Traditions of Argumentation in Teachers’ Responses to Multilingualism in Early Childhood Education

In this study, we investigate how professionals in early childhood education (ECE) reason about multilingualism. Empirical data are analyzed in terms of ‘traditions of argumentation’ which proposes that we cannot argue for something without, explicitly or implicitly, arguing against something else. The analyses use transcribed data from two focus groups conducted with teachers in two preschools in Sweden. These teachers had experience teaching culturally and linguistically diverse groups of children. The reoccurring rhetorical strategy used by the teachers to talk about their work with multilingual children used a set of contrasts. Three contrasts were identified: (1) I/we versus them (others); (2) here-and-now versus there-and-then; and (3) building ECE on research versus personal experience. The study has implications for teachers and students in preschool teacher education to understand the possible tensions and contrasts inherent in teaching culturally and linguistically diverse children. Rather than simplifying professional practice to either side of a dichotomy, teachers should be encouraged to understand and verbalize the bases of their professional knowledge, and understand the different positions from which they draw knowledge to inform practice.


Introduction
The interests of this study are twofold: empirically, how do early childhood education teachers respond to contemporary multilingualism and methodologically, in research, how to analytically account for what is implicitly stated in teachers' responses rather than what is explicated. The first issue is a pressing issue for contemporary early childhood education (ECE), because there can be a multitude of languages represented among children in ECE groups. How ECE teachers respond to this challenge in their practice and their reasoning about multilingualism are important to study empirically. One theoretical premise is that how teachers speak about the use of language by children in ECE groups has consequences for how they relate to the children. A second theoretical premise is that in communication individuals communicate much more than what they explicitly say. In most communication, we respond not necessarily to what is explicitly said, but what we take to be the intended meaning. In research, this poses a challenge about how we can examine analytically what is metaphorically said, beyond the words that are spoken.

Argumentation and the Example of Multilingualism in ECE
In this paper, we use a theoretical perspective of argumentation to study the processes of how teachers explain and justify their beliefs about multilingualism and their practices in their preschools. Studying argumentation related to multilingualism in ECE is essential for several reasons that are explained in this section.
First, this is a complex issue with important implications for supporting multilingual children's learning as a part of the professionalism of ECE. On a general level, it relates to the organization of education, both in terms of monolingual and bi-/ multilingual education in which several languages are used as tools of instruction (Skutnabb-Kangas and McCarty 2008). Organizing education is not only a linguistic and pedagogical issue but also a political and socio-political one (Freeman Field 2008;García and Otheguy 2020;Lo Bianco 2008;May 2008). For example, the choice of terminology in discussing multilingualism is value-laden. Therefore, the terms and concepts used by teachers need to be considered because they make some aspects of meaning more visible, beneficial, problematic, interesting or uninteresting (Freeman Field 2008;Skutnabb-Kangas and McCarty 2008).
Second, a study from the project from which the present data are drawn implies that teachers' professional work with multilingualism and language learning in the context of ECE is more likely to be based on personal experiences of teachers than on educational theory or research evidence (Kultti andPramling Samuelsson 2016, 2017). Multilingualism is a key feature at an overarching level in Swedish ECE. Facilitating children's opportunities for language development in both their first and second language is explicitly stated in national education law (SFS 2010: 800) and in the curriculum for preschool (The Swedish National Agency for Education 2019). In addition, personal beliefs about multilingualism and language development may be based on misunderstandings (May 2008;Tobin, Arzubiaga and Adair 2013). For example, a common understanding is that language learning is easy for children (e.g., , for a critical review). There is also a tendency to stress grammatical and semantic skills rather than pragmatics when discussing the teaching and learning of language(s) in ECE. In response to these conditions, Tobin et al. (2013) argue that preschool teacher education needs to emphasize ways in which to relate to and engage with recently migrated children and their families.
Third, multilingualism and language learning require collaboration between preschool teachers and the child's caregivers with a first language other than the majority one. Using and learning a majority language, on ideological and practical levels, are major issues debated in relation to linguistic and cultural diversity in ECE, as shown by Tobin et al. (2013). In their study, caregivers expressed a wish for the child's fluency in the home language as well as in the majority language. Despite this wish, the question of bilingual education was not a focus of discussion. Instead, the caregivers voiced their opinions about pragmatic concerns for the child which was commonly about learning a second language for academic purposes. They also expressed a fear that using the child's first language in education settings might result in a lack in skills in the majority language which could lead to stigmatizing the child. This is something that early childhood professionals can contribute to counteracting through building on and talking about research-based education practice that highlights the importance of multilingualism and second language learning during the early childhood years.
In this study, we analyze the reasoning of teachers in ECE in Sweden working in multilingual settings. In Sweden, preschool denotes settings for children up to 5 years, guided by a national curriculum and with personnel including preschool teachers with a 3.5-year university education but also adults without specific qualifications. In this study, we explore what and how participating preschool teachers, who work in a multilingual area of Sweden, argue about the challenges posed by multilingualism. The data from teachers are analyzed in terms of what we, and Billig (1996), refer to as traditions of argumentation.

Researching Communication: Theoretical Premises
A basic premise of human communication is that we always communicate more than what we say. In everyday life, we all respond to what is intended, rather than merely processing what is explicitly said. This is fundamental in being able to communicate and interact with others. However, to research this 'extra meaning,' which is implicit in communication poses some challenges. How can we analytically account for what is communicated in the extra meaning which may often be in contrast to what is explicitly said (as is the case with irony)? There are different approaches to this analytical problem. One is to have particular transcription conventions, for example, being able to show how speakers change their tone of voice to indicate irony or humor (e.g., Psathas 1995). Another is to present people with alternative suggestions and see how they respond. One example of this latter approach is presented by Kövecses (2000) who asked people with different languages to respond to sentences, such as 'He was angry, but he didn't lose control' and 'He was very angry, but he lost control' (p. 24). The conjunction 'but' is used to counter expectations. Hence, respondents natural use of language can reveal implicit perspectives. In this example, the different sentences revealed how respondents understood the nature of anger and how people who are angry are expected to behave. Such a method can be very useful to identify taken-for-granted, cultural perspectives on phenomena.
If instead, we wanted to analyze communication of individuals' reasoning, we may need different analytical tools. One such analytical approach is referred to as traditions of argumentation (Billig 1996;Shotter 1993). This concept refers to the fact that we cannot argue for something without, explicitly or implicitly, at the same time argue against something else. For example, to argue that teaching in preschool needs to be understood in different terms than in how teaching in schools is performed, we make this distinction explicit. However, often what we argue is implied rather than explicitly stated. For example, to argue that learning in the era of digital technologies poses particular challenges to education implicitly counters the argument that education can proceed as it did before the advent of these technologies. Even if this idea is not stated explicitly, it is implied by the structure of the argument.
One reason why people communicate with implied assumptions rather than stating explicitly what they mean is that we adhere to a principle of communicative economy in order not to bore the listener with extensive explication. We make assumptions that others will understand what we mean. Getting at these taken-forgranted viewpoints are important when studying how individuals communicate about debated and contested issues. The following example illustrates this idea: This study provides empirical support for the view that Spanish-speaking children from low socioeconomic status (SES) families in the US benefit from L1 maintenance and enrichment, and are not disadvantaged in their L2 English progress compared to peers in monolingual English programmes (Hickey and de Mejía 2014 p. 135).
The authors point out that the children referred to 'benefit' from L1 maintenance and also that the children 'are not disadvantaged' compared to 'peers in monolingual English programmes.' The researchers seemed to be implicitly arguing against a tradition whereby some 'disadvantage' for progress in L2 English language was implicitly presumed or claimed. Hence, the authors do not merely present their conclusions, they also argue against an alternative view. This duality lies at the heart of argument, if this communication practice is understood in terms of traditions of argumentation (Billig 1996). Hence, we can analytically account for what is takenfor-granted in communication which has also been described as, what goes without saying (Wittgenstein 1969). In this study, we will analyze the responses of early childhood education (ECE) teachers to multilingualism, in terms of such traditions of argumentation.
The overall structure of our reasoning is: (1) in communicating with others, people communicate more than what they (explicitly) say; (2) this is an empirical phenomenon; (3) this poses an analytical challenge to research on how we identify what is 'taken-for-granted' in participants' discourse; and (4) this phenomenon and the analytical challenge become salient when studying discourse on debated and controversial issues. This is the focus of this research on exploring the responses of early childhood education teachers to multilingualism.

Methodology
The study is part of a research project which aimed to develop new knowledge about collaboration between teachers and children's caregivers from diverse linguistic backgrounds in the context of preschools in which Swedish is the language of instruction (see Kultti andPramling Samuelsson 2016, 2017). More specifically, the research project investigated how the teachers and caregivers talk about, reflect on, and develop participation in preschool in order to promote young children's participation and learning during a settling-in period. The settling-in period, in this case, refers to a child's first three days in preschool as well as the introduction and followup meetings between teachers and caregivers.

Participants
This project was conducted in two preschools where the teachers have experience with culturally and linguistically diverse groups of children. The context for the present study is focus group conversations between teachers and researchers after the settling-in period for new children at the preschool. There were three to five educators and one to two researchers participating in these conversations which utilized readings from a popular-scientific book (Kultti 2014), which teachers had received for preparatory reading for the research project. This book introduces a sociocultural perspective on language learning; language use in different contexts; communication and communicative resources; institutionalized activities and tools for participation in play; guided participation, scaffolding and teaching. This book also contained questions for discussion related to each of these themes.

Data Generation
The research data consisted of 6 h of audio-recorded conversations. At each meeting, the teachers were asked, based on their reading (see above) to introduce a discussion of one of the themes and connecting it to their experience. The focus group conversations were audio-recorded, transcribed and translated into English for publication purpose, by the researchers. When translating the participants' utterances, great care was taken to mimic the nature of their talk, for example their use of figurative speech. The examples presented in the analyses are intended to cover the variety of argumentative work subsumed under each identified theme in the data.
The ethical guidelines of the Swedish Research Council for social science research concerning informed consent, anonymity and confidentiality were followed. In the analysis, we use acronyms to identify participants (e.g., PA refers to Preschool teacher A; PB refers to Preschool teacher B; etc.).

Data Analyses
We analyze the data as dynamically evolved sense-making, rather than with an interest in what a particular participant says/argues. We have therefore pooled all the data into one data set, not separating out groups/preschools in our presentation. This is of analytical as well as of ethical importance. The emerging themes from the analyses (see below) were generated through repeatedly reviewing the data and transcripts (cf. Derry et al. 2010).
Analyzing talk ('reasoning' in this research) as social practice means to see how meaning is negotiated over consecutive turns of speech (cf. Wells 1999), as responsive to previous utterances as well as potentially anticipating coming ones (e.g., to preclude an expected counter argument or critical remark) (cf. Bakhtin 1986). Analyzing talk as social practice also means to approach empirical data in terms of the participants' concerns, that is, what they say is seen as what they perceive to be relevant responses even if they do not talk about what the researcher perceives to be directly relevant to the conversation topic or aligned to the researcher's perspective.

Findings
In this section, we present the findings through three excerpts from focus group conversations on preschool teachers' reasoning about multilingual issues. Each excerpt is presented indicating the speaker (e.g., teacher PA or teacher PB, etc.). The number of the conversational turns is also noted, as recorded in the transcriptions/translations of the focus group conversations.

Traditions of Argumentation in Response to Multilingualism
Across different topics discussed in the focus groups over time, three emerging tensions or contrasts were identified across excerpts and these are illustrated across the presented excerpts: (1) I/we versus them (others); (2) Here-and-now versus there-and-then; and (3) Building ECE on research versus personal experience.

Excerpt 1: Reasoning About the Importance of Children's First Language
A starting point in the focus groups was questions concerning teachers' knowledge about the different languages that children speak in their group: How many languages do the children in your group speak? What are these languages? What languages are spoken by the work team/teachers at your preschool? What is possible in your work team to find out about any languages that you are not familiar with? 10: PC: I think, we have the hang of what languages the children speak and where they come from. I remember at my, at the start of my career, then it was a lot 'I don't know what languages they speak, don't know where they're from.' And then one wasn't as interested as today. At least in the house we care about where the parents come from, what languages are spoken at home. It has changed, gotten better. 11: PA: Precisely. 12: PB: Very good. 13: PA: I also think that we've gotten much better when it comes to us adults being more curious about other languages and want to use single words, and use in the activities.
Speaking about linguistic diversity among children in contemporary Swedish preschool, PC argues, 'I think we have the hang of what languages the children speak and where they come from.' She then creates a contrast, suggesting 'I remember at my, at the start of my career, then it was a lot "I don't know what languages they speak, don't know where they're from"' (turn 10). Her reasoning is grounded in personal experience, and presented in somewhat tentative terms ('I think') rather than stating 'how it is.' Through her reasoning, she establishes her and her present colleagues' position in contrast to what happened previously (before). She summarizes her reasoning in terms of '… and then one was not as interested as one is today.' Hence, when put in contrast to her and her colleagues' present understandings, she speaks about, 'I' and 'one.' These are also placed in the past. Finally, she formulates a conclusion, 'It has changed, gotten better.' In this way, she argues for the professional group of which she is a member as: 'having the hang of' the linguistic diversity among children in their settings. At the same time, she implicitly and explicitly argues against the professional knowledge and interest that was previously evident. Rhetorically, her reasoning is constituted through a number of distinctions: between us and 'I' in the here-and-now and as knowledgeable, while they ('one' and 'I,' in the past) as there-and-then, were less informed. This reasoning can be understood as talking one's profession into being (cf. Shotter 1993), as part of narrating the identity of this ECE professional group. This identity in-the-making is in part argued through distancing oneself as 'other' and also distinguishing between 'now' and 'before.' This focus group session was held at the start of the preschool teachers' participation in the research project and was focused on general considerations for working with multilingual children.

Excerpt 2: Reasoning About Supporting Usage of the Mother Tongue
This conversation addresses the support given for using mother tongue, rules regulating this support, and whether the staff know this information, as well as informing the children's caregivers about the nature and availability of this support. 35: PC: Then it's important to raise research and experience. Research says, experience says that children best learn Swedish if they have grounding in their mother tongue, not that we think, I think but that research says that it's good.
[---] 37: PA: Only when I worked twelve years ago, or ten years ago, almost forbidden for children to talk their mother tongue. And when two children, for example with Somali as mother tongue or Bosnian start talking their mother tongue, 'no, we talk Swedish here.' It was like that at first. But now it's changed. You challenge the children to play and talk, and you're like this really curious about this. Those limitations are no longer there. 38: PC: Attitude and approach have changed. Here everyone is very interested in language and curious: what's it called in your language? In my language it's called. Our children are early language aware. They know that when I talk to you I cannot talk my home language. But I can talk to her, she understands my home language. They know that it's different languages and that's not obvious to all children. 39: PA: But at in-schooling they may start talking their mother tongue. We listen and try to interpret. Respond in Swedish. But still listen. It has changed greatly this, the approach. 40: PC: But I don't know if it's like that everywhere. I have worked in a preschool where there were only Swedish-speaking children and Swedish-speaking families. There was an entirely different approach. They weren't at all as favorable or curious. There it was only Swedish. [---] It's not obvious that it's like that everywhere.
In arguing the importance of children being given the opportunity to build on their first language when learning a second language, PC makes a distinction, 'then it's important to highlight research and experience. Research says, experience says […], not that we think, I think but research says…' (turn 35). According to Swedish School Law (SFS 2010: 800), school-including preschool-should build on 'research and practiced experience.' PA emphasizes this grounding in contrast to personal opinion ('not that we think, I think'). This reasoning implicitly connects to a long tradition of debate about Swedish ECE, whether it is to be grounded in research and/or other forms of knowledge, for example appropriated through being socialized by more experienced members, or grounded on philosophical/ideological bases.
In this excerpt, PA argues that things have changed through contrasting current practice to previous practice, and by implication how she (and unspecified others) have changed in expected ways: 'It was like that in the beginning, but now it's changed' and '… such limitations do not remain' (turn 37). PC confirms this idea by suggesting, ' … here everyone is very interested in language and curious' (38) and, after PA adds '…it has changed very much this, the approach' (turn 39), '… but I don't know if it's like that everywhere' (turn 40); for example, where she used to work, '…there was an entirely different attitude' (turn 40). Hence, the argument that this group of ECE teachers do their work adequately is made through creating a contrast between 'here ' and 'there,' and 'everywhere [else].' There was also a shift in tempo, from 'here is…' to 'there it was…' (turn 40). In this way, through a set of contrasts, the teachers talk their profession 'into being' by discussing the expectations of what the professional role entails.

Excerpt 3: Reasoning About How Best to Support Language Development
The next conversation starts with teachers' reflection about communication practices. 98: PB: I work with the youngest children and I notice that it's very important as pedagogue not to be quiet, we're not used to, we talk with the children who talk. It's even more important for the children who do not speak, to give them a language and words for what they do. [--] And unfortunately, there are many who do not speak with the children. Who are quiet when they work with the children 'cause they do not speak. And I can feel sometimes, that I'm nagging, talk, talk, name, tell, put word on. 'Cause if you're not aware of it and our children, it's important to our children who don't have Swedish at home. […]. If I don't have these Swedish words, when am I gonna talk Swedish then? So, it's really important. And then when you have older children, I sometimes notice that we're attracted to the talkative children.
[ ---] 104: PA: Mindful way of working. But it can always be better. We think a lot about this, how we are with, conversing with children, how do we communicate? To be in a smaller group. To really be part. To interpret, see what the children do, put words on their actions. 'It's very easy to see who has this approach, who doesn't'. Like this, what the children do, and do not support the children in communication and interaction with each other.
[ ---] 109: PB: Then I think also about the parents we have today, I notice when I work with the youngest children, many parents regardless of cultural background, it may be Somalia, it could be any country. They do not have this tradition like in Sweden, that we talk with our children and tell and name what will happen. Now we're going shopping and you need to put on your shoes. That is, they don't talk with their children, they don't talk with them. They think that if they don't talk, it doesn't exist. So, it's even more important to lay the foundation somewhere, 'cause language is not so important. 110: PA: It's important with role models, I think. When I came to Sweden and parents talked with their children I got an 'a-ha.' It feels like, I began to think that I want to talk like this, in this way with my children. And then your colleagues can also be role models for each other. We can be role models for the parents. I see a change. When you work with awareness that you convey something, important to talk with the children, you sit down and talk with the children, the parents listen. And then you see a change also with the parents when they meet their children, I think.
Talking about how to facilitate children's language development, PB argues for the necessity to talk particularly to those children who do not yet speak by themselves (turn 98). This argument is constituted in contrast to another tradition of argumentation where 'everything should come from the child' and the teachers' to follow the children's lead. PA continues, talking about 'mindful way of working. But it can always be better' (turn 104), that is, again, indicating that they do what they should, while at the same time modesty hinting that they can develop their practice further. Emphasizing, 'to really be a part,' 'one can very easily see who has this approach or who does not', implies a critical difference between themselves (who can still develop what they do) and others (who do not have such an approach).
Constituting themselves as a professional ECE group is done in contrast to (some) parents: '… they think that they [i.e., their children] do not talk so it doesn't exist' (turn 109). PA reconnects, arguing, '…it's important with role models, I think. When I came to Sweden and parents talked with their children, I got an 'a-ha'…' (turn 110). Personal experience is used to ground the teachers' reasoning. In this case, parents as a group are constituted as role models; hence, a distinction which is not clarified is made between some parents and other parents; one group that the ECE teachers need to inform and another group of parents from whom they (she) can learn. Hence, in the teachers' arguments, groups of caregivers (parents) and their perceived characteristics are not uniform and fixed. Rather, these groups are invoked as rhetorical strategies in arguing one's point of view, about one's work as a preschool teacher (i.e., parents as a group can be invoked rhetorically as supporting one's reasoning but also as providing a counter case to one's own stance).

Discussion
In this study, we have analyzed the reasoning of preschool teachers in focus group conversations as they discussed their work with multilingual children in preschool through traditions of argumentations (Billig 1996;Shotter 1993). This meant analyzing not only what the teachers argued for but also at the same time, explicitly or implicitly, argued against. Through our analysis, we have shown how the teachers argue their point of view through constituting three relationships between differently valued poles: • I/we versus them (others); • Here-and-now versus there-and-then; and • Building ECE on research versus personal experience.
These patterns, traditions of argumentation, were prevalent throughout the data, regardless of more specific topics of the conversation. In this final section, we briefly review the findings and the implications for the developmental support provided to multilingual children.
A reoccurring rhetorical strategy in the preschool teachers' argumentation was that they talked their profession into being, through a set of contrasts (cf. Shotter 1993). In this way, they constitute themselves as a knowledgeable and developing group of professionals who are responsive to perceived expectations. However, the image of their profession generated through their arguing is ambiguous. On the one hand, they emphasized their collective knowing, on the other hand, they talked, almost exclusively, in terms of personal experiences. Only on rare occasions did they make explicit the need to ground professional decisions in research (Excerpt 2).
The starting point for the conversations about multilingualism and language use and learning, was rather invisible, unless it was pointed out by the researcher. Instead, the discussions were carried out in rather general terms without grounding the conversations in the contextual conditions in which the majority of the children have first languages other than Swedish. In addition, language learning was referred to in general terms rather than specifying what it includes, such as grammatical, semantic and pragmatic skills, as well as communicative skills and competencies and how these can be supported through pedagogical actions.
An implication for ECE teacher education from these findings and the nature of the reasoning expressed by the preschool teachers is to highlight and provide teacher education students and teacher educators with opportunities to understand how teachers' reasoning is expressed about the complexities in ECE practices and children's learning and development in multilingual settings. On the one hand, this could be seen to maintain and reinforce the inherent professional tensions that are evident in current preschool teacher education. On the other hand, gaining an understanding of these dialogical processes, which can be a basis for learning to think critically, can inform teacher education. Students and teacher educators can understand and be responsive to the inherent complexities which characterize different aspects of professional work in ECE.

Conclusions
In communication, we co-constitute what we talk about and how we speak about phenomena will have material consequences (Säljö 2002). The traditions of argumentation that we have focused on testify that ECE can be a field of tensions between different ideologies and traditions of knowing (e.g., personal experiences versus research-based education). The ambiguous character of argumentation is not necessarily an inherent problem. It can provide realization that pedagogical practices are not clear-cut or one-dimensional. This idea builds on a long-standing but not a dominant tradition of argumentation in educational theory (e.g., Dewey 1916Dewey /2008. We would argue that managing the tensions and dilemmas evident through these research findings lies at the heart of the professional work of early childhood teachers (cf. Pramling and Pramling Samuelsson 2018). Rather than precluding to either side of the dichotomies, often expressed in discussions on professional work in early childhood education, it is important to simply recognize the complexities.
Another contribution of the present study is the insight provided about the reasoning of ECE teachers. It illustrates how traditions of argumentation can be a useful analytical tool to identify what features underpin professional thinking and which tend to remain implicit. These features are crucial elements in shaping the professional work of preschool teachers, for example, the finding that participants' approach to argumentation drew on personal views rather than research findings. However, we only had access to participants' reasoning through group conversations and we do not make any claims beyond this about the basis of their knowledge.
The findings may also illustrate teachers' unfamiliarity with verbalizing their ways of working when they are in conversations with others. Against a background in which multilingual issues in ECE have become a political issue (Freeman Field 2008;García and Otheguy 2020;Lo Bianco 2008;May 2008), skills for verbalizing professional understandings are crucial in current times when it has become important for teachers to base their professional understandings on research evidence and policy documents, as a means to define the profession. Tobin et al. (2013) had also identified that caregivers were more likely to express pragmatic concerns rather than describe theoretical understandings about second language learning.
Another theoretical premise underpinning this study was, in Shotter's (1993) terms, 'to talk in new ways is to "construct" new forms of social relations' (p. 9). From this premise, a conclusion of the present study is that in order for ECE to support the development of children with a first language other than the majority language in any context is for preschool teachers to reconnect with the traditions of argumentation. Assigning a role for preschool teachers on their active participation in mutual activities with children as a means to facilitate learning is also important.