Learning environment and social inclusion for newly arrived migrant children placed in separate programmes in elementary schools in Norway

Abstract Global migration has led to increasing numbers of children encountering schools as newcomers in their new countries. Statistics reveal that large groups of migrant children tend to perform poorly academically; thus, education for newcomers has become an urgent issue for host countries. The aim of this study is to explore how Norwegian schools facilitate education and social inclusion for newly arrived migrant children, based on qualitative data from separate programmes for newcomers. The findings show a large variation as to how the schools emphasise the teaching of subjects and how they facilitate active learning processes for newcomers. The analyses suggest that this variation, at least partly, is linked to a lack of clarity in the official Norwegian guidelines for the education of newly arrived migrant children. As far as inclusion is concerned, the schools, however, show quite similar results: none of them succeed in creating arenas for interaction between newcomers and mainstream peers. This may have an impact on the newcomers’ well-being, as well as on how fast they learn Norwegian—and hence school subjects—because the verbal and social interaction between language learners and target language users is crucial to the development of L2 skills.


PUBLIC INTEREST STATEMENT
In the wake of global migration, large numbers of migrant children encounter schools in their new countries as newcomers, normally without knowledge of the language of instruction, nor with knowledge of their new communities´ cultural codes and practices. Research and statistics reveal that it may be challenging for the host countries to provide high quality education and to create inclusive environments for the newly arrived migrant pupils.
This article highlights the education of newcomers placed in separate programmes at three mainstream elementary schools in Norway and analyses how learning is organised as well as how the schools facilitate social interaction between the newcomers and the mainstream pupils.
The results show considerable differences in how learning activities are organised and this may have impact on the newcomers´ learning outcome. The schools showed similar approaches to issues of inclusion: none of them succeeded in creating sustainable arenas for interaction.

Introduction
In the decade from 2006-2016 Norway experienced a huge increase in the number of children and youth who fled or migrated to the country within compulsory school age (6-16), together with their families or as young unaccompanied. According to the Norwegian Ministry of Education and Research (Utdanningsdirektoratet, 2012/16) these pupils are defined as newcomers, a category that also comprises foreign born children who arrive in the country shortly before compulsory school age.
In a Norwegian context newcomers belong to the overarching category pupils with immigrant background but differ from the other group within this category: pupils born in Norway with foreign born parents. 1 Currently, 18 % of the pupils in elementary schools in Norway have immigrant background. Out of these, around 8 % are foreign born pupils (Education, n.d.o, 2020).
In Norway, as well as in other receiving countries, the low educational progress of newcomers (Steinkällner, 2013) has become an urgent issue for educators, school owners, politicians and policymakers. Here, the Migrant Integration Policy Index, MIPEX, (Huddleston et al., 2015) measures migrants' opportunities for participation in society on an international scale. Success in targeting the educational needs of migrant pupils and to what extent intercultural approaches to education are offered are among the dimensions investigated. The key findings state the following: Countries respond to large numbers and poor outcomes of immigrant pupils with many new, but weak targeted education policies, which are not always well implemented or effective in practice, (Huddleston et al., 2015). (p. 30)

Separate programmes for newcomers
A change in the Norwegian Education Act in 2012 opened up for education of newcomers in separate groups for a duration of up to two years, and in 2016 around 65% of the newcomers in Norwegian elementary schools were placed in separate programmes located at mainstream schools (Rambøll, 2016). This calls for special attention to how separate programmes function both academically and socially for newcomers. According to the Core curriculum-values and principles for primary and secondary education in Norway (Curriculum, 2017) Norwegian schools are supposed to develop an inclusive environment for all pupils, which implies both the pupils´ right to be integrated as participants in the school community, as well as the pupils´ right to have their diversity acknowledged. While research shows that newcomer pupils are positive to being taught in separate groups or classes parts of the school day (Manzoni & Rolfe, 2019), pupils also emphasize a strong wish to make friends early and to interact with mainstream peers (Manzoni & Rolfe, 2019). Being placed in a separate programme for a period of 1-2 years, as might be the case in Norway, is potentially challenging for the newcomers´ academic and social development, if arrangements for interacting with mainstream peers and for participation in mainstream lessons are not provided.
The current study's purpose is to provide new insights into the learning and inclusion processes for newcomers placed in separate programmes in elementary schools in Norway in order to contribute to a knowledge base from which policy makers can make research based decisions. As education for newcomers in Norway is considered an under-researched area, (Norozi, 2019;Thorshaug & Svendsen, 2014) there is a need for additional knowledge in the field.
The overarching research questions in this study are as follows: • How are the separate programmes organised, and what characterises the learning environments, as to the learning of L2 (Norwegian) and of school subjects?
• To what extent are the pupils in the separate programmes included in the school community at large?
Following the analyses of these questions the study will also discuss actions that schools can take to enhance the progress and success of L2 newcomers.

Selected research on education for newcomers relevant for this study
Newcomer pupils are not to be considered a homogenous group, and there are of course differences and nuances among the pupils, as to how well they perform at school in their new countries. Still research document that the general picture, at least for Norway, is alarming. According to Steinkällner (2013), migrant and refugee pupils are overrepresented in the category of those who leave compulsory school in Norway without having accomplished their courses. In addition, those newcomers who complete school achieve significantly lower results than pupils with immigrant background born in Norway and pupils with majority background. These differences persist when the educational level of the parents is taken into consideration.
There are several factors underlying this, such as the level of education achieved before arrival in Norway, the way the migration process works and the way the uncertainties regarding their stay in Norway influence their health and well-being (Frater-Mathieson, 2004;Loewen, 2004). Nevertheless, one must assume that a crucial factor is how the Norwegian school contributes to the academic success of the pupils. Following Steinkällner's research, the Norwegian school seems to be struggling with creating learning environments that are of a good quality for these pupils.
In the first decade of the 21 st century Hamilton and Moore (2004) in their literature review concluded that international research on migrant children and education was limited: While there is a large and diverse body of literature on refugees that addresses social, medical, political, linguistic and educational issues, there is a paucity of material specifically concerned with refugee children: of this, only a small proportion is about school-based interventions and programmes. (p. 2) Rutter (2006) drew much of the same conclusions in her work (p. 5): 'In comparison with the volume of studies on the traumatic experiences of refugee children there is little research about refugee children's educational experiences'.
In their analysis of educational provisions for newcomers in Norway, Thorshaug and Svendsen (2014) conclude that there is an explicit need for more research in the field and in particular on the effects of the measures taken. One of their recommendations is a clearer curriculum design for newcomers, hence securing the learning of both L2 and subjects. Hilt (2016) analyses the inclusion and exclusion processes for students in separate programmes in a Norwegian upper secondary school, showing how the educational practices meant to be inclusive in fact exclude pupils from participation. In the same vein, Chinga-Ramirez (2017) focuses on the minority pupils' self-understanding in relation to Norwegian majority pupils. In line with Hilt (2016), she argues that having a separate class, which is supposed to provide for an adapted education and integration, in fact disempowers the pupils and defines them as deviant from their majority peers. Hilt (2016) also underlines the absence of clear guidelines for educational practices in the separate programmes. Interestingly, De Wal Pastoor (2017) similarly points to the absence of regulations for unaccompanied minor refugees who live in group homes in Norway. She shows how newcomers benefit educationally and socially when group home staff emphasise education and work-life relations. However, group homes for unaccompanied minors in Norway are not subject to quality checks by the relevant authorities. Thus, it may be incidental whether the group home enhances the unaccompanied minors' educational and social life or not. Norozi (2019) in a study based on data from two Norwegian reception classes, claims that newly arrived pupils´ schooling in Norway is an under-researched area, in particular at the elementary level. Norozi investigates the professionals´ (teachers, coordinators and head teachers) perceptions of and experiences from working with reception classes for newcomers, and shows that the Norwegian speaking teachers found their task complex and difficult with regard to language challenges, to lack of suitable learning resources and as far as cooperation with their pupils´ families was concerned. Norozi concludes that mainstream teachers need professional support to be able to teach in more culturally responsive ways, In a Swedish context, two research-based books on the education of newcomers were published in 2015-2016 (Bunar, 2015;Kaya, 2016). However, even with this, Bunar states that researchbased knowledge in the field is quite limited. Nilsson and Axelsson, (2013) who study newcomers' experiences of their transition from separate programmes to mainstream classes in Sweden, find that although pupils describe a strong wish to enter mainstream class, they also describe social isolation and how they miss the teachers who target their needs in the separate programme. This study offers an important description of the tension between the newcomers´ need for tailored educational activities in smaller groups, and the need for being socially included into mainstream school culture. The authors therefore conclude that social and pedagogical resources should be seen as interlinked and interdependent when organising education for newcomers.
In the OECD report Helping Immigrant Students to Succeed at School and Beyond (OECD, 2015) first-and second-generation immigrant students' school performances are compared with those of nonimmigrant students, and the conclusions are in line with those drawn by Steinkällner (2013) for the Norwegian context: In most countries, first-generation immigrant students [. . .] perform worse than students without an immigrant background, and second-generation immigrant students [. . .] perform somewhere between the two (OECD, 2015), (p. 2) This report also measures immigrant students' sense of belonging at school (first and second generation) compared with non-immigrant students. Interestingly, the results are quite varied. In some countries, for example, the UK and US, migrant children reported a sense of belonging at the level of the other two groups. In the second group of countries, for example, Denmark, Mexico and France, the second-generation immigrants scored the lowest on sense of belonging, while in the third group of countries, for example, Italy, Norway, Spain and Sweden, the migrant children reported a lower sense of belonging than the two other groups. Manzoni and Rolfe (2019) in their study of how British schools are integrating new migrant pupils and their families, collected interview data from 15 schools throughout England (9 primary schools and 6 secondary schools). Among the interviewed were school leaders, teachers, parents and pupils. In the schools studied migrant pupils were usually placed in mainstream classes corresponding to their age from the start, often combined with language tuition in English in separate groups for parts of the day. The school managers and the teachers emphasized that segregation of newcomers normally was avoided because it was considered counterproductive to integration and that it potentially could limit the pupils´ access to school subjects other than English. The pupils clearly stated that they profited from English tuition in separate groups, but at the same time they wanted to make friends among the 'ordinary' pupils as soon as possible. Some of the schools had mentoring schemes and/or buddy schemes where both teachers/teacher assistants and peers could act as mentors for the newcomers, a practice that was highly appreciated by the pupils who had experienced it. The study shows that while the pupils clearly wished to be part of mainstream classes, they also needed language instruction in separate groups as well as support from mentors or buddies when placed in mainstream classes. This corresponds highly with what Nilsson and Axelson (2013) describes for Swedish newcomers in the transition process from separate programmes to mainstream classes: they strongly wish to transfer to mainstream classes but they also need scaffolding to stay there. This, however, is financially demanding, and Manzoni & Rolfe also report that financial constraints in British schools put limits on the support they were able to offer to migrant children.

Theoretical framework
Education for newly arrived migrant pupils is demanding because it requires competences both within second language learning and intercultural pedagogy in order to support the pupils´ academic and social development. To be able to understand the complexity of the activities in the newcomer programmes studied here, the theoretical perspectives in this study are selected from an interdisciplinary base: research on multicultural and intercultural education and research on second language learning. Banks (2009) uses the concept of multicultural education, which is common in the US, whereas the concept intercultural education is frequently used in several European countries (C.O.E., 2008), including Norway. In their discussion of the two terms, Holm and Zilliacus (2009) point out that in Europe, multicultural education has come to represent a static and descriptive entity that does not comprise the complexity of modern plurality, while the term intercultural education is understood as a dynamic concept that emphasises interactions within and between the members of different cultural groups. This corresponds to the way the two concepts are presented in the UNESCO guidelines for intercultural education as well (UNESCO, 2006). Holm and Zilliacus (2009) show that this perceived dichotomy between the concepts lacks nuance, furthermore outlining how both concepts are characterised by internal variation: there are traditional and progressive versions of multicultural and intercultural education. Here, Banks and Banks' (2001) definition of multicultural education shows that their point of departure is a broad, dynamic concept used for the cultural complexity of modernity, comprising male and female students, exceptional students and students who are members of diverse racial, ethnic, language and cultural groups. Thus, there is a correspondence between Banks and Banks' use of the concept of multicultural education and the way intercultural education is perceived and described in European contexts. This makes it possible to ascribe more or less the same meaning to the two concepts, a stance that is taken in the current study. Banks (2009) emphasises the role of L2 learning and development of academic skills as essential for the education of minority pupils but simultaneously claims that the measures preventing marginalisation processes must be included as well. Banks (2009) presents his holistic paradigm, which views the school as a social system where teaching activities, curriculum, staff attitudes, the languages and dialects of the school and community participation are all intertwined dimensions. This is much in line with Ladson-Billings (2014) call for an understanding of the socio-political context in which educational activities are embedded. In the same vein, Nieto (2017) points to the importance of explicitly addressing power relations to avoid focusing on the superficial aspects of diversity. Nieto (2017) refers to Banks as one of the researchers who made power relations visible early on. Indeed, Banks' theoretical approach provides a fruitful lens for analysing the data in the current study because they contain both processes of learning in an L2 environment and the interplay between learning and social inclusion processes.
As mentioned, Banks (2009) includes the attitudes of the school staff in his approach to multicultural education. In her discussion of how teachers' attitudes, perceptions and beliefs influence their instructional judgements and actions, Gay (2010) highlights that through their training, teacher students must learn how to consciously analyse their attitudes and beliefs towards diversity in classrooms and schools. This corresponds to Norozi´s (2019) analyses of the Norwegian reception class teachers´ need for more culturally responsive practices. Similar views are emphasised by Skrefsrud and Østberg (2015) who, in a Norwegian setting, discuss the concept of context-based teacher education where school subjects, didactics and pedagogy are integrated in ways that make them relevant for classrooms and schools that are characterised by diversity.
The development of L2 skills is crucial for newcomers' opportunities to obtain access to the curriculum, which is a prerequisite for performing well at school. In line with Banks' interdisciplinary approach to minority pupils' schooling, the stance towards second language learning taken in the present research is a sociocultural one. For instance, in the wake of Long's interaction hypothesis (1980), much substantial and convincing research on L2 learning has documented how learners profit from active participation in negotiated interaction. The knowledge developed in this field shows how important it is to create learning environments that allow pupils to interact with each other and with target language users as a way to enhance second language learning and the learning of school subjects.
How L2 learners achieve increased proficiency through participation in collaborative dialogues is the issue of the works of, for example, Swain (2000), Swain and Lapkin (2003), Kim and McDonough (2008) Storch and Wiggelsworth (2007) and Swain and Watanabe (2013). A central theoretical backdrop for research on collaborative dialogue in second language acquisition (SLA) is Vygotsky's sociocultural theory (1978), where language production is seen as a tool in cognitive activity and where the interactions between learners are considered vital for the development of both thought and language and, furthermore, as a prerequisite for learning.
One of the earliest works on sociocultural SLA comes from Frawley and Lantolf's (1985) research, where the authors argue, with reference to Vygotsky, that [. . .] the origin of the ability to engage successfully as an individual in strategic processes (self-regulation) lies in social interaction. (p. 20) Block (2003) labels the period starting in the mid-1990s 'the social turn in SLA', of which an increasing interest for research within the sociocultural paradigm is a part, resulting in research by, for example, Lantolf and Pavlenko (1995) who consider the locus of learning being situated in dialogic interaction, Gibbons (2006) who has classroom research as her point of departure, and analyses second language development through teacher scaffolding as well as peer interaction, and Duff (2007) who investigates the similarities as well as the differences between the research field Second Language Socialization (SLS) and Sociocultural Theory, describing how 'both approaches have a social, cultural, interactional and cognitive orientation to language learning' (p. 312), and how linguistic scaffolding through interactions help novices reach their goal.
Based on previous research on how learners profit from participating in learning environments that stimulate conversational interaction, a sociocultural theoretical perspective on L2 learning has been chosen for the present study. Here, the focus on collaborative dialogue comprises both the interactions among second language learners as well as those between second language learners and mainstream peers. The latter kind of interaction interestingly links second language learning to issues of inclusion in the sense that arenas for interaction need to be provided if L2 learners in separate programmes are to interact with mainstream peers.

Data and methods
Data were collected at three schools for two months in 2017. The schools, which are situated in two municipalities in South-Eastern Norway, were selected based on the fact that they were reception schools for newcomers in their area, organising education in separate programmes at mainstream schools. A reception school in a Norwegian municipality is a school for migrant pupils from different home schools in the area. The pupils are placed at the reception school for a period of up to two years before they either continue in mainstream classes at the same school or are transferred to their home schools or proceed to lower secondary/secondary education. A reception school is supposed to be sufficiently pedagogically equipped to provide for quality education for the newcomers.
The research design is within the qualitative, interpretive paradigm, combining different field research methods for a triangulation of the data: nonparticipant observations in classrooms, (Hennink et al., 2011), a combination of factual and narrative interviews with school leaders (Kvale & Brinkmann, 2009) and focus group interviews with teachers. To contextualise the data generated through the observations and interviews, a document analysis (Bowen, 2009) of the guidelines for the education of newcomer pupils in Norway (Utdanningsdirektoratet, 2012/16) was carried out.
Observations of teaching and learning activities can give first-hand information of what is going on in the classrooms; hence, this method can shed light on the learning environment for newcomers. To be able to make comparisons in the learning environments between the three schools, observations-which may be characterised as semistructured (O´Leary, 2014)-were carried out with four main themes in mind: • What work methods were used?
• The approach to L2 learning • The teaching and learning of subjects • The use of learning resources The choice of nonparticipant observation made it possible to take field notes during the observations. There was only one researcher in the classroom during the observations, and the teacher explained the researcher's presence with a wish to become familiar with the learning activities in the classroom. Of course, an unknown person attracts attention in the beginning, but in this case, the pupils quickly went back to their normal activities, and there was little or no contact between the researcher and the pupils during the observations. The recorded interviews with the teachers and management were designed to fulfil at least two purposes: obtaining factual information about how the school organised its education for newcomers and to bring forth the leaders' and teachers' understanding of the learning environment and the social inclusion processes for the newcomers at their schools.
The core data thus consist of field notes from classroom observations and transcriptions of audio-recorded interviews, in addition to the document analysis of the official guidelines for the education of newcomer pupils in Norway (Utdanningsdirektoratet, 2012/16).
A total of 16 lessons in classrooms were observed, and four hours of group interviews with the teachers were conducted, as well as 1.5 hours of individual interviews with three school leaders (cf. Table 1) who were department leaders for the newcomer programmes at their respective schools. The teachers in the group interviews were as follows: at school A, the two class teachers responsible for the teaching of all subjects from grades 1 to 10; at school B, two teachers from each age group in the school's separate programme, totalling six teachers. Like the teachers at school A, they were responsible for all subjects in class; and at school C, the social sciences and Norwegian language teachers participated in the interview. The reason why the observations and group interviews at school A (cf. Table 1) were more extensive compared with those at schools B and C was because of the complexity in the group composition at school A (see below).
The active participants in this project were teachers and school leaders, who had agreed to participate in the interviews and who gave their consent. The pupils were not approached as individuals and were only observed as groups of pupils in class. No individual pupil participated in any interview situation, and no individual pupil is possible to trace through the published data.
Likewise, no individual teacher or school leader is identifiable by name. Furthermore, the schools were anonymised.
The classroom observation data were analysed with reference to the four pre-established categories mentioned above: work methods, approach to L2 learning, the teaching and learning of subjects and the use of learning resources.
The interview data was transcribed and analysed with a focus on thematic content linked to the abovementioned categories for classroom observations and to organisational features (group size, group composition, teacher competences, pupils' possibilities for staying in mainstream classes, etc.) Furthermore, the interviewees reflections and thoughts around issues of social inclusion were highlighted in the analytical process.
A document analysis is seen as complementary to observations and interviews, being part of a triangulation in qualitative research. The document analysed is the official guidelines for the education of newcomers in Norway (Utdanningsdirektoratet, 2012/16), a document frequently referred to by the interviewees. The point of departure for the document analysis was the categories used to structure the current study's classroom observations (cf. above). However, the question of using bilingual resources and issues regarding the inclusion of newcomers into mainstream classes and schools were also central to the analysis.

Organisational features
Regarding the organisational features, the observations and interviews unveiled significant variations between the schools: school A gathered its 16 newcomers from grades 1 to 10 in the same class, resulting in a wide age span among the pupils (from 6 to 16 years). For parts of the day, the pupils were divided into three groups if there were rooms available. In contrast, at school B, the 30 newcomers were divided into three age groups covering grades 1-2, grades 3-4 and grades 5-7. There were 10 pupils in each group, and they had permanent classrooms of their own. School C had chosen to keep the newcomers from grades 8-10 together in one group of 18 pupils, all in their own classroom; however, they were separated from the main school building.
All three schools had between five and six pupils per teacher. However, at school B, the newcomer programme was staffed with three teacher assistants as well, thereby offering the pupils support from a larger number of adults.
All teachers had basic qualifications, although not all the teachers at school C were trained in L2 pedagogy. Regarding access to bilingual instruction, which is a highly relevant issue in the education of newcomers in Norwegian schools, this was absent at school C and present only to a limited degree at the other two schools: school A had one teacher in a part-time position and who had Arabic as a first language, while school B had a teacher assistant who spoke Thai, which was the first language of some of the newcomers.

Work methods, the teaching and learning of subjects and approach to L2 learning
At school A and C, the teachers taught subjects through a combination of class plenary and individual activities. Choosing individual work methods is understandable when teachers are faced with larger and more heterogeneous groups, as was the situation at these schools. At school B, however, exploratory talk was facilitated. For instance, in a mathematics lesson for the group covering grades 3-4, the nine pupils, who were in groups of three, moved between different tasks engaging them in play, as well as in oral reflections with the teacher. The study does not provide enough data to generalise about the differences in work methods. Still, the different organisational frames at the schools may have influenced the way teaching was allowed to be organised. It is easier to facilitate group activities with a limited number of pupils and with a group consisting of pupils who are more or less of the same age, like at school B. Research on second language learning, for example, Swain (2000), Storch and Wiggelsworth (2007) and Swain and Watanabe (2013), show that verbal interaction is crucial for the development of knowledge and language, and it is therefore likely to assume that pupils benefit from being exposed to learning activities where peer interaction is combined with teacher guidance.
The observational data revealed that in general, the teachers at school B systematically structured learning activities in subjects like mathematics and science through thematically organised approaches that combined language and content learning. At the two other schools, the teaching and learning of subjects appeared less clear and structured.
When it comes to L2 learning, the teachers at schools A and B showed quite similar approaches: the learning processes contained the conscious use of everyday episodes, hence linking linguistic development to situations the pupils had experienced. Formal aspects of language were presented much in line with Long's focus on the form approach (1991) and were raised when the pupils made them relevant. For instance, the group at school A had put up nesting boxes in trees surrounding the school yard. A lot of oral interaction in Norwegian was motivated through this project, and eventually, written texts were produced, too. At school B, the teachers for the smallest children continually tested out whether the pupils understood what was going on; they consciously provided verbal transitions between activities, thereby repeating what had been done and preparing what was to come. At school C, the observations revealed less structured L2 teaching, for instance, regarding the classroom sequences containing attention to the formal aspects of language. In one lesson, the pupils had read a piece of fiction they clearly were interested in talking about. Instead of exploring their interest, they were asked to answer detailed questions about the pronunciation of consonants in Norwegian. A similar episode occurred when the class read a text about different professions and the educational training needed to work within these professions. The pupils were engaged in the issue but not encouraged to develop the theme. Instead, they were expected to outline the rules for using definite and nondefinite forms of nouns in Norwegian. Here, the teachers missed taking advantage of the momentum lying in the natural interest created through reading. The sequences that were meant to raise awareness on form appeared as random rather than incidental (Ellis, 2005) and were certainly not linked to the questions that the pupils had raised. Alternatively, in both cases, the teachers could have opened up exploratory talk in small groups based on the pupils' expressed interest, creating arenas for negotiation of meaning, thereby providing valuable training in communicative skills. Attention to form could have been raised during these interactions as a result of communicational needs, following Long's focus on the form approach (1991).
It is likely to assume that the differences we see between school C and the two other schools here, at least to some extent, is due to the differences in teacher competences. As mentioned, not all teachers at school C were trained in L2 learning, which left the teachers at this school with fewer tools to structure the learning environment for L2 learners.
In Norway, schools have access to digital, bilingual learning resources that are especially designed for L2 learners, for example, Skolekassa (Nasjonalt Senter for flerkulturell opplaering (NAFO), N, 2016). However, insufficient equipment and technical problems in the rooms where the separate programmes were offered made such access difficult. This was reported by the teachers at both schools A and C and to some extent also at school B.

Issues of inclusion-possibilities to interact with mainstream peers
As mentioned, at schools A and B, the newcomer programmes were located in separate classrooms in the main building but had some important differences: The pupils at school B had their permanent three classrooms, while at school A the pupils had one classroom at their disposal and otherwise had to move between different locations whenever they were vacant. They used the rooms in the day-care facilities, school kitchen and library when the class was divided. At school C, however, the newcomer programme was not located in the main building at all; the class was placed in its own house outside of the building where the mainstream classes had their rooms.

Social inclusion in the school as a whole
In the interviews, the teachers from all three schools pointed to the absence of systematic interactions between the newcomer programmes and mainstream classes. At school B, for instance, a winter activity day had been arranged, but the newcomers had not been invited. At school C, the teachers talked about a democracy camp recently arranged for eighth graders, where the integration of immigrants was the chosen theme. The newcomer class was not informed about the event. One of the teachers at school C expressed it in the following way: We have talked about creating arenas for inclusion, without getting any further. The pupils express that they feel excluded and wonder why they are treated differently from their mainstream peers. The pupils feel this is wrong, and we as teachers do, too.
These are narratives of invisibility that are hardly intended but are a reality. The teachers mentioned how they have tried to invite themselves and their pupils into mainstream classes, how the teachers in mainstream classes have responded positively at first but eventually have failed to follow up. At all three schools, it appears that the newcomer programmes are not considered fully worthy of being included in the school society.

Fragile arrangements for visits to mainstream classes
As part of the newcomer programme, the students, especially at the lower secondary level, were supposed to become familiar with teaching and learning practices and social life in the mainstream classes they eventually would attend after their stay in the separate programme (Utdanningsdirektoratet, 2012/16). Therefore, plans for the newly arrived pupils' presence in mainstream classes during their time in the separate programme normally should be worked out and operationalised as soon as the pupils are able to understand and communicate enough in Norwegian to benefit from a visit to mainstream classes. However, the interviews revealed that none of the three schools had developed sustainable frameworks for such visits.
At school A, the intention was that the pupils at the lower secondary level should participate in mainstream classes one to two days a week, but both the leaders and teachers confirmed that this turned out to be difficult to arrange systematically. At school C, the plan was likewise to let the pupils stay in the mainstream classes one day a week-but according to the staff and leaders, the routines for this were weak.
School B had only pupils at the primary level and were, as such, not expected to have systematic routines for newcomers' visits to mainstream classes. Both staff and leaders were, however, eager to develop a system for making newcomers' participation in mainstream classes possible. So far, they had not succeeded in materialising this idea.
During the interviews, especially with the leaders, it became clear that the official guidelines for running education programmes for newcomers in Norway (Utdanningsdirektoratet, 2012/16) was a text that they had read, although interpreted somewhat differently. They frequently referred to this as guidelines to lean on but simultaneously characterised the document as a text with a certain lack of clarity.

Analysis of the guidelines for education of newcomers in Norway
The guideline document was first published in 2012 and revised in 2016 by the Norwegian Education Directorate. Its main target group is school owners, that is, municipal or regional school managements.
The political intention is clearly described in the document: newcomer programmes should be of a temporary character-limited to a maximum of two years-and the pupils should be transferred to mainstream classes as soon as possible. Simultaneously, the guidelines emphasise the freedom of school owners to organise the programmes. No limitations are set as to how many subjects, hours taught per subject or competence goals in the standard curriculum that the newcomer pupils could get exemption from, nor are there any limitations as to the exemptions from evaluation in subjects.
Furthermore, the guidelines put few constraints on the composition of classes regarding the age span of the pupils, the number of pupils in the same group or the teacher-pupil ratio. As to teacher qualifications, the guidelines underline that teachers should be trained in second language learning and intercultural pedagogy. It is also stated that it would be advantageous to have bilingual teachers available. However, these are recommendations kept at the 'ought to' level. As shown above, the interviews revealed that the school leaders interpreted these guidelines differently, and the programmes varied correspondingly. As one of the school leaders put it: I am not sure whether we do things right here -I have asked for inspection, actually.
No doubt, the guidelines contain well-founded, research-based recommendations for the education of newcomers. The problem is that the measures are formulated in a nonbinding manner, thereby making it possible for school owners to decide whether to implement them or not. For example, crucial issues like creating good routines for interactions between newcomers and mainstream pupils are described at the 'ought to' level. The same goes for the mapping and assessment of the competences the newcomers previously have obtained in languages other than Norwegian and in other school subjects, before they are placed in the newcomer programme. This lack of binding formulations seems to be the result of an attempt to balance the need for national standards, on the one hand, and the autonomy of municipal and regional school management, on the other hand. In Norway, the freedom of school owners is traditionally strongly emphasised when regulations for education are worked out. This is also discussed in the white paper Diversity and Mastery-Bilingual Children, Youth and Adults in the Educational System 2 (NOU 2010(NOU :7 (2010 In fact, it is reasonable to link the origin of the National guidelines for the education of newcomers to the 2010 white paper, where the education of newly arrived pupils in Norway was addressed and where the need for explicit guidelines was stated, backed up by the argument that it was important to secure equal educational opportunities for newcomers, regardless of where they lived in the country.
The committee writing the 2010 white paper described the education of newcomers in Norway at that time as challenging and characterised by variation, giving special attention to two major issues: • Subjects from the standard curriculum should hold a central position in the education of newcomers, in addition to intensive language instruction.
• Newly arrived pupils represent a group where there is a huge need for mother tongue and bilingual instruction to get access to the curriculum. (NOU 2010(NOU :7, 2010 In the present guidelines, however, these issues are not put to the fore to the same extent. Regarding the teaching of subjects, this is mostly described with reference to the procedures for exemptions from the standard curriculum, thus the white paper's urge for a clearer focus on school subjects is not followed up in the guidelines. The same goes for mother tongue and bilingual instruction, even though these rights are regulated by the Education Act (The Education Act, 1998).
Based on this, the guideline document does not take a clear stance when it comes to the two major issues underlined by the white paper. This, combined with a general lack of clarity, which are exemplified by the formulations at the 'ought to' level, are the crucial factors in explaining why the schools in these data have different approaches regarding how systematic subjects are taught and how systematic bilingual resources are being used. The same goes for the importance of creating strong and systematic arenas for the interaction between newly arrived pupils and mainstream peers.
Earlier in the current article, I referred to publications by De Wal Pastoor (2017) and Hilt (2016), who point to the lack of clear guidelines for education of newcomers in upper secondary school and the lack of quality assurance in Norwegian youth homes for those who are young and unaccompanied. A lack of clearness was also found as a theme in the interviews with teachers and leaders at the schools studied here. One of the leaders expressed it in the following way: Neither teachers nor school leaders are satisfied with how this works today -but there does not seem to be enough political will to make a change.

Discussion
The findings reveal organisational differences between the three newcomer programs studied, and it is likely to assume that these differences may influence the quality of the education offered. The age span between the pupils in a group may limit the teachers' possibilities to facilitate learning for all. Where bilingual instruction is provided, it will facilitate newcomers' access to the curriculum, and the pupils will probably learn faster than in classrooms where bilingual instruction is absent. Furthermore, teachers trained in L2 learning will probably create better learning environments than those without. Stable access to adequate rooms at school is likewise important, as is functioning internet connections, when learning resources are web-based.
While these differences to some extent can be linked to a certain lack of clarity in the official guidelines, the challenges linked to inclusion seem still more complex. The lack of inclusion concerns all three schools studied, more or less to the same extent; they were all facing great challenges regarding the inclusion of the newly arrived pupils into the school community. One of the teachers described the situation at her school as follows: We haven't created any arena for them to be included into the school community -we just haven't done that.
From this, it is likely to assume that improved conditions for interactions with mainstream peers would enhance the second language development of the newcomers, as well as their sense of well-being, which is an important condition for learning.
Two central questions emerge here: What are the obstacles? Why is it so difficult to initiate measures of inclusion in the school community? I will argue that the way newcomers are made 'invisible' in the schools studied here can be linked to marginalisation processes in Banks´ sense (Banks, 2009). Not being invited to events meant for the school community is one example. Being localised in a building outside the mainstream area of the school is another. To be delinked from functioning technical solutions while the mainstream classes are online is yet another. Surely, these examples are not the results of intentional actions, but what they reveal is a lack of consciousness in the school community at large of the responsibility to include groups of pupils who are unfamiliar with the language, culture and practices in Norwegian schools. Banks (2009) argues that, for instance, changing the curriculum alone will not help a school improve the conditions for minority pupils. The teachers' and majority pupils' attitudes, perceptions and beliefs will have to be addressed as well. This fits well with the approaches taken by both Ladson-Billings (2014), who underlines the need for a broader, sociopolitical understanding of educational activities, and Nieto (2017), who points to the importance of addressing power relations when dealing with diversity issues in education.
A common, identified challenge at all the schools studied here is that staff outside the newcomer programmes and the pupils in the mainstream classes are not relating to the newcomers in any significant way. This influences the situation for the newly arrived pupils in a number of ways: they miss valuable opportunities to develop their L2 competences and to proceed academically. In addition, we can assume that being part of a separate programme that is in fact excluded from the mainstream community at the schools negatively influences the pupils' self-understanding (cf. Nilsson & Axelson, 2013;Chinga-Ramirez, 2017). As shown, the OECD report measuring immigrant students' sense of belonging at school (OECD, 2015) documents that in Norwegian schools, migrant pupils feel less included than second-generation immigrant pupils and mainstream pupils. Here there is an interesting contrast to the OECD-results reported by newcomer pupils in British schools: they report a sense of belonging at the level of the two other groups: second-generation immigrant pupils and non-immigrant pupils. These differences might be the result of the different practices in British and Norwegian schools with regard to the organisation of education for newcomers. Manzoni and Rolfe (2019) show how British schools usually avoid placing newcomer pupils in separate programmes on a regular basis as opposed to the situation in Norway, where separate programmes are quite common (Rambøll, 2016). Furthermore, Manzoni and Rolfe (2019) report that teachers and pupils are engaged as mentors and buddies who act as facilitators for integration of migrant pupils in several of the British schools studied. They also show how the schools engage migrant pupils in extracurricular activities together with mainstream pupils. Initiatives like these probably play an important role in developing the migrant pupils´ sense of belonging and, as we have seen, such initiatives are absent in the Norwegian schools studied here. In Norway a voluntary organisation, Ung Inkludering, Inclusion of Youth (Inkludering, 2017), was established in 2017 as an initiative to create arenas for interaction between migrant and mainstream pupils in Norwegian schools through similar mentor and buddy schemes that Manzoni and Rolfe (2019) refer to from Britain. Ung Inkludering is however a young organisation with limited resources and for the time being operating only in three regions in Norway. It goes without saying that a small voluntary organisation is not capable of providing for integration of migrant pupils in Norwegian schools alone. To make a difference here, the school-owners and the schools themselves must take on a greater responsibility for the integration of migrant pupils.
Making a change and using a holistic paradigm in Banks´ sense, would imply initiating what Banks (2009) andØstberg (2017) label an empowering school culture, for example, with arenas where the newly arrived pupils could interact with mainstream peers and where newly arrived pupils could visit mainstream classes in subjects they feel competent in. This is also related to the contextual approach to the education of refugee children noted by Rutter (2006) and Hamilton and Moore (2004), an approach that takes into account the social and ecological context surrounding the educational activities (Bronfenbrenner, 1992). These approaches, however, presuppose a school environment where the staff and pupils in mainstream classes interact with the staff and pupils in the separate programmes, in other words, that the newly arrived pupils and the separate programme are seen as a common responsibility for the school community, not as a part of the school left to the separate programme teachers and coordinators to run.
At a superordinate level, this is linked to the larger discourse on integration of immigrants into Norwegian society. Gullestad (2002) discusses how the Norwegian public discourse on immigration almost exclusively deals with 'them', that is, the immigrants, and not with 'us', the majority population. Lidén (2017) discusses how the content of the concept of integration has gradually changed in the Norwegian discourse, from describing a two-way process, one where the majority and minority in some state of reciprocity adjust to each other, to being conceived more like a one-way process, one where the minority assimilates. Thus, the role of the majority population in the integration process is erased, much like the role of the staff and pupils in mainstream classes appears to be vis-vis the separate programmes in the current study. Teachers and pupils in mainstream classes are not challenged to involve themselves in the daily lives of the newly arrived pupils in the separate programmes, either. The data in the current study indicate that teachers and pupils in mainstream classes will have to if the quality of the education of newcomers in Norwegian schools is to improve.

Conclusion
The present study has investigated organisational practices, teaching and learning practices and issues surrounding the social inclusion of newcomers in separate programmes at three elementary schools in Norway. The findings document a variation between the programmes regarding the age span of the pupils put together, how systematically subjects are taught, the approaches to work methods in general, to second language learning and to the use of bilingual resources to get access to the curriculum. The data analyses indicate that this variation, at least partly, is because of a lack of clarity in the official guidelines for the education of newcomers in Norway.
In common, the schools struggle with issues of inclusion; they do not succeed in facilitating interactions between newcomers and their mainstream peers, nor among the teachers in separate programmes and teachers in mainstream classes. The data analyses indicate that this is related to a lack of consciousness among leaders and teachers in looking at the newly arrived pupils as fully worthy members of the school community.
The present study is a small-scale design based on qualitative methods, a fact that calls upon cautiousness when it comes to generalisations. Still, the insights gained from the current study could represent a step towards more elaborate research on how the low educational attainment of newly arrived pupils should be understood and analysed in a Norwegian context and in other receiving countries as well. Furthermore, the knowledge produced in the current study could inform the development of more targeted programmes, thereby forming a valuable basis for discussing policy changes and actions. For example, the analyses here strongly indicate that if the practice of organising education for newcomers in separate programmes should continue, arenas for interaction between newcomers and mainstream peers should be secured, as should intensive language instruction, bilingual support, and the teaching of subjects from the standard curriculum.

Funding
The author received no direct funding for this research.