Children’s agency in interactions: how children use language(s) and contribute to the language ecology in Swiss bilingual German-English daycare centres

ABSTRACT

Although Switzerland is a plurilingual country, most early education and care (ECEC) institutions are monolingual. Yet, new institutions have recently established English as a second language of instruction, addressing economically advantaged families. Despite the growing body of international research on language policy and practice in multilingual ECEC, only few have addressed such ‘privileged’ institutions, and the role children play in dealing with multiple languages. We investigate the language policies and practices in daycare centres with a bilingual language policy, and particularly the children’s agency in dealing with German, English, and other languages in interactions with each other, and with teachers. We draw on the concept of children’s agency and view children as actors who contribute to the construction of the social and cultural world in which they live together with adults. We ask how children use languages in daycare centres, how they contribute to the centres’ language practices, and how thereby different forms of agency manifest. We draw on data from a focused ethnography conducted in three daycare centres in German-speaking Switzerland. The results show that children’s multilingual agency is not only enabled and limited in ECEC settings, but also actively developed in concert with language learning by children themselves.

KEYWORDS:

• Children’s agency
• multilingualism
• early childhood education and care
• language policy
• language ecology

I. Introduction

Switzerland is a plurilingual country, with the four national languages German, French, Italian, and Romansh. Yet, its linguistic landscape is strictly separated based on the canton’s official language(s) resulting in a monolingual German-speaking majority. The German-speaking regions can further be characterised by diglossic features with both Swiss German (an umbrella term for myriad local dialectal varieties) and Standard/High German co-existing. While Swiss German is the everyday spoken language, the use of High German is typically restricted to written and formal contexts. In early education and care (ECEC), often only the local official language is spoken. Meanwhile, new ECEC institutions are establishing English, a non-national language, as the second language of instruction and can therefore be described as conceptually bilingual. They primarily address (socio-)economically advantaged families with high expectations regarding their children’s English language learning from an early age (Knoll & Kuhn, 2022). While research on language policy and practice in multilingual ECEC institutions is still thin on the ground in Switzerland, there has been a growing body of international research (e.g. Cekaite & Evaldsson, 2017; Schwartz, 2018; Simoes Lourêiro & Neumann, 2020). However, only few have addressed the role children play in dealing with multiple languages.

In this paper, we investigate the language policies and practices in day care centres with a bilingual language policy, and particularly the children’s agency in dealing with (Swiss) German, English, and other languages in interactions with each other and with their teachers. We thereby aim to investigate children’s and teacher’s language use in situ and to contribute to an ongoing theoretical debate and research agenda in new childhood studies (e.g. Schwartz et al., 2023; Varpanen, 2019). To do so, we draw on a conception of children’s agency in which children are seen as actors who contribute to the construction of the social and cultural world they live in together with adults. We address agency not as a (ontological) ‘feature’ children naturally have, but as being constructed in daily interaction in relation to social conditions. Therefore, children’s agency can be enabled and favoured as well as limited due to given possibilities and restrictions, for instance by institutional rules and teachers’ professional practice. Agency manifests in interaction on the micro-level when children put themselves in a social position and are socially positioned, depending on the opportunities they take, open up and are opened up to them. We ask how children make use of different linguistic resources in daycare centers, how they contribute to the centers’ language ecology, policy, and practice, and how thereby different forms of agency manifest. To answer these questions, we draw on data from a research project conducted 2019 in day care centres in German-speaking Switzerland (Becker & Knoll, 2021).

In the following chapter (II) our theoretical perspective is elaborated, consisting of children’s agency and the ecology of language learning. We then introduce our field of research, the ethnographic methodology of the study, and the strategy of analysis in section III. The subsequent results contain three separate chapters: how teachers monolanguage and children translanguage (García, 2009) in the daycare centres (IV), how they both adapt their linguistic resources (V), and which context factors may enable or limit children’s multilingual agency (VI). We conclude by discussing our results in concert with existing research findings.

II. Agency and the ecology of language learning

Agency, as an empirical concept in education, sociology, and political sciences (Biesta & Tedder, 2006), addresses the question of how actors (critically) respond to different (challenging) situations based on their resources and capacities. To contribute to the reconceptualisation of agency, moving away from viewing it as a merely internal characteristic to emphasising its relationality and embeddedness in various external structures, Emirbayer and Mische (1998, p. 963) define ‘human agency as a temporally embedded process of social engagement, informed by the past … oriented toward the future … [and] the present.’ According to this definition, and as reiterated by Biesta and Tedder (2006), individuals’ agency is expressed through habits and practices ‘in transaction with context’ or ‘by means of an environment.’ That is, it is important to understand how agency is possible in certain contexts and how the environment can enable or impede agentic behaviour (Emirbayer & Mische, 1998).

These conditions are particularly relevant when it comes to children’s agency, referred to as ‘one of the most important theoretical developments’ by James (2009, p. 34) and concerned with ‘children as social actors’ (James, 2009). In van Nijnatten’s (2013) concept of ‘interactive agency,’ linking identity and welfare to the development of successful agentic behaviour, conditions which can increase children’s agency are explored. Children only develop interactive agency throughout their socialisation processes and especially in interactions through and with language (ibid.). As he points out, ‘Language forms the basis of human development; it is at the core of personal and social identity … [and] a vital function for becoming a person and for making coherent past experiences, actual feelings and acts, and future projections’ (van Nijnatten, 2013, p. 18).

More recently, early childhood education and care (ECEC) research with a focus on bilingualism has applied the concept of agency to children’s interaction in daycare centres. Boyd et al. (2017) analysed children’s agency in Swedish conceptually bilingual preschools. Children used code alternation to show alignment. Yet, unlike the teachers who prioritise the non-local language, children also relied on Swedish. Thus, children’s agency influenced the language policy and their multilingual practices. Almér (2017) found that monolingual bias is reproduced through children’s beliefs and in form of institutionalised one-language-one-person policies although children used different languages. Children have a functional interactive agency and they can make their voices heard in interactions with adults to shape the dialogue, take on different roles and positions, and make use of varying linguistic resources. Examining two children’s language-based agency in Luxembourg and Israel, Schwartz et al. (2023, p. 819) ‘identified 10 types of agentic behaviour, including engaging in repetition after peers and the teacher, creatively producing language, translanguaging, and self-monitoring,’ while calling for more research in exactly those forms. As further convincingly argued by Kirsch (2018), such agentic behaviour, and especially one that emerges out of multilingual contexts, is highly dependent on teachers’ interpretation and implementation of language policies and their own identification as multilinguals (cf. also Becker, 2023). Simoes Lourêiro and Neumann (2020) similarly focused on daycare centres’ language practices and policies in Luxembourg and found that children shape the existing language ecology significantly despite the missing representation of children’s voices in ECEC curricula. As they put it, ‘ … children … actively contribute to the constitution of the institutional language practices by affecting practitioners’ language use [and] … thus undermine the institutional language policy … ’ (Simoes Lourêiro & Neumann, 2020, p. 172).

Contributing to this existing, but thus far rather selective literature especially on multilingual ECEC contexts and children’s agentic behaviour, this study investigates children’s agency in interaction and their impact on the local language ecology. Adopting an ecolinguistic perspective, this study employs Hult’s (2009, p. 89) definition of language ecology as a tool ‘to map aspects of multilingualism by tracing how the specific language choices of individuals construct and are constructed by the social environment as it takes shape across nested ecosystems.’ Given the embeddedness of agency in sociocultural structures and conditions (Ahearn, 2001), the study draws on the concept of ecology of language learning to analyse children’s agency in interactions with peers and teachers. This is considered important since conceptual language policies determined by the daycare centre management and their active implementation through teachers and more flexibly also through children always exist within a sociocultural environment.

Language policy, according to Bergroth and Palviainen (2017) is interpreted by ‘policy agents’ with different consequences for those individuals who are perceived as rather ‘powerless,’ without, or with limited agency. This can refer to the classroom and typical teacher-student relationships, the institution or wider society, adding multiple layers of interdependencies to the study of children’s language learning and a more critical understanding of agency. van Lier (2010, p. 3) employs the concept of ecology to situations of language learning to analyse ‘the learning process, the actions and activities of teachers and learners, the multilayered nature of interaction and language use … among all the elements in the setting.’ Underlying constituents of such settings, according to van Lier (2010), are relationships, quality, and agency, which enable learning and teaching and are therefore crucial for language learning from an ecological perspective. For this study, specifically language practices in the classroom are relevant. Yet, as cautioned by van Lier (2010, p. 3), ‘in classrooms the types of interaction that predominate often strip the multilayeredness away from the discourse, resulting in utterances that do not allow the learners’ voices to develop and diversify.’ That said, such settings can also be seen as language-conducive contexts (Schwartz, 2018) when the constituents of relationships, quality, and agency are beneficial and foster learning and development (van Lier, 2004). Schwartz (2018, p. 6) defines such language-conducive contexts ‘as contexts rich in multisensory activities with a wide array of semiotic resources and diverse teacher–child and peer interactions, encouraging the child’s engagement in the novel language learning.’ Such conditions can contribute to the children’s natural language acquisition process not only on a linguistic level, but also to their openness to and appropriation of other languages as indicated by their agentic behaviour (van Lier, 2010). It is the present study’s aim to examine to what extent language-conducive contexts are created through which meaningful peer-child, teacher–child interactions, activities, strategies, and policies to ultimately enable and increase children’s agency meaningfully.

III. Study, field, and analysis

In this article, we draw on data from a 2019 research project located at the University of Fribourg, Switzerland. The aim of this project was to investigate children’s and teacher’s language use in situ in conceptually multilingual daycare centres in German-speaking Switzerland (Becker & Knoll, 2021). We conducted an ethnographic study, primarily observing everyday life in daycare centres and talking to teachers and the centres’ managers. Additionally, we collected internal documents and website content from the centres (linguistic concepts/pictures). For data collection, we spent 1–3 days weekly on average in three daycare centres located in Swiss urban areas from March to June 2019. These were selected from a total of 87 multilingual preschool institutions we identified based on a systematic online search.¹

Within these three daycare centres we examined in depth, there are groups of ten to twelve children, three months to four years old, and educated by three to four teachers simultaneously, one of whom usually speaks English and the others German. They do not provide publicly supported places and therefore primarily address economically privileged families. Children’s first languages include, besides German: Bosnian, Finnish, French, Greek, Hebrew, Italian, Korean, Portuguese, Romansh, Russian, Spanish, and Turkish.

Apple Tree daycare centre is situated close to the main train station of a big city. The building is surrounded by a small park where children can spend time in free play sequences. Generally following a Montessori approach, the centre subordinates multilingual education to this concept. Unlike in the other two centres, the English-speaking teacher’s first language (L1)² is not English, but another European language; therefore, she speaks English with a non-standard accent. The other teachers speak Swiss German with the children, and in some particular situations in the children’s heritage languages (e.g. Italian or Spanish). The children’s language use is not restricted – they can use any linguistic resource.

Butterfly daycare centre is in the suburbs of the same city. Multiple daycare groups and a kindergarten class are housed in the building with a playground next to it. Teachers are only allowed to speak in their L1: one teacher speaks English; three others speak Swiss German with the children. As in Apple Tree, children are allowed to speak as they want/can.

At Little Birds daycare centre, however, children’s language use is restricted: as reported by the manager, if children use a language other than English or German, they are told not to do so. Little Birds is in a city district of another Swiss city which recently was subject to urban upgrading and gentrification. Multilingualism plays a dominant role in the centre’s pedagogical concept. One teacher, an English-German bilingual native speaker, is responsible for speaking only English with the children, while the other three teachers speak Swiss German.

As the study is conceptualised as a ‘focused ethnography’ (Knoblauch, 2005), we limit our analysis to the examination of how children’s agency manifests in different forms, how they make use of language in daycare centres and thereby contribute to the centres’ language ecology, policy, and practices. Data basis for the analysis are fieldnotes from participant observation of multilingual interactions among children and teachers, including transcripts of verbal interactions and door-and-fishing-talks with teachers. Using a coding strategy based on Grounded Theory (Corbin & Strauss, 2015), we first identified relevant sequences from fieldnotes. Second, we subjected these sequences to a detailed sequential analysis (Breidenstein et al., 2013) with the aim of obtaining the widest possible range of forms of children’s agency.³

IV. Translanguaging children and monolanguaging teachers

In the three centres we investigated, German and English are used by teachers and expected to be used by the children as well, which they both use. Hence, children quickly learn that more than one language can be used. Whereas some children, particularly young ones or new to the daycare centre, may still be in the phase of tentative attempts to speak in a new language, others already use multiple languages and navigate between them:⁴

Stephanie (4):

‘I’m Rubble.’

Seraina (4):

‘Ryder, what is your name?’

Celine (4):

‘Stephanie, ich bi de Marshall.’ (‘I’m Marshall’)

Seraina:

‘Ond ich bi de Skye.’ (‘And I am Skye.’)

Stephanie:

‘D’Seraina isch de Skye.’ (‘Seraina is Skye.’)

Celine:

‘Nei ich be de Papa [incomp.]. Ond du besch d’Chase.’ (‘No, I am daddy [incomp.]. And you are Chase.’)

Seraina:

‘Ryder, come.’

Stephanie:

‘To the monkey. Here’s Rubble. Du besch Rubble.’ (‘You are Rubble.’)

Celine:

‘Monkey.’

Seraina:

‘I don’t know who is this. It will be a monkey.’

Stephanie:

‘Oh no, this will be Rocky. Are you all right Rocky?’

Celine:

‘Ich ben de Rocky.’ (‘I am Rocky.’)

(Little Birds, audio transcript from 22 May 2019)

While Stephanie, Seraina and Celine are playing the dog characters known from a cartoon/movie series, they integrate Swiss German and English and translanguage between and within a turn, manifesting their multilingual communicative agency. They can understand each other and act across linguistic barriers. In acting multilingually, they stimulate others to do alike, thereby fostering their multilingual agency.

However, multilingual agency is not restricted to children’s verbal communication in German and English:

Philip (3), Taro (3) and Dina (4) are playing at the doctor’s corner. Philip takes a toy phone and speaks something into it that I first do not understand, then puts it down again and imitates an incoming call sound: ‘ring-ring-ring-ring.’ Then Taro and Philip both take a pillow and throw them at each other, each saying ‘Hello Philip’ and ‘Hello Taro’, and laughing together. After a few times back and forth, they lie down on the couch and start talking. Again, I don’t understand what they are saying at first, but then I realize that it must be fantasy language. (Butterfly; fieldnotes from 2 May 2019)

In this sequence, the children include nonverbal acting into their play: laughing, telephone sounds, and a pillow battle. The use of fantasy language shows that they can extend their linguistic resources to non-existing languages and use it for playing and interacting.

While many children translanguage and show a substantial multilingual agency, the teachers largely monolanguage. We observed this linguistic practice in all three daycare centres in a plethora of interactions between teachers and children, and it is specified in the existing centres’ linguistic concepts: each teacher is required to either speak German or English. This practice is followed to a large extent, most strictly at Little Birds, and rather consequently at Butterfly’s, but much more loosely handled at Apple Tree’s. However, while each teacher only speaks one language at a time, multiple languages are used across all teachers who are present at the groups. Therefore, with the terminology of Gort and Pontier (2013), one can describe these settings as a bilingual instructional language context, combined with a teachers’ monolingual language mode and a one-teacher/one-language concept.

V. Language adaptation

Teachers sometimes adapt their language use according to their opposite. For example, the researcher observed how a child was brought by her father in the morning:

I see the father speaking with Cynthia (teacher; Swiss German). Then Cynthia says goodbye to him: ‘Have a good day, bye’. Without a break, she addresses the child (3)⁵ that is now in her care: ‘Was häsch debii?’ (‘What do you carry with you?’) (Little Birds, fieldnotes from 29 April 2019)

As she turns from father to daughter, teacher Cynthia adapts her use of language by moving from English to German.

Children on their part can adapt their language use (if needed), and conversely also know how to maintain their language(s). This flexible language use, depending on the competencies in German and English, and knowledge about appropriate usage, is itself a competence and manifestation of a practical multilingual agency. Children sometimes adjust their language practice to a teacher, as Francesca (2; High German) does at Little Birds:

Francesca:

‘JOHN HAT MI-MICH GE-E-GE-GHÄLFT.’ (JOHN HE-E-HE-HELLPED M-MYSELF.) (sic!)

Zoey (teacher):

‘Did he help you?’

Francesca:

‘Yes.’

Zoey:

‘Did you say < thank you John>?’

Francesca:

‘Thank you John.’

(Little Birds; audio transcript from 6 May 2019)

Francesca seems to be influenced by Zoey’s linguistic traction and adjusts her language use to English, partly producing an actual repeating of what her teacher said. Possibly Francesca felt relieved that she was offered to continue in English, as she experienced linguistic uncertainty in High German in the previous turn. In any case, Zoey leaves Francesca the option to express herself in German or English.

At Apple Tree, teacher Maria is employed to speak English with the children. She understands (and occasionally speaks) High German, but not so much Swiss German. Hence, we observed several times that children usually speaking Swiss German use High German when addressing Maria to ensure mutual understanding:

Maria is helping children to get dressed and speaking with them in English. She turns to Ronja (3; Swiss German) to put a jacket on her. But Ronja, who does not yet speak English, shakes her head and says to Maria: ‘Neeei, SIND NICHT MEINE.’ (Nooo, ARE NOT MINES.’) Maria replies: ‘These are not yours?’ and hangs the jacket back up on the hook. (Apple Tree; fieldnotes from 9 April 2019)

Ronja, similar to other children who do not yet speak English, seems to anticipate that Swiss German will not help communication with Maria reliably, and that she has to adjust to be sure that the teacher understands her.

However, children do not always adjust their language to their teacher. There are many interactions we observed where children maintained their language, as in the following interaction with Zoey again:

Chiara (3; High German) is sitting on a bench between the teachers Zoey and Vivienne. Zoey is about to tell me about her upcoming training course when Chiara suddenly scares Zoey and says:

Chiara:

‘BUUH!’

Zoey (teacher):

‘I’m so scared.’

Chiara:

‘DU BIST EIN DINOSAURIER!’ (‘YOU’RE A DINOSAUR!’)

Zoey imitates a frightened face.

(Little Birds; fieldnotes from 12 April 2019)

Unlike Francesca in the sequence above, Chiara is not influenced by Zoey's English. She seems to have no reason here to deviate from her usual High German, also knowing that Zoey understands her.

Children not only sometimes tend to adjust their language practices to the ones of teachers, but also of other children. In the next sequence, Letizia (3; Spanish, English), Dina (4; Bosnian), and Samira (4; Swiss German) are playing in the sandbox, with Klara (teacher; Swiss German). Samira is playing with a small shovel from a doctor play case when Letizia intervenes:

Klara (teacher):

‘Hesch de du Schüfeli überhoupt, Samira? (…) Geits mit däm chline Schüfeli?’

(‘Do you even have a shovel, Samira? (…) Can you do it with the small shovel?’)

Samira:

‘Ja.’ (‘Yes.’)

Klara:

‘Ja?’ (‘Yes?’)

Dina:

‘You heavy (incomp.).’

Letizia:

(directed to Samira) ‘No, that is not (incomp.), it’s doctor!’

Klara:

‘Öh, das macht doch nüt.’ (‘Eeh, that doesn’t matter.’)

Letizia:

‘Samira!’

Klara:

‘Es isch nid schlimm, Letizia, nid schlimm.’ (‘That’s not bad, Letizia, not bad.’)

Letizia:

‘That one doctor.’

Klara:

‘Ja, das chame glich oh chli bruche zum im Sand zspile.’ (‘Yes, one can also use this to play in the sand.’)

Letizia:

‘(incomp.)’

Klara:

‘I ga mau ga luege, villäch hets ja no es chlises Schüfeli, villäch fingi irgendwo eis.’ (‘I'll go see, maybe it still has such a small shovel, maybe I can find one somewhere.’)

Letizia:

‘Samiraaa, Neeei! Tu nid! Das isch von die Doktor! (…) Aber das isch von die Doktooor!’ (‘Samiraa, nooo! This is from doctor! (…) But this is from doctooor!’)

Samira:

‘(incomp.)’

Letizia:

‘Mou, von die Doktooor!’ (‘Yes, from doctooor!’)

Klara:

‘Lue mau i ha glich no es chlises Schüfeli gfunge, wosch ds?’ (‘Look I found another small shovel, wanna have it?’)

(directed to Samira) ‘Geit villäch scho chli gäbiger aus ds Doktording.’ (‘Maybe it still works better than the doctor thing.’)

Letizia:

‘(incomp.)’

(Butterfly; audio transcript from 22 May 2019)

Letizia first tries to address Samira in English, the language she mostly uses in Butterfly daycare centre. After some attempts and no reaction from Samira, Letizia tries in Swiss German, aware that Samira does not understand English yet. Although Samira’s reaction is incomprehensible on the recording, she reacts verbally to Letizia’s intervention. This in turn gives Letizia the feedback that it is worth adapting one's own use of language to be better understood.

In sum, adjusting language use and translanguaging helps children to ensure communication and mutual understanding with teachers and peers. The awareness of language ecology, and who speaks and understands which languages, points to an important aspect of children’s multilingual agency: the practical ability to distinguish situations where speech adaptation is useful from those where it is not (necessary), or where you can just speak the way you want.

VI. Enabling and limiting children's multilingual agency

Children’s multilingual agency depends largely on enabling and limiting factors. Those can take on different forms: structural (e.g. the daycare centre’s bilingual concept, language ecology and instructional learning context), situational (e.g. interaction dynamics), and personal (e.g. children and teachers). In this chapter, we mainly focus on teachers and their role in facilitating and restricting children’s agency.

In all three daycare centres we observed teachers encouraging children to use different linguistic resources than the ones they usually use or to translanguage. This encouragement can take different forms, to wit: it can be implicit or explicit, gentle or rather insistent. For example, Rebecca, a teacher at Little Birds, asked Stephanie (4; English and Hebrew) what ‘flamingo’ means in Hebrew, and Stephanie said the word. On the one hand, this can be seen as an opportunity the teacher gives to the child to show and practice his/her multilingualism. But, on the other hand, such situations can also go wrong, namely when the child cannot or does not want to ‘provide’ the translation and may have a shameful experience (Akbaba, 2014).

In total, we found such teachers’ invitations for children to speak in their heritage languages (other than German and English) to be quite seldom (Becker & Knoll, 2021). At Little Birds, the centre manager told us in an interview that children would not be encouraged to use other languages than German and English. Even to the contrary: if they did, the employees would be instructed to remind the children to go back to the two official languages spoken in the centre (and this rule applies analogously to the parents). From a perspective of children’s agency, this rule and practice has to be treated as a limiting factor since it prevents children from expressing themselves linguistically in the way that suits them best. Hence, the language policy of this institution is taking a limiting role in children’s multilingual agency. To the contrary, Apple Tree and Butterfly centres, where policy does not prohibit the use of children’s first languages, rather encourage it. In sum, on the part of daycare centres and their linguistic concepts regarding children’s first languages, there are factors both enabling and limiting children’s linguistic agency.

As long as it concerns the two languages German and English, however, teachers are mostly well aware of the children’s multilingual abilities and try to promote them. Rebecca, in this example, is heading a circle sequence, the children sitting on the floor. In a repetitive welcome sequence, a child's name is drawn from a basket, which is welcomed by everyone with a short clap. The child is asked by another child how he or she is doing, and after his or her answer Rebecca passes on to the next child. After some turns, the name ‘Noémi’ is drawn, but the five-months-old girl is just having lunch with Zoey, the other teacher, in the next room. So, Rebecca asks four-year-old Stephanie:

Rebecca:

‘Magsch mal d Noémi go fröge also d Zoey wies de Noémi gaht?’ (teacher) (‘Would you like to ask Noémi, that is, Zoey, how Noémi is doing?’)

Stephanie leaves the room without a word, I can hear that she is asking Zoey something, the door is open, but I do not understand anything. After about 15 s, Stephanie returns.

Rebecca:

‘Und?’ (‘So?’)

Stephanie:

‘Mega mega guet und (unv.).’ (‘Very very good and (incomp.).’)

Rebecca:

‘Okay.’

Rebecca then turns to me for a moment and says quietly, in addition to a previous conversation between us:

Rebecca:

‘Isch jetz villeicht au grad spannend gsi d Zoey hät e Englischi Antwort gäh und d Stephanie hät sie wien ich dir gseit han mir uf Dütsch übersetzt also.’

(‘That was maybe also interesting Zoey gave an English answer and Stephanie translated it like I told you into German for me so.’)

(Little Birds; audio transcript from 29 April 2019)

First, Stephanie’s ability to translate Zoey’s English answer into Swiss German for Rebecca needs to be highlighted, a competence, which Stephanie can show in this situation. Second, Rebecca is well aware of this ability Stephanie demonstrates, she seems to use the situation (that Noémi was not in the room) to offer Stephanie a linguistic exercise. And third, the sequence also shows that children’s linguistic agency is treated in dependence of age: while Noémi is too young to express herself, Zoey (teacher) has to speak instead of her, and older Stephanie can deliver the message.

In another sequence at Butterfly daycare centre, language policy and the teacher’s language practice clearly prove to be limiting factors for children’s agency. Teacher Klara (Swiss German) is playing a counting game with around half a dozen children, including four-year-olds Pascal (English and Polish), Mandy (Swiss German), Lena (Swiss German) and Dina (Bosnian), in which they can choose their language. After they already had English and Italian, it is now up to Pascal to decide:

Klara (teacher):

‘MÖCHTEST DU schwiizerdütsch ODER HOCHDEUTSCH?’ (‘WOULD YOULIKE Swiss German OR HIGH GERMAN?’)

Pascal:

‘HOCHDEUTSCH … ODER VIELLEICHT POLNISCH?’

(‘HIGH GERMAN … OR MAYBE POLISH?’)

Klara:

‘WAS MÖCHTEST DU? Eis, zwöi, drü, ODER EINS, ZWEI, DREI?’ (‘WHAT WOULD YOU LIKE? One, two, three, OR ONE, TWO, THREE?’)

Lena:

‘Isch Polnisch Schwizerdütsch?’

(‘Is Polish Swiss German?’)

Klara:

‘NEIN, ICH GLAUB PASCAL MACHT MANCHMAL EIN BISSCHEN EIN DURCHEINANDER MIT SCHWEIZERDEUTSCH UND POLNISCH.’ (‘NO, I THINK PASCAL SOMETIMES MAKES A BIT OF A MESS WITH SWISS GERMAN AND POLISH.’)

[…]

The teacher decides to count in Swiss German.

Klara (teacher):

‘SCHWEIZERDEUTSCH, WIR PROBIEREN ES, DAS TÖNT NÄMLICH FAST WIE POLNISCH.’

(SWISS GERMAN, LET’S TRY, IT SOUNDS ALMOST LIKE POLISH.’)

all:

‘Eis, zwöi, drü, vier, füf, sächs, siebä, acht.’

(‘One, two, three, four, five, six, seven, eight.’)

Mandy:

‘ABER SO GEHT DOCH GAR NICHT POLNISCH?!’

(‘BUT POLISH DOESN’T WORK THAT WAY?!’)

Klara:

‘ICH KANN EBEN GAR NICHT AUF POLNISCH. WIR MÜSSEN MAL DIE MAMA FRAGEN VON PASCAL, DAMIT SIE UNS HELFEN KANN UND SAGEN WIE’S GEHT.’

(‘I JUST DON’T SPEAK POLISH. WE HAVE TO ASK PASCAL’S MOM SO THAT SHE CAN HELP US AND TELL US HOW TO GO.’)

Lena:

‘Aber i weiss nid was Polnisch isch (incomp.) d Letizia will immer ämm Spanisch, de Pascal wett immer Polnisch und wär wett … ’

‘(But I don’t know what Polish is (incomp.) Letizia always wants ehhm Spanish, Pascal always wants Polish and who wants …)’

Klara:

‘IST DOCH SUPER, DANN KÖNNEN WIR IMMER VERSCHIEDENE SPRACHEN ÜBEN.’

(‘THAT’S GREAT, THEN WE CAN ALWAYS PRACTICE DIFFERENT LANGUAGES.’)

Lena:

‘Und was wett immer de Philip?’

(‘And what does Philip always want?’)

Klara:

‘ICH GLAUB PORTUGIESISCH MEINST DU, ODER?’

(‘I THINK IT’S PORTUGUESE YOU MEAN, RIGHT?’)

Lena:

‘Ja.’ (‘Yes.’)

(Butterfly; audio transcript from 8 May 2019)

Pascal’s choice to count in Polish is ignored by the teacher and as a result children’s multilingual agency is strictly limited by the centre’s language policy and the teacher’s language practice, that is, limited to legitimate languages such as Italian and English, but not Polish (see Becker & Knoll, 2021, pp. 8–9). While Pascal does not verbally react to the ‘miscounting’, other children do: Mandy deconstructs the apparent analogy of Swiss German and Polish postulated by the teacher. And Lena's statement manifests that she is overwhelmed by the situation with many different languages present. This feeling of being overwhelmed may be triggered by the centre’s perfectionist approach to language: only native speakers should speak a language and should only count if they can say the numbers correctly. With her intervention, Lena is indirectly questioning an extended use of several languages by different children and teachers in the centre. However, she remains unheard by her teacher.

VII. Discussion and conclusion

Drawing on the concepts of children’s agency and ecology of language learning, the present study investigated how children’s agency manifests in different forms, how they make use of multiple languages in daycare centres and thereby contribute to the centres’ language policy and practice. Generally, the results depict that the three daycare centres provide language-conducive contexts (Schwartz, 2018) in order for children to successfully develop language competences in the locally spoken language (German) and English. The institutions mainly incorporate language learning in playful interactions and other non-language-focused activities. Given that children are free to choose the language – with the limited choice of German and English – the conditions to engage in multilingual practices and thus to interpret the existing language policies according to their own desires and needs are very favourable. In fact, children can employ translanguaging strategies including German and English, depending on their interlocutors, situation, or activity (as captured in the interaction among Klara, Samira, Dina, and Letizia above, for instance). By so doing, they not only demonstrate that they can flexibly use their linguistic resources, but also that they determine which ones are more appropriate given the linguistic task at hand, therefore demonstrating their agency. What is more, by adapting their way of speaking to their interlocutors and the linguistic resources those possess, they express not only multilayered agency, but also a profound awareness of the daycare centres’ language ecology. Most importantly, this behaviour indicates that children engage in multilingual agency, that is, they appropriate the language-conducive contexts and therefore use them to their favour to both develop their language repertoire and agency through multilingual practices. Additionally, through those practices, children actively shape the daycare centres’ language ecology, as similarly reported by Simoes Lourêiro and Neumann (2020) for multilingual children in Luxembourg. The interaction captured among Celine, Seraina, and Stephanie at Little Bird’s, for instance, indicates that they even go one step further and incentivise their peers to also translanguage and – as we would argue – express pride in being able to do so, which again showcases their agency. Interestingly, according to Morton and Evnitskaya (2018), such language practices are rather rarely captured in ECEC research and demonstrate that children in the present study’s context generally feel at ease in the Swiss daycare centres’ language ecology. As Morton and Evnitskaya (2018) further argue, typically, peer interaction is conducted in the children’s first or the local languages whereas the second language – in this study’s case English – would be more strictly reserved for instructional purposes.

Our results have also shown that children can expertly distinguish between situations in which such behaviour is necessary and/or beneficial and when relying on the locally spoken language – which remains the status-quo language in all daycare centres – is sufficient to achieve mutual understanding and meaning-making. This situation is exacerbated by the imposed and practiced one-teacher-one-language policy, which creates the impression that monolingualism is the norm and that each person should only speak one language with some teachers applying ‘double monolingualism’ (García, 2009). In their position as the daycare centres’ authorities and role models for language acquisition, such enacted language policies can be detrimental to children’s agency and language learning. As our results have shown, the one-teacher-one-language strategy tends to impede children’s translanguaging but does not prohibit it. Similar to previous results (cf. Boyd et al.’s, 2017; Simoes Lourêiro & Neumann, 2020), children adapt their language practices to a certain extent to the prevalent monolingual standard but interpret it more flexibly according to their communicative needs (see for instance the interaction between Francesca and Zoey above). As Panagiotopoulou (2016, p. 23) put it succinctly, ‘Young children thus experience that their inherited linguistic repertoire must first be corrected by monolingualising language promotion measures [our translation].’ Depending on the daycare centres, the ‘monolingual habitus’ (Gogolin, 2008) is reproduced rather consequently at a personal level at Little Birds, for instance, and less strictly at Apple Tree. It also tends to be reproduced more generally at group level at all three locations. Such monolingual or conceptually bilingual practices have recently been criticised in the field of ECEC since the artificial separation of languages through promotion concepts does not mirror the existing linguistic diversity and therefore advantages speakers of the languages included in the daycare centres’ concepts while excluding others (Becker & Knoll, 2021; Jaspers & Verschueren, 2011; Kirsch, 2018; Neumann, 2012).

To be sure, given the heavy emphasis on a conceptual bilingual German-English language policy, children’s language acquisition and practice is well promoted in those languages – but limited to those languages only. Children’s (or teachers’, for that matter) heritage languages are excluded from institutional language policy and practice, as vividly depicted in the quoted conversation among Klara, Pascal, and Lena about Pascal’s L1 Polish above. Klara tries to ignore Pascal’s wish and forces him to choose between Swiss German and High German. Interestingly, the children understand that Swiss German is not the same as Polish, deconstructing the teachers’ attempt to solve her ‘problem’ and putting her in need of explanation. While counting in Italian, a language that is not part of the centre’s bilingual concept, is possible and seems legitimate, it is not in the case of Polish. Hence, a gap among different HLs becomes evident. Pascal’s L1, Polish, an Eastern European language, is not recognised as a legitimate, in-class language, while Italian, a Western European language spoken in a neighbouring country and in parts of Switzerland is, although it is not one of the children’s L1.

Finally, as the results of our analysis have shown, multilingual agency is not only possible in ECEC settings, but also in fact actively developed in concert with language learning by children themselves. To facilitate these processes and integrate children as bottom-up, ‘bilingual policy agents’ (Bergroth & Palviainen, 2017), this study advocates the recognition of children’s voices and agency as well as their diverse linguistic repertoires. Through a better understanding of what requirements are necessary for language-conducive contexts in ECEC settings, more research is needed taking into account overt and covert language policies and the hierarchies created around ostensibly ‘prestigious’ and ‘less prestigious’ languages. It is a contribution to ‘the ways in which children themselves can be understood as active participants in society’ (James, 2009, p. 34) and to the promotion of multilingual agency through innovative, bottom-up ECEC policies and practices for an inclusive and equitable language ecology.

Acknowledgements

We would like to thank the three daycare centers, namely the managers, teachers, children, and parents, for participation and support of our study, as well as to our master students and interns, Sarah Frenz, Lea Roos, Lea Schneider, Andrea Schweizer, and Michelle Willen, for their contributions to data collection and analysis. For valuable feedback on a former version of this paper, we owe special thanks to two anonymous colleagues who also contributed to this special issue.

Disclosure statement

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

Notes

1 Sampling criteria: expected language use, size of centre/groups, accessibility.

2 Although (strong versions of) translanguaging theories tend to use linguistic repertoires and idiolects to do away with the idea of fixed linguistic entities, we employ the term language here to showcase the institutional diversity and complexity.

3 Some sequences from fieldnotes we use were already subject of analysis in a former publication which examined the use of children’s heritage languages in daycare centres (Becker & Knoll, 2021). For this contribution, we re-analysed these sequences based on the current research question.

4 We translated all the excerpts from our fieldnotes and transcripts into English (regular letters). Original language is displayed in direct speech: High German in CAPITALS, Swiss German dialect in italic.

5 Numbers in brackets indicate the age of the child in years.