communicate [sic.]. [...] It's more important that we understand each other and then we use the Swedish [sic.].

5.2.2. Teachers’ perceptions of their pupils’ language use

All teachers that participated in the interviews perceive that the pupils use Swedish (their L1) in the classroom during the English lesson. Although the result is the same, the reasons behind the usage of Swedish are perceived to be somewhat differentiated. The amount of the pupils’ English use is by the teachers associated with diverse reasons such as proficiency and self-confidence. Teacher 5 explains her opinion of the correlation:

“De som har ett språkligt självförtroende de pratar kanske 50/50 [engelska/svenska]. De som inte har det språkliga självförtroendet med sig de som är svaga i ämnet [...] de pratar 100 procent svenska” [Those who have a linguistic self-confidence might speak 50/50 [English/Swedish]. Those who do not have linguistic self-confidence or are weak within the subject [...] they speak 100 percent Swedish].

27

“de som är osäkra föredrar att ta svenskan, jag tror nog att de hade kunnat säga vissa saker på engelska om de bara hade försökt, men de är nog litet osäkra…” [those who are insecure prefer to use Swedish, I do believe that they could say some things in English, if only they tried, but they are probably a bit insecure…].

Another explanation provided by Teacher 3 is that the use of Swedish is predominant, since this is the language used between them outside the lesson:

“I had two students whose English is phenomenal, it’s... and they do really well, but as soon as I give them the topic they’ll start speaking in Swedish and I just heard, it’s like… uh English.... and they just went : ‘ohhh yes’ - and then they went.... and they can float in both languages, it is just that their natural inclination is with that person is that they normally speak that language so it sometimes can be hard”.

The above-mentioned statement also shows how the teacher feels the need to remind the pupils to speak English. This is corroborated by three of the other teachers (Teachers 1, 4 and 5). Teacher 2 also expresses the need to remind the pupils of speaking English, but partly takes another view where the pupils choose their own quota of English versus Swedish:

“Eleverna använder främst svenska. Det är individuellt varierat hur mycket eleverna använder svenska och engelska i klassrummet. De som vill använder mer engelska - inte bara när jag påpekar att engelska ska användas. Det är ju viktigt med övning och träning. Men eleverna väljer automatiskt oftast att använda svenska – jag måste påpeka att de ska använda engelska.” [The pupils use predominantly Swedish. It varies how much the individual pupils use Swedish and English in the classroom. Those who want to use more English [use more English], not only when I point out that English should be used. It is important with practice and training. But the pupils most often automatically choose to use Swedish - I have to indicate that they should use English.]

The pupils’ use of Swedish is not always seen as negative. One of the teachers (Teacher 2) displays a pronounced positive view on the pupils’ use of Swedish within the English class. The other four teachers demonstrate an ambivalence, where both benefits and drawbacks are

28

discussed. There are specific examples where the teachers can see Swedish as an asset. For instance, one teacher - Teacher 4 - highlighted a benefit using parallel teachings within grammar. At one time period, the teacher taught English grammar in both Swedish and English. During the same time period the equivalent within Swedish grammar was taught during the Swedish lessons. The teacher perceived that the pupils’ comprehension of grammar was heightened due to the dual use of terminology in both languages. The result was a positive interaction, not only from Swedish to English, but also from English to Swedish. Concerns are, however, voiced that using too much Swedish is detrimental. Teacher 5 expresses a wish to speak only English, but summarises the draw-back she perceives with English-only use:

“Ja, svenskan hjälper dem även om det kanske stjälper dem också om man pratar rent forskning [sic.] att skapa ett gemensamt språk kring engelskan - [...] jag hade nog tappat 50 procent av mina klasser om jag hade pratat bara engelska. Det är svårt.” [Yes, the Swedish helps them, even if it might also trip them up, if we are talking research [sic.] and to create a common language around English - [...] I believe I would have lost 50 percent of my classes if I had only spoken English. It is difficult.].

Teacher 1 wants her pupils to communicate more in English and states that she wants them to push through their insecurities, so they learn more. Teacher 3 conducts almost all her teaching in English, in an immersive program, but states some apprehension concerning only-English use, “[...] if you only do immersive teaching, you value one language stronger than the other and I think it's important for them to value the Swedish heritage that they have as well with them”.

Finally, as an answer to the question if the teachers had heard or knew anything about translanguaging, all the teachers replied that they had not come across translanguaging as a pedagogical method. Two of the teachers - Teachers 3 and 4 - were positive towards the possibilities and learning more about it. Furthermore, Teacher 3 stated that she had experience of the concept of translanguaging - but not the term - from research and literature.

29

5.3. Results of four observations

During the four classroom observations, all the teachers used Swedish during their English teaching to varying degrees. Two of the teachers used it sparsely, one of the teachers used it intermittently and one of the teachers used Swedish the majority of the time.

When issuing instructions, all but one teacher used similar patterns of English-Swedish or English-Swedish-English. One example of this is Teacher 4: “Gather in groups - samlas i gruppen” or Teacher 2: “Don’t forget your homework – glöm inte er läxa – have a nice meal and don’t forget to wash your hands”. Furthermore, this parallel use of English and Swedish was sometimes combined with more Swedish adding extra information, for instance “It still has to go with the story – det måste ändå höra ihop – annars blir det inte roligt” and “You have five minutes – ni har fem minuter på er, vi kanske kan fortsätta senare.” (both quotes Teacher 4). In some instances, it was even emphasised that Swedish was used for the sole purpose of making clear that everybody has understood the instructions: “Jag tar det på svenska så att alla vet vad vi ska göra…” [ I’ll say it in Swedish, so we all know what to do…] (Teacher 1).

The choice of language outside of instructions varied. When answering the pupils two of the teachers (Teacher 3 and 4) spoke English when receiving questions in Swedish. One of the teachers (Teacher 1) switched between answering in English and answering in Swedish. On occasion, the same teacher also pretended not to understand when the pupils spoke Swedish to her, coaxing them to speak English. Teacher 2 almost exclusively spoke Swedish to the pupils.

Translation of glossary occurred during most observations. The most common form was translations from English to Swedish. Three of the teachers also translated from Swedish to English, with the purpose of assisting pupils to find the correct word. In some instances, it also consisted of supplying and translating a word during conversation, when the pupils’ own vocabulary was insufficient. The translation of glossary was at some times at the behest of the pupils and sometimes initiated by the teacher. Teacher 3 only used Swedish when translating glossary from English to Swedish, when a pupil pointed out the need to translate the word. In all classroom contexts Swedish was used by the teachers as a means to quickly clarify for one or more students without disrupting the activity. Swedish was also used, when the teacher

30

deemed the subject too demanding in comparison with the pupils’ knowledge of English - such as grammar: “Would you say that this is written in present tense or past tense – Jag tar det på Svenska nu när det är grammatik….”(Teacher 4). Moreover, Swedish was used to keep order - for instance: “Kan du hålla reda på denna – in your orange…” (Teacher 4) and “Här ska vara tyst – tyst nu!” (Teacher 1). However, the teachers seemed to prefer to start with English to keep order and supplemented with Swedish to emphasise.

Finally, Swedish was also used by two of the teachers (Teacher 1 and 2) as a way to build bonds with their pupils. One example of this is when Teacher 1 states: ”Ni börjar bli trötta och det förstår jag – och ni har varit superduktiga!” [you are starting to get tired and I can understand that - and you have been really, really good!] or Teacher 2 says: “Det är därför du tränar, vännen” [That’s why you are practicing, my friend]. Feedback could also be considered a bond-building activity, but here both Swedish and English were used by all teachers to varying degrees.