Language education policies for young learners in Europe

Abstract:Language education policies are policies that inform the ways in which languages should be taught, learned, used and assessed in educational institutions. They are usually created with a view to support the development or the maintenance of particular languages, as it is often believed that schools play an important role in shaping language practices in society at large. This paper aims to give an overview of the major types of language education policies for young learners currently available in Europe, including "foreign" language education, bilingual and multilingual education. It also aims to highlight how these language education policies are underpinned by various ideologies and understandings of "language" and "bilingualism". We argue that Language Education Policies (LEPs) need to be conceptualised as being multi-layered and dynamic processes, which need to involve all educational actors for them to lead to students' success and sustained bi/multilingual development. Finally, this paper discusses two examples of multilingual language education policies for young learners in Europe, one in Luxembourg for pre-primary education and one in Ireland for primary education, with a view to highlight possible trajectories towards the creation and implementation of multilingual language education policies. It is hoped that this paper will inform language education policy makers, researchers and actors on current developments and challenges related to language education policy in Europe.Resumé:Les politiques linguistiques éducatives concernent les choix de différentes modalités d'enseignement, d'apprentissage et d'évaluation des langues dans les institutions éducatives. En général elles sont élaborées afin de soutenir ou de maintenir le développement de langues spécifiques, puisqu'il est admis que l'école joue un rôle central sur les pratiques linguistiques à l'oeuvre dans la société. Dans ce chapitre nous visons à donner une vue d'ensemble des principaux types de politiques linguistiques éducatives élaborées ces dernières années en Europe pour les jeunes apprenants, tel que l'enseignement des langues dites étrangères et l'enseignement bilingue et multilingue. Nous montrons également comment ces politiques linguistiques sont empruntés de différentes idéologies et conceptualisations du langage et du bilinguisme. Nous défendons l'idée que les politiques linguistiques éducatives doivent être comprises comme des processus dynamiques fonctionnant à de multiples niveaux, et impliquant la participation de nombreux acteurs si leurs objectifs sont un apprentissage réussi et un développement durable du bilinguisme et du multilinguisme chez les élèves. Dans une dernière partie, l'article analyse deux exemples de politiques linguistiques en Europe, l'une pour la petite enfance au Luxembourg et l'autre pour l'école primaire en Irlande, afin d'illustrer comment créer et mettre en oeuvre des politiques linguistiques éducatives visant une éducation multilingue pour tous. Nous espérons que cet article trouvera un écho auprès d'auteurs de politiques linguistiques éducatives, de chercheurs, d'enseignants et de différents acteurs concernés par les développements et les défis liés aux politiques linguistiques éducatives en Europe.


Introduction
Language Education Policies (LEPs) are policies that inform the ways in which languages should be taught, learned, used and assessed in educational institutions. Globalisation and the increased transnational mobility of people in Europe and in the world are contextual factors that have had a major impact on the way language policy and language in education policy are being researched, conceptualised and implemented. This means that LEP is not a neutral phenomenon in terms of social intents and consequences. It has a political dimension of ideological control (e.g. Tollefson 1991) that requires critical scrutiny.
In this paper we adopt Shohamy's (2006: 76) definition of LEP as: a mechanism used to create de facto language practices in educational institutions, especially in centralized educational systems. LEP is considered a form of imposition and manipulation of language policy as it is used by those in authority to turn ideology into practice through formal education. Yet, at times, LEP is also used as a bottom-up, grassroots mechanism to negotiate, demand and introduce alternative language policies. Thus LEP is another mechanism through which ideology is meant to turn into practice or practice into ideology.
While LEPs are not necessarily at the forefront of teachers' awareness, teachers are on the front line of LEPs because classrooms are key sites where these policies are enacted. Through the decisions teachers and learners take every day in the classroom, they implement, shape, interpret and negotiate LEPs (Menken and García 2010). Thus, LEP is highly relevant for educators, parents and students, from early childhood education care (ECEC) settings to higher education.
In this paper, we recognise the diversity of practice across Europe and provide a synthesis of the major types of LEPs for early years and primary education currently available in Europe. We acknowledge that the linguistic diversity and historical contexts of North, Central and South America, Africa, Asia and the Pacific are different and beyond the scope of this paper. We start by defining the notion of language education policies and discuss in particular issues related to the labelling of "languages" in education policies. We then turn to the Council of Europe's model of language education policy and present two recent LEP developments, namely foreign language education in early years and bilingual education. Finally, we argue for the need to move towards multilingual language education policies and present two examples of multilingual LEPs, one implemented at national level in Luxembourg and the other in a primary school in Ireland.

Defining language education policies in Europe: typologies, key concepts and ideologies
LEPs are first and foremost about language choice in the educational sphere.
They usually set what language(s) should be used as mediums of instruction or taught as subjects, the age at which "foreign/second" languages should be taught, the contents and objectives of the curriculum, and what language(s) should be used in tests and assessments. They tend to serve national languages, to promote the support of dominant languages and to regulate the use of, or silence, minoritised languages (home, regional, heritage or indigenous languages). They also involve decisions as to how much time is allotted to the use or the teaching of a particular language and whether a language is optional or not. LEPs also regulate language practices; they indicate what language(s) or language variety is "correct", which one is appropriate or not in a given educational context, and when they should be used and for what purposes. They are complex and dynamic processes, involving multiple layers and policy agents. Ricento and Hornberger (1996) have compared the multi-layers of language (education) policy to the layers of an onion (see also Hornberger and Johnson 2007). To acknowledge the stratification of language policy, Bonacina-Pugh (2012; 2020; see also Bonacina-Pugh, Barakos and Chen, 2020) builds on Ball's (1993) conceptualisation of policy as "text" and "discourse" and proposes to conceptualise language policy as "text", "discourse" and "practice". In educational contexts, language policy as "text" refers for instance to statements about what language(s) should be used as a medium of instruction or in assessment. Language education policy conceptualised as "discourse" refers to the beliefs and ideologies held by any educational actors, such as a ministry of education, a head of school, a teacher or a pupil that might influence language choice practices and decisions about what language(s) should be used in a particular educational context for teaching, learning and assessment. Language education policy as "practice" refers to "what is usually done" (Spolsky 2004), that is, a set of implicit norms of language choice (Bonacina-Pugh 2012). It is what Bonacina-Pugh (2012; 2020) calls the "practiced" language policy. These implicit norms of language choice are seen as policy because they influence future language choice acts. In other words, when a speaker is deciding what language(s) they will use in a particular context, they might draw on their implicit knowledge of what is "usually done" rather than on their explicit knowledge of a declared policy. For instance, in her study of a classroom for newly arrived migrant pupils in the French educational system, Bonacina-Pugh (2020) argues that the "practiced" language policy in that classroom is multilingual, that is, that the use of multiple languages is seen as legitimate and appropriate. Therefore, when deciding what language(s) to use and when, participants in that classroom refer predominantly to the "practiced language policy" rather than to the language policy at the level of texts, which remains in the French educational system largely monolingual in French (Ministère de l'Education Nationale 2012).
Another important aspect of language education policies is the notion of "language" or rather the naming of the languages targeted by these policies. Traditionally, language education has been conceptualised in a non-integrated fashion, with on one side the national language being taught as a subject and medium of instruction for other school subjects, and additional languages being taught alongside other academic disciplines during a limited number of hours per week. Thus, these "additional" languages have been referred to as "foreign" in most European countries, or "modern" (as opposed to classical languages, in the UK for example), or as "second" (in North America), all these terms illustrating the unequal relationship between these languages and the national one. Lately, the meanings of such terms have been questioned and new terms such as "world" languages have been proposed to reflect the fact that in our globalised world a language such as English is no longer "foreign" in the strict sense of the term to most learners; neither is it a "second" language in the case of a growing number of families who choose to speak English to their children from birth, in countries where English is not the national language (Bouches 2017;Curdt Christiansen 2018). Other terms qualifying languages in education are also used by policy makers and researchers, reflecting political categorisations that give more or less status and recognition to these languages in curricula. These terms further reflect the different power of their speakers, often invisibilising their bi/plurilingualism, as for example with the terms "heritage", "migrant", "of origin", "regional", "oriental", "exogenous", etc.
In order to give policy makers and educators some support to conceptualise LEPs in Europe, the Council of Europe has been prominent in proposing, as early as the 1970s, an impressive body of reports, documents and frameworks to develop more efficient and integrated approaches to the teaching of languages (Council of Europe 2007;. The Council of Europe's projects are based on the two key notions of plurilingual and intercultural education.1 For instance, a recent Council of Europe's recommendation (2022) clearly states the importance of plurilingual and intercultural education. It invites member states to see intercultural education as a major strand of LEPs and gives nineteen policy measures to be considered by those responsible for 1. A terminological distinction must be explained: for the Council of Europe (based in Strasbourg, forty-seven member states), multilingualism is a societal phenomenon whereas plurilingualism is an individual characteristic and both multilingualism and plurilingualism are central features of a European identity based on democratic citizenship. The other European institution dealing with LEPs in Europe is the European Commission of the European Union (twenty-seven member states); Multilingualism is the only term used by the European commission and it includes both perspectives. national, regional and institutional policy in all educational sectors with a view to further develop and implement a plurilingual and intercultural education across the curriculum (Council of Europe 2022: 7). Th e Council of Europe's projects consider that the central feature of language education is the learner and their plurilingual repertoire of linguistic and cultural resources (family languages, heritage languages, foreign languages, etc.), represented by the following model: Th e model is based on the notion of "repertoire", which allows for the recognition of the primary identity of the learner as a basis for further language learning, including fi rst of all the language of schooling. Th e language of schooling includes two dimensions: the language as subject (with its own components -literacy, literature, language awareness) and the language of other school subjects as the main means of acquisition of content knowledge. Competence in the language of schooling is considered a priority for academic achievement and learners' future professional development. Languages other than the language(s) of schooling can include all the denominations we mentioned above, depending on the LEP of the country concerned. This said, "foreign" languages are now legitimate school subjects integrated in the regular curriculum of most countries in Europe, whereas regional, migrant and heritage languages often remain optional and marginalised. The term "foreign" is still widely used in European policy document and statistics. The Eurostat website for example gives the following information: in 2020, almost three fifths of pupils in the EU learnt two or more "foreign" languages in upper secondary education and 94 % of pupils learnt English (Eurostat 2022).
However, learning "foreign" languages as subjects has long been considered as not very efficient because of the limited amount of exposure to the language in the non-authentic context of the classroom. In order to address this time constraint, two main LEPs were developed: 1) early language teaching, that is introducing "foreign" languages at primary level and even at pre-primary level, and 2) offering bilingual education under the various models we will explain below.

The two main LEPs in Europe: early foreign language education and bilingual education.
The European Commission has been strongly supporting early language learning for the past thirty years, based on research showing it would result in more efficient language learning, and improved mother tongue competence. In 2011, a European Commission and British Council funded project led by Janet Enever gave a detailed insight of the policy and implementation processes for early foreign language learning programmes in seven countries in Europe (Croatia, England, Italy, the Netherlands, Poland, Spain and Sweden), offering different descriptions of learner experiences and contexts for learning.2 Indeed, today, foreign languages are taught at primary level in all European countries, although the diversity of languages on offer is rather limited, English being the dominant choice in all countries where it is not the official language (Enever 2012). The Eurostat website shows that a number of EU member states have close to 100% of pupils learning English in primary education and that all or nearly all (99-100 %) primary school pupils in Malta, Cyprus, Austria, Spain and Italy learnt English as a foreign language in 2016, as was also the case in Liechtenstein, Norway and Macedonia. In addition, more than nine out of every ten primary school children learnt English in Latvia, Poland, France and Croatia.
The European Commission also supports language learning at pre-primary level and published a policy handbook in 2011, entitled 'Language Learning at Pre-Primary School Level: Making It Efficient And Sustainable' (EC 2011). The guidelines and recommendations reflect examples of good practice and academic evidence on how to ensure the quality, effectiveness and sustainability of language learning in pre-primary settings, i.e. in any kind of setting where learning takes place before primary school (e.g. nursery, kindergarten, etc.). Whether early or very early language teaching has improved the competence of learners is an issue for debate, one of the main criticisms residing in the same pedagogical approach being used over ten years and thus decreasing the motivation of learners (e.g. Bolster et al. 2004;Graham et al. 2016).
This brings us to the second LEP that is strongly supported by European institutions, i.e. bilingual education (BE). Included in the model above, one should note the use of the plural to refer to the language(s) of schooling. This means that children can be educated with more than one language of schooling, thus referring to bilingual and plurilingual education, whereby all or some of the school subjects are instructed through two languages (or more) instead of only one, the national language. In BE programmes, the teaching time, the curriculum and the classroom need to make space for two languages. In some cases, teachers are bilingual and teach through both languages and, in other cases, each language is taught by a separate teacher. One of the key characteristics of BE lies in the fact that "foreign" languages are not learned as subjects but are used as mediums of instruction to teach other subjects (Baker and Wright 2017).
Instruction through two languages has taken so many forms that such a definition falls short of the complexity involved in describing and analysing the many different types of BE. More comprehensively, García (2009: 6) defines BE "as a way of providing meaningful and equitable education as well as an education that builds tolerance towards other linguistic and cultural groups". She adds that BE for minoritised children often questions the hegemony of the dominant language, that it challenges national language borders and fashions multiple identities. Therefore, the general goals of BE are far broader than those of second language teaching: "BE focuses not only on the acquisition of additional languages but on helping students to become global and responsible citizens as they learn to function across cultures and worlds, that is beyond the cultural borders in which traditional schooling often operates" (García 2009: 6).
BE is a highly ideologised and politicised domain of research and practice where different conceptualisations of language and bilingualism have played a major part in the way programmes have been implemented in schools in different parts of the world. BE serves many students at different levels in different languages with different societal goals. In some contexts, BE reinforces existing social hierarchies; in others, it serves the educational need of minoritised speakers; yet in others it helps to revitalise community languages. Today, bilingual education including English is even offered as early as three months in private day-care institutions in many European countries, attesting to the growing commodification of BE when it includes the English language (Caporal Ebersold 2018).
Hélot and García (2019) analyse BE according to the status of the languages involved, equal in the case of French and English in Canada for example, or unequal with Spanish and English in the US, because Spanish speakers have been minoritised and expected to assimilate. BE has also been known to give a new legitimacy to some languages. This is the case, for instance, of indigenous languages in former colonial contexts (e.g. New Zealand, Hawai'i), "regional" languages in Europe (e.g. Irish, Welsh, Basque, etc), and sign language (especially as it has been a long struggle for deaf communities to access bi-modal BE; Swanwick 2017). BE with "migrant" languages, however, as explained by Sierens and Van Avermaet (2017: 489) for Europe, "has nowhere been able to establish itself as a fully valued teaching model", the main reason being that, in order to integrate, "migrant" populations are expected to acquire the official language. On the other hand, BE in dominant European languages is very strongly supported by European institutions that stress the economic advantages of bi-and multilingualism, and the value of linguistic and cultural diversity for European identity. Again, similarly to the dominance of English in "foreign" language learning, BE throughout the world has grown as a direct result of the way in which English is being constructed as essential for participation in the global economy (Martin Rojo 2018).3 As a result, BE programmes in minoritised languages can be at risk because of the demand for English as social and cultural capital. That said, there are examples where English has been successfully integrated into BE models alongside minority languages, as in the Basque Country in Spain for example (Cenoz 2015; see also Lasagabaster and Huguet 2007, who analysed multilingualism in nine different contexts of BE, where minority, majority and foreign languages were integrated).
There are many different models, types and frameworks of BE which have been extensively described in the scientific literature (e.g. Baker and Wright 2017;Hélot and Erfurt, 2016 for France, etc.). Various typologies (e.g. Genesee et al. 2006) have been drawn up according to the goals of the programmes, for example maintenance, transitional or enrichment aims; some are mainstream models, others are not; some serve minoritised students and others majority students. Multiple terms are used such as "immersion" (most well-known because of the success of the Canadian BE programmes) and "two-way dual language BE" implemented in the US where speakers of English are educated with learners of English through two languages (e.g. Fialais 2019).
In Europe, since the 1990s (Nicula 2017), a widely implemented type of BE is known under the denomination Content and Language Integrated Learning (CLIL). The main goal of CLIL is concerned with devising more efficient teaching approaches of additional languages in mainstream education (Dalton-Puffer 2011), for the most part dominant European languages and, again, English is the most widely language on offer. CLIL means that one additional language is used as a medium of instruction to teach subject content such as history, geography or mathematics and it involves a specific pedagogy to teach both language and content simultaneously. In some cases such as in France, one subject, or at the most two subjects, are taught through the medium of the additional language, rarely more, making CLIL a less ambitious bilingual approach than immersion programmes (Hélot and Cavalli 2017). In Spain, the CLIL model is very widely used to replace traditional second language teaching and can involve almost half of the teaching time in a language other than Spanish, and in most cases that language is English (Lasagabaster and Ruiz de Zarobe 2010). The conceptualisation of CLIL by the researchers who first devised this model (Mehisto et al. 2008;Coyle et al. 2010) drew attention to the didactic potential to rethink second language acquisition through its integration into content teaching. Indeed, learners are engaged in a different experience of learning and becoming pluriliterate, that is, literate in the ways in which knowledge and meaning are constructed and articulated in various subject disciplines, across languages and cultures (e.g. Meyer and Coyle 2017). However, some criticisms have been raised regarding the CLIL model, mainly because it focuses mostly on dominant languages rather than considering the many linguistic profiles of learners. For instance, CLIL programmes often assume that all learners are monolingual, and thus do not address the specific needs of many minoritised language speakers who are already bi-or multlingual. In other words, as argued by Hélot and Cavalli (2017), CLIL does not always challenge the monoglossic ideology of mainstream schooling. For example, a recent thesis carried out in the Basque Autonomous Country (Villabona Perurena 2020) analyses the practices of four teachers responsible for CLIL English classes in schools implementing the D model of BE in Basque.4 The classroom observations show that content teachers do not conceptualise the CLIL approach as a bilingual (or multilingual) way of acquiring content knowledge but merely as a means for the students to learn English.

Towards multilingual LEP for social justice in Europe
Because there are so many understandings of what constitutes bilingual education, some researchers have moved away from using typologies of BE models to focus instead on exploring the ideologies and conceptualisations of language(s) underpinning BE. From the point of view of LEP, researchers now make a distinction between monoglossic and heteroglossic ideologies of bilingualism. A monoglossic view of bilingualism sees monolingualism as the norm and adopts a dichotomous interpretation of languages as being either an L1 or a L2. It understands bilingualism as being simply the addition of two languages, which leads to a pedagogy where these two languages remain separated at all times.
This monoglossic conceptualisation of bilingualism has been challenged by the necessity to acknowledge that our classrooms have always been multilingual, whether the students are allowed to use their multiple languages or not. In today's global era, mobility has shaped the way we study communication (e.g. Blommaert and Rampton 2011). A new focus has emerged, which consists in understanding the complexity and fluidity of multilingual speakers' practices rather than describing the functions of each language in use.
In fact, following what some researchers (May 2014;Conteh and Meier 2014) consider as the "multilingual turn" in applied linguistics, recent research draws on a reconceptualisation of languages as no longer bounded autonomous systems and describes the fluid practices of bi/plurilingual speakers' as 4. There are three modes of bilingual education in the autonomous Basque Country (Spain). Model A: almost all teaching is completed in Spanish; Basque is taught as a subject. Model B: teaching is completed half in Spanish and half in Basque; both languages are thus medium as well as subject. And Model D: almost all teaching is in Basque. There is no letter C in Basque.
languaging (a term first coined by Mignolo 1996 then later adopted by many sociolinguists such as Blackledge and Creese 2010 etc.).
Research perspectives on bilingual and multilingual education today attempt to move away from a monoglossic ideology that insisted on the separation of languages to adopt instead a heteroglossic perspective to bilingualism. This has given rise to a reimagining of bilingual pedagogy no longer centred strictly on languages but on the practices of bilinguals, on their languaging as normal modes of communication, and on using the full extent of their plurilingual repertoires to learn. At the heart of this heteroglossic approach to bilingual pedagogy is the concept of translanguaging (e.g. García and Wei 2014; for a review of translanguaging in education see Bonacina-Pugh et al. 2021). Translanguaging means that bilinguals deploy semiotic features that would traditionally be seen as belonging to different named "languages" but that form, in practice, one integrated system. Translanguaging as pedagogy is often adopted in "practiced" language policies (e.g. Bonacina-Pugh 2020) in bilingual educational contexts, but has so far rarely been acknowledged and legitimised in language policy texts and discourses. As a consequence, many bilingual students are seen to learn today through translanguaging, but are still assessed monolingually. As explained by many researchers (García 2009;Cummins 2000;Genesee et al. 2006;Grosjean 2022) the bilingual child's language proficiency is not like that of a monolingual speaker, as both languages integrate to some extent. Therefore assessment procedures should not evaluate competence in two individual languages but should look at skills across both languages. However, outside of the domain of special education (Rose et al. 2022) there is little empirical evidence of new approaches to assessment in bilingual programmes. Whilst the notion of language separation continues to be relevant in some contexts such as in translation and interpreting, the notion of translanguaging has gained ground in the education of bilinguals and multilinguals and has opened the door to multilingual education.
In terms of LEP, multilingual education is mainly conceived as a more flexible and dynamic approach to language education because it takes into account all the learners' languages in a classroom and integrates them in the teaching approach. It is socially transformative because it also challenges raciolinguistic ideologies (Flores and Rosa 2015) and values all languages as equally helpful for teaching and learning, regardless of their status in the wider society. In other words, the aim of multilingual education, as it is formulated today, is to achieve greater equality and equity, or social justice (Piller 2017;Skutnabb-Kangas et al. 2009).
In the twentieth century in Europe (as well as in North America), BE served the cultural interests of the elites who wanted their children to acquire more linguistic capital, or as an approach to compensate the deficit perspective on the competence of minoritised speakers, whereas in the twenty-first century, BE is being promoted throughout the world in order to find more efficient approaches to language learning and to prepare students for the needs of a global world (e.g. Martin Rojo 2018). Extensive research was also carried out in the twentieth century on multilingual education and addressing the need of minority communities by researchers like Skutnabb-Kangas (1984) and Cummins (1996), among others. Following on their research, the central objective of multilingual education today is social justice (e.g. Arshad et al. 2012;Conteh 2017;Piller 2017). In the words of Piller (2017: 7) for example, "language is used as a means of exclusion, discrimination and disadvantage, therefore linguistic diversity intersects with social justice". Consequently, it behoves all teachers, including "foreign" language and bilingual educators to rethink their approach to language education in ethical terms. How can this be achieved? Through the implementation of strong forms of multilingual education offered to all students in all classrooms.

Empirical examples of multilingual language education policies
We now review two examples of multilingual language education policies in Europe. The first is an example of a new policy for multilingual education in the early childhood education sector in Luxembourg, and the second is an example of a multilingual language education policy in a primary school in Ireland.
In Luxembourg, a new LEP has recently been implemented in the ECEC sector in order to rethink the existing model of multilingual education in schools. Luxembourg is a small highly multilingual European state counting 645,397 inhabitants (STATISTIQUE.lu 2023), of whom 47% are foreigners. The language law of 1984 declared three official languages, namely Luxembourgish, French and German (English is widely spoken too; STATISTIQUE.lu 2023). A multilingual approach was implemented with these three languages each used successively as languages of instruction. In Luxembourg, 52.7% of children attending schools are multilingual, out of which 28% are Portuguese speakers (MENJE and SNJ 2017). Horner and Weber (2008) have analysed extensively the challenge for Portuguese speakers to be educated through Luxembourgish first, then German and finally French. Gómez-Fernández (2014) in his doctoral thesis explained the loss of linguistic capital of a young first grader Brazilian child who had to learn through Luxembourgish and German on his arrival in Luxembourg with no support in his home language. The law of 27 August 2017 changed the policy towards multilingualism for children aged one to five in order to improve the pedagogical quality of ECEC ('Loi du 29 août 2017'). The new policy targets the development of young children's plurilingual competence, the acknowledgement and promotion of all family languages and the development of linguistic and intercultural awareness. Multilingual education is also described in the new LEP as a means to develop intercultural competence, to support self-esteem and to combat ethnocentrism. Then the 2018 national framework for non-formal education insists that multilingualism be considered as an important societal resource and be conceptualised in terms of experience, meaning it should be both lived and learnt starting in ECEC settings. The focus of the policy is therefore on two of the societal and school languages (French and Luxembourgish) and on the inclusion of the home languages of the children. In other words, the previously common practiced policy of discouraging children from using their family language in ECEC settings (Seele 2016;Neuman 2015) is being reversed. The aim of the new LEP is that children should familiarise themselves with the languages (in the plural) they will speak in the future. It is about implementing multilingualism in the first socialising settings children encounter, to help them to be better prepared for their formal education, and to give more prominence to multilingualism as a societal resource in Luxembourg. From the point of view of research on language policy, this new plurilingual framework for ECEC education in Luxembourg is an interesting example of a LEP that has transformed a monoglossic regime into a truly multilingual one and made educators as well as society at large aware that linguistic inequality and discrimination must be challenged from the earliest age. In other words that all children should be empowered to express themselves using all their plurilingual resources and should never again be made to feel linguistically insecure or ashamed of their language practices (Weber 2014). At present, the policy has not yet been evaluated but it is monitored by regional agents who visit the ECEC settings, and a thirty-hour professional development course is being offered free to all ECEC personnel responsible for pedagogical development within their settings.5 Our second example (Little and Kirwan 2019) illustrates how a multilingual LEP can lead to pupils' success if it includes all school subjects, all teachers 5. See the details of the policy and how it is implemented at "Éducation plurilingue", enfance jeaunesse.lu, https://www.enfancejeunesse.lu/fr/education-plurilingue#items=6. and all parents. In their book Engaging with Linguistic Diversity (2019), Little and Kirwan analyse how a primary school in Dublin (Ireland) with 70% of students of migrant background speaking fifty different languages managed to achieve well in the national evaluation and some even above average. The authors explain that the policy consisted first and foremost in giving equal status in the school to all the languages spoken by the students. It also included the parents who were strongly encouraged to continue speaking their languages at home with their children. Once the policy was decided and explained to the staff, a multilingual pedagogy was developed insisting on the integration of the students' plurilingual repertoires both in oracy and literacy activities, not only concomitantly with English but also with Irish and French, two other languages included in the curriculum. As the title of Little and Kirwan's book shows, implementing a LEP that transforms the language regime of a school and allows students to use their plurilingual repertoires to access school knowledge as well as the dominant language of instruction (English in this case) and two other languages (Irish and French) implies a level of engagement from educators who must be prepared to challenge their representations of language and multilingualism, to accept to work with a multiplicity of languages they do not know, and to put children's needs first (see also Weber 2014). What can be learned from this fascinating ethnographic piece of research is that the integration of a clear LEP of multilingualism at the start of the project, with thorough teacher education on multilingual pedagogy for all teachers, was the key to students' success.

Concluding remarks
We have presented in this paper a review of how the notion of LEPs is currently defined and how it is conceptualised as a multiple and complex phenomenon, comprising mainly of "texts", "discourses", and "practices". We discussed key issues related to the naming of languages in LEPs in Europe, including the ideological and political implications of using terms such as "foreign", "additional", "regional" and so forth. We also discussed the "terminological puzzle" (Dalton-Puffer et al. 2014) of various models of LEPs for pre-primary and primary education. We then presented the guidance from the Council of Europe and how two main models of LEPs emerged as a result, namely early foreign language education and BE, including CLIL. Finally, we argued for the need to develop multilingual language education policies for all, and presented two successful examples of multilingual LEPs for young learners, one in an early care setting in Luxembourg and one in a primary school in Ireland. We have shown how LEPs can offer helpful models for educators to decide how language(s) should be used, taught and learned in educational contexts. However, and most importantly, we have shown that LEPs are also "a very strong form of language manipulation" (Shohamy 2006: 78). They impose particular language(s) or language varieties and exclude others, in a context attended by all children and young people for many years. They therefore contribute to the maintenance of the status and prestige of the languages chosen by those in power and to the continuing exclusion and silencing of other languages, whether home, minority or migrant languages. In researching, designing and implementing LEPs, it is therefore crucial to adopt a critical and empirical lens to unravel the ideological and political choices underlying all LEPs as well as the agency and creative practiced language policies of all educational actors.
To conclude, we call for the need to have the domain of LEP research as an integral part of language teacher education programmes as well as professional development programmes for in-service teachers, since language education has been shown to be strongly linked to social justice. In other words, it is crucial for teachers to understand that throughout their interactions with learners, they are implicitly or explicitly implementing a language regime which prioritises some languages at the expense of others, and that experiencing language discrimination from a very young age, silences and disempowers children (Piller 2017;Hélot and Rubio 2013). In line with policy makers of the European Commission (Council Recommendation of 22 May 2019: point 4h), we agree that support should be given to schools "to define their own approach to language learning, while respecting national legislation and helping schools to actively value and use their linguistic diversity". We hope to have shown in this article that existing LEPs can be questioned, and new ones can be created, to support the multilingual development of all young learners in Europe.