‘Refugees here and Finns there’ – categorisations of race, nationality, and gender in a Finnish classroom

ABSTRACT Schools represent a central meeting place where societal inequalities are reproduced and questions of social justice become important. This study focuses on categorisations related to race, nationality, and gender in interactions in Finnish teaching environments, as well as teacher reflections on these situations. We discuss the implications of the categorisations on social justice and the role of the teacher in these situations. We conducted video observations of a sixth-grade teacher in a Finnish primary school. The study employs both critical multicultural education approaches and Conversation Analysis. Results show that the pupils use categories race, nationality, and gender in ways that limit the agency and positioning of some of the pupils. The extensive and intersecting categorisation in teaching situations makes it demanding for teachers to address and challenge unequal norms attached to the categories. Results also indicate that teachers need an understanding of othering and normativity in order to allow spontaneous critical discussion and problematising categorisations that pupils use. Also, the results highlight the importance of involving pupils in the process of questioning norms that do not provide all pupils with the same agency or sense of belonging.


Introduction
Equality and equity are foundational pillars of the Finnish education system. Education with the same quality is guaranteed free of charge for all pupils, and achievement differences are relatively small between schools and municipalities (OECD 2016). However, the notion that all pupils are equal does not guarantee equal treatment on a school or individual level (Juva & Holm 2017;Riitaoja 2013). Several studies have shown that school is a place of reproduction of difference between the 'Finns' and 'immigrants' (Lappalainen 2009;Juva & Holm 2017;Riitaoja (2013) ;Souto 2011), as well as between boys and girls (Gordon, Holland, and Lahelma 2000;Lahelma 2014;Öhrn 2019), which limits opportunities and reproduces inequality among different groups of pupils.
The aim of this paper is to analyse categorisation related to race, nationality, and gender in teaching situations and subsequent teacher reflections on teaching, and to discuss their implications for social justice among the pupils and the role of teachers in these situations. We understand social justice as a core aim of education, comprising both equal access to quality education, full participation, and relevant education for all pupils (cf. Ayers, Quinn, and Stovall 2009, p.xiv). First, we introduce critical multiculturalism and Conversation Analysis (CA). Second, we explain the data and method used: video observations and stimulated recall interviews analysed with CA. Third, we explain the results of analysing categorisation related to race, nationality, and gender, before addressing the implications for teachers and teacher education.

Theoretical framework
The analytical framework for this study is both critical approaches to multicultural education (Gorski 2008;May and Sleeter 2010), and Conversation Analysis (CA) (Kitzinger 2000;Schegloff 2007). Critical multicultural education was developed as a response to, and critique of, conservative and liberal multicultural education that has an asymmetrical, othering, or even deficit perspective regarding non-dominant groups (McLaren and Ryoo 2012). It derives from approaches such as antiracist education, critical race theory, critical pedagogy, and postcolonial theory (May and Sleeter 2010). McLaren and Ryoo (2012, 66) argue that: 'Critical multiculturalism moves beyond essentializing concepts of 'sameness' or 'difference' -emphasising how sameness or difference are produced between or among groups and must be understood in terms of the specificity of their production -and stresses the need to transform social, cultural, and institutional relations that produce asymmetrical relations of power based on a politics of difference or identity in politics (as opposed to identity politics).' In line with critical theories behind critical multiculturalism, as well as nonessentialising views of identity in CA, we regard all three central categories in this study -race, nationality, and gender -as social and political constructs. Previous studies have shown the relevance of adopting a critical multiculturalist approach to education in Finnish schools. Race has been, and is still, used to make sense of physical differences in an othering and derogatory way, both on an individual and societal level (cf. Lentin 2008). In (Juva & Holm 2017) Finnishness was constructed as normality in a seemingly neutral way, but still constructing those with migrant backgrounds as others. In Finland and Finnish schools, there is a persistent denial of racism (Alemanji 2016), and race is often hidden behind labels that seem less problematic, such as culture or nationality. All three concepts are nevertheless tightly connected and the construct of race is sustained by the idea of a nation state with people of one ethnicity and one uniform culture (Lentin 2008). In the Finnish context, Finnishness is often associated with ethnicity and whiteness (cf. Lappalainen 2009;Rastas 2009). In her ethnographic study, Lappalainen (2009) found that Finnishness was more connected to skin colour than to citizenship or language for pre-school children. Children perceived as immigrants who wanted to claim Finnishness had to do this repeatedly. Further, Lappalainen (2009) found that gender and ethnicity intersect, since non-white boys encountered more racism than non-white girls, and she argues that 'white Finnish masculinity operates in a defensive way against non-white masculinities, which are defined as "other"' (ibid., p.72). In their study of Roma mediators in Finnish schools, (Helakorpi, Lappalainen and Sahlström 2019) also found that Roma mediators had to make themselves tolerable in the eyes of the non-Roma, in order to promote tolerance towards Roma in general. Lahelma (2011) points out that talk about gender is muted in order to seem 'gender neutral', but Finnish education is gendered in many ways. Boys' achievement is remarkably lower than girls', which in turn fuels a discourse of natural gender binary (Lahelma 2014). Girls encounter more control of their embodiment and use of space and voice in school than boys, hence limiting their agency (Gordon, Holland, and Lahelma 2000;Gordon 2006). We define agency as the perceived capacity to carry out decisions and actions (cf. Gordon 2006). Huuki (2010) found that boys, using humour, often move on the border between violence and caring or entertainment to maintain their status in school. In their interview study from Sweden, Eliasson, Isaksson, and Laflamme (2007) found that verbal abuse worked as a resource for boys to a higher degree than for girls. Verbal abuse was directed from boys to boys, or from boys to girls, to maintain hegemonic masculinity. Also, Lahelma and Öhrn (2017) state that being the ''right kind of boy includes pressure to act tough and bullying others.
These studies describe how inequalities based on race, ethnicity, nationality, and gender are reproduced in Finnish schools, and thus play a substantial role in pupils' agency and positioning. These studies also show that critical multicultural education has not been implemented sufficiently in Finnish schools. Many Finnish teachers embrace a colour-blind approach (Mansikka & Holm 2011), leaving issues related to race, nationality, gender, and sexuality unaddressed. Also, the multicultural education taught in Finnish teacher education is mostly conservative or liberal, addressing issues of social justice superficially (Hummelstedt-Djedou, Zilliacus & Holm 2018). Scholars of critical approaches criticise liberal approaches of multicultural education for focusing on celebrating diversity, which means essentialised differences and unequal structures are reproduced instead of challenged (Gorski 2008;May and Sleeter 2010). Multicultural education is also often understood as only intended for immigrants or 'the culturally different' (Nieto and Bode 2018;Riitaoja 2013;Hummelstedt-Djedou, Zilliacus & Holm 2018). Critical approaches have in common that they ask teachers and students to critically examine the power relations and structures in society and everyday life, and to act for social change (May and Sleeter 2010;Sleeter 2015;Nieto and Bode 2018). Critical approaches have an intersectional perspective (de Los Reyes, Molina, and Mulinari 2006), in which several socially constructed categories affecting people's positions and relations to power are taken into account simultaneously (May and Sleeter 2010). The gap between the aims of critical multicultural education and existing inequalities in the Finnish school shows a need for a closer investigation of what happens in classrooms regarding crucial categories such as race, gender, and nationality, and what the teacher does and could do in these situations.
We approached this task by using Conversation Analysis (CA), which has developed from ethnomethodology (Garfinkel 1967;Sacks 1992). CA studies naturally occurring talk-in-interaction and analyses, turn by turn, the social actions of speakers and recipients. Within CA, the everyday micro-organisation of human sociality is the core focus and is considered as constitutive of societal institutions and structures (Kitzinger 2000;Schegloff 1997Schegloff , 2007. Kitzinger (2000) and Stokoe (2006) argue for the compatibility of feminist research and CA, and show how issues like power and oppression can be studied as something that is maintained and distributed by the participants in interaction. Both critical multiculturalism and CA rely on a nonessentialist account of identities and identity categories (Antaki and Widdicombe 2008;Kitzinger 2000;Stokoe 2006;May and Sleeter 2010), where they are seen as fluid, and as 'continually reconstructed through participation in social situations' (May and Sleeter 2010, 10). Within CA, identity categories can be described as resources, which the participants use in the interaction to accomplish the actions they want (Stokoe 2006).
We rely on the compatibility of critical multicultural education and CA to empirically analyse what categories are made relevant in a particular context and situation, and how they are used in the interaction to enable and constrain certain actions and belongings that affect social justice (cf. Kitzinger 2000;Stokoe 2006). Another study that has used CA to study categorisation and social justice is Whitehead's (2012) study on how, in post-apartheid South-Africa, racial categories were used as resources or constraints in interaction.

Data and method
The data consist of four days of video observations of teaching in year 6 at a primary school in the capital area in Finland and a stimulated recall interview (Dempsey 2010) with the class teacher who had specialised in multicultural education as a minor during his studies to become a primary school teacher. Approximately half of the students had a migration background and Finnish as a second language, while half of them (including the teacher) were white, ethnic Finns. The mother tongues of the pupils in the class that had Finnish as a second language and that were registered at the school were Somali, English, Persian, Kurdish, Arabic, Russian. This heterogeneous group of students had in common that they become racialised in different ways and were regarded as not being Finnish.
The observations were done with a camera and a wireless microphone attached to the teacher. The focus was mostly on the teacher and the interactions around him, but on some occasions, on the whole class. Field notes with exact time information were conducted during the observations, with a focus on categorisations relevant for the realisation of social justice in the classroom. The stimulated recall interview was done with the teacher a week after the video observations. Five situations where categorisation relevant for social justice took place were selected and shown to him to comment on. Two of these situations were then transcribed for further analysis, since they were representative of the intersectional categorisation going on in the focus class, as well as provided a possibility for thorough analysis of the actions made by the categorisations.
The analysis focuses on identity categories made relevant by the participants in the data, and how they are ascribing them to themselves or others when accomplishing an action (Antaki and Widdicombe 2008). Only categorisation related to race, nationality and gender has been included in the analysis.
Within this body of data, we have focused on who is categorising whom, the actions triggered by the categorisations, and what responses the categorisations receive from the teacher and the classmates. The teacher's reflections on the situations were analysed in the same way.

Ethical considerations
The video observations focused on the teacher, but they also captured interactions between the pupils. We first asked permission from the teacher for the video observations, and then asked permission from the education department in the city, the principal of the school, the pupils, and their parents. Pupils and their parents were informed beforehand about the project by the teacher and in a letter accompanying the consent form. We gave them three options: to not participate at all, to participate with restrictions (with the video material only being used for transcriptions but not shown outside the project), or to participate with full permission, with the video material being used in research settings including publication and teaching at the university. The teacher agreed that we could use the video material in research settings and teaching. Four of the pupils in his class did not want to participate, while the rest, 21 pupils, gave either restricted or full permission. We filmed the teacher and surrounding pupils and avoided filming the pupils who had said no. Situations in which they happened to move into the picture were not used for analysis. We used code names in all transcripts to protect the identity of the pupils and the teacher, as well as the name of the school.

Results
In what follows, we present the results of the analysis of the two chosen situations with categorisations related to race, nationality, and gender. These categorisations are discussed from an intersectional perspective (de Los Reyes, Molina, and Mulinari 2006).

The refugees here and the Finns there -dividing teams by orienting to race
The first situation took place during a PE lesson when the class was about to play football, and gender and race ended up being relevant categories for team allocation. The teacher Mika started organising the game by asking everybody to divide into groups, depending on whether they wanted to play at 'full speed' or 'easy-going', and to then go and stand in different parts of the football field. Eight girls and five boys went to the 'full speed'-group, and six girls and one boy went to the 'easy-going'-group. In the 'easy-going' group (six girls and one boy) the teacher Mika orients to gender as a relevant category when he asked students to form pairs and then assumed that the six girls would create pairs with each other, and that the one boy Markku would be the third person in one team.
When Mika started to divide the teams in the 'full speed' group, he relied on two groups of girls, one with white ethnic Finnish girls and one with racialised girls. He did not ask each girl which team she preferred, which he did do with most of the boys. This can be seen as an act of free choice for the girls, since they formed the groups themselves. However, they were not given the possibility to choose, and race and gender were taken for granted as categories that affected the creation of teams. Mika then asked (Supplementary Excerpt 1a, line 1) the white girl Krissu, who was standing on the side of the three racialised girls, to go to the group of white girls, and the racialised girl Tamara, also standing next to the three racialised girls, to go to the group of racialised girls. Thus, both gender and race were (non-verbally) used as relevant categories for team composition. As can be seen in lines 2-3 (Supplementary Excerpt 1a), Mika asked the racialised boy Faysal which group he wanted to be a part of, giving him the agency to choose, with respect to both gender and race.
When Faysal was about to choose, Petteri commented 'the Finnish there (.) the refugees there' (Supplementary Excerpt 1a, line 5), highlighting that one group consisted of racialised pupils -'the refugees' and the other white ethnic Finns -'the Finns'. Another pupil, Beni, who was racialised himself, responded to this by asking 'what refugees' and lifted his hand, saying 'we're not racist' (Supplementary Excerpt 1a, line 6). By stating this he rejected the use of the term 'refugee' and claimed that the 'us' was not racist, implying that calling somebody refugee would be a racist categorisation. At the same time, Faysal chose the group of racialised pupils where he was already standing. Beni went on rejecting that a particular pupil would be a refugee (Supplementary Excerpt 1a, line 8). The teacher Mika gave the next boy Tommi, a white ethnic Finn, the freedom to choose (Supplementary Excerpt 1a, line 7), but Tommi asked him for help since he didn't know where to go, physically standing in the middle of the two groups taking form. In contrast to the girls who were put in teams, Mika emphasised that Tommi had a free choice by saying: 'you can go wherever just (.) decide (.) just (.) you can freely choose'. Right after Petteri commented 'the Finns there', and Tommi, categorised as Finnish in the categories oriented to by Petteri, went to the white, 'Finnish', group (Supplementary Excerpt 1a, line 13). The way Tommi at first hesitated about what group to choose, and then chose, after Petteri's comment, suggests that the comment made by Petteri affected his decision.
Mika then asks Petteri where he wanted to go (Supplementary Excerpt 1b, line 14). He chose the 'Finns' group of white pupils, where he was already standing, without hesitation. Mika asked Komppu, who was racialised, where he wanted to go, and in the middle of Mika's question Petteri commented again: 'refugees' (Supplementary Excerpt 1b, line 17). Komppu answered 'wherever' and Mika told him to go to the group where the racialised pupils stood, again without explicitly mentioning race but also not resisting the categorisation proposed by Petteri. Komppu, who was standing close to the 'Finnish' group, then went to the other side. For the last two boys, the situation was the same as for the two girls Krissu and Tamara. Mika told Balavdi (Supplementary Excerpt 1b, line 19) to go to the group where the racialised pupils stood, and Beni to go to the group where the white pupils stood. Here, he suggested a division that did not follow categorisation by race, since Balavdi was white and Beni was racialised. In contrast to the boys, these two did not have the possibility to choose for themselves. The division of teams was completed and while everybody moved to their team, Petteri commented once more 'REFUGEES' (Supplementary Excerpt 1b, line 27). Overall, it seemed that although race was verbally foregrounded by just one pupil (Petteri), such categorisation restricts the agency of choosing one's team for several pupils. This can also be regarded as a way for Petteri to have reinforced his position in terms of hegemonic masculinity by commenting on identities of others (Eliasson, Isaksson, and Laflamme 2007) and behaving defensively towards non-white masculinities (Lappalainen 2009) by calling them 'refugees'. Gender was used non-verbally as a relevant category by both pupils and the teacher, and therefore also limited the options for both Markku and the other girls in the 'easy-going' group, and for Krissu and Tamara in the 'full speed' group. Although the last two boys did not get to choose their teams, overall the girls had fewer options to choose than the boys, which is in line with Gordon's (2006) analysis, in which the agency of girls is more controlled.
In the stimulated recall interview, we asked Mika about the pupils using the categories Finns and refugees. He told us that they had been doing this for some weeks, and that he had intervened earlier but that he now ignored it, since he had become accustomed to it. He focused more on getting teams that were evenly skilled. There had been an earlier situation where he had divided them into teams and a pupil had commented that 'the guy divided us like refugees vs Finns'. Mika had answered that he had 'obviously' done this and asked if it was a problem and proposed to make new teams. However, the pupils did not find this problematic, but rather humorous and somebody had said that it was a 'good joke'. Mika continued by explaining that there were two pupils with an immigrant background 'in the group of Finns, but the contrast between the teams was still quite large'. The use of the word 'contrast', as well as the comment regarding the division into Finns and refugees as 'obvious', also legitimises the use of 'Finns' as a category for the white pupils, and 'refugee' for all racialised pupils. Viewing the categories of Finns and refugees or migrants as fixed or useable, since they were used by the pupils, suggests a logic through which categories that are already fixed and in use do not need further problematisation. Regarding the use of 'refugee' in particular, Mika stated that he thought it came from the media (the data gathering was done in spring 2016 right after the arrival of a large number of refugees), and that nobody from the class had arrived as a refugee. Mika explained that all pupils were Finnish citizens, but not all identified as Finns. Our observations showed that those considered to have a migration background talked about white ethnic Finnish pupils as 'Finns' and about themselves according to their country of origin or their parents' origin. However, most of the white ethnic Finnish pupils did not talk about themselves being Finnish in the same explicit way, which shows how it is a category of norm and privilege (Juva & Holm 2017). Although the teacher described the pupils' identification as Finnish or not as an individual issue, these sports-related situations and other situations during the observation week, indicated that race matters for being able to claim Finnishness (cf. Lappalainen 2009;Rastas 2009;Tuori 2009). Gender was not verbalised and explicitly oriented to, to the same extent as race, but gendered positions of boys and girls still affected the agency and freedom of the pupils. This seems to be an example of how 'gender neutrality' can mute gender impact, which is described by Lahelma (2011) as a general feature of Finnish education.

Why are some not laughing? -gender, nationality and race during a lesson on Russian history
A second situation took place during a lesson on Russian history, which was given by one of the pupils in class, Krissu, who had a Russian background. She brought in a book about Russian tsars and showed pictures of them to the class. One pupil asked why they were so ugly and Krissu showed a picture of a tsar's wife, who she thought was really ugly. This starts a gendered objectifying and othering of tsar wives and girls in the class. Petteri compared the photo with Tamsu, a girl in the class, and said 'looks totally like Tamsu'. Tamsu responded by asking if she was 'white nowadays', referring to the fact that she was non-white. Here she used her racialised position as a resource to reject the claim of looking like someone who had been described as ugly. The 'nowadays' reference refers to the circumstance that she had not been viewed as white before in this class and school context. Krissu continued to explain that the tsar's wives were ugly and several pupils laughed at the photos. The comparison with girls in the class continued when Petteri commented that another wife looked 'totally like Janina'. She responded with a simple 'don't ha ha ha'. Another pupil continued by saying that this tsar's wife was related to Balavdi, a boy in the class. Abdul continued by asking if it was his 'mother's tribe'. Here, nationality was intertwined with gendered categorisation, since Balavdi was ascribed as being potentially related to this woman because of his Russian background. Krissu showed yet another picture and a pupil commented: 'it is Markku's sister'. This shows that to be able to make fun of a boy in the class, the comparison is between his female relative and the tsar's wife, and not directly between him and a tsar. These comments of comparison with tsar wives served to make fun of other students and to give oneself a higher status. At this point, the teacher Mika interrupted the laughing and asked if they felt powerful when they laughed at the pictures. Petteri first agreed by saying yes, and then changed his response to 'no you don't'. Mika continued by stating that not everybody was laughing and asked why (Supplementary Excerpt 2, line 1).
Some of the pupils started answering 'because' and Balavdi stated 'because some here are Russian and they don't want to laugh at Russian women', implying that belonging to a national group would include loyalty by not laughing at pictures representing that group. Krissu, however, immediately contradicted this claim by saying that she laughed at the pictures even though she was Russian (Supplementary Excerpt 2, line 7). So she acknowledged her Russian identity, but did not agree with the ascribed qualities of being Russian. Balavdi responded by calling it all a 'good joke'. Abdul continued by asking if there was 'just one Russian' and Krissu answered 'yes I am' (Supplementary Excerpt 2, line 10). Abdul then asked if there is one Somali and got 'no' and laughter as an answer from several pupils, especially Petteri who said 'NO' in a loud voice. Somebody commented 'Faysal is also' (Supplementary Excerpt 2, line 15), which included the already known information that Abdul considered himself to be a Somali as well. Balavdi stated that 'sometimes there are so many Somalis' and Abdul played hurt by putting his head on the desk and saying something about 'only me Somali' and 'hurt' (the rest was inaudible due to several pupils all talking at the same time). Balavdi responded that 'I didn't say now but sometimes', which indicated that he agreed it could be interpreted as negative to say that there would be 'so many Somalis'. Abdul continued to play hurt and said 'then I won't come to school' (Supplementary Excerpt 2, line 20). Petteri followed by saying: 'Finland is (.) an invaded (.) country' and repeated 'after a year Finland will be invaded'. Since this represented a response to the number of Russians and Somalis, Petteri was insinuating that Finland was being invaded by groups of other nationalities found in the school. Abdul proceeded with 'by the way, it isn't'. Krissu then jumped in and asked if Petteri believed that Russia would invade Finland (Supplementary Excerpt 2, line 26). The conversation went on with pupils comparing natural resources in Russia and Finland, and Krissu continued by showing pictures of clothing and buildings during Russian tsarist times.
When Mika was asked why he interrupted the laughing, he said he wanted to have pupils reflect on why they laugh at people that they think look ugly. He did not comment on the gender aspect of some boys comparing girls in class to tsarist wives. Regarding the comments about numbers of Russians and Somalis, Mika noted that he thought these comments came from the pupils' 'personal history, parents, and media portrayals, partly from school and the kinds of categorising taking place during school lessons. He then stated that a common new culture was needed that could be categorised as its own. He also stated that for new culture to be born, one needed to come to 'terms with one's own', and therefore pupils needed to have the opportunity to categorise themselves. He mentioned that, in his view, self-categorisation was fine, but that other-categorisation was potentially problematic, depending on the situation. He argued that categorisation of others is often used as a tool by the powerful against the weaker, and this made him angry. Mika also mentioned that pupil categorisations often took place without teachers hearing them. When asked about nationality as a relevant category for the pupils, he claimed that it's a bit 'funny', since he didn't think that teachers referred to national categories with the pupils. Mika stated that the most evident group divisions used in school were class divisions into A and B groups, based on Finnish as a first language or Finnish as a second language. When Mika was asked if those who consider themselves as Finns used the category Finn about themselves, he said 'rarely'. He also mentioned that they sometimes discussed political issues and that the 'True Finns' political party had 'succeeded in spoiling their reputation. So [laughs] it's [refers to the category "Finn"], (not) a very hot category either'. The interviewer commented that it is interesting that those who are not considered to be Finns used the category 'Finn' about others, but not many of those considered to be Finns use it about themselves. Mika had not thought about this much and said that perhaps one can get used to it when you are in the class.
In the situation described above, categories of gender, nationality, and race intersected. The manner in which boys commented on girls in the class could be compared to boys using verbal abuse against girls to maintain a hegemonic masculine position (Eliasson, Isaksson, and Laflamme 2007), as well as controlling girls by commenting on what is appropriate (Gordon 2006). The way the discussion continued maintained an understanding of national categories as dichotomous, and an understanding of Finland as homogenous and ethnically Finnish, with Russians and Somalis as potential invaders. As in the first situation, it was the white ethnic Finnish boy Petteri who was active in commenting on racialised pupils and took a defensive stance towards non-white masculinities (Lappalainen 2009). What Petteri did, and other boys also participated in, could also be described as status work by using humour, making it hard to draw the line between entertainment and violence (Huuki 2010). In both, boys were more involved in category negotiations, which can be seen as what is demanded from them in order to claim being the right kind of 'tough' boy (Lahelma and Öhrn 2017). However, the way the teacher ignored the gender dimension, or the question of the exclusive way that belonging to Finland was expressed, shows how such issues easily become implicit in daily school activities.
Although masculinity and Finnishness (whiteness) shine through as privileging norms, the analysis also makes it visible that the gender and nationality categories were used in different ways as both resources and constraints (cf. Whitehead 2012), by the way both Tamara and Krissu demonstrated resistance. Both cases show that though girls were more controlled and commented on, they simultaneously sought agency by claiming both space and voice (Gordon 2006).

Discussion
There is extensive, intersecting and simultaneous categorisation regarding gender, race, and nationality in peer and teaching interaction in the Finnish school class analysed for this study. From the point of view of social justice and critical multicultural education, the implications of the categorisations for pupils' agency, use of space and enabled positions are considerable. The examples show that gender limits the agency of girls when choosing teams, and becomes a resource for control and objectification when comparing girls to ugly wives of tsars (Gordon 2006). They also show that for several boys to maintain or achieve status, frequent categorisation of others is demanded, and that the white male position offers many opportunities to do so. However, while categories of race and nationality are explicitly topicalised in the discussions, gender is more non-verbal and implicit, which might be a result of overall Finnish 'gender neutrality' muting the impacts of gender in education (Lahelma 2011). Racialisation limits the agency of boys when choosing teams, and limits who can take on the position of being Finnish (Juva & Holm 2017;Rastas 2009). Being considered as having a nationality other than Finnish limits belonging to Finnishness and Finland and is even described as a threat. In the examples there is also resistance to certain racial categorisations, as well as redefinition of supposed limitations of national belonging. All these are examples of pupils not only conforming to surrounding discourses but also using the contingency of the discourses to claim agency and becoming agentic subjects (Gordon 2006).
The analysis of the teacher interview shows that he mostly commented on the categorisation from the outside, not reflecting on his own privileged position as a white male. He considered race and nationality categorisation as something pupils should be free to do as long as they did not use it as power over others. He described himself as concerned about using categories that harm others. Also regarding the laughter that took place about tsar wives, he questioned the action of laughing at others. However, what can seem to be harmless and normalised discourse can limit the positions and choices more for some than others (Juva & Holm 2017). The teacher's description was that it was up to the pupils to identify as Finnish or not, which made it seem like a matter of choice. He did not comment on whether it was possible for the racialised students to identify as Finnish in the way Finnishness is constructed at school, closely connected to whiteness (Juva & Holm 2017;Lappalainen 2009) and being ethnically and linguistically Finnish. In his own comments he legitimised the use of migrants and Finns as separate groups. He described himself as being used to discussing themes like immigration and racism in other situations, and said he did not intervene in everything every time.
More often it was other classmates who responded to categorisation processes rather than the teacher, since the teacher was also busy with pedagogical tasks and solutions, such as dividing pupils into equally competent teams to enable soccer to commence, and being concerned with social relations among the pupils, and his relationship with particular pupils. Pupils were often left alone to cope with the consequences of categorisation. Overall, the teacher acknowledged that there was a good deal of categorisation, that he either did not hear or had become blind to. He stated from the outset that this kind of research was positive because it enabled one to become more self-conscious about how one communicates and acts.

Implications
The analysis shows how many parallel dimensions of teaching -pedagogical, social, relational -take place simultaneously, and how this makes it demanding for teachers to address the categorisation that takes place, and to challenge the unequal norms associated with categorisation. Categorisation has several layerseven when the teacher intervened when laughing took place, he seemed unaware of the gender dimension. Though the teacher had studied multicultural education as a minor, and was concerned with unequal use of power on a general level in the interview, he did not address the unequal norms behind the categorisations in the analysed situations. Teachers, also when already committed and insightful, need further insight to see the full picture of structural inequalities. To achieve this, the multicultural education taught in Finnish teacher education needs a stronger critical and social justice-oriented approach (Hummelstedt-Djedou, Zilliacus & Holm 2018). Teachers need more insight into manifestations of othering and normativity, in order to recognise and problematise categorisation. This entails understanding how processes of gender, sexuality, race, nationality, class, and ability operate in society and school (Gorski 2008;May and Sleeter 2010). Teachers need to realise that the categories that seem the most fixed and commonly used by the pupils could be the ones most in need of deconstruction in order to challenge a hegemonic order of us and them. This is particularly important regarding the category 'Finn' and 'Finnish'. Such a category needs to be broadened for all pupils, to make the school truly inclusive. Teachers could use the daily references to race and nationality as a resource to initiate discussions about social justice, to provide possibilities for pupils to become agents for social justice (Nieto and Bode 2018), and possibilities for what Sleeter (2015) calls 'democratic activism', also where teachers fail to recognise prevailing norms. Additionally, teachers need to more widely acknowledge their own position and privilege of being able to choose when to categorise themselves and when to avoid categorisation altogether. Pupils do not have the privilege to stay outside of these processes in similar fashion. Developing bonds of trust with pupils (Sleeter 2015) demands a humble attitude, willingness to admit mistakes, and sensitivity to the vulnerable position of the pupils.

Transcription key
TextEnglish translation of talk in Finnish (.) micropause, less than 0.2 seconds [ indicates the start of overlapping talk ] indicates the end of overlapping talk TEXT emphasis or talk that is louder than normal ((text)) transcriber's description of non-verbal actions in particular (text) uncertain transcription ((inaudible)) inaudible talk The transcription key is a modified version of the conventions developed by Gail Jefferson.

Disclosure statement
We have no conflicts of interest to disclose. We have no financial interest or benefit arising from this research.

Funding
This work was supported by University of Helsinki, strategic funding (2015-2017)

Notes on contributors
Ida Hummelstedt is a doctoral student in the Faculty of Educational Sciences, University of Helsinki. Her dissertation focuses on multicultural education in Finnish teacher education and in primary schools. Hummelstedt's research interests include critical multicultural education, norm-critical education, and anti-racism education. She is currently part of the project Multilingual and Intercultural Education in Finland and Sweden at University of Helsinki and Stockholm University.
Gunilla Holm is Professor of Education in the Faculty of Educational Sciences at the University of Helsinki and director of the Nordic Centre of Excellence in Education 'Justice through Education'. Her research focuses on issues in education related to race, ethnicity, class and gender as well as on photography as a research method. Professor Holm's current projects include Racism In Lower Secondary Education (RILSE).
Fritjof Sahlström is a professor in education at Åbo Akademi University in Finland. His