‘You don’t know how to say cow in Polish’. – Co-creating and navigating language ideological assemblages in a linguistically diverse kindergarten in Germany

ABSTRACT This study examines how language ideologies are negotiated and navigated in a linguistically diverse kindergarten group in Germany, focusing on the multilingual language practices of teachers and children. Drawing on data generated during 3months of focused linguistic ethnographic fieldwork, I analyse situations in which children and teachers actively include languages other than German into the kindergarten discourse through, e.g. translation requests, switches to family languages, and references to family languages. An ethnomethodological approach is adopted to trace how participants locally assign meanings to different languages and language use in interaction. The findings show that teachers and children express various, at times opposing language ideologies, leading to the dynamic formation of language ideological assemblages. Children position themselves in these assemblages by reworking them and/or foregrounding different aspects of their own multilingual identifications.


Introduction
Kindergartens in Germany are attended by children with a great cultural and linguistic diversity: 22% of kindergarteners primarily speak another language than German at home (Autorengruppe Bildungsberichterstattung, 2020, p. 8) and 29% have an ascribed 'migration background', meaning that they themselves or at least one of their parents migrated to Germany (Autorengruppe Bildungsberichterstattung, 2020, p. 87).Linguistic diversity in Early Childhood Education and Care (ECEC)-settings is, much like in any other social context, a dynamic phenomenon arising from emergent interactional language practices that are mediated by language ideologies, rooted in sociocultural contexts.
At their core, language ideologies are beliefs about language that link to wider social and political dimensions (Woolard, 2020).
Children, in their play, may orient toward dominant language ideologies regarding context-sensitive appropriateness, relationships, and sociocultural meanings of language resources starting as early as kindergarten age (Cornips, 2020;Paugh, 2012).Even though multilingualism in ECEC is a growing field, there has been little research on how children actively engage with language ideologies in ECEC environments and, as such, negotiate social meanings of diverse linguistic resources, as well as on how teachers take part in such negotiation processes (but see, e.g.Bergroth & Palviainen, 2017;Cekaite & Evaldsson, 2017;Zettl, 2019).Therefore, the main interest of this paper concerns the ways in which children and teachers navigate multilingual language practices and engage with language ideologies in the context of a linguistically diverse kindergarten group in Germany.
Understanding multilingualism as more than just a matter of resources, but also as an 'interactionally-framed practice' (Blommaert et al., 2005), leads me to take into account social processes of (power) positionings in relation to languages and (non-)speakers of these languages.

Language ideologies as a lens on language practices
Language ideologies refer to the beliefs that individuals and groups hold about language and its use (Woolard, 1998).Such beliefs mediate social processes of language use and the co-creation of meaning (Silverstein, 1979).Language ideologies emerge and manifest in interrelation with wider social, cultural and political frameworks and processes, as Irvine (1989, p. 255) pinpoints: '[Language ideologies are] the cultural […] system of ideas about social and linguistic relationships, together with their loading of moral and political interests'.For example, an ideology of language purism that favours a standardised variety, has been particularly salient in nation states, where language has typically been linked to national identity (Wright, 2016).
Language ideologies manifest in social interaction in various ways, ranging from explicit talk about language to implicit metapragmatics, marked by, for example, contextualisation cues as part of language in interaction, which signal interaction partners how to interpret language (Woolard, 1998).As socially constructed objects, language ideologies are dynamic and multiple.Kroskrity has captured this dimension with the notion of 'language ideological assemblages'.In any social setting, various language ideologies might be at stake, alongside other ideologies that simultaneously contribute to social processes of meaning-making within that setting (Kroskrity, 2021).Kroskrity highlights that these become meaningful in assemblages in interrelation with actors' positionalities in political economic structures, as well as their own awareness of language ideologies, linked to their experiences of socialisation (Kroskrity, 2021).In an ECEC setting, ideas about language use might, for example, emerge in interrelation with educational ideologies (Cekaite & Evaldsson, 2008).
Language ideologies have been used as a fruitful approach in understanding dynamics related to language use in multilingual settings (e.g.Cornips, 2018;Karrebaek, 2013).Through language ideologies, using specific languages or certain ways of speaking, come to index, for example, societal positions which get linked to certain identity characteristics.Closely related to that, indexical orders come into being, for example when specific accents or word choices get to signal someone's ethnicity or regional background, which is then associated with, e.g.social status or level of education (Silverstein, 2003).As part of that process, language hierarchies can emerge, constellations in which one language gets associated with a higher status than another language, and these can consequently shape interaction norms in a given social context, like in ECEC settings (Cekaite & Björk-Willén, 2013).

Children's engagement with language ideologies
Language practices are considered by an underlying ideological dimension which manifests in language regimes, shaped by language hierarchies, in which children, throughout their socialisation, actively take part (Purkarthofer & De Korne, 2020).For example, children in Dominica (Eastern Caribbean) show sensitivities toward power dynamics between languages and their distinct domains of use during play.Paugh (2012) describes that they use English when playing school and use Creole Patwa when enacting adult male roles associated with places beyond the home-/school-context, e.g.farmers or bus drivers.Patwa is the local variety depicted as vulgar by their caregivers.However, children can also resist linkages between domains of use and languages in peer play and rearrange indexical orders, as shown for the case of Mayan Tzotzil siblings by de León (2019).These children created ludic spaces in which they commented on and challenged conventional links between Mayan Tzotzil, Spanish, and respective domains of use.For example, as part of their pretend-play of mothers responding to an earth quake announcement on the radio, they created parallel constructions in Spanish and Tzotzil, challenging domain-specific patterns and levelling the two languages.Laughter was used to mediate an awareness of disparities between domains and according lexical gaps.
Taking a look beyond the home context, children often make their first institutional experiences in Early Childhood Education and Care (ECEC).Also in ECEC, language use is structured in specific ways, often corresponding to dominant language ideologies expressed in language policies and daily interaction.Children actively take part in these language policies (Bergroth & Palviainen, 2017).They do so in resonance with the institutional policies, yet, in contrast to simply enacting them, they are also engaged in maintaining, undermining and alternating them (Simoes Lourêiro & Neumann, 2020).These results link to the findings in bilingual communities outside of ECEC, that I reported on above.While children across contexts thus co-construct dominant language ideologies as part of their own language socialisation, they also demonstrate agency through challenging these ideologies at times.
Evaldsson and Cekaite (2022), for example, show how children in multilingual peer groups in primary school 'talk monolingualism into being' through corrective practices oriented toward the norm of correct Swedish as well as through explicit display of Swedish competences among peers.Children's corrective practices in peer interaction had social functions such as asserting powerful positions within the peer group (Evaldsson & Cekaite, 2022).Another study by the same authors (2019) highlights, however, the importance of paying close attention to situatedness and context of interaction.Reporting on a pre-school attended by children with diverse family languages, Evaldsson and Cekaite conclude that, while the on-site discourse mostly reproduced a dominantly monolingual language ideology that relied on Swedish, teachers included children's multilingual competences in very specific instructional contexts.The children, on the other hand, managed to shape their own spaces for family language use in more spontaneous, ludic contexts, thereby resisting the dominant language ideologies.When doing so, they playfully transformed lexical resources, exploited sounds and used smiles and laughter to co-create multilingual peer language play with linguistic and embodied means.
Taken together, the reviewed literature thus highlights that children's language (ideological) socialisation in multilingual settings is connected to wider power dynamics and indexicalities of language in sociocultural contexts.At the same time, it ties in with local processes in which children agentively use a variety of semiotic resources in creative ways to make meaning in complex settings, interconnected with the dynamic construction of social positionings within the setting.The study at hand further develops this aspect in the context of language-profile ECEC in Germany by zooming in on interactional ideological work as it emerges when children and teachers actively include languages other than German into the kindergarten discourse.

Methods and ethics
Data for this study stems from three months of focused linguistic ethnographic fieldwork in a kindergarten in an urban environment in North-Rhine Westphalia, Germany.62 h of audio recordings and 11.5 h of video recordings were generated and complemented with fieldnotes documenting participant observation including informal engagement with children and teachers.Information about children's family languages was gained both from the teachers as well as during consent negotiation and when parents brought and picked up their children from kindergarten.As a researcher, I took part and conducted observations both in formal and non-formal activities in the day-to-day of the kindergarten.During fieldwork, I constantly moved between learning through playing with the children to taking an observing role to assisting the teachers with easy tasks.The ways in which these shifting positions impacted the data were taken into account in the fieldnotes.
The study is part of a project on linguistic diversity in ECEC in the German-Dutch border region which got ethical clearance from the Ethical Review Committee at Maastricht University.Parents/Legal guardians of all participating children were informed about the study in a personal conversation with me, in addition to a written description of the study prior to giving written consent.I also explained my presence and reasons for note-taking and audio and video documentation to the children, mentioning my intention to write a book 'about what children do, play and say' all day in kindergarten.I often asked the children if I could join them for an activity and respected if they situationally disapproved of my presence or documentation of the activity.

Data analysis
An ethnomethodological approach was adopted in order to investigate how participants locally assign meaning to specific language resources associated with the children's family languages as well as the majoritised language German.The data was analysed in an iterative process of going back and forth between fieldnotes and recordings alongside the generation of new memos.First, any descriptions of instances where teachers or children engaged in multilingual language practices or referred to family languages were identified in the fieldnotes.In most cases, these contained an approximate time stamp of a matching audio or video recording, which allowed to listen to the situations again and transcribe them.If no recording was available, the fieldnotes were used and reworked into ethnographic vignettes.The analysis of fieldnotes and transcripts concerned attitudes that interaction partners express toward the family language in question across turns as well as the way the interaction unfolds across time.Going back to the fieldnotes allowed to understand the interactions in the wider sociocultural context of the kindergarten.

The kindergarten & the 'green' group
The kindergarten, where data for this study was generated, is located in an urban environment in the federal state of North-Rhine-Westphalia in Germany, and is attended by children between three and six.The high degree of linguistic and cultural diversity in its neighbourhood is also reflected in the kindergarten.Around two-third of the children are bilingual or multilingual, and around half of the children exclusively speak other languages than German at home when entering the kindergarten at the age of 3. A wide range of languages is represented.Turkish, Arabic varieties, Russian, English, Kurdish and Polish are the most frequent languages that children speak beside German.The majority of staff is of German background and knows foreign languages that they learnt in school (especially English). 1  The kindergarten in question is a language-profile kindergarten, i.e. at the time of data generation, it was part of a programme of the German government promoting a focus on language. 2As main part of the programme, it was granted one additional part-time staff member whose main responsibility was to create awareness for the domain of language development amongst the staff.In practice, this meant that she provided information about language development and support in the teachers' common room, and assisted in the individual groups from time to time to observe and give feedback, but also to provide examples of how to support language development as part of the day-to-day of the kindergarten group.The additional staff member also organised a wide range of language based-activities for the children.She was trained as a kindergarten teacher and developed an affinity with the topic of multilingualism and language development throughout her professional career.Subsequently, she followed different small-scale trainings.
The dominant language of the kindergarten was German and the teachers saw it as an important task to help the children speak and understand German well in order to prepare them for elementary school.Children also mainly spoke German amongst each other.However, teachers sometimes showed interest in the children's home language, specifically through asking them for translations.In addition, teachers sometimes used English individually with children who speak English at home.In general, an open attitude to multilingualism was also encouraged through the additional staff member for language support.

Findings
In the following, empirical examples from interactions in which multilingualism was at stake in the kindergarten are presented and analysed as to how teachers and children navigate multilingual practices and (re-)produce and engage with language ideologies in interaction.

Co-creating language ideological conflict
As part of social interactions, the children negotiated correctness of names for specific things, undergirded by pertinent language ideologies.In the following sequence, Inga 3  (5;1) comments on an image of a watermelon naming it with its Russian name arbuz, which is followed by negotiations of the validity of the name.Inga was one out of two Russian speakers in the kindergarten group.In the day-to-day of the kindergarten, Russian was usually not very visible/audible and it was only situationally brought up by Inga or the teachers.
Extract 1: This is arbuz 4   Emil and Tom empty tissues from their package into a bowl.One of the packages has drawings of a watermelon.
Inga sits at the table and observes the two boys.The interaction is kicked off by a juxtaposition of two linguistic resources for the same food item: Melone and arbuz.Integrated in his language stream in German, Tom (5;1) verbalises what he sees 'One with melon' (l.1).Inga, then, brings in the Russian name for Melone, drawing on her family language.In so doing, she presents the name arbuz as a relevant form of knowledge in the kindergarten, and Tom opposes her (l.3). 5 After Inga reinforces the validity of her knowledge ('This is arbuz, arbuz', l.4), their peer Emil (3;4) joins Tom's side, contesting 'No, melon!' (l.5).
The emerging conflict leads Inga to give an account for her multilingual language practice based on her own identifications: 'I ne-I not in German, I am not in German.This is arbu-arbuz' (l.6).Here, it becomes clear that the conflict revolves around naming practices on the surface and relies on language ideologies and identifications as a foundation.Clarifying that she does not identify as German, which she links to naming the item arbuz, Inga demonstrates bilingual awareness about the item having two different names in the two different languages.Tom and Emil, however, follow a monolingual logic and keep delegitimizing arbuz (l.8, l.11).Inga next goes on to expand the expression of her bilingual awareness, identifying arbuz as Russian (l.12).When I, who is also present, ask for confirmation again, Inga aligns with me and the content of my question about 'arbuz' (l.14) and Tom misaligns (l.15).Inga reacts by reinforcing the link between her use of arbuz and her own identification: 'I am not this in German' (l.16).
While Tom thus seems to produce strong ideologies of monolingualism, Inga, being as young as five years old, manages to resist the monolingual ideology that devalues her multilingual language practice situationally and, thereby, diversifies the discourse in the kindergarten.She does so not only through her code switching but also through her metalinguistic action of justifying arbuz as one possible legitimate name for the food item.The language ideological assemblage that the children co-create situationally is incoherent, which manifests in conflict.
Simultaneously, Inga uses the Russian name to assert the legitimacy of her identity as a Russian speaker, while her peers orient to a monolingual logic that presents German as the only legitimate standard in the kindergarten.Given Inga's demonstration of bilingual awareness, she does not seem to question the correctness of Tom's name Melone.Yet, she remains in a defensive mode to justify her name arbuz, which gets strongly opposed by Tom and Emil.Furthermore, my own participation in the interaction seems to reinforce the conflict: It leads both Inga and Tom to reconfirm their individual standpoints to me and afterwards go on to oppose one another directly again.Hence, the ideological clash remains unsolved in the interaction.
While in this situation the idea that one thing can have two different names was contested, the bilingual affordances of specific items were exploited as an opportunity to prompt and foreground translation skills on other occasions, as will be discussed in the next section.

Translation as a skill
In educational contexts, teachers might, for example, ask children for translations to their home languages, intending to recognise them as multilinguals.While such occasions can provide children with the opportunity to demonstrate their multilingual skills, it can also have opposite effects, e.g. when children do not know the translation (Akbaba, 2014;Knoll & Becker, this issue) or when children do not want to be singled out as someone with a different home language (Thomauske, 2017).Akbaba (2014), therefore, speaks of ethnic differentiation and highlights students' agentive strategies, including humour, to handle such situations.
In the next extract, teacher Iris asks Inga about a translation to Russian, after she lacks the word for 'mermaid' in German: Extract 2: I am still child Inga asked teacher Iris to draw a mermaid for her to colour.She did not know the word for 'mermaid' and described it as someone without legs.Teacher Iris understood from the context, told Inga the word for mermaid and drew one.After Inga coloured the picture, Iris asks again how the creature would be called: Teacher Iris' first question 'Was ist das?' ('What is this?', l.1), asked in German, departs from the interaction norm of the kindergarten which prioritises German.When child Inga expresses that she does not know (l.2), teacher Iris asks for the word in Russian (l.3), considering that Inga might, as is common in bilingual development, know words in one of her two languages but not in the other.Inga states that she does not know the answer (l.4), so that teacher Iris asks once again for the word in German (l.5).While another child answers (l.6), Inga justifies her unknowingness by foregrounding another aspect of her identity, i.e. that she is still a child (l.7).Thereby, she implicitly highlights that her bilingualism is still developing.Teacher Iris accepts her reasoning and confirms: 'Yes, you don't need to know it' (l.10).While German remains the unspoken norm, teacher Iris constructs Russian skills as a valuable resource by including them in the educational format.In so doing, Iris also classifies child Inga as a competent Russian speaker.How ascriptions by others are handled, e.g.confirmed or resisted, interrelates with self ascriptions and experiences of the self, including the own sense of agency (Akbaba, 2014).Inga actively re-classifies herself, highlighting her identity as a child, when she does not fit the category of a fully proficient Russian speaker.Throughout this process, she and teacher Iris navigate different language ideologies: While a dominant language ideology that favours German is the starting point of the interaction, bilingualism is subsequently recognised by both participants through its inclusion in the educational format, initiated by the teacher Iris.However, Iris initially rather follows a competency ideology here, while Inga's reply emphasises a processual view on bilingualism, which Iris eventually also proceeds to orient to.The situation, therefore, shows the moment-to-moment evolving nature of language assemblages, which get reworked and navigated both by teachers and children in the kindergarten.
While in this situation, a teacher situationally made knowledge of a family language relevant, there were also some situations in which children did so themselves.The following extract is an example thereof.Child Martin's (5;3) parents speak Polish, but he himself was raised as a German-speaker and has rather limited knowledge of Polish.Yet, he identifies with Polish and recognises his knowledge as one of his characteristics.
Extract 3: You don't know how to say cow in Polish!Lunch time conversation.Some of the children invent new names for themselves.Martin tries to convince his peers to call him 'Fuzzi'.
Really, what does it mean?24 Martin Sag ich dir nicht!I won't tell you!
In this situation, Martin valorises Polish, positioning himself as superior due to his knowledge of the language.In the beginning of the extract, he takes the lead in a game in which the peers teasingly invent new names for each other while sitting at the lunch table and eating pears for dessert.Child Martin repeatedly tells his peers to call him 'Fuzzi' (l.1, l.3, l.7), a downgrading name for a person who cannot be taken seriously (see Duden, n.y.).His peers do not follow his rule and either call him Elsa (l.2, l.4) or Martin-kanu, a made-up variation of his name (l.6).After Martin appeals again to his peers to call him Fuzzi (l.5), teacher Eleonore chimes in, first with a question for clarification (l.8) and then with a negative evaluation of the resource 'Fuzzi' as a name (l.10).Next, I, who is present as a researcher, display my curiosity.With the questions 'Where do you know the name from, Martin?' (l.12), 'Do you know a Fuzzi?' (l.14), and 'So what does it mean?' (l.16),I go along with Martin's conceptualisation of Fuzzi as a name.Yet, I disrupt the play just as teacher Eleonore previously did.
The disruptions which come in the form of critique and questioning index Eleonore's and my positionalities as teacher and researcher, which are, from an institutional perspective, more powerful statuses in the kindergarten than those of the children.As a reaction, Martin first takes a ludic stance when giving a wrong explanation of the meaning of Fuzzi ('Fuzzi means Hello', l. 17), followed by 'FOOLED!' (l.19).He next refers to his family language when telling me: 'You don't know how to say cow in Polish!' (l.20).Yet completely unrelated to 'Fuzzi', the reference to Polish enables Martin to situationally position himself as a more knowledgeable subject than me.Martin confirms that he himself knows how to say cow in Polish (l.22) and, upon my question for the translation (l.23), states that he does not tell me (l.24).
In this situation, Martin uses his knowledge of Polish, more specifically, his ability to perform specific translations, to distinguish himself from me.Resorting to his family language provides him with the opportunity to empower himself by foregrounding specific multilingual knowledge.He highlights that he is the only one to have this specific multilingual knowledge at the table and expresses this during language play in German, which underlines his bilingual competence.
In this situation, Martin leveraged the particular linguistic repertoires of the individuals involved in the interaction for his own interactional aims and, in so doing, brought forward an ideology that recognises and appreciates multilingual skills.While knowledge pertinent to Polish gained meaning as a unique feature of him, this was not the case with children whose family language included English.Given that Martin was more competent in Polish than the teachers or me, Polish had specific language ideological affordances of uniqueness that English did not have, since the teachers (as well as I a researcher) occasionally used English in the kindergarten.The next section will focus on language (-ideological) practices which involved the use of English by teachers and/or children.

Negotiating the legitimacy of English
The kindergarten group hosts two children whose family language is English.While Grace (4;10) exclusively uses German in the kindergarten, Amy (3;2) does not actively speak German yet and primarily addresses teachers and peers in English.While Amy's peers answer her in German, the teachers sometimes switch to English when interacting with her.According to them, they do so especially when conveying something that they find important for Amy to understand.Teacher Mareen indicated that she often automatically switches to English but that she wants to remind herself to speak more German to Amy so that Amy has better chances to learn the majority language.
The following example shows how the teachers commonly switch between the two languages with Amy: Huh?So that you don't can see me 7 Amy ((takes hand up in front of her face again)) 8 Iris (teacher) Hey! Amy I don't like that!I really don't like that.9 Amy ((puts down her hand and holds up her drawing up, looks at Iris)) 10 Iris (teacher) Hast du was gemalt?Did you draw something?11 Amy ((nods head)) Teacher Iris first addresses Amy in German, asking her what she said (l.2).Amy reacts in English ('No no no', l.3) and teacher Iris repeats Amy's English words with a questioning intonation (l.4).When Amy shows a behaviour that the teacher disapproves of, i.e. holding her hand in front of her face to block a (l.5),Iris mirrors her behaviour (l.6) and explains to her in English 'Hey!Amy I don't like that!I really don't like that' (l.8).When Amy puts down her hand again and shows her painting to Iris, the teacher switches to German again, asking her if she drew something (l.10).Amy reacts by nodding her head (l.11).
The switch from German to English occurs exactly for the part with educational purposes where the teacher disapproves of Amy's behaviour.This educational strategy fits in with the teachers' reasoning that they use English when finding it especially important that Amy understands.In practice, however, teacher Iris uses bodily touch, gestures and facial expression to express her disapproval as well, rendering the switch to English not necessary for ensuring understanding.When talking about Amy's painting (l.9-11), it furthermore becomes clear that Iris and Amy also manage to engage in meaningful conversation through the use of German in combination with other material resources (the painting), spatial resources (holding the painting, l.9) and embodied resources (gaze, l.9, nodding, l.11).Consequently, the switch in the interaction rather serves the aim of enforcing the disapproval, as using English makes the disapproval sequence distinct from the rest of the interaction.
Simultaneously, the teachers invoke a higher value of English in comparison to other family languages not only through using it with Amy, but also especially through deploying it in relevant instructional situations.Morillo Morales and Cornips (2022) have shown that teachers' language choice across educational and more playful situations is linked to children's understanding of the social meaning of a given language.As can be seen in the picture in l. 6, also Amy's peer Fiene (3;6) looks at teacher Iris when she speaks in English.Thereby, she takes the position of an overhearer in the embodied participation framework (Blum-Kulka & Snow, 2002;Goodwin, 2003).In a multi-party conversation, children may take on the role of a bystander who overhears talk not directed towards them (Blum-Kulka & Snow, 2002).Peers who overhear the teachers' and Amy's use of English also take note of how English is legitimised as a valuable language resource in the kindergarten.Overhearing is an important mode to socialise children into indexicalization processes of societal meanings of different language varieties (Rickert, 2023).
Yet, as active agents of language ideological processes, children also negotiate the meanings of English in relation to German, as becomes visible in the next extract:

Extract 5:
Fieldnotes 1/12/2021 Preparations for Christmas during circle time.Amy opens the advent calendar and pulls out a paper with a story.Teacher Mareen repeatedly tells her in German that she should not pull out the paper.Child Nura observes the teachers attempt and tells her: 'Die spricht nicht' ('She doesn't speak').Teacher Mareen: 'Die spricht Englisch'.(She speaks English).Teacher Mareen repeats once more, in German, that Amy should not take out the paper.Child Elise intervenes in English: 'No Amy, no' and her peer Tom joins her, also in English: 'No, please, no!'.Teacher Mareen and her colleague Iris look at each other and smile.
In this situation, teacher Mareen speaks to Amy in German in front of the whole group.Amy's peers intervene as the situation unfolds: When Amy does not react in the way that Mareen intended (i.e.stops to take the paper), peer Nura (5;1), orienting to German as the language norm in the kindergarten, gives the account that Amy does not speak.Following a monolingual ideology, Nura portrays Amy as 'languageless' (Rosa, 2016).In so doing, she also positions Amy as 'communicatively incapacitated (…) [and] "out of place"' (Blommaert et al., 2005, p. 198).Teacher Mareen counters by explaining that Amy speaks English and, thereby, positions Amy as a speaker.However, she proceeds to present German as locally and situationally more important by continuing to address Amy in German.Thereby, while recognising Amy's English language competences on a more general level, she clearly sets interaction norms that require a passive understanding of German.Here, a language ideological assemblage emerges, which links an educational ideology focused on instructed participation in circle time activities with a language ideology that positions German as the required language for such participation.However, within this assemblage, it is also acknowledged that Amy speaks another language that may be relevant in other contexts.
Amy's peers Elise (4;5) and Tom react to this by engaging in what de León (2019) calls 'playing at being bilingual' and, thereby, rendering English locally and situationally relevant.Elise and Tom only know a few words of English but use these purposefully in their communication with Amy here.By addressing Amy in English ('No Amy, no'; 'No, please, no') when she does not react to Mareen's German, the peers Elise and Tom do not only display their own multilingual competence but also, in contrast to Nura earlier, orient to Amy as a competent interaction partner.
In the extracts presented in this section, it became apparent how teachers and children co-create language ideological assemblages in which English was ultimately legitimised as a valuable resource, subject to situational contingencies in the ECEC setting.As such, English had different functions such as ensuring understanding or adding emphasis.Yet, English was always positioned in relation to German, which remains the dominant language in the kindergarten.

Discussion and conclusion
Just like in many kindergarten groups in Germany, the 'green group' where this research has been conducted, is a social space in which children with diverse linguistic backgrounds come together.The main interest of this paper was the question how children and teachers navigate multilingual language practices and engage with language ideologies in the context of a linguistically diverse kindergarten group in Germany.
It became clear that children actively bring in languages other than German into the kindergarten discourse in many forms and on different occasions, e.g. when drawing on their family languages in interactions with peers (like child Inga in 5.1), or through 'playing at being bilingual' (de León, 2019) like child Elise and child Tom when they meaningfully mobilise their English resources (5.3).
Also teachers bring in languages other than German, mainly in the scope of specific educational interactions.More specifically, this is usually preceded by a child's restricted language capabilities in German.For example, teacher Iris gives emphasis to her disapproval of child Amy's behaviour by switching to her family language English, the language that Amy also actively uses (section 5.3).On another occasion, she inquires about a word in Russian, when child Inga does not know it in German (section 5.2).In conclusion, the teachers' use of, or questions about, languages other than German brings these languages into play as meaningful resources when German does not suffice , so that these languages always get positioned in relation to German.
A domain-specific use of home languages within early educational contexts has been discussed in different lights in the literature.While Cekaite and Evaldsson (2017) concluded that in their study in Sweden, family languages were also mainly valued by teachers in instructional contexts, Morillo Morales and Cornips (2022) found different dynamics in Limburg in the Netherlands.In pre-schools there, the dominant language variety Dutch was used in instructional contexts in contrast to the regional minority language Limburgish in interactions geared towards care, the difference being that these kindergartens were bidialectal.Taken together, a domain-specific acknowledgement of a certain language promotes language ideologies which limit the legitimacy of a specific language to a specific context (e.g.towards educational ends, or towards care ends) and, as such, pertinent ideologies unfold in parallel with educational ideologies.
In the kindergarten in question, such simultaneously unfolding ideologies took shape in the formation of language ideological assemblages, and teachers and children expressed and negotiated various, at times opposing beliefs about language use.These negotiations were complex and dynamic, and the meaning of using the children's home languages was constantly redefined by both teachers and children on a moment-to-moment basis.At the same time, German rather had a normative position, being the stable dominant language in the kindergarten.These findings align with the findings of Zettl ( 2019), who has concluded that language ideologies in German kindergartens move on a continuum.
Consequently, multilingual children were confronted with varying attitudes from interaction partners.Interrelated with this, they had to position themselves on the spot, which they did through reworking the language ideological assemblages in different ways, foregrounding different aspects of their own linguistic identity and taking up or rejecting sociolinguistic identities that were assigned to them by other children and teachers.The children got to experience that the language ideologies their current interaction partners evoke are not necessarily aligned with other present ideologies in the kindergarten and that they are subject to negotiation.For example, child Inga experienced a resistance to her multilingual language practices from peers who positioned a Russian word as illegitimate and presented only the German form as valid, which she, in turn contested (section 5.1).On yet other occasions, Russian was situationally presented as relevant by the teacher in the kindergarten context if a German word was not known (section 5.2).
All in all, the findings highlight that teachers should be aware of the potential tensions that might arise for children as they need to navigate emerging language ideological assemblages in the kindergarten.To foster a language-inclusive space, which ultimately benefits the children's wellbeing and language development (De Houwer, 2015), teachers could aim to increase the awareness about multilingualism within kindergarten groups.This could entail explaining that different children have different languages and that things have different words/names in different languages.All children, whether monolingual or multilingual, can benefit from such a basic understanding, as it can not only lead to a more constructive peer environment, but also opens up new venues for language exploration and language play, like the peers Elise and Tom as well as Martin demonstrated.
up and takes down Amy's hand)) If you ask me something, I make this ((holds hand in front of her face))