Interaction and meaning making in basic adult education for immigrants the case of Swedish for immigrants in Sweden (SFI)

Abstract

In this article, focus is on students’ oral production in two classrooms in Swedish for immigrants (SFI). The study focuses on practices with interaction patterns where students are involved in negotiation of meaning. Theoretical basis is the importance of interpersonal interaction for language development, with a focus on students’ use of varied speaker roles and speech actions. Data were obtained through observations, field notes, and audio and video recordings. The examples presented here demonstrate that students were involved in the negotiation of meaning and had space to try different speaker roles and speech actions, such as (among others) initiating, agreeing, dissenting, arguing, interrupting, and taking the floor. These examples only constituted a restricted part of class time, and most of the teaching was of a type where little oral or written interaction took place. Overall, we think that SFI education could be improved by developing teaching that stimulates and allows for negotiations of meaning and language production and also allows for use of different types of digital media, both for oral and written interaction.

Keywords:

• Classroom interaction
• Swedish for immigrants
• SFI
• adult education
• speaker roles
• speech actions

Introduction

Human migration as part of globalisation processes create a need for education for adults in basic skills in national languages. Processes of migration have also resulted in the emergence of new perspectives of language as multiple resources which are object to continuous change (Blommaert 2010, Conteh and Meier 2014, May 2014). However, research following these perspectives has rarely studied spaces for language development, which is our focal point here. This article focuses on adult education in Swedish for immigrants (SFI), specifically in what is called study path one and two. SFI is organised at the municipal level and study path one includes students with less than a few years of earlier schooling. Path two is for students with less than nine years of previous formal education. SFI’s focus on functional language skills is an important key for an immigrant’s labour market and social integration. The goals of SFI are to offer support and stimulation for the development of knowledge and competence, to strengthen the students’ positions socially and in the labour market, and to support personal development (School Ordinance 2010:100, 20, Ch. 2).

A perceived problem expressed in current public debates and among Swedish teachers is that it takes too long for some students (specifically those in study paths one and two) to pass the final SFI-test, which comes after the final course, D. Earlier studies from SFI (Wedin et al. 2018, Wedin and Norlund Shaswar 2019) showed that students are given little space for their own language production in situations similar to what they meet outside the classroom, as well as little variation regarding students’ language use in classrooms. These findings show one potential reason for the low passing rate.

The importance of interaction for language development is well documented in research that utilises various theoretical frameworks and thus, various arguments for the importance of interaction exists (see Swain 1985, Cummins 2000, Gass and Mackey 2015). There are, however, few studies on interactional patterns in SFI and of practices where adult students are given space to negotiate meaning by using varied interaction patterns which they, as adults, need on a daily basis. Thus, this study looks at classroom practices focussing on speaker roles and speech actions in such classroom interaction. It focuses on students’ negotiation of meaning by performing varied speech actions and varied speaker roles. Speech actions based on speech act theory (Austin 1962, Searle 1979, Varga 2013) is used to name and describe students’ talk in class. While speech actions refer to what students do with language, the action, speaker roles refers to positions taken, such as expert, novice and a friend. The study investigates students’ negotiation of meaning in two classrooms, and the aim is to explore practices where students are involved in varied speech actions and speaker roles in their negotiation of meaning. The following research questions are used:

1. What speech actions do students use?

2. What speaker roles do they play?

3. In what types of classroom practices does variation in these appear?

Theory and earlier research

The study takes its theoretical ground in research in a sociocultural view of language and education (Lantolf and Pavlenko 1995, Cummins 2000) and understanding of language as multiple and changing resources (Blommaert 2010, Canagarajah 2018). A sociocultural perspective on learning of a second language (L2) suggests language is perceived as communication and meaning making, with interaction, mediation and scaffolding as central concepts (Lantolf and Pavlenko 1995, Swain et al. 2015). Research on the importance of classroom interaction for students’ development of language (Ellis and Shintani 2015) shows that this development is to a high degree related to students’ own language production (Wedin and Norlund Shaswar 2019). From a perspective on language as multiple and changing resources (Blommaert 2010, García and Li 2014, Canagarajah 2018) we here start from the participants’ speech actions and the speaker roles they play for their meaning making in the classrooms.

Interaction and language development

Early researchers that focussed on interaction, like Long (1981) and Swain (1985), created theories about the importance of interaction for the individual learner in the form of input and output in the target language. Vygotskij [1934]1999 and Cummins (2000), however, from sociocultural approaches, have rather stressed the social importance of interaction as motivation for language development.

Through the Output Hypothesis (Long 1981), which was further developed in The Comprehensible Output Hypothesis (Swain 1985), focus was directed towards the individual’s language production. Learners were stimulated to talk about their own language and were thus assumed to develop consciousness about their gaps in their learner language, or interlanguage, and to simultaneously develop a metalinguistic awareness about rules for the target language. The need to create space for students to express extended and more complex thoughts has been stressed by many researchers, for example Ellis and Shintani (2015) and Wedin (2010). Wedin showed that such spaces were less common in classrooms in the Swedish compulsory school, where teachers focussed instead on giving the floor to every student, which resulted in interaction patterns with short speaking turns and few opportunities for students to either meet or produce more advanced and extended speech. Ellis and Shintani (2015) claim that extended talk occurs more often in student-initiated interactions in which students need to find their words, and when students are to carry out tasks that offer spaces for performing varied speech functions, such as initiating, agreeing, and arguing, and when taking on various speaker roles, such as the role of expert or interested listener (see also Wedin and Norlund Shaswar 2019). Exercises in which students are stimulated to produce output because of high demands for exchange of information (see for example Doughty and Pica 1986, Nakahama et al. 2001) often include what are called information gaps. These have proved to have positive influence on L2 acquisition, which is useful as authentic interactions are hard to create in a classroom setting.

Researchers with a sociocultural approach highlight the importance of meaning making. Cummins (2000, 2017) stresses the importance of teacher–student interaction in classrooms, and the importance of focussing on linguistic form, language use, and the social aspects of language. The traditional interaction pattern for whole-class education, Initiation – Response – Evaluation (IRE), that was first described by Sinclair and Coulthard (1975), is often stereotypical: the teacher asks a question that is answered by a student, and then, the teacher assesses the answer. This pattern is commonly perceived as being ineffective for language development due to the restricted space it offers to students for extended talks and the playing of various speaker roles, and it is perceived of as far from natural, authentic interactions. In the context of Swedish education, Lindbladh and Sahlström (2001) and Wedin (2011) have shown that whole-class education of this type has decreased in favour of students’ individual works. However, Lindberg (2004) and Rosén and Wedin (2015) have shown that teacher-directed whole-class teaching may be organised in ways that can both support language development and offer space for advanced student speech.

However, the understanding of what language development varies among sociocultural researchers. Some view interaction as in itself constituting language development. Swain and Lapkin (1998) write that ‘what occurs in collaborative dialogue is learning. That is, learning does not happen outside performance; it occurs in performance’ (p. 321, emphasis in the original). To our understanding, this definition of learning underlines the importance of studying interaction. From this perspective, it is central to investigate the interaction that L2 learners take part in, and the speech actions and speaker roles in that interaction.

Previous research has shown the importance of interactions with natural and authentic characteristics in the language classroom, and particularly of patterns that offer students possibilities to take on different speaker roles and use speech actions (see for example Tornberg 2000, Ellis and Shintani 2015, Rosén and Wedin 2015). The importance of variation in oral interaction is particularly relevant for SFI with its functional focus. Our own earlier studies in SFI-classrooms showed a dominance of whole-class education sandwiched between individual work with various exercises (Norlund Shaswar 2014, Wedin et al. 2018), and that group activities in the form of conversation were unusual. Wedin and Norlund Shaswar (2019) present interaction patterns from SFI-classrooms that to a certain extent differ from earlier studies in classrooms with children and adolescents (Lindbladh and Sahlström 2001, Rosén and Wedin 2015), because teachers in these SFI-classrooms left more space for the adult students. For example, it was not usual for SFI students to claim the floor by holding up their hand, and as a result of low teacher control, some students (together with the teacher) dominated in teacher-led whole class education.

Interaction and migration

We understand the participants in interaction as human actors involved in negotiation of meaning, and ‘orchestrating language and other semiotic resources to their advantage’ (Canagarajah 2018: 5). Our interest is directed towards the results of the interactions rather than formal correctness in terms of standard Swedish, although we are aware of the function that correctness may have in a language classroom, for example in terms of pronunciation and idiomatic use of lexcon. Verbal resources may obtain new meanings through negotiation and are here understood as mediated and nested in a variety of semiotic resources, and to that end ‘situated in expanded social, material, historical and geographical scales’ (Canagarajah 2018: 7). Processes like increasing global mobility have transformed sociolinguistic research about multilingualism, and directed focus from stability to mobility (Heller 2007). Blommaert (2010) argued for changing perspective from a focus on immobile languages to mobile resources. In what has come to be known as the multilingual turn (Conteh and Meier 2014, May 2014), earlier dominant monolingual perspectives in L2 research were contested, and instead language became understood as multiple resources that were situated, ecological and negotiated and that were an object for continuous change (see for example Heller 2007, Makoni and Pennycook 2007).

This perspective of language as multiple and changing resources will be used in the study of spaces of language development. Particularly the possibilities of practicing varying speaker roles and speech actions that students in SFI, as adults, need to use in their everyday life will be in focus. By concentrating on functional language in whole-class interaction, we direct our interest towards an important space for SFI-students’ development of spoken Swedish. We recognise language as resources, situated and embedded in material life and including complex and varying linguistic, semiotic, material and social features. However, here we will turn our gaze towards oral exchange, motivated by our wish to understand SFI as a space for developing Swedish. While we are aware that external factors such as social networks and physical positioning are involved in social interactions, we direct our focus towards the verbal negotiation of meaning.

Materials and methods

Material for this article was created through an action research project¹ including two municipal schools for adult education, the S-school and the O-school, in two parts of Sweden. The epistemological shift in sociolinguistics, on language as multiple and changing resources (Blommaert 2010, Conteh and Meier 2014, May 2014) mentioned above is here reflected methodologically through ethnographic and critical approaches (Creese 2008, Copland and Creese 2015, Martin-Jones and Martin 2017). Copland and Creese show that linguistic ethnography relates ‘the micro to the macro, the small to the large, the varied to the routine, the individual to the social, the creative to the constraining, and the historical to the present and to the future’ (Copland and Creese 2015: 26). The use of linguistic ethnography here permits a perspective on language as communicative actions in ongoing routines through the study of the linguistic sign as a social phenomenon.

The action research aimed at developing methods for the teaching in SFI, study paths one and two. Classroom practices were observed during various teaching methods and documented through field notes and parts of the observations were audio or video recorded. Participant observation in 18 lessons from two SFI-classes and two different Swedish schools, was used as a primary tool for data collection. Thus all students were adults. One class observed was in study path 1, courses A-B, and the other in study path 2, courses C-D. The two teachers in the chosen classrooms were both experienced teachers with certificates to teach SFI. The teachers were chosen because they expressed an interest in developing their teaching.

Ethical considerations

SFI is a complex activity in which student groups may be changed on a short notice, and participating students changed throughout the project. This fluidity placed high demands on getting consent from incoming students, who in course A-B may have restricted knowledge of Swedish. We tried to make sure that all involved students could understand information about the study and give informed consent to being a part of the study. For example, letters explaining the study were available in languages that the students could understand, in both written and oral (recorded) form.

Because of the vulnerable situation that some students in SFI may experience, particular respect was paid to the integrity of the individual student. The video camera was not directed towards individual students unless a student consciously turned towards the camera, and students who did not want to be filmed were respected by the camera being turned in another direction. Audio recordings were acquired both in the whole class and by placing the recorder close to one or a few students after further information was given and a request was made for consent. All material has been handled in ways that will prevent unauthorised access, and presentations have been made in ways that prevent student identification. All names given here are pseudonyms.

Analysis

In the analysis, the concept speech action was used following the interpretation of speech act theory by Varga (2013). With roots in linguistic philosophy, speech act theory was developed by Austin (1962) and developed by Searle (1979). In this study, speech action is used as the actions performed through students’ utterances, described as what students do, while speaker roles is used to identify how the speaker positions him- or herself orally.

The material was analysed step by step, first by transcription of recordings. We see the action of transcribing talk as an analytical process that involves decision-making about how to reduce oral language into writing (Cameron 2001: 43). The transcription key used is a modified version of a key presented by Poland (2004: 279; see footnote for transcription key²). In a second step, the transcriptions were analysed together with field notes to identify a) interaction patterns in the classrooms, b) sequences where students used varied speech actions and played varied speaker roles, and c) episodes with high range of variation. The analysis of these sequences was then in a third step used to answer the research questions.

Results

First an overview is presented of interactional patterns in these classrooms, in which sequences with variation in students’ oral production are identified. These patterns will then be used as a frame for a detailed analysis of the selected episodes, which will then be used to answer research questions one and two about speech actions and speaker roles. Finally, this will be used to answer the third question about classroom practices where variation appears.

Interaction patterns identified in earlier studies of SFI-classrooms (Wedin and Norlund Shaswar 2019) were also found here. For example, students who talked without raising their hands were generally not reprimanded (as is common in classrooms with children and adolescents). Generally, the students paid attention to behave as expected and showed interest in following the teachers’ instructions. Teachers showed respect to the adults students by rarely reprimanding student initiatives, like conversations between students about things other than the task at hand, or simultaneous talk with the teacher during whole class teaching.

In both these classrooms, education was mainly whole class teaching or individual work, with work in pairs on two occasions in one of the classes. Individual work included very little interaction, usually only in the form of the teacher’s explanations to an individual student, and as the interest here is directed towards students’ linguistic production in interactions, we choose not to analyse individual work. The identified interactions were linked to teaching approaches, and of three types: 1) whole-class conversations; 2) pair exercises; 3) student-initiated student-student interaction. Whole-class conversations were teachers led, while pair exercises were teacher-initiated and had the form of question answer. Student-initiated conversations between students were more or less related to given tasks.

Whole class conversations

Apart from the teacher, there were 5–8 students participating in one of the classrooms, while the other classroom had 12–16 students. The latter classroom also sometimes included a teacher assistant to enable division of the group. In the whole-class teaching, the teacher led talk about a theme that could be related to a subsequent teaching moment or to something that the teacher found relevant. In many cases, students answered at first with single words (yes/no) or short phrases, and the teachers sometimes followed up with new questions or asked students to extend their answers in other ways. On some occasions, students initiated a topic. For example, they talked about cooking a favourite dish. Both the teachers’ and the students’ own experiences were used, and it could be observed that the teachers had the ambition to choose topics that they assumed were relevant for students, by frequently questioning about students’ everyday lives, and by being attentive towards student-initiated themes. Thus, conversations were often about participants’ families, cooking, children, students practical work, the weather, clothes, and similar topics.

For example, in the initial gathering one morning in S-school, the teacher, Yohana, involved the five participating students in a talk about food and groceries by telling them about her own visit to an Asian grocery store in a town nearby. The students, who on this occasion were from Syria, Palestine, and Afghanistan, knew the store and the talk started with groceries and continued to dishes and cooking. Many words came up both in Swedish and Arabic that were not understood by all, and the participants together used digital media to make sure that everybody understood, looking for Swedish and Arabic words as well as illustrations. Yohana used a projector to show how she searched via Google and to show films that they had stored in DropBox. Several students took out their mobile phones to search and show pictures to explain words and concepts that came up. Thus, images and films were combined with oral and written language. The conversation was ended when Yohana handed out questions from a teaching aid about breakfast, lunch, snacks, and dinner, and after that she reviewed the questions and encouraged subsequent discussion, with students working in pairs to answer the questions first orally and then individually in written form. In this way, experiences were related to teaching content based on a teaching aid, and we observed this method in both classes on several occasions. Although the teachers dominated, on such occasions some students took the opportunity to contribute with their experiences and opinions.

This is seen in another example from one O-class, when the teacher presented the celebration of Halloween in relation to the All Saints’ festival through a film. In a recording of two students, Ahmed and Bilal, it can be heard that they repeat some of the words said in the film. When the teacher afterwards asks the students what they think, the following interaction between Ahmed, Bilal, Fatima, the teacher, and the teacher assistant occurs³:

Example 1

Table

1	Ahmed:	Alla barnen kläder ha kläder på sig och går	All children clothes wear clothes and go
2		till knicke dörren grannar tricker e alla	to knick door neighbours trick uh all
3		grannar tri knicker dörren	neighbours tri knick door
4	Teacher:	Knackar på dörren	Knock on the door
5	Ahmed:	kni knicker på dörren e ja e dom e dom ger e till barnen och äpple och bullar	Kni knick on door uh yes uh they uh they give uh to children and apple and buns
6
7	Teacher:	Och godis	And sweets
8	Ahmed:	Och godis	And sweets
9	Teacher:	Det är inte alla	It’s not all
10	Ahmed:	Det är inte alla en en kvinna inte en kvinna är arg en kvinna	It’s not all a a woman not a woman is angry a woman
11
12	Fatima:	Ja	Yes
13	Ahmed:	En granna	A neighbour
14	Teacher:	Alla barn gör det inte	Not all children do
15	Bilal:	Inte göra ha några barnen	Not do have some children
16	Ahmed:	Några barnen fyra barn	Some children four children
17	Teacher:	Mina barn dom har aldrig gjort det mina barn dom har aldrig vi firar inte	My children they have never done it my children they have never we don’t celebrate
18
19		Halloween	Halloween
20	Ahmed:	Ja	Yes
21	Teacher:	Vi har aldrig firat Halloween så de har aldrig de har klätt ut sig någon gång i skolan så alla skulle men dom har aldrig gått och knackat	We have never celebrated Halloween so they have never dressed themselves up sometimes in school so all were to but they have never gone knocking
22
23
24
		(25-36 uteslutet)	(32-43 excluded)
37	Teacher:	Ja det är en ganska ny inte gammal inte gammal tradition	Yes it is a quite new not old not old tradition
38
39	Assist:	Den kanske har varit	It may have been
40	Bilal:	Inte gammal ja	Not old yes
41	Teacher:	Så ganska ny när vi Birgitta och jag	So rather new when we Birgitta and I
42		var då då fanns det inte	were then then there wasn’t
43	Bilal:	Inte gammal ja	Not old yes
44	Ahmed:	Fanns	Was
45	Teacher:	I Sverige det har kommit från USA	In Sweden it has come from the USA
46	Bilal:	Från Amerika	From America
47	Assist:	Ganska kanske tio femton år och så har det blivit mer och mer	Rather perhaps ten fifteen years and then it has become more and more
48
49	Bilal:	Också dom kommer från Iran Iran eller Halloween	Also they come from Iran Iran or Halloween
50
51	Teacher:	E	Uh
52	Bilal:	Halloween Iran också kommer	Halloween Iran also comes
53	Teacher:	Nej Irland e	No Ireland uh
54	Bilal:	Irland	Ireland
55	Ahmed:	Ingles	Ingles
56	Bilal:	Ja jag vet när London UK	Yes I know close London UK Great Britain
57	Teacher:	Storbritannien
58	Ahmed:	Ja precis	Yes exactly
59	Teacher:	Irish Irish	Irish Irish
60		(visar på kartan) Här är Sverige här är Irland	(shows on the map) Here is Sweden here is Ireland
61

Here, Ahmed and Bilal exhibit their knowledge by retelling what they have understood from the film. They frequently repeat what the teacher or the teacher assistant says, add new information and extend it. In lines 1 and 10–11, Ahmed initiates new content, and in line 59, he displays knowledge in relation to Great Britain. Also, Bilal initiates new content in line 49 and Ahmed shows in line 55 that he understands the relation between Ireland and England (‘Ingles’). Here, one may understand Ahmed and Bilal as active participants in negotiation of meaning, where the repeating may function socially as the listener’s support to the person talking, as showing active listening and a benevolent attitude, while repetition of linguistic features may be a conscious learner strategy. Students’ speech actions in this case are repeating, adding information, initiating, extending the topic, displaying competence, and positioning oneself as an active and benevolent participant in the interaction.

If, on the other hand, we look at the teacher’s turns (for example on lines 7, 9, 14 and 41–42), the main part consists of arguing, to some extent against the film — that is., that this was not something that she or her own children had taken part in except to a restricted extent, and that the feature is new to Sweden. Furthermore, she takes on a traditional teacher role by verifying what one student has said (line 58), and when Ireland is mentioned, to show where Ireland is on the map (lines 60–61).

The main topic is initiated by the teacher and the students actively show their interest, while simultaneously displaying their own knowledge and participating in the negotiation of meaning that takes place. Here, the teacher may be perceived as positioning herself in relation to what is said in the film, and simultaneously positioning herself as a teacher offering what she herself perceives to be correct knowledge. She also takes on the role as support for the students, for example in line 44 where she confirms Bilal’s clarification.

This example is similar to what we have witnessed in other SFI-classrooms (Norlund Shaswar and Wedin 2019) and we find that space here was created for students to exercise different speech actions and to take on various speaker roles. However, we also saw that the teachers tended to take a large part of speech space, and that some of the students hardly said anything during the whole lesson.

Pair work

Pair work did not occur often, and such exercises were often directly taken from the teaching aid used. Here we will go through one example from an instance in the O-class, where a form of pair work took place in which students read some questions and then answered them. During the observations, both students tried to read the question together, after which one of them asked the other one the question. On some occasions, the teacher passed by and encouraged them to both answer the question. After a question about their own families, the following conversation took place:

Example 2

Table

1	Issam:	Berätta om din familj	Tell us about your family
2	Rami:	Vad betyder?	What mean?
3	Issam:	Hm	Ehm
4	Rami:	Vad betyder?	What mean?
5	Issam:	Den berätta om din familj	That tell about your family
6	Rami:	Vad betyder berätta? Det samma prata din familj	What does tell mean? That same talk your family
7
8	Issam:	Den prata familj pappa mamma och barnen fyra barnen x x	That talk family daddy mummy and children four children xx
9
10	Rami:	x x	xx
11	Issam:	fyra barnen och två pojkar två syster e två pojkar tre flicka den flicka stor flicka stor e tretti år e den flicka liti liten fem år och pojken e stor sex år Sirwan, vad heter Dldar elva år den gå skolan Sirwan skolan Berget	Four children and two boys two sister uh two boys three girl that girl big girl big uh thirty years uh that girl litl little five years and the boy uh big six years Sirwan, what’s called Dldar eleven years that go school Sirwan school The Mountain
12
13
14
15
16
17	Rami:	Ja	Yes
18	Issam:	Hon skolan den Tara går dagis Mimer förskolan Björken	She school that Tara go nursery Mimer pre-school The Birch
19
20	Rami:	Baba Sirwan kommer?	Sirwan’s baba comes?
21	Issam:	Baba Sirwan kommer också sfi och skol här	Sirwan’s baba also goes SFI and
22		ja	school here yes

In this example Rami asks what the question means — the word ‘berätta’ (tell) is the problem, and he himself suggests ‘pratar’ (talks). Issam agrees by repeating ‘Den prata familj’ (That talk family) and then gives a suggestion, which is Issam’s own answer to the question. Issam extends by telling about his own children at which point Rami asks (line 20)

‘Baba Sirwan kommer?’ (Sirwan’s father comes?), and then Issam extends the answer to tell that “Baba Sirwan” also is an SFI-student.

For students who still have relatively little experience of reading and writing in Swedish, the reading is often laborious and students have difficulties understanding the question in itself. In the following example, the questioning was done in the form of a game in which students threw dice to decide which question to answer, followed by reading the question out loud.

Example 3

Table

1	Issam:	När	When
2	Rami:	När	When
3	Issam:	fyller	is
4	Rami:	fyller du år?	is your birthday?
5	Issam:	Vad betyder	What means
6	Rami:	När fyller du år	When is your birthday
7		(tystnad 22 sek)	(silence 22 sec)
8	Issam:	Den sho (vad) pratar	That sho (what) talk
9	Rami:	Kan inte förstår	Can’t understand
10	Issam:	Den jag vet inte. Spel!	That I don’t know. Play!

Here they give up, and Issam urges Rami to throw the dice again to continue with the next question, and thus avoid the question that they do not understand. In the following example, the student pair Zahra and Muhammad share Arabic. Zahra has higher competence in Swedish, and both turn to Arabic now and then to support Muhammad in the interaction.

Example 4

Table

1	Zahra:	Vad du gör på kvällarna	What do you do in the afternoons
2	Muhammad:	Kvinna?	Woman?
3	Zahra:	Kvällarna kväll ar na illha shu tamul bi- lel (säg till henne vad du gör på kvällen)	Afternoons aft er noons illha shu tamul bi- lel (say to her what you do in the afternoon)
4
5
6	Muhammad:	Bi-lel (på kvällen) e jag e TV	Bi-lel (in the evening) uh I uh TV
7	Zahra:	Lyssna du nyheter? Al-akhbar (nyheter)	You listen news? Al-akhbar (news)
8	Muhammad:	(arabiska ohörbart)	(Arabic, unhearable)
9	Zahra:	Titta på TV lyssna på nyheter	Watch TV and listen to news
10	Muhammad:	Jag bs2018wi akl shu besaawi (jag lagar mat vad gör jag)	I bsaawi akl shu besaawi (cook food what do I do)
11
12	Zahra:	Laga mat	Cook
13	Muhammad:	Laga mat lagar mat bedrus (studera) lyssna shu (vad) lyssna shu (vad)	Cook cook bedrus (study) listen shu (what) listen shu (what)
14
15	Zahra:	E studera	Uh study
16	Muhammad:	Studera studera illi badu (vad kommer sedan)	Study study illi badu (what comes next)
17
18	Zahra:	Vad du ska göra i morgon shu biddak tamul bukra (vad ska du göra i morgon helg)	What are you going to do tomorrow shu biddak tamul bukra (what are you going to do tomorrow weekend
19
20
21	Muhammad:	I morron e jag e bint shu isma el-bint (dotter vad heter dotter)	Tomorrow uh I uh bint shu sima el-bint (daughter what is daughter)
22
23	Zahra:	Dotter	Daughter

Zara helps Muhammad express himself in Swedish through Arabic. Because both speak Arabic, Zahra with her more developed Swedish can use Arabic to explain, and during the question-answer interaction, Muhammad turns to using more and more Arabic and less Swedish. Zahra’s here takes on a role which is similar to the teacher, scaffolding among other things by urging (lines 3–5), giving a Swedish expression (15), giving a question as support (18–20), repeating-confirming (23), and clarifying (18).

On some occasions, a question leads to a demand for further information, such as when Issam and Rami are to answer the question about what they do after school, when the observing researcher passes by and the conversation turns into a conversation about interaction through social media:

Example 5

Table

1	B:	Du använder internet	You use the Internet
2	Issam:	Ja jag jag gå alla dagar gå e	Yes I I go every day go uh
3	Rami:	Vi sitter	We sit
4	Issam:	Nej Pepsi e och tea och kaffe	No Pepsi uh and tea and coffee
5	Rami:	Dricka	Drink
6	Issam:	Och internet	And Internet
7	B:	Fikar och surfar	Coffee break and surfing
8	Issam:	Pratar e Internet pratar family pappa mamma syster bror. Ja	Talk uh Internet talk family daddy mummy sister brother. Yes
9
10	B:	Vad använder du när du pratar är det Skype eller är det	What do you use when you talk is it Skype or is it
11
12	Issam:	E Whatsapp	Uh whatsapp
13	B:	Whatsapp?	Whatsapp?
14	Issam:	Ja mycket ja	Yes a lot yes
15	B:	Aha	Umhm
16	Rami:	Ja IMO Messenger	Yes IMO Messenger
17	Issam:	Jag har IMO Messenger	I have IMO Messenger
18	Rami:	Vi pratar ja	We talk yes
19	B:	Ni pratar	You talk
20	Rami:	Kanske det alla pratar Whatsapp det samma Skype och det samma IMO	Maybe that all talk Whatsapp that same Skype and that same IMO
21
22	B:	Det är samma	That’s the same
23	Rami:	Det samma	That same
24	B:	Lika bra	Equally good
25	Rami:	Ja	Yes
26	B:	Alla är lika bra	All are equally good
27	Rami:	Lika bra det är ja Skype det är titta det är video	Equally good yes Skype it’s look that’s video
28
29	B:	M man ser	M you see
30	Rami:	Och det är också Whatsapp det är samma. Whatsapp det här (visar med mobilen) Whatsapp Whatsapp det här det är också video	And that’s also Whatsapp that’s the same Whatsapp this (shows on the mobile)Whatsapp Whatsapp this is this is also video
31
32
33
34	B:	Det är också video	It’s also video
35	Rami:	Ja. Det är också video.	Yes. That is also video.
36	B:	Mhm så du kan se	Mhm so you can see
37	Rami:	Ja	Yes
38	B:	Familjen och kompisarna	Family and friends
39	Issam:	Också Snapchat Snapchat också Snapchat prata video	Also Snapchat Snapchat also Snapchat talk video
40
41	B:	Ok	OK
42	Rami:	Det samma Abu (father) Sirwan audio	That same Abu (father) Sirwan audio
43	Issam:	Va? Ja den Whatsapp ok video	What? Yes that Whatsapp OK video

As in earlier examples, this conversation builds on questions given in the teacher aid, but while earlier examples were based on the home sphere, interest is here directed towards other uses of spare time, like communication with relatives in other countries through digital tools. In this case, one of the students has been to a café at Coop Forum and has had something to drink while communicating with relatives and friends in the home country. Here, students answer, explain and exemplify, and when the researcher asks about the tool used suggesting Skype (line 10-11) students extend the talk to other digital tools and demonstrate them (line 30-33).

Interaction in all these examples may be understood as relatively similar to interactions in everyday contexts. Follow-up questions often extended the conversations and supported the speakers through repetitions of what was said. These exercises with predetermined questions offered opportunities for speaking dialogues of various lengths, and for negotiating both language and content. Although the given exercises in these cases were simple, variation occurred in both speaker roles and speech actions.

Students’ use of other languages, here Arabic and some English, sometimes functioned as a scaffold in which the students themselves initiated translation to Swedish, and sometimes as a strategy to escape expressing oneself in Swedish. Skills in one language is an important base for the development of any new language. The adult may perceive him- or herself as incompetent in the new language in ways that may be felt as threatening (face-threatening, Goffman 1955), and then the possibility of using skills in another language may be a way to escape or overcome the perceived threat.

Student-student talk

Students also talked together outside the teacher’s plans in the classroom, both in individual work and during whole-class education. This private talk, secret and intimate interactions outside the teacher’s supervision, is similar to what Gilmore called sub rosa (Gilmore 1986). While the class listened to the film about Halloween, Ahmed and Bilal had their own private conversations, which a tape recorder left on their table recorded. These two students have different language backgrounds and used only Swedish when they talked together, such as in Example 6, concerning visits to discotheques and drinking alcohol.

Example 6

Table

1	Ahmed:	Du gick inte disco? Va? Du gick?	You not went disco? Eh? You went?
2	Bilal:	Nej jag gick inte disco	No I didn’t go disco
	Ahmed:	Jag har gick alla e lördagar	I have went all uh Saturdays
3	Bilal:	Nej nej jag gick inte jag gillar inte	No no I didn’t go I don’t like
4	Ahmed:	Nej jag vet klockan sju	No I know seven o’clock
5	Bilal:	Nej vi dricker inte alkohol jag har sett vi dricker inte	No we don’t drink alcohol I have seen we don’t drink
6
7	Ahmed:	Jag har sett du	I’ve seen you
8	Bilal:	Nej alltid det inte bra bara sitt hemma läsa den hitta jobb bra sen tjäna mycket pengar kolla	No always that not good only sit home study that find job good then earn a lot of money look
9
10
11	Ahmed:	Du bara kolla i disco	You only watch in disco
12	Bilal:	Det är bra jobba bra eller hur?	That’s good work good isn’t it?
13	Ahmed:	Det är bra	That’s good
14	Bilal:	Det är inte bra	That’s not good

Here Ahmed initiates a conversation about going to a discotheque, something that Bilal says he does not like, and he refers to “we” who do not drink alcohol. Bilal instead argues for staying home, studying, finding work and (thereby) earn a lot of money. Both stand by their respective views of the value of discos and alcohol. The interaction starts with a question but develops to arguing where each participant defends his own standpoint.

Shortly after, during the same lesson, after the talk with the teacher about the relation of Halloween with Iran or Ireland, Ahmed initiates a conversation in a joking tone with Bilal about finding a boat.

Example 7

Table

1	Ahmed:	Vi måste köpa e köpa en båt (.) båt	We have to buy uh buy a boat (.) boat
2	Bilal:	Båt ja jag vet vi måste köpa en båt sen	Boat yes I know we have to buy a boat then
3	Ahmed:	Jag och du går till där	Me and you go to there
4	Bilal:	Irland	Ireland
5	Ahmed:	Ja i vatten	Yes in water
6	Bilal:	Okej	OK
7	Ahmed:	I vatten	In water
8	Bilal:	Det är bra jag köper en båt sommaren vi köp jag och du köper en båt tillsammans en båt liten båt inte inte mycket kosta kanske köper tillsammans	That’s good I buy a boat summer we buy you and I buy a boat together a boat small boat not not much cost perhaps buy together
9
10
11
12	Ahmed:	Ja jag vill köpa jag vill köpa min jag min och hon henne	Yes I want to buy I want to buy my I my and she her
13
14	Bilal:	Henne	Her
15	Ahmed:	(skrattar)	(laughs)
16	Bilal:	Men hon kan båt	But she can boat
17	Ahmed:	Inte komma inte komma	Not come not come
18	Bilal:	Hon inte	She not
19	Ahmed:	Nej inte komma inte kommer	No not come not come
20	Bilal:	Jag kommer inte	I don’t come
21	Ahmed:	Inte kommer?	Not come?
22	Bilal:	Kommer kär jag inte kommer inte	Come in love I not come not
23	Ahmed:	(skrattar) Du gillar hon	(laughs) You like she
24	Bilal:	Henne jag gillar	Her I like

This conversation is an extension that continues from the talk in the whole class of Ireland, and here Ahmed suggests that they get themselves a boat to go to Ireland. The conversation then turns to be about a female acquaintance, and Bilal says that if she will not come then he will not go. This makes Ahmed laugh and say that Bilal likes her. Even if the meaning here becomes a bit difficult to interpret, it may be understood as an example of joking between Ahmed and Bilal, similar to their behaviour exhibited previously.

On one occasion at the beginning of a lesson in the O-class, Yordanos initiated talk about haircutting, while students kept coming in one after the other and the teacher counted them to plan division of the group⁴.

Example 8

Table

1	Yordanos:	Vem klappar du?	Who cit you?
2	Karim:	Vad klappa	What cit
3	Rami:	Det håret det klappa håret	That hair that cit hair
4	Karim:	Chi yaani (kurmanji: vad alltså?)	Chi yaani (Kurmanji: What then?)
5	Rami:	(kurmanji ohörbart)	(Kurmanji inaudible)
6	Muhammad:	bil arabi (på arabiska)	bil arabi (in Arabic)
7	Karim:	Kompis in hemma	Friend in home
8	Yordanos:	Jaha det bra alltid klippa	Oh yes that good always cut
9	Amiina:	(somaliska ohörbart)	(Somali inaudible)
10	Yordanos:	Nej kompis	No friend
11	Muhammad:	Nej e	No uh
12	Karim:	Kanske det håret det klipper håret också Inte jag gratis jag inte nej betala här dyr 200 nej varför nej nej inte betala alla kvinna min land	Maybe that hair that cut hair also Not I free I not no pay here expensive 200 no why no no not pay all woman my country
13
14
15	Yordanos:
16		Jag betala 200	I pay 200
17		M till jobb har mycket pengar själv	M to job have much money self
18	Rami:	Ja ja	Yes yes
19	Yordanos:	själv inte 200 300	Self not 200 300
20	Karim:	200 200 mycket	200 200 much
21	Yordanos:	Nej nej dyr	No no expensive
22	Karim:	300?	300?
23	Yordanos:	Ja	Yes
24	Muhammad:	Enta hal’ta shirak (Har du klippt håret) två?	Enta hal’ta shirak (Have you cut the hair) two
25	Yordanos:
26	Muhammad:	300?	300?
27		Ja	Yes
28	Rami:	200?	200?
29	Yordanos:	Ja	Yes
30	Amiina:	Ja kanske har kanske 2, 300	Yes maybe have maybe 2, 300
31	Yordanos:	200?	200?
32	Amiina:	Varför? Alla mannen hemma klappa	Why? All man at home cit
33	Karim:	Där?	There?
34	Yordanos:	Exempel min barnet jag köpa 300 maskinen	Example my child I buy 300 machine
35	Rami:
36	Yordanos:	Ja	Yes
37		jag klappa	I cit
38	Karim:	Ja	Yes
39	Yordanos:	Jag	I
40	Muhammad:	Familj	Family
41	Yordanos: Muhammad: Yordanos:	Jag min barnet	I my child

Five students participate in this discussion, initiated by Yordanos, about haircuts and prices. Simultaneous talk is going on around them, such as the teacher counting the students as they entered the classroom, and other students having other conversations in several languages. Thus, this conversation may not directly be understood as sub rosa because it was not secret — instead, it took place parallel to the teacher’s less formal introduction of the lesson. In order to make the excerpt above more understandable, we have chosen to only transcribe speech from the five students involved in this specific discussion. The interaction was partly difficult to interpret, and the participants had difficulties understanding each other. Individual students used words in Somali, Tigrinya, Kurdish, and Arabic, for example in the beginning when students try to figure out the meaning of what Yordanos is saying. Yordanos herself highlights that she has bought herself a haircutting machine at the cost of one haircut, and that she now cuts the hair of her husband and their child and thus saves money. She argues for charging friends when she cuts their hair and discusses a reasonable price. She also stresses that all women in her country work and earn money. Thus, the first part of the lesson was turned into a space for Yordanos to position herself as a noteworthy woman among other women from her country, self-reliant and economically capable.

In these three examples of student-initiated interactions outside the teacher’s planning, a form of negotiation of meaning becomes visible, with different characteristics than the earlier examples of interactions more directly initiated by the teacher. Here, students initiated positioning themselves through arguing, and by using sophisticated speech including joking, teasing and spontaneous self-expression. Their spaces for exploring speaker roles and speech actions were widened, and meaning was negotiated with fewer instances of scaffolding through clarifications and explanations, or the need to search for a certain word. In classrooms with children and adolescents, this type of talk often takes place outside the teacher’s control, such as when students whisper or in other ways consciously prevent the teacher from hearing. Here, such conversations took place outside the teacher’s planned teaching, but nothing indicated that students were trying to hide their conversations, and in the third example, the teacher indirectly took part by continuing to silently count and place students, which allowed the conversation to continue and develop.

Classroom practices with variation in speaker roles and speech actions - findings

This study set out to explore classroom practices where students are involved in negotiating meaning by investigating students’ oral language production in classrooms. Analysis of the selected sequences will be used to answer the research questions.

The first research question concerns what speech actions the students use. Of importance for the students’ use of both speech actions and speaker roles is how the teacher acts in relation to the students. As mentioned above, both teachers showed the students respect as adults and gave priority to making students talk and to base their teaching on their perceived needs and experiences. In our analysis of the selected sequences, students practiced varying speech actions such as showing knowledge, socialising, scaffolding, initiating, defending one’s own standpoint, arguing, teasing, joking, positioning oneself and others, negotiating meaning, clarifying, explaining, and searching for a word or grammatical form. However, these selected sequences only constituted a minor part of the lessons, and a large part of the speech space was used by the teacher. It was also the case that some students only produced talk to a very limited extent during the lessons.

The second research question examined what speaker roles the students play. Speaker roles are closely related to speech actions, as roles are played in actions, and examples of speaker roles which we identify are played by students in these sequences are as an ambitious and good student, expert, novice, joking friend, supporter, learner and an understanding and supporting classmate. Both the teacher and some students positioned themselves by claiming certain roles, such as a woman earning her own money.

The third research question concerns the types of educational practices where varied speech actions and speaker roles occur. In this study variation in interaction did not occur at all during students’ individual work. Although the teachers dominated the talk, roles were sometimes shifted among the participants between expert and novice and there was some room for negotiation of meaning. During the selected whole class sequence, the teacher sensitively listened to students’ comments and suggestions and thus the interaction had similarities to interaction outside the classroom where people share experiences with each other. In this sequence and in the student-initiated sub-rosa talk, focus was on content, on what was said, and not on form, how it was formulated, and all participants made an effort to understand each other through the negotiation. In the sequences of students’ private talk, the interaction afforded them expanded spaces for trying out speaker roles and speech actions, with less scaffolding such as clarifications and explanations. In the pair work, the task was quite controlled and simple, but to some extent resembled interaction in everyday contexts and it gave students opportunities to negotiate language as well as content.

Discussion

In the classroom talk presented here, including whole-class interaction, pair work, and private talk among students, we see that the students got space to negotiate meaning in ways that are close to talk outside the classroom in daily life. This negotiation of meaning offered considerably more space for different speaker roles and speech actions than is allowed for in the traditional question-answer interaction of the IRE-type. These sequences also demonstrated the social importance of classroom interactions (see also Cummins 2000, Lindbladh and Sahlström 2001, Wedin and Norlund Shaswar 2019). The student-initiated interactions in which students positioned themselves with jokes and arguing were important for identity development and for exploration of new roles in new contexts, and also offered such linguistic challenges which Ellis and Shintani (2015) stress are important for language learning.

However, these examples only constituted a restricted part of the classroom time, and a dominant part of the teaching was of types that included nearly no written or oral interaction. The range of participation in the talk also varied among students, with some dominating the talk and some saying very little. Based on this study and earlier research (such as Lindberg [1996]2005, Ellis and Shintani 2015, Wedin and Norlund Shaswar 2019) it may be argued that exercises and spaces in SFI that promote verbal negotiation of meaning are important to expose students to various types of interaction. Furthermore, diverse tasks and changing group constellations are also important to allow and expand student-student interaction. However, in our observations, we saw no use of role play, prepared arguing, or other exercises that put higher demands on students’ language use, which would be particularly important for more advanced learners (Lindberg [1996] 2005, Tornberg 2000). More demanding conversations might be stimulated through other contexts that provide linguistic challenges while still being relevant for students’ work and social lives. One example of a speaker role that requires a high level of communicative competence is that of parents, who need to be able to discuss their children’s time at pre-school and school with teachers and other parents. The need for communication is strong for newly arrived parents, who thus need to develop that communicative competence quickly (Duek 2017).

A potential appears for students’ use of earlier linguistic resources. While students in these examples sometimes used other languages than Swedish to support interactions or to escape using Swedish, such interactions could be more consciously planned to support their L2-learning (García 2009, Paulsrud et al. 2017, 2018). Furthermore, we also see a potential for the development of use of various digital tools, both for oral and written interaction (Buendgens-Kosten and Elsner 2018, Vollmer 2018, Norlund Shaswar, in press. With the help of digital tools to look up the phrase ‘fyller år’, the two participants in example 3 might not have given up.

In conclusion, the study shows that teaching in SFI-education could be developed to stimulate and give more space for negotiation of meaning and students’ own language use. Strong pressure for students to quickly pass the final exam, may lead to a risk that teachers devalue the more time-consuming practices that include students’ negotiation of meaning, even though such practices are important for students’ language acquisition and subsequent language development. The study also shows a need for further studies of adults’ oral production and of classroom interaction in second language classrooms. As the need for immigrants to develop proficiency in the new language is emphasised in public debate, much more knowledge is needed of the characteristics of language use in classrooms.

Notes

1. Omitted for reasons of anonymity.

2. Transcription key.

3. The transcription has been made in a form that is intended to be a close representation of student speech, which means that there are some deviations from what can be seen as standard Swedish, and some adjustments have been made in order to facilitate readability. In some recordings, there are parallel conversations going on, as is usual in classrooms, but here, only speech that was assessed as important to the respective examples were included.

4. In this situation, there are several conversations going on simultaneously and there is a lot of small talk around the interlocutors. We have chosen not to mark all parallel speech in order to render the example more readable, and we use small letters for speech that is more or less simultaneous or in someone else’s turn of speech.

Disclosure statement

This research does not include any financial interests or benefits arisen from the direct applications.