Comparing language ideologies in multilingual classrooms across Norway and Zambia

This article compares the language ideologies (Kroskrity 2000) of pre-and in-service teachers in Norway and Zambia respectively. Despite their historical, political, and linguistic differences, both countries struggle to adapt their educational systems to students’ multilingualism. Thus, it is interesting to see how pre-and in-service teachers from the two countries consider the role of multilinguals within their respective education systems. The data are from two qualitative studies about multilingualism in education from Norway and Zambia that explore pre-and in-service teachers’ language ideologies (Kroskrity 2000). Based on focus group interviews with 24 Norwegian pre-service teachers and 36 Zambian in-service teachers, the current article shows that the Norwegian pre-service teachers and the Zambian in-service teachers expressed convergent descriptions of the challenges associated with multilingualism in education. Yet the teachers revealed divergent language ideologies in relation to how to solve these challenges. While the Norwegian pre-service teachers conveyed rather monoglossic language ideologies, the Zambian in-service teachers aligned themselves with more heteroglossic ideologies. In line with these language ideologies, they positioned themselves differently towards the current language policies in the two countries. This divergent pattern is discussed in light of the specific language ecologies of the two states.


INTRODUCTION
The present study compares pre-and in-service teachers' language ideologies (Kroskrity 2000) in two radically different contexts: Norway and Zambia.Norway has become increasingly multilingual over the past few decades due to recent immigration (Statistics Norway 2019), while Zambia has a long history of multilingualism (Makalela 2016).Nonetheless, because of Zambia's colonial history, Zambian education does not have an equally long history of multilingual education (Tambulukani and Bus 2012).As both countries' education systems are struggling to meet the needs of their multilingual students, it is important to ask which language ideologies can be identified within the respective education systems and how they relate to local conditions.Hence, this article presents a comparative analysis of the findings from two qualitative research projects about multilingualism in education involving focus group interviews with pre-and in-service teachers from Norway and Zambia respectively.This comparison is considered in light of the particular language ecologies of the states in question.
It is valuable to compare language ideologies across the Global North and Global South in order to learn from each other.As Pennycook and Makoni (2020) note, there is a tendency to universalise research findings from the Global North, while excluding the majority of the world from scientific theorising.Heeding to the call for Southern perspectives, this article presents insights from in-service teachers in the Global South that can benefit pre-service teachers in the Global North.In the following, we first present the language ecologies of Norway and Zambia and our understanding of language ideologies.We then present our methods and materials.Next, we introduce the comparative analysis of the data from the two research projects, before we discuss the underlying causes behind the convergent and divergent language ideologies in the two countries in connection with the particular language ecologies.

LANGUAGE ECOLOGY OF NORWAY AND ZAMBIA
Although classroom practices are situated and localised, they are also part of larger and more political ideologies (Creese and Martin 2008).With regards to language ideologies, Blackledge (2008: 30) argues that 'language ideologies are positioned in, and subject to, their social, political and historical context'.Hence, in order to conduct meaningful comparisons in educational research, it is imperative to provide comprehensive descriptions of the contexts involved and their historical determinants (Phillips 2011).In the following, we will therefore present the language ecologies in the two countries involved.The term 'ecology of language' was first introduced by Haugen (1972).Since then, language ecology has been a theoretical framework frequently applied in sociolinguistic research (Creese, Martin, and Hornberger 2008;Pennycook 2010), although its application has often been inconsistent (e.g.Creese and Martin 2008;Pennycook 2004).However, in this article 'ecology of language' is used interchangeably with 'language ecology', and refers to the following: A conceptual orientation to critical thinking about multilingualism that calls upon researchers to focus on relationships among languages, on relationships among social contexts of language, on relationships among individual speakers and their languages, and on inter-relationships among these three dimensions' (Hult 2013: 1).This particular approach to investigating linguistic situations is used to describe the contexts of Norway and Zambia because it accentuates the connection between the different languages used in the two countries and the relationship these languages have to the societies at large (e.g.Pennycook 2010).Pennycook and Makoni (2020: 45) argue that multilingualism is not 'a universal category; indeed, the very idea that multilingualism could refer to the same thing in diverse contexts of communication is revealed as an absurdity'.Therefore, what follows explores the concept of 'multilingualism' in the contexts of Norway and Zambia.
Norway has traditionally been a rather homogenous country in terms of religion, ethnicity, and language.Norwegian has been the official language of instruction in public schools and universities for more than a century.Through the nation-building period in the 19th and early 20th century and the establishment of the welfare state, Norwegian education developed into an 'undifferentiated fellowship' (Engen 2010) where equality was understood as sameness (Chinga-Ramirez 2017).Today, policy documents describe Norwegian as 'a common language' for the population of Norway (Ministry of Education and Research 2008), although the indigenous Sámi languages have also been recognised as official languages and equal in status to Norwegian in certain parts of the country.For the past decades, increased immigration has brought even greater linguistic diversity to Norwegian classrooms.Thus, approximately 17% of all students within Norwegian education speak a language other than Norwegian at home (Statistics Norway 2019).These students are expected to quickly acquire proficiency in Norwegian in order to follow instruction through the 'common language' of Norwegian education (e.g.

Ministry of Education and Research 2008).
In a literature review of Norwegian research on language ideologies and language beliefs, Kulbrandstad (2015) points out that there is still much research to be done on this topic.Yet he concludes that research has so far indicated that Norway is influenced by a 'monolingual ethos' promoting the idea of one language-one nation (Kulbrandstad 2015: 271).Haukås' (2016) qualitative study of Norwegian 'foreign language teachers' (e.g.French, German, and Spanish), found that although the teachers believed that their own knowledge of different languages had been beneficial to their language learning, they did not come to the same conclusion regarding their students.Hence, they did not involve other languages than Norwegian and English in their instruction, since they believed they would have to know these languages in order to include them.
Zambia is currently and historically at its core a multilingual state (Banda and Jimaima 2017).Currently English is the official national language, and seven Zambian languages (Bemba, Nyanja, Kaonde, Lozi, Lunda, Luvale, and Tonga) are recognised as official regional languages (Banda and Jimaima 2017).Moreover, there are numerous minority languages and language varieties without official recognition (Banda and Jimaima 2017;Mwanza 2017).During colonial rule, local languages were tolerated or even promoted as media of instruction in Zambian schools, particularly at the elementary level (Chimbutane 2012).In 1928 four local languages (Bemba, Lozi, Nyanja, and Tonga) were selected by the colonial administrators as media of instruction from Grade 1 to Grade 4 (Manchisi 2004).Albaugh (2014) points out missionaries' need to 'save souls', colonial administrators' need for local auxiliaries, and Africans' wish to qualify for jobs through education as reasons behind the use of various local languages in education during colonial rule.However, at independence in 1964, Zambia adopted the backwards policy of abandoning the use of local languages and instead adopt an English-only policy at all levels of education (Chimbutane 2012).
The English-only policy proved to be a challenge, especially for early grade learners.Hence, Zambia experienced low literacy achievements among early grade learners (Manchisi 2004).Since the late 1990s, this policy has therefore been challenged, so that the seven official local languages were first introduced to instruct initial reading skills in Grade 1, while English was the medium of instruction from Grade 2 onwards (Mkandawire 2017;Tambulukani and Bus, 2012).Since 2013, the use of the official local languages has been extended to Grade 4 (Chileshe, Mkandawire, and Tambulukani 2018;Tambulukani 2015).Nonetheless, Tambulukani and Bus (2012: 142) note that 'there may not be any great overlap between the local Zambian language that is officially designated as the language of instruction in a particular district and the language spoken at home and in the playground'.This is due to a situation where language policies do not take into consideration the other language varieties that students bring with them to class (Tambulukani and Bus 2012).Although there is high enrolment in primary education in Zambia (Masaiti and Chita 2014), the education system still struggles to achieve desirable literacy levels in the population (Tambulukani and Bus 2012).Some researchers have pointed to the linguistic diversity as an important reason behind the challenges related to the country's literacy levels (Tambulukani and Bus 2012).
Zambian studies of teachers' language ideologies or language attitudes is limited to a study by Mwanza (2017), based on interviews with 18 teachers of English from six secondary schools.Particularly relevant for the current article, this study investigated the teachers' views about the place and value of Zambian languages in the teaching of English.Mwanza found that that the participants in his study held negative attitudes towards Zambian languages and that the teachers had a monolingual approach to English teaching.However, no studies have to our knowledge explored the language ideologies or attitudes of primary school teachers in Zambia, nor teachers of school subjects beyond English.
The language ecologies of Norway and Zambia are very different.However, both countries struggle to meet the needs of their multilingual students.How different language policies in education have been developed and how teachers approach multilingualism in the classroom is also influenced by language ideologies (Hélot and Ó Laoire 2011;Jaffe 2009;Jaspers and Rosiers 2019).García (2009: 84) argues that 'attitudes, values, and beliefs about language are always ideological, and are enmeshed in social systems of dominance and subordination of groups, relating to ethnicity, class, and gender'.Hence, there are no apolitical or neutral ways to manage languages in a society.In the next section, we consider language ideologies and how these are expressed in semiotic processes.

MONOGLOSSIC AND HETEROGLOSSIC LANGUAGE IDEOLOGIES
The study of language ideologies is concerned with the 'linkage of microcultural worlds of language and discourse to macrosocial forces' (Kroskrity 2000: 3).Conteh and Meier (2014: 4) claim that 'which languages are taught, and through which languages content is taught … in schools are based on sociopolitical discourses and ideology'.In other words, language ideologies define which languages are prestigious and valuable.Often the valuable languages belong to the dominant groups of society, while the languages of minorities and suppressed groups are disvalued (Flores and García 2014), as we have already seen from the historical management of linguistic diversity in Norway and Zambia presented above.Yet it is important to note that on a societal and an individual level one will not be able to identify only one single ideology (Kroskrity 2000).Rather, everyone is influenced by multiple discourses and will display different language ideologies depending on time and place.Thus, research participants frequently align themselves with different ideologies over the course of a single focus group discussion (Iversen 2019).García (2009: 120) distinguishes between two competing theoretical frameworks regarding multilingualism in education: Educational programmes founded on monoglossic language ideologies and educational programmes founded on heteroglossic language ideologies.Whereas the first language ideology only considers linguistic practices enacted by monolinguals to be legitimate, the second language ideology embraces the fluid and dynamic linguistic practices of multilingual communities.In accordance with this division of language ideologies, Zambia's transitional programme from local language to English is influenced by a monoglossic language ideology.Similarly, Norway's transitional programmes for newly arrived students and limited opportunities for linguistic minorities to develop their community languages are also founded on a monoglossic ideology.However, research suggests that educational systems founded on monoglossic ideologies do not necessarily produce monoglossic ideologies (Iversen 2019;Palmer 2011) and practices among its teachers (Bailey and Marsden 2017;Jaspers and Rosiers 2019).
Which language ideology that gain political support in a given context is determined by the particular language ecology, including historical and political developments (Kroskrity 2000).As with most other European countries, Norway was heavily influenced by an ideology of nation-building throughout the 19th century, extending far into the 20th century (Engen 2010).As part of this process of nation-building, it was necessary to establish the idea of one nation, one language.This led to the implementation of monoglossic policies, which consequently suppressed linguistic minorities, such as the Sámi (Engen 2014).However, during the colonial rule in Zambia, the British colonisers were not particularly interested in nation-building (Albaugh 2014).Thus, their language policies were more heteroglossic compared to the monoglossic policies adopted after independence.Following Zambia's independence, the government was preoccupied with the idea of nation-building under the slogan 'One Zambia, One Nation' and English was considered to be the most functional language for consolidating the new-born Zambian nation (Chimbutane 2012;Simwinga 2014).Based on the historical developments in the two countries, it is relevant to explore which language ideologies pre-and in-service teachers express in Norway and Zambia respectively, and in what way they converge and diverge from each other.Irvine and Gal (2000) identify three semiotic processes involved in language ideologies.First, the process of iconisation involves the process where certain linguistic features or characteristics are depicted as a social group's inherent nature or essence.Second, fractal recursivity describes either the process of projecting differences between groups based on linguistic features or the process of uniting subdivisions into supercategories against new oppositions.Finally, erasure is the semiotic process where language ideologies are applied to simplify linguistic realities, either by being ignored or by being actively removed.Hence, monoglossic or heteroglossic language ideology in education will be expressed and exercised through processes of iconisation, fractal recursivity, and erasure.
Pre-and in-service teachers' language ideologies are developed in their particular context; in a particular language ecology.In the analysis of the focus group discussions, we linked the participants' utterances to either monoglossic or heteroglossic language ideologies and analysed how the particular language ideologies were connected to the wider language ecologies of the two countries.In the next section, we elaborate on how this analysis was conducted.

METHODS AND MATERIALS
In the present study, the authors compare the language ecology of Norway and Zambia, and the findings from two qualitative research projects on pre-and in-service teachers' language ideologies, from the respective countries.An important rationale behind comparative research is that it provides researchers with new knowledge not only about other contexts, but also about one's own (Bray 2014).By contrasting the language ideologies of Norwegian pre-service teachers and the Zambian in-service teachers, we hope to accentuate insights from in-service teachers in the Global South that can benefit in-service teachers in the Global North.
The data from the Norwegian project consists of seven transcribed focus group interviews with 24 pre-service teachers from two teacher education institutions in Norway.The pre-service teachers had a Norwegian language background and used Norwegian as a home language in their upbringing.They had participated in field placement in primary schools (n=6) characterised by linguistic diversity just before or while the focus group interviews were conducted.The focus group interviews were conducted in Norwegian and transcribed and translated by the first author.
The data from the Zambian project consists of focus group interviews with 36 in-service teachers working in primary schools (n=10) characterised by linguistic diversity.The focus group interviews were conducted in Nyanja, and transcribed and translated by the second author.The fact that this study compares Norwegian pre-service teachers to Zambian in-service teachers is of course a limitation to this study.The Norwegian pre-service teachers had much less teaching experience compared to the Zambian in-service teachers.Nonetheless, we believe that the convergent and divergent patterns in the reported language ideologies between the two groups contribute new insights into the field of language ideologies and multilingualism in education.
After the focus group interviews had been conducted, the transcripts were object to a qualitative content analysis, through a process of meaning condensation (e.g.Brinkmann and Kvale 2015: 233-235).In line with this approach, we first identified statements in the transcripts relevant for the research question: Through a thorough reading of the transcripts, utterances expressing certain views on multilingualism in education were identified.Next, we restated the content of the different meaning units as simply as possible (Condensation).Each statement was assigned a code according to what Brinkmann and Kvale (2015: 228) describe as 'data-driven coding'.This implies that 'the researcher starts out without codes and develops them through readings of the material' (Brinkmann and Kvale 2015: 228).After the relevant units had been coded, the codes were clustered into three main categories: Comments on multilingualism as a challenge in education, comments on solutions to the challenges associated with multilingualism, and comments on language policies in education.
In order to facilitate a comparative analysis, the researchers agreed on certain categories that captured the codes that had been developed separately.Following the coding and categorisation of the two data sets, certain patterns from the two contexts emerged, making it possible to identify convergent and divergent patterns in the two data sets.
The Norwegian Centre for Research Data approved the Norwegian study.In line with their recommendations, all of the participants have been given pseudonyms and the names of schools have not been mentioned in order to ensure the participants' anonymity.Furthermore, the participants freely consented to participation and were informed about their right to withdraw from the study.The same principles were followed in the Zambian study.This study was approved by the Humanities and Social Science research Ethics Committee at the University of Zambia.Further consent was obtained from the Permanent Secretary from the Ministry of General Education and from the District Education Board Secretary.

FINDINGS
The content analysis revealed three main themes in the participants' conversations about multilingualism in education.First, they described the challenges they faced when working in multilingual classrooms.Second, they discussed solutions to these challenges, and finally, they provided comments on current language policies in place.Although there were differences in language ecology, including historical developments and language policies, both the pre-and in-service teachers described similar challenges associated with multilingualism in education.However, their solutions were quite different.In addition, their comments on language policies also diverged.Below, we first describe how their comments converged or diverged and relate them to Irvine and Gal's (2000) three semiotic processes of language ideologies.We then discuss the language ideologies in light of the two countries' language ecologies.Finally, we consider implications of our findings in the final section of this article.

Describing challenges
Both the Zambian in-service teachers and the Norwegian pre-service teachers described several challenges associated with multilingualism in education.In particular, both groups mentioned issues related to limited proficiency in the language of instruction and emphasised the need for a common medium of communication in the classroom:

Masiye (Zambia):
There are other pupils like that one who just came from Congo and do not know any English or any Zambian languages.He only speaks French and kiSwahili, so it is hard for him to interact and learn from others.Unfortunately, I have scant knowledge of kiSwahili and French, I just know numbers and a few words which are not enough for teaching.Sofie (Norway): I also agree that before any progress has been made with the Norwegian language; before they have learnt anything at all, it's difficult to consider multilingualism only as something positive.Because it's challenging to have students who don't know the language in the classroom.
In the examples above, one can see that both Masiye and Sofie agreed on the challenges associated with multilingualism.
However, Masiye regretted his limited proficiency in the languages of his students, while Sofie expected the students to have sufficient proficiency in Norwegian before starting in her class.In both countries, some of the participants felt that students should be proficient in the language of instruction before they could be included into mainstream education.In the case of Zambia, the inservice teachers complained about the lack of suitable learning material and requested some form of support from the government.One teacher expressed frustration over the government's lack of attention towards multilingual students in mainstream education:

Mwangala (Zambia):
There is no deliberate policy by either the government or this school to help children that speak other languages to learn the language of literacy instruction.
Because of this lack of support from the government, the Zambian in-service teachers had to solve the challenges in the classroom themselves, while the Norwegian pre-service teachers pointed to other resources already available in many schools, such as mother tongue teachers, special education teachers, and different introductory programmes:

Madeleine (Norway): If their Norwegian had been so bad that they couldn't understand anything, they I would've-No. I would've gone to the principal and said that they had to-that they had to establish a class in basic Norwegian or something. Because I can't have them in my class if they don't understand anything and I can't understand them. That doesn't work. That's why we have differentiated instruction and special education and things like that. It's to take care of the students that the teacher can't handle in the classroom.
In this extract, Madeleine is arguing for an erasure of multilingualism from her classroom, since multilingualism does not fit her iconic classroom (e.g.Irvine and Gal 2000).In the Norwegian focus group interviews, the pre-service teachers tended to iconise a classroom where all students understood each other and could communicate effectively with each other through the medium of Norwegian.Furthermore, through fractal recursivity, they drew a clear line between the Norwegian 'us' and the multilingual 'other', as evident from Madeleine and Sofie's comments.In the Zambian focus group interviews, the in-service teachers also iconised a classroom where all participants could communicate effectively.Nonetheless, they were open to engage with the multilingualism present in the classroom to a greater degree than the Norwegian pre-service teachers.As evident in the next section, the Zambian in-service teachers refrained from fractal recursivity as opposed to the Norwegian pre-service teachers.

Describing solutions
There was a distinct difference in how the Zambian in-service teachers and the Norwegian pre-service teachers described solutions for the challenges associated with multilingualism.In case of the Zambian in-service teachers, most of them seemed to hold a pragmatic attitude, where the main objective was for the students to understand the content:

Mwangala (Zambia):
What is important is that pupils understand what they need to learn using a language that they know and appropriate methodologies.
From this statement, one can see that Mwangala's focus is on the students' learning, rather than the particular language used.In fact, many of the Zambian in-service teachers described how they frequently drew on several languages within their own language repertoire in the classroom: Belita (Zambia): Sometimes we use different languages, like Bemba, Nyanja, English and others that are known to me as a teacher.Provided there are learners in class that speak those languages.

Masiye (Zambia): This one [student]
just came from Chingola and when I deliberately use Nyanja throughout without switching to Bemba, she would not understand anything.So I have to switch to Bemba and emphasise the task that she has to do.
The teachers demonstrated an openness to draw on a wide repertoire of their linguistic resources in order to facilitate students' learning.In cases where the teachers were not proficient in the languages in question, they would not hesitate to involve other teachers, parents, or other students in the class to make sure that the student could comprehend the content: Mwangala (Zambia): I had a child who was speaking Tonga in Term 1, she did not understand the Nyanja language of instruction, and I did not know the Tonga language either.Whenever this child spoke or asked a question, I would go to a certain teacher within the school to help me interpret what the child was saying, and eventually, this helped the child learn.
In this quote, Mwangala displays a flexibility to meet the needs of the Tonga-speaking student and use the resources available at the school in order to support this particular student.Furthermore, Mwangala avoided any fractal recursivity by describing the classroom as a multilingual space, where different languages are frequently used as a natural part of her teaching practices.
In case of the Norwegian pre-service teachers, there was greater variety within the group about the degree to which they were open to involving students' languages into their teaching.They all agreed that their limited proficiency in students' languages prevented them from using those languages themselves.Moreover, they considered it an almost absurd idea that they should learn languages, such as Polish or Somali: Nora (Norway): You can't just tell the teacher, well, you need to learn Somali.That doesn't work [laughs].Madeleine (Norway): When they move to Norway, we are a Norwegian country, we speak Norwegian here.If I had moved to a different country, I would've-I couldn't expect them to speak my language.I would have to change, I would have to adapt to the new country I had arrived in.So, I feel like those who come to Norway also need to adapt to the Norwegian.
As one can see from Madeleine and Nora's statements, the pre-service teachers did not think that it was their responsibility to learn the languages of their students; rather they expected the students to assimilate.If implemented as classroom practices, this would contribute to the erasure of languages beyond Norwegian (e.g.Irvine and Gal 2000).One group was asked about the potential use of different languages in communication with parents.They expressed hesitancy to use languages they were not sufficiently proficient in when communicating with parents.Yet one of the participants suggested that she would be willing to contact parents in English, if necessary.She explained: Madeleine (Norway): Well, I feel that it's easier because English is like the language.It's the world language.
Madeleine (Norway): English is my limit.And Swedish.Swedish, English and Norwegian [laughs].
Harald (Norway): Yes, I agree with that.Because… No. English has to be my limit for my work as a teacher.
This brief exchange illustrates how the Norwegian pre-service teachers also expected parents to adopt to the language norms of the classroom.Bjarne's laughter can be interpreted as an opposition towards Madeleine's restrictive approach to multilingualism, although he does not voice any verbal support of any particular position.Nonetheless, there was an agreement among the Norwegian pre-service teachers that students could be allowed to use their own languages for learning as long as it did not affect others in a negative way.For example, they generally agreed that students could use a wide repertoire of their linguistic resources for information searches online, taking personal notes, and sometimes to discuss content with peers in languages other than Norwegian.However, this should not be allowed to affect the teacher or other students in class (Iversen 2019).This reflects previous studies from Norway that shows how teachers are able to create spaces for multilingualism despite monoglossic ideologies (Beiler 2019;Danbolt and Hugo 2012).

Comments on policies
There was a striking difference in how the two groups positioned themselves regarding the language policies in education in the two different countries.On the one hand, the Zambian in-service teachers repeatedly criticised the current language policies in Zambia.With regard to the authorities' strict regulation of literacy instruction, several teachers mentioned this as an issue in relation to the support of multilingual students: Tamara (Zambia): I recommend that teachers should not be restricted in terms of the method to be used to teach literacy in Grade 1.We should be allowed to choose or mix methods for teaching literacy so that the weaknesses of one method is made up for by another to help pupils read.
Belita (Zambia): All pupils are treated the same in class.This is why sometimes I combine teaching methodology and classroom organisation with that of the NBTL, where we put pupils in ability groups.

Soko (Zambia):
Actually we have a problem on teaching and learning materials.Because I have 61 pupils in class, but only three copies of the pupils' book for the whole term and we do not have a teachers' handbook at this school.
These statements obviously refer to the general limited resources in many Zambian schools, but the Zambian inservice teachers also link this to their struggle to provide multilingual students with quality education.As authorities request teachers to follow a strict programme for literacy instruction, teachers argue that they have to use different methodologies due to the large number of students, limited resources, and different language backgrounds.Several participants admitted that they did so 'illegally', yet they found it necessary in order to support students' learning.
On the other hand, the Norwegian pre-service teachers tended to express support of current policies and frequently referenced official policies in support of their own practices.In one focus group interview, the participants discussed the language rights awarded the indigenous Sámi language, one participant said: Thora (Norway): And then Sámi is supposed to be on equal terms with Norwegian.It says so in some laws and things.That they are supposed to be on equal terms and [the Sámi student] has the right to instruction in Sámi.
Hence, in this statement, the iconic classroom had room for the Sámi minority and this minority seems to be included in the wider 'us'.In another focus group interview, the pre-service teachers used the same justification for the differentiation between the Sámi language minority and other language minorities within Norwegian education: Josefine (Norway): It says so in the law, kind of, that they are entitled to [instruction in their mother tongue] and that the same rights do not apply to those with another mother tongue.
As one can see from these examples, the Norwegian pre-service teachers tended to have great confidence that the Norwegian Education Act was fair.Moreover, they were certain that they would receive the support from various actors within the Norwegian education system if they should receive multilingual students in their classes in the future, such as mother tongue teachers, special educators, and basic Norwegian instructors.This is contrasted with the Zambian in-service teachers' absent confidence and expectations towards their own authorities to provide them the support they needed.In the following, we discuss the participants' comments on challenges, solutions, and policies related to multilingualism in education in light of language ideologies and language ecologies.

DISCUSSION
The present study set out to investigate the language ideologies of Norwegian pre-service teachers who have participated in field placement in primary schools with a linguistically diverse student population and of Zambian in-service teachers teaching in primary schools characterised by great linguistic diversity.Despite the differences in language ecologies, including history and education systems in Norway and Zambia, this comparative study of the two research projects has shown that there are several similarities in the language ideologies among Norwegian pre-service teachers and Zambian in-service teachers.Particularly, the two groups provided similar descriptions of the challenges associated with multilingualism in education.These descriptions presented an iconic classroom where all participants were able to communicate effectively with each other.However, there was a clear distinction in how the Norwegian and Zambian participants envisioned how this classroom communication should take place.
In the case of the Zambian in-service teachers, they conceived of the iconic classroom as a multilingual classroom where teachers would be able to draw on a wide repertoire of their linguistic resources in order for communication to function effectively.Moreover, they would like to have access to teachers with proficiency in other languages.In other words, the Zambian in-service teachers described an iconic classroom where languages were used in a flexible manner in order to support students' learning.Although the language policy (Ministry of Education, Science, Vocational Training, and Early Education 2013) recommends the use of a regional language for classroom instruction, some teachers were using multiple languages in multilingual classes especially when they knew and understood the languages spoken by their students.This semiotic practice aligns with what García (2009) describes as a heteroglossic language ideology, which implies an embrace of the fluid and dynamic linguistic practices of multilingual communities.Nonetheless, this language ideology is in conflict with current language policies for education in Zambia (e.g.Mwansa 2018;Ministry of Education, Science, Vocational Training, and Early Education 2013;Tambulukani 2015).This tension leads the Zambian in-service teachers to criticise the current language policies in education and call for change.
The Norwegian pre-service teachers generally described an iconic classroom where students and teachers were able to communicate effectively through the medium of Norwegian.In their discussions, they developed a fractal recursivity that contributed to contrast the Norwegian-speaking students with multilingual students with a migrant background.If the practices they described were to be implemented as classroom practices, they would most likely lead to the erasure of the multilingualism present in the classroom.These semiotic processes in the Norwegian pre-service teachers' discussions aligned with a monoglossic language ideology (e.g.García 2009), where only the linguistic practices enacted by monolinguals are considered legitimate.Since such a language ideology to a great extent is reflected in current language policies in education in Norway (e.g.Chinga-Ramirez 2017; Engen 2010), there was no need for the Norwegian pre-service teachers to criticise policies as the Zambian inservice teachers did.Hence, no such opposition was reported in the focus group interviews.
There can be many reasons why the Zambian in-service teachers displayed a greater openness to include a wider repertoire of students' linguistic resources in the classroom.It can be a question of classroom experience.As teachers who have considerable teaching experience, they might be more pragmatic when it comes to supporting their students or more flexible in their teaching approaches.However, the Norwegian pre-service teachers' language ideologies are also reflected in previous studies of in-service teachers' beliefs or attitudes about multilingualism from Norway (Haukås 2016;Kulbrandstad 2015;Pran and Holst 2015).Alternatively, it can be a question of language competence.The Zambian in-service teachers demonstrated a more varied linguistic competence in a number of languages, while the Norwegian pre-service teachers' more restricted language repertoires also contributed to limit their opportunities to engage with the multilingualism present in the classroom.Nonetheless, we would argue that the Zambian in-service teachers' openness should be linked to language ecology.
On the one hand, the Zambian inservice teachers' heteroglossic language ideologies should be understood as a response to Zambia's multilingual reality, where people interact across language boundaries on a daily basis, and are used to engage in translanguaging practices in order to secure comprehension (Banda and Jimaima 2017;Banda, Jimaima, and Mokwena 2019;Makalela 2016).This language ecology gives rise to more heteroglossic language ideologies, where communication through the medium of various languages is considered appropriate and acceptable.On the other hand, the Norwegian pre-service teachers were reluctant when it came to the idea of using other languages than Norwegian (and English).Influenced by what Kulbrandstad (2015) describe as a 'monolingual ethos' in Norway, the pre-service teachers expected students to adapt to the current language policies and did not question current policies.

CONCLUDING REMARKS
The findings indicate that the Zambian in-service teachers are more open to including students' multilingualism in the classroom than the Norwegian preservice teachers.While the Zambian in-service teachers display a pragmatic ideology, where the main purpose of education is for students to learn regardless of language, the Norwegian pre-service teachers are concerned with students' proficiency in the language of instruction, Norwegian.Consequently, in order for their ideologies to be reflected in educational policies, the Zambian in-service teachers see a need for policy extensive reforms.Whereas the Norwegian pre-service teachers tend to support and defend current monolingual educational policies.
These findings suggest that there is much to be learnt from the Zambian inservice teachers' openness to engaging with multilingualism.Their willingness and reported ability to draw on their own and their colleagues' linguistic repertoires can serve as an example for teachers in other contexts.The Norwegian pre-service teachers can also learn from the Zambian in-service teachers' critical attitude towards the current situation in their education system.All teachers should be able to assess education policies and practices in a critical way.At the same time, there is need for more resources and government support to enable Zambian teachers to provide all students with quality education.
Educational authorities should listen to the Zambian in-service teachers' insights and their multilingual teaching practices should be encouraged and supported.