Are education and nationalism a happy marriage? Ethno-nationalist disruptions of education in Dutch classrooms

Abstract

Following Gellner, citizenship education has often been framed in terms of nationalism. This framing is supported by methodological nationalism that legitimizes nationalism as either functional (civic nationalism) or natural (ethnic nationalism). Based on a triangulation of qualitative and quantitative data, this study of the dynamics in the classes of a Dutch faculty of social professions highlights the disruptive impact of nationalism on citizenship education, spilling over to other courses as well. Ethno-nationalist discourses in Dutch media and politics as well as in multiculturalism approaches used in citizenship education fuel conflicts between non-migrant students and students with a migration background that disrupt education. It is argued that in globalized settings like these classes, a more viable approach to citizenship education would take an institutional instead of communitarian perspective.

Keywords:

• Citizenship
• civic nationalism
• education
• ethnic conflicts
• ethnic nationalism
• the Netherlands

Introduction

Following Ernest Gellner’s (1983) argument that the development of industrial society requires homogeneity and identity formation on the national level, citizenship education has been understood as a prime instrument for achieving national homogenization and identity. However, Riyad Shahjahan and Adrianna Kezar (2013) urge us not to take such nationalist understandings of education for granted and to leave assumptions of methodological nationalism behind. Their call resonates with a wider literature that questions methodological nationalism (e.g., Wimmer and Glick Schiller 2002). Is a nationalist framing of citizenship education still functional and adequate in the current context of globalization and transnationalism or has it even become counterproductive? The aim of this study is to contribute to our understanding of the contemporary impact of nationalism on citizen education.

This study has been carried out among students of social professions programmes at a Dutch university of applied sciences and highlights the destructive impact of nationalism, as it fuels polarization and disrupts education in class. The findings of interviews, focus groups and questionnaires among students and teachers in 2011 indicate¹ that the dispersion of ethno-nationalist categories stemming from dominant discourses in Dutch media and politics as well as from dysfunctional diversity theories triggers ethno-nationalist conflicts in classes made up of students with diverse origins. An argument is made for notions of citizenship education that foster institutionally nested instead of communitarian identities.

Nationalism and education

Ernest Gellner (1983) has conceptualized nationalism as a requirement of industrial society that needs translocal normalization and homogenization of uniform knowledge and discipline. Several studies have shown how the state fulfils this requirement by constructing ‘the nation’ and propagating national identity through public schooling, especially citizenship education (e.g., Bénéï 2005 and 2008; Bourdieu and Passeron 1990; Hahn 1998; Jaskulowski, Majewski, and Surmiak 2018).

However, other studies found that the effects of national identity constructions may be limited or even problematic. Students in Cluj (Romania) reproduce nationalist rhetorics in questionnaires, but without taking such rhetorics seriously in everyday life (Fox 2004). Some students in mainland China (Fairbrother 2003) and Japan (Parmenter 1999) resist such rhetorics. National identities clash with indigenous identities of Aymara students in Bolivia (Luykx 1998). Nationalist notions of ‘white’ and Christian Englishness emphasize differences with ethnic or racial minorities in English citizenship education (Gholami 2017). Moreover, education may also promote cosmopolitan instead of nationalist views, like in Israeli (Yemeni, Bar-Nissan, and Shavit 2014), Japanese (Saito 2011), Irish (Tormey 2006) and Norwegian (Osler and Lybaek 2014) education.

In her study on citizenship education in vocational schools in Berlin, Cynthia Miller-Idriss (2009) found that teachers explicitly take an anti-nationalist approach against the backdrop of Nazi horrors in the past (see also Parmenter 1999 on Japan), but that does not keep some students from feeling attracted to nationalist views. Despite her constructivist positioning, however, she tends to naturalize students’ nationalist views. She makes no conceptual differentiation between citizenship, belongingness and national identity, which suggests that citizenship and belongingness require national identity.

Next, she does not address the question where these students get their nationalist ideas from, whereas her own material provides clues about those sources. Respondents referred to official nationalist framings of migrants who wound need ‘integration’ and to the recurrent use of national categories in German media, despite their civic outlook (Miller-Idriss 2009, 103). The teachers take an anti-nationalist stance associating ‘the nation’ with feelings of guilt, but thereby they keep notions of ‘the nation’ alive as something important.

Methodological nationalism and beyond

Miller-Idriss’ otherwise excellent work illustrates how difficult, but nevertheless indispensable, it is to escape from methodological nationalism (Jaskulowski, Majewski, and Surmiak 2018; Shahjahan and Kezar 2013). Andreas Wimmer and Nina Glick Schiller (2002, 302, italics mine) have defined methodological nationalism as ‘the assumption that the nation/state/society is the natural social and political form of the modern world’. That is when we take ‘nations’ as units of analysis for granted and assume them to be unitary actors within assumedly self-evident processes of nation-state building.

Methodological nationalist views are supported by trends in the literature on nationalism itself. As said, Ernest Gellner (1983) saw nationalism as a functional requirement of modern industrial society. National identity would allow citizens to lead a meaningful life, enjoy civic rights and socialize themselves into a national community. This civic nationalism contrasts with an ethnic understanding of nationalism by, for example, Anthony Smith (1986). ‘The nation’ preceded modern times and favoured the development of the state, he argued. Specific cultural characteristics would mark ‘the nation’. Individuals would adopt a national identity as they are born into a national community.

This conceptual distinction between civic and ethnic nationalism is useful to analyse cases of nationalism (see Dutch nationalism below), but risks legitimizing nationalism in a functionalist or naturalizing way. Civic nationalism risks functionalism: there is a modern necessity for ‘the nation’ so it will emerge. Ethnic nationalism risks naturalizing ‘the nation’: it has been there since ancient times thus it is normal and natural. In other words, ‘the nation’ has to be so it is, versus ‘the nation’ is so it has to be. However, with such a functionalist or normalizing legitimization, how can we maintain a critical distance towards ‘the nation’ and nationalism as well as towards their relations with education? Getting away from methodological nationalism in our approach is necessary to critically discuss the empirical relationship between nationalism and education. Several studies have fleshed out the convergence between institutions and nationalism (Goode and Stroup 2015), but very few have looked at the impact of nationalism on the functioning of institutions like education.

Regarding nationalism, the literature also distinguishes between elite nationalism and everyday nationalism. Scholars point at the ways in which elite nationalist claims are adopted, reworked and adapted by people in everyday life (Fox and Miller-Idriss 2008; Goode and Stroup 2015), highlighting ‘ordinary’ people’s agency. Anthony Smith (2008) emphasized that everyday nationhood constructions need to be studied in interactions with official constructions of ‘the nation’. Discussing the influences of such official constructions in an encompassing way, i.e., not limited to teachers’ views, is important for understanding everyday nationalism among students. The main traits of official nationalist discourses in media and politics need to be outlined first.

Dutch ethno-nationalism

In the Netherlands, discourses in media and politics have taken an ethno-nationalist turn since 2000 (Scholten and Holzhacker 2009; Van Reekum 2012), manifested in many official practices, like the campaign for the revival of ‘Dutch norms and values’ by Prime Minister Jan Peter Balkenende (2002–2010). The assumed ‘threat’ to ‘Dutch national identity’ has become a major topic of national elections campaigns. Jan Willem Duyvendak (2011) writes about a ‘thickening’ of ‘Dutch culture and identity’ and Semin Suvarierol (2012) calls it ‘nation-freezing’, i.e., a particular and substantive definition of ‘the nation’ is imposed on individuals.

Paradoxically, however, the prime components of such ethno-nationalist notions are exactly civic and liberal values. Dominant constructions of ‘Dutchness’ portray progressive, civic and liberal values like emancipation, equal rights for homosexuals, sexual freedom, gender equality, individualism and tolerance as typically Dutch (Van Reekum 2012; Van Reekum and Duyvendak 2012). Universalistic civic values are transformed into particularistic ethnic building-blocks of ‘the Dutch nation’.

In this ethno-nationalization of civic values (Laegaard 2007), ‘the nation’ is opposed to those who are believed to threaten it (Halikiopoulou, Mock and Vasilopoulou 2013), particularly migrants with so-called ‘non-Western’ backgrounds. They are obliged to pass ‘civic’ integration exams that test their adoption of assumedly Dutch cultural traits (Suvarierol 2012). Contemporary ethno-nationalist constructions of an opposition between assumedly Dutch traits and those represented by migrants were made possible by the culturalization of migrants in multiculturalist discourses in the 1980s and 1990s (Siebers and Dennissen 2015; Duyvendak, Geschiere, and Tonkens 2016).

Official constructions of ‘the Dutch nation’ reflect typical ethno-nationalist traits, i.e., they are particularistic, substantive, deterministic and totalizing. Rogier van Reekum and Marguerite van den Berg (2015) studied the incorporation of civic values into ethnic nationalism in parenting courses. Theirs is the only study that focuses on the role of Dutch ethnic nationalism in education. The current study contributes to filling this gap.

Research setting and methods

The research was carried out by the author, supported by assistants,² at the faculty of social professions of a Dutch university of applied sciences in 2011. In 2008, the Equal Treatment Commission, the official Dutch tribunal for filing a discrimination complaint,³ found the faculty guilty of discriminating against two migrant staff members. The following unrest in the whole university caused its board to initiate a university-wide research project to get an overview of the overall role of ethno-migrant⁴ diversity in class and ethno-migrant inequality in study success. This article is based on a subproject that focused only on the faculty of social professions. It was singled out for an in-depth study because of the occurrence of conflicts between non-migrant and migrant students with a so-called ‘non-Western’ background⁵ in its classes. The subproject’s aim was to identify the dynamics and factors behind such conflicts that particularly emerged in citizenship education courses.

The faculty offered three programmes: socio-cultural education and training CMV (Cultureel-Maatschappelijke Vorming), socio-pedagogical work SPH (Sociaal-Pedagogische Hulpverlening) and social work and services MWD (Maatschappelijk Werk en Dienstverlening). The CMV programme instructed students to develop cultural activities with a social development perspective. SPH and MWD trained social workers to work with, respectively, groups and individuals, both have full-time and part-time programmes. Only a few of CMV’s 180 students had a migration background. Over 20% of SPH’s 800 students had a (mostly ‘non-Western’) migration background, whereas over 75% of MWD’s 730 students had such a background (data provided by the programme coordinators in 2011).

Citizenship education was provided in the courses Ethics and Citizenship, and Diversity and Identity. They focused on the basics of civic behaviour, social responsibility, democracy, and politics as well as on ethical problems that social workers are confronted with in their work. Driven by the unrest of 2008, diversity became one of the main topics of citizenship education. A complete minor programme (30 ects) was dedicated to diversity. The main focus was on diversity cases students may find difficult to deal with, but also some diversity theories were discussed (see below).

The author and an assistant studied relevant documents and held seven focus group meetings with either students or teachers. Applying non-purposive sampling, we put a message on the intranet to invite them to participate. We also organized discussions in five classes that showed interest in the subject and discussed preliminary findings in all five teachers’ teams. These meetings were spread over the various programmes and included both migrant and non-migrant respondents. Teachers and students who often took the floor in these meetings were invited for an individual interview (purposive sampling). In total, 24 interviews were conducted with the director, four programme coordinators and with students and teachers. We tried to spread student and teacher respondents over relevant categories, like the various programmes and ethno-migrant backgrounds (see Table 1).

Table 1. Interview respondents and backgrounds.

Background^a		Teachers		Students
Non-migration		5		2
‘Western’ migration		1		1
‘non-Western’ migration		3		5
Indonesian		–		2
Total		9		10

a See notes 4 and 5. ‘Western’: Germany and Macedonia. ‘Non-Western’: Surinam, Curaçao, Turkey and Kenya.

Table 2. Variables, items and reliability students’ questionnaire.a

Variable		Items		Cronbach’s alpha
Ethno-nationalist conflicts in class		Sometimes there are tensions and conflicts between students of my class with different cultural backgrounds and origins Sometimes, students of my class with different cultural backgrounds and origins make nasty remarks towards each other that are related to their background		.869
Impact ethno-nationalist discourses in media and politics		Discussions on issues of crime, terrorism, homosexuality, Islam, Islamization, migration, minorities, Geert Wilders’ statements and 9/11 as represented in the media lead to fierce discussions and tensions between students with different cultural backgrounds and origins in class In discussions on issues of crime, terrorism, homosexuality, Islam, Islamization, migration, minorities, Geert Wilders’ statements and 9/11 as represented in the media, fellow students feel addressed as a member of their cultural group and defend their own group If I recognize viewpoints from the media and politicians in the statements of students, they trigger radical reactions from other students		.780
Ethno-migrant composition in class		In my class, the share of students with a first- or second-generation foreign background (also called ‘allochthone’) is about: None =1; 20% = 2; 40% = 3; 60% = 4; 80% = 5; Everybody =6.
Ethno-nationalist groupist behaviour		In my class, groups of students with the same cultural backgrounds and descent hardly socialize with students with other cultural backgrounds
Ethno-nationalist tensions between students and teachers		Sometimes, there are tensions in class between students and the teacher in which differences in cultural backgrounds between these students and the teacher play a role.
Recognition work		I am concerned about what friends think of me I am concerned about what teachers think of me I do my best to make sure that fellow students consider me to be one of them I do my best to make sure that teachers see me as a good student		.771
Acceptance by fellow students		I have the feeling that all fellow students in class accept me as I am
Treated well at internships		At my internships, I have been treated well by everybody
Sharing leisure interests with other students		With fellow students I mingle with most, I share the same interests in music, having a night out, sports…
Laziness		I find it oké if I can follow the lessons without having read the book or the texts for this lesson
Teachers assess exams properly		Teachers assess papers and exams based on professional criteria and not on political or personal views

^a Translated from Dutch original. Items measured on a 5-point-Likert scale from totally disagree to totally agree, unless indicated otherwise.

Table 3. Variables, items and reliability teachers’ questionnaire.a

Variable		Items		Cronbach’s alpha
Ethno-nationalist conflicts in class		The formation of assignment teams with students with the same ­cultural background leads to tensions between the teams: 0 = in none of the classes in which I teach, 1 = in a minority of classes in which I teach, 2 = in half of the classes in which I teach, 3 = in the majority of classes in which I teach, 4 = in all classes in which I teach The formation of assignment teams with students with the same cultural background leads to conflicts between the teams: 0 = in none of the classes in which I teach, 1 = in a minority of classes in which I teach, 2 = in half of the classes in which I teach, 3 = in the majority of classes in which I teach, 4 = in all classes in which I teach I have the impression that students search for safety in their own subgroup due to tensions between the subgroups		.794
Students’ ethno-nationalist groupist behaviour in class		Students with the same cultural background prefer to socialize with each other: 0= in none of the classes in which I teach, 1 = in a minority of classes in which I teach, 2 = in half of the classes in which I teach, 3 = in the majority of classes in which I teach, 4 = in all classes in which I teach Students with the same cultural background prefer to constitute teams for doing tasks and assignments together: 0= in none of the classes in which I teach, 1 = in a minority of classes in which I teach, 2 = in half of the classes in which I teach, 3 = in the majority of classes in which I teach, 4 = in all classes in which I teach		.780
On professional ground rejecting students’ mutual ethno-nationalist identification		Apart from whether students experience it as a problem, I think it is inappropriate to address each other in ethnic terms [Turk, Moroccan, Dutch…] considering the professional practice as social professional
Favouring a professional approach to cultural differences		Personal sensitizing and societal problematizing of cultural diversity issues are meant to help students to pose the right questions to clients and citizens in their professional practice It is important at moments of fierce discussions about gender- and sex-related issues that, as a teacher, you teach students to separate person and opinion: the opinion of a student about a particular gender or sexual orientation should not be an attack on the person with that gender or sexual orientation		.705
Ethno-nationalist radical viewpoints expressed in class		It is my experience that radical viewpoints are especially expressed in class in discussions involving students with various cultural backgrounds
Essentialist diversity theories fuel radical viewpoints in class		It is my experience that theories that understand individuals as members of a cultural group incite students to take radical standpoints

^a Translated from Dutch original. Items measured on a 5-point-Likert scale from totally disagree to totally agree, unless indicated otherwise.

The interview and group meeting scripts focused on general dynamics in class, experiences with ethno-migrant diversity, topics that produce tensions and events related to backgrounds. Such issues were questioned in an ethnographic sense asking about when, context, who were involved, who said and did what, how it evolved, what were the consequences, etc. For qualitative data analysis, Matthew Miles and Michael Huberman’s (1994) steps of data reduction (selective, open and axial coding), data display and drawing conclusions were used, starting with interview topics as initial codes.

These qualitative data served as input for two questionnaires using non-purposive sampling. The questionnaire for teachers had an N = 58 that represented 66% of the teachers. They were spread over the programmes, 50 had a non-migration background and eight a migration background with various origins. For students, 42 relevant questions were included in a university-wide questionnaire as part of the wider project on study success (see above). Of the social professions faculty, 110 students responded (CMV 15, MWD 38 and SPH 57), i.e., the response rate was 6.4%. Of those 110 students, 26 had a migration background with various origins. Questions were about what actually happened in class, not about personal views or opinions.

For quantitative data analysis, AMOS structural equation modelling software version 22 was used. We applied two categories of fit indices (Curşeu et al. 2007, 132). Absolute fit indices show the general fit between the theoretical model and the data. Relative fit indices compare the tested model with the null model. The null model suggests that the variables in the model are mutually independent and that there is no covariance among the variables (Widaman and Thompson 2003). The indices used here measure the degree to which constructed models fit real relations between variables in the data file. Constructed models (Figures 1 and 2) match the norms set for such fit (see Browne and Cudeck 1993; Widaman and Thompson 2003).

Figure 1. Model students’ questionnaire.

The response rate of the students’ questionnaire is quite low, as in most university-wide questionnaires without a communication campaign addressing a particular faculty. Numbers of respondents are limited, but for data analysis the significance threshold of p<.05 was applied. Validity is mainly ensured through data triangulation of various stakeholders, of various research instruments and of various kinds of data on the same subjects. All findings presented below are supported by several data sources, including quantitative and qualitative data, unless indicated otherwise.

Ethno-nationalist conflicts in class

Three SPH students accounted of the following event that occurred in a citizenship education class. A film was shown about homosexuality when a male student with a Moroccan background started to make vomiting gestures to express his disgust of homosexuality. He typified it as ‘dirty’, as ‘an illness’. Thereupon, a female bi-sexual student with a non-migrant background became angry and said that she felt offended and hurt by his comments. Other non-migrant students started to back her up, migrant students did the same with the ‘Moroccan’ student and fierce accusations and insults were exchanged between the groups. This conflict had continued for a year already and recurrently disrupted education.

Conflictive practices like these included hate and threat emails, sometimes the police had to be called in. It is not easy to assess the frequency of occurrence of such conflicts precisely, but 19.1% of the teachers indicated in their questionnaire that such conflicts occurred in half or more of the classes in which they teach. Half of them answered that such conflicts and tensions took place in a minority of their classes. Thus, most likely these conflicts and tensions happened in a substantial minority of the faculty’s classes.

A comparison of the focus group and interview data on those conflicts revealed that they resemble the ethno-nationalist characteristics discussed above: deterministic, totalizing, particularistic and substantive. First, students and teachers indicated that in these conflicts students socialize in subgroups on an ethno-nationalist basis. Members of those subgroups identify each other in ethno-nationalist terms. Then, they call each other ‘Dutch’, ‘Turk’, ‘Surinamese’, ‘Moroccan’ and so on and often address each other as representatives of their ethno-nationalist subgroup. Identities like sex and gender also play a role in such incidents, but subordinated to the basic ethno-nationalist ones.

Second, these conflicts have ‘Dutch’ versus ‘non-Dutch’ as the main opposition and sustain themselves in a deterministic way, i.e., independently from students’ intentions. All three students involved in the conflict discussed above told us that they very much regret these conflicts, but are unable to stop them. They feel that those ethno-nationalist subgroup formations, categorizations and polarizations are imposed on them.

Third, such conflicts tend to become totalizing. They are mostly sparked in citizenship education but continue in other courses as well. Students and teachers told us that, when a student enters in these classes, everyone watches whether s/he sits with the subgroup s/he is ascribed to. This polarization harms the learning process since a student brings forward views s/he is supposed to voice as ethno-nationalist group member instead of those s/he learned at school, teachers said.

Fourth, the qualitative data on these conflicts show that substantive and particularistic characteristics are attributed to and represented by members of these ethno-nationalist groups. ‘Dutch’ students bring forward civic values like tolerance, liberal views on sexuality and gender roles, emancipation, equal rights, sexual freedom and law abiding, whereas ‘Turkish’, ‘Moroccans’ and ‘Antilleans’ students voice opposite views. The ‘Dutch’ particularistic appropriation of universal civic and liberal values (see above) forces other ethno-nationalist groups to either take illiberal and uncivic views or to say that many ‘Dutch’ violate those liberal and civic values just as badly.

Conditions

Several conditions tended to favour the occurrence of such ethno-nationalist conflicts. Both students and teachers emphasized that an important motivation for MWA and SPH students to choose these programmes is to help people who suffer from similar problems they themselves experienced. Having a refugee background or living in ‘lower’ class neighbourhoods, many migrant students had experiences of broken families, trauma, crime and drugs and wanted to help clients with similar problems. This close connection between personal experiences and social issues discussed in class, especially when related to diversity like the overrepresentation of migrants in crime statistics, easily leads students to take these issues very personally, giving a particularly emotional and conflictive potential to discussions about these diversity-related issues.

Moreover, the didactic approach common to social professions that emphasizes personal reflection also favours ethno-nationalist conflicts. Miriam, a 23-year-old full-time SPH student without a migration background, said about sexuality-related issues that trigger ethno-nationalist conflicts:

You do not know how to deal with these issues. You do not have the theory plus your own norms, values and feelings hinder you, perhaps because you hardly know anything about it.

She argued that one’s own norms, values and feelings impede a professional way of dealing with these matters. However, students were encouraged to draw on exactly those norms, values and feelings due to the emphasis on personal reflection, teachers emphasized. Social professions’ didactics want to create empathy, sensitivity and openness for the feelings and experiences of clients. Therefore, students have to make their own assumptions explicit. Only then, they will be able to relativize and problematize them.

However, regarding diversity-related issues, this second step hardly happened, as Miriam and other students argued. Teachers did not provide much input in terms of theory and knowledge. Their hesitation partially stemmed from the condemnation of the faculty by the Equal Treatment Commission in 2008. The subject of diversity is too delicate for them to discuss openly. ‘We have found a fearful way of talking about these matters’, a teacher said. Thus, students were free to express whatever view that comes to their mind, hardly disciplined or regulated by professional expertise.

Factors

It is argued that these conditions hamper the development of a professional attitude towards diversity-related social problems, i.e., they do not encourage students to learn the knowledge and skills they need as future social professionals, and thus give free way to factors that produce ethno-nationalist conflicts. These factors are discussed below.

Ethno-nationalist discourses in Dutch media and politics

With its R² of .799 (p = .000), the students’ questionnaire model developed in the structural equation analysis (Figure 1) provides a strong explanation of the occurrence of ethno-nationalist conflicts. Two remarks need to be made here. First, the questions in the questionnaire refer to individual observations regarding ethno-nationalist conflicts, whereas those conflicts are a group phenomenon occurring in class. Thus, variables like socio-economic class, gender and age do not appear in the model, since they refer to individual respondent’s traits. It may well be possible that the average age, gender composition and socio-economic background of the class as a whole may influence the occurrence of ethno-nationalist conflicts. Ethno-migrant composition, i.e., the percentage of students with a migration background in class, does refer to the class as a whole and shows that the higher this percentage, the more likely ethno-migrant conflicts occur (see Figure 1, Tables 4 and 5).

Table 4. Regression weights of the students’ questionnaire model.

Variable		Variable		Estimates		Sign.
Impact ethno-nationalist discourses in media and politics	➞	Ethno-nationalist conflicts in class		.413		***
Ethno-nationalist groupist behaviour	➞	Ethno-nationalist conflicts in class		.225		***
Ethno-migrant composition in class	➞	Ethno-nationalist conflicts in class		.168		.002
Recognition work	➞	Ethno-nationalist conflicts in class		−.154		.011
Ethno-nationalist tensions between students and teachers	➞	Ethno-nationalist conflicts in class		.254		***
Acceptance by fellow students	➞	Ethno-nationalist conflicts in class		−.152		.003
Treated well at internships	➞	Ethno-nationalist conflicts in class		.194		***
Laziness	➞	Ethno-nationalist conflicts in class		−.119		.007
Teachers assess exams properly	➞	Ethno-nationalist conflicts in class		−.123		.049
Sharing leisure interests with other students	➞	Ethno-nationalist conflicts in class		−.189		***
Impact ethno-nationalist discourses in media and politics	➞	Ethno-nationalist tensions between students and teachers		.560		***
Impact ethno-nationalist discourses in media and politics	➞	Recognition work		.198		.042
Ethno-nationalist groupist behaviour	➞	Recognition work		−.144		.049
Impact ethno-nationalist discourses in media and politics	➞	Ethno-nationalist groupist behaviour		.464		***
Ethno-migrant composition in class	➞	Ethno-nationalist groupist behaviour		.387		***
Acceptance by fellow students	➞	Impact ethno-nationalist discourses in media and politics		−.195		.007
Laziness	➞	Impact ethno-nationalist discourses in media and politics		−.160		.014
Teachers assess exams properly	➞	Impact ethno-nationalist discourses in media and politics		−.275		.002
Sharing leisure interests with other students	➞	Impact ethno-nationalist discourses in media and politics		.177		.007
Ethno-migrant composition in class	➞	Impact ethno-nationalist discourses in media and politics		.293		***
Treated well at internships	➞	Sharing leisure interests with other students		.244		.028
Ethno-migrant composition in class	➞	Teachers assess exams properly		−.221		.002
Acceptance by fellow students	➞	Treated well at internships		.304		.010

Table 5. Ranking of variables explaining ethno-nationalist conflicts in class (students’ questionnaire model).

Rank		Variable		R^2 change after erasure from model^a		Sig. F change		Percentage^b
1.		Impact ethno-nationalist discourses in media and politics		.060		***		20.13
2.		Treated well at internships		.042		.001		14.09
3.		Ethno-migrant composition in class		.037		.002		12.42
3.		Sharing leisure interests with other students		.037		.002		12.42
5.		Teachers assess exams properly		.027		.007		9.05
6.		Ethno-nationalist tensions between students and teachers		.023		.012		7.72
7.		Acceptance by fellow students		.021		.017		7.05
7.		Laziness		.021		.016		7.05
9.		Ethno-nationalist groupist behaviour		.018		.026		6.04
10.		Recognition work		.012		.048		4.03

a The R² of the model on Ethno-nationalist conflicts in class drops this value after deleting the indicated variable from the model. The larger this difference, the stronger the effect of this variable is on Ethno-nationalist conflicts in class.

b The percentage that the R² change of this variable contributes to the sum of R² changes after the variables have been deleted from the model. The total of R² changes in the third column is .298. The .060 of the first variable, for example, represents 20.13% of this .298.

Second, some variables in the model probably shape the observation of the occurrence of ethno-migrant conflicts instead of indicating this occurrence itself. If a student is treated well at internship institutions, the occurrence of ethno-nationalist conflicts in class will contrast very much with such treatment and will lead to higher scores on such conflicts. That will influence his or her observation, not necessarily the occurrence of conflicts themselves. Likewise, if a student does not care too much about the study (Laziness), s/he may be less likely to notice such conflicts. The same holds true when students try to create a positive self-identity (Recognition work, see Figure 1, Tables 4 and 5).

Then, the model contains variables that may influence both the observation of the occurrence of ethno-nationalist conflicts and this occurrence itself. If a student shares interests in music, in having a night out or in sports with other students, that may lead him or her to pay little attention to ethno-nationalist conflicts in class. However, if such interests are shared by most of the students in class, conflicts themselves will be less likely to occur. The same holds true for feeling accepted by fellow students in class.

The model suggests that the strongest predictor of ethno-nationalist conflicts in class is the impact of ethno-nationalist discourses in media and politics. It explains 20.13% of the model’s variance in ethno-nationalist conflicts in class (Table 5). This impact refers to fierce discussions and tensions between students triggered by issues that Dutch media and politics associate with ‘non-Western’ migrants, like problems of crime, terrorism, homosexuality, ‘Islamization,’ minorities, radical right politician Geert Wilders’ discriminatory statements and the 9/11 events. Of course, strictly speaking we cannot talk about causality here based on cross-sectional instead of longitudinal data. However, the interview and focus group data confirm these quantitative findings. They show that conflicts in class are verbalized in terms of this ethno-nationalist discourse in media and politics with frequent references to statements by politicians in the media.

The conflictive event about homosexuality discussed above exemplifies how this works. Students reproduce the association of migrants with illiberal views on homosexuality and the association of liberal views with ‘Dutchness’ in discourses in Dutch media and politics. They translate the notion of incompatibility between liberal and illiberal views in these discourses into conflicts between each other. Their conflicts on the same issues resonate the provocations in ‘civic’ integration packages for migrants that show Dutch pictures of kissing homosexuals and topless women that would depict ‘Dutch’ liberal values (Suvarierol 2012). Ethno-nationalist discourses in Dutch media and politics trigger (Siebers and Dennissen 2015) such conflicts around topics like crime, sexuality and gender that are particularly relevant to the work of social workers, much less so to other programmes like ICT or economics. These ethno-nationalist discourses provide the meanings that students take over in their conflicts in class. These meanings essentialize migrants and non-migrants in categories of background and descent, and stigmatize migrants by attributing negative connotations to them (see Siebers 2017 and Siebers and Dennissen 2015 for similar processes in Dutch work settings).

Tables 4 and 5 furthermore show that these ethno-nationalist discourses in media and politics not only fuel ethno-nationalist conflicts between students directly, but also ethno-nationalist tensions between students and teachers as well as ethno-national subgroup formation in class that, in turn, also feed ethno-nationalist conflicts between students (7.72 and 6.04% respectively).

Multiculturalist theories in citizenship education

The model constructed on the basis of the teachers’ questionnaire (Figure 2) also delivers a strong explanation of the occurrence of ethno-nationalist conflicts with an R² of .737 (p = .000). It mainly contains variables that focus on teachers’ observations in class with some variables expressing teachers’ views that, I assume, they enact in class and thus influence ethno-nationalist conflicts.

Figure 2. Model teachers’ questionnaire.

The teachers’ model confirms the impact of ethno-nationalist group formation and radical viewpoints, explaining 39.65 and 25.90% respectively of the model’s variance in ethno-nationalist conflicts in class (Tables 6 and 7). Regarding those radical views, teachers indicated topics that stem from ethno-nationalist discourses in Dutch media and politics, similar to the ones indicated above.

Table 6. Regression weights of the teachers’ questionnaire model.

Variable		Variable		Estimates		Sign.
Ethno-nationalist radical viewpoints expressed in class	➞	Ethno-nationalist conflicts in class		.304		***
Essentialist diversity theories fuel radical viewpoints in class	➞	Ethno-nationalist conflicts in class		.269		.005
Students’ ethno-nationalist groupist behaviour in class	➞	Ethno-nationalist conflicts in class		.507		***
On professional ground rejecting students’ mutual ethno-nationalist identification	➞	Ethno-nationalist conflicts in class		−.213		.008
Favouring a professional approach to cultural differences	➞	Ethno-nationalist conflicts in class		−.379		.005
Essentialist diversity theories fuel radical viewpoints in class	➞	Favouring a professional approach to cultural differences		.209		.023

Table 7. Ranking of variables explaining ethno-nationalist conflicts in class based on the teachers’ questionnaire model.

Rank		Variable		R^2 change after erasure from model^a		Sig. F change		Percentage^a
1.		Students’ ethno-nationalist groupist behaviour in class		.297		***		39.65
2.		Ethno-nationalist radical viewpoints expressed in class		.194		***		25.90
3.		Essentialist diversity theories fuel radical viewpoints in class		.112		***		14.95
4.		On professional ground rejecting students’ mutual ethno-nationalist identification		.083		.001		11.08
5.		Favouring a professional approach to cultural differences		.063		.004		8.42

a See Table 5.

Tables 6 and 7 also show that, although the emphasis in these courses is on self-reflection instead of theory, some essentialist and multiculturalist theories are actually discussed in citizenship education. They culturalize migrants and non-migrants in different communitarian categories and fuel ethno-nationalist conflicts in class (14.95%). Their effect is exactly the opposite from what is intended in these courses. Teachers and students referred to theories by Geert Hofstede (2001) on national cultural differences, by John Berry (e.g., 2005) on acculturation and by David Pinto (2000) on intercultural communication. These theories reproduce a similar essentialization and stigmatization as discourses in Dutch politics and media discussed above do. Several students indicated their strong discontent with such theories.

The teachers’ model (Tables 6 and 7) also indicates a way out of ethno-nationalist conflicts. Teachers effectively countered such conflicts when they discourage students to use ethno-nationalist identifications (11.08%) or when they encourage students to take a professional approach to diversity issues (8.42%). That involves discussing cultural diversity issues not as an aim in itself, but as a way to help students to pose the right questions to clients and citizens in professional practice. It requires separating person and opinion and to avoid taking critique on cultural issues personally. Ina, a 61-year-old teacher with a ‘Western’ migration background, said: ‘You must take a professional distance, step over your own experiences, background and socialization’ (see also Miriam’s statement above).

However, the legitimacy of social professions is undermined by the fact that most teachers and students tend to classify their professions as typically ‘Dutch’ or ‘Western’. Isaac, a 56-year-old male teacher with a Dutch background, said: ‘We do all this from a Western perspective’. Thus, social professions are ethno-nationalized as a cultural particularity, which encouraged some students with a ‘non-Western’ migration background to reject what teachers tell them on the basis of respect for diversity.

Conclusions

After leaving methodological nationalist assumptions about a supposed naturalness or functionality of ‘the nation’ behind, the destructive impact of citizenship education framed along ethno-nationalist lines becomes visible. In a globalized world in which students with various backgrounds get together in class, ethno-nationalist discourses in media and politics as well as in specific diversity theories trigger ethno-nationalist polarization and conflicts in class of this faculty. They put forward a substantive and particularistic understanding of citizenship that excludes students with migrant origins. Subsequent conflicts disrupt the learning process in citizenship education and in other courses in these MWD and SPH programmes.

These programmes are particularly vulnerable to those conflicts as key elements of these ethno-nationalist discourses, like the association of migrants with crime and illiberal views on homosexuality, are directly relevant to the problems these students prepare themselves to deal with in their future work. That is much less the case in CMV and other programmes at this university that, for example, focus on ICT or technique, where students seem to ignore the effects of the ethno-nationalist framing of citizenship education.

This study contributes to the literature by highlighting the destructive impact of nationalism on citizenship education and overall education in the classes of this faculty (see also Duemmler 2015). Its findings entail a critique not only on ethnic but also on civic nationalism. Gellner’s (1983) argument about the functional need for civic nationalism in industrial society may no longer be valid in post-Fordist Dutch society marked by the service sector instead of industry. Moreover, civic values have become incorporated into Dutch ethno-nationalism and thus can hardly serve as an alternative to ethno-nationalism as long as they maintain a link with nationalism as such. These findings align with cosmopolitan understandings of citizenship education that define citizenship not only on national but also on local, global or European levels (Banks 2008; Osler 2011; Shahjahan and Kezar 2013).

The question is, though, what kind of identities citizenship education is supposed to promote on these and other levels. Both civic and ethnic nationalism define national identity in terms of (imagined) community membership (Anderson 2006), but already in 1887 Ferdinand Tönnies pointed to a modern shift from community to a society differentiated into various fields or institutions like education. As an identity framed in terms of community membership, ethno-nationalist identity disrupts the proper functioning of education in the globalized classes of this faculty within which it erects an opposition between students considered to belong to this community and those who do not. That may apply to all identities framed in terms of community membership (Etzioni 2001), including the multiculturalist theories taught in this faculty.

In globalized settings, a more fruitful approach to citizenship education is perhaps to teach students how to navigate effectively and responsibly through institutions like education, the labour market and politics. Here, sameness can be constructed and recognized focusing on institutionally ‘nested’ identities of fellow students, colleagues and citizens, providing the conditions for the subsequent recognition of individual uniqueness (Brewer 1991; see also Shore et al. 2011 and Siebers 2009 on work settings). Without the need for an overarching community, the basic universal nature of civic and liberal values like human rights can be restored. These concluding thoughts may perhaps inspire future research and reflection on the directions in which to develop a fruitful understanding of citizenship education, especially in times of rising nationalism.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 .There are no reasons to believe that the impact of the conditions and factors we found to trigger ethno-nationalist conflicts in class in 2011 will be any different today. Moreover, the findings of this study will be ever more relevant with ethno-nationalism currently on the rise in many parts of the world.

2 .The author is very grateful to Paul Mutsaers, several student assistants and reviewers of previous drafts.

3 .Now called The Netherlands Institute for Human Rights.

4 .The term ethno-migrant (Siebers and Van Gastel 2015) is used here to refer to relations between people with and without a migration background, relations that may or may not become the object of ethno-nationalization. In the Netherlands, someone is defined as having a migration background if at least one of his or her parents was born abroad. Origin matters here, not nationality. Dutch studies (e.g., Siebers 2009; 2017) show that tensions and conflicts may evolve around the opposition of first- and second-generation migrants with a so-called ‘non-Western’ background versus non-migrants. Differences between people with various migration backgrounds become subordinated to this main opposition, see also section Ethno-nationalist conflicts in class

5 .Latin America, the Caribbean, Africa, the Middle East and Asia, except for Japan and Indonesia, are officially labelled as ‘non-Western’ parts of the world.