‘You can’t escape from it. It’s in your blood’: Naturalizing ethnicity and strategies to ensure family and in-group cohesion

Ethnicity, between constructivism and primordialism

By living and acting in specific social groups and specific fields, individuals internalize certain cognitive schemes of social and symbolic boundaries common for members of these social groups into their personal and unique identity (Brubaker et al., 2004; Wacquant, 1998). These boundaries are constructed around the interplay of (perceived) cultural differences and similarities (Erickson, 2010; Lamont and Molnar, 2002; Wimmer, 2008). In turn, these shared dispositions generate actions and strategies that are adapted to the specificities of a specific social domain. Furthermore, the macro and micro level are linked in such a way that the ideological paradigm in a given field or society generates specific and institutionalized practices, which become embodied as a second nature into individual dispositions that are felt to be ‘in your blood’ (Bourdieu, 1989; Gil-White, 1999; Sallaz, 2010).

This bears strong resemblance to the discussion in the literature on the two leading theories on ethnicity: constructivism versus primordialism (Verkuyten, 2005). Although socially constructed and heavily depending on social interaction (Barth, 1969), Roosens (1994) argues a fundamental aspect of ethnic identity is that it feels like it is in your blood: you have family-like ties with your co-ethnics (Bentley, 1987; Jenkins, 2008). This ‘imagined’ blood relatedness can therefore be considered the difference between a ‘regular’ social group and an ethnic group (Roosens, 1994). Some call it the ‘social construction of primordiality’ as it specifically stresses the (often unconscious) practices and representations that are felt to be self-evident and natural, although they are socially constructed (Gil-White, 1999; Yelvington, 1991).

This is not solely an intra-individual but also an interindividual process, as individuals to some extent share complex cultural schemes of meaning, or mental structures that influence ‘the way individuals perceive, interpret and act in the world, and how knowledge is acquired, stored, recalled, activated, and extended to new domains’ (Brubaker, et al., 2004: 41; Wacquant, 1998). As these schemes are to a large extent internalized as ‘second nature’, identities and social boundaries become naturalized and taken for granted (Yuval-Davis, 2006). In particular, the internalization of ethnic identities is one of the most basic elements of one’s personal identity, which is often felt to be primordial and biological, and thus innate (Jenkins, 2000; Roosens, 1994). It is believed to be a first-order dimension of human experience (Jenkins, 2008: 78). Belonging to an ethnic group assumes the same emotional importance as belonging to a family does, as the narrative in the beginning of this article illustrates (Roosens, 1998). At the same time, by ‘doing being ethnic’ – others refer to it as ‘doing race’ (Sallaz, 2010) or ‘doing difference’ (West and Fenstermaker, 1995) – in everyday life through sharing a language or religion, content is given to this identity. It is this interplay between praxis, representation and perception that constitutes one’s ethnic identity. As ethnic identities are inherently social, processes of social reproduction and continuity with the past are important as individuals are socialized to become knowledgeable about their ethnic background. However, these interaction processes are influenced by power differences between individuals and between groups, which we discuss now.

Dominant ethnic groups and whiteness theory

Processes of power or domination not only influence interindividual interactions but are also located on the level of social groups, institutions and fields (Lukes, 2005; Swartz, 2007). Social reality is differently constructed and perceived according to one’s social location, and this depends heavily on and interrelates strongly with the accumulation of (social, cultural and economic) capital by individuals and groups (Bourdieu, 1985). The accumulation of these forms of capital generates symbolic capital and power, which helps individuals to occupy more dominant positions in a specific field. However, one always has to be conscious of the fact that specific types of groups tend to hold dominant positions across different social fields (Bourdieu, 1989, 1991). Moreover, an important privilege of dominant groups is that they have more power and resources to influence the representation of social reality and in particular the definition of what is ‘normal’ and ‘valuable’, and what is a ‘pathology’ and ‘problem’ in that social reality (Bourdieu, 1991; Lukes, 2005; Yuval-Davis, 2010).

Starting from a similar concern, scholars studying ethnic (or race) relations developed whiteness theory precisely to stress the omnipresence of relations of domination between specific ethnic or race groups (Garner, 2007). This perception of whiteness – ‘the idea that there exists a certain group of people who are, by virtue of their heredity, entitled to greater rights and privileges than other groups’ (MacMullan, 2009: 168; 2009) – is internalized by individuals of different groups and is very hard to challenge due to, among others, its legitimization in everyday practice (Garner, 2007; Lewis, 2004). Therefore, scholars such as Sullivan (2006) and MacMullan (2009) stress the (un)conscious and habitual processes related to whiteness. As Sullivan (2006: 4) argues, ‘understanding white privilege as habit explains how oppressive structures such as white domination take root in people’s selves’ and become self-evident features of social reality and interaction processes.

Finally, various scholars point to the crucial role of the cultural and state apparatus in perpetuating these (white) privileged representations and practices as individuals become attached and committed to these powerful social institutions (Hannerz, 1992; Bourdieu, 1997; Leonardo, 2009; Reay, 2010; Sullivan, 2006). When MacMullan (2009: 170) writes about the habit of whiteness as a set of ‘pre-conscious responses to an environment … that rely upon certain inherited categories’, he refers to the transmission of these categories, which can be, for example, state-imposed (Bourdieu, 1997; Brubaker et al., 2004) but can also be part of a variety of socialization processes (Jenkins, 2008). Therefore, specific attention has to be paid to the ethnic identity and social categorization processes in individuals’ formative years (Bourdieu, 1990; Sallaz, 2010). One approach is to study how children deal with these issues and already internalize at a very young age specific social categorizations and identity features (Connolly, 2011; Jenkins, 2000). As the internalization of these dispositions is strongly influenced by paternal discourse and practice, some scholars focus on the ethnic socialization strategies developed by parents (for an overview see Hughes et al., 2006).

As a final note, we need to stress that each of the concepts mentioned above – culture, ethnicity, whiteness – is prone to essentialization and is sometimes represented as fixed and innate in human nature (Abu-Lughod, 1991; Carter and Fenton, 2010). Research can tackle this problem by focusing ‘upon the active work of agents who shape their life in accordance with the opportunities and constraints surrounding them’ (Crossley, 2001: 117). Even with respect to processes of social reproduction scholars can shift the focus from social structures to the discourses of individuals by using narrative analysis, thereby gaining deeper insights into meaning-giving processes (Yuval-Davis, 2010). Exemplary for the relative malleability of social and mental structures is the transition of social groups across the social hierarchy and from an out-group status to an in-group status. Various scholars point to the malleability – at least to some extent – of fault lines with respect to whiteness. Ethnic minority groups such as the Irish and Italians in the US that were previously perceived as non-white and sometimes even perceived more negatively than African-Americans are nowadays perceived as part of the ‘white group’ (Ignatiev, 1995; Bronwen, 2001). However, this highlights at the same time the endurance of specific ethnic (or racial) fault lines with respect to the African-American group, which remains the fundamental other for the ‘white majority group’ (Lamont, 2002; Oriol and Hily, 2002). Research in Flanders and Europe shows that Muslims in Europe tend to hold a similar position with respect to the dominant majority, being ‘the other’ that is inherently different and therefore eternally different (Snauwaert et al., 1999; Said, 1997). We discuss Flanders now more in detail as the context of this research.

The field of Flanders: Ethnic relations in the Belgian-Flemish equilibrium

Flanders and Belgium are complex contexts for conducting research on identity construction and ethnic relations, given amongst other elements the tense Flemish-Walloon-Brussels relations (Roosens, 1998). One can legally distinguish Belgium as a country and a Belgian nationality, but there is no such equivalent on the level of the semi-autonomous region of Flanders. However, individuals and groups can still construct a Flemish subnationality or identity irrespective of these ‘formal structures’ (Bail, 2008). Moreover, Flemish policy-makers developed civic integration courses – obligatory for third-country nationals – which precisely aim at the transmission of ‘Flemish values and norms’ to these newcomers. This Belgian-Flemish ethnicity – the dominant ethnic group in Flanders – is primarily constructed upon cultural emblems such as the Dutch language and the Christian-Catholic religion (Roosens, 1998). This accumulates, for example, in a general integration policy stating that successful participation in society is only possible if migrants learn Dutch as soon as possible. ¹ This conception of social reality lies at the heart of Flemish society. At the same time Christianity occupies a dominant position in the political system (a strong Christian Democratic Party), in the educational system (Catholic schools enrol around 65% of all pupils in basic and secondary education) but also in the family as more than 70 per cent of Belgian-Flemish children were baptized in 2000 (Van Meerbeeck, 2001).

Next to this dominant group various minority groups are present in Flanders, of which Moroccans represent 15 per cent of all minorities and Italians 5 per cent (number based on country of birth; Noppe and Lodewijckx, 2012). Italians mainly share the Christian-Catholic religion with the dominant group, whereas Moroccans often self-identify as Muslim. Islam is nowadays in general quite negatively perceived in Flanders. The presumed incommensurability of ‘Islamic values and practices’ with the ‘Western’ democratic system is often the subject of intense and emotional debates (Billiet and Swyngedouw, 2009; Arnaut et al., 2009). As a consequence, and although Italian (mainly in the period 1945–60) and Moroccan (mainly in the period 1960–73) migrants to some extent share a similar labour migration background, political processes such as the construction of the European Union led to the construction of a European identity of which Moroccans and Muslims do not seem part (Roosens, 1998). Although Italian migrants were initially perceived very negatively, their social status seemed to change over time, especially with the inflow of large groups of Moroccan and Turkish labourers (Beyers, 2007). A telling example of this whitening process of Italians is that of a local representative of an extreme-right and anti-Islam political party addressing Italian origin citizens with the message (written in Italian) that Turkish people were taking over the Italian pizzerias (Het Belang van Limburg, 2006).

As is often the case, the presence of more symbolic differences and similarities goes hand in hand with social and socioeconomic differences (Roosens, 1998). Most importantly, whereas 10 per cent of the Belgian community risks living below the poverty line, this is the case for more than 50 per cent of the Moroccan migrants and around 20 per cent of the Italian migrants (Van Robaeys and Perrin, 2006). As expected, this lack of economic (and social) capital has detrimental effects on the average position of these communities within other fields, such as the educational system (higher drop-out rates for Moroccan and Italian youth) and the labour market, and thus on their cultural, economic and social capital (Duquet et al., 2006). It is within this broader social context that this article discusses the parental perceptions and discourses.

Method and sample

As the narratives and perspectives of individual parents are central in this article, an interpretative phenomenological approach is crucial (Kvale, 1996) and ‘attempting to make sense of, or interpret, phenomena in terms of the meanings people bring to them’ is the primary goal (Denzin and Lincoln, 2005: 3). To understand how people interpret certain phenomena and give meaning to their daily life world, a qualitative research methodology is relevant (Maso and Smaling, 1998). Semi-structured in-depth interviews were conducted by the researcher, guided by a general topic list containing questions related to identity construction, family life and intergenerational relations. The start-up question was always ‘Could you give me an overview of your family’s history starting from the family you were born in up until your own family now, with special attention for your family composition, residence status, and other life events you find important?’ This proved to be an interesting question (of course, sometimes more sub-questions were necessary) and the information collected often gave a good overview of events important to these parents, which could be elaborated on further into the interview.

The research data are obtained via 42 in-depth interviews with fathers and mothers of Belgian-Flemish, Italian and Moroccan origin. Belgian-Flemish respondents did not stress their Flemish but predominantly their Belgian ethnicity. ² As 15 couples were interviewed, in most cases it was possible to interview both parents from the same family separately. The respondents have diverse backgrounds, but they all have at least one child and identify themselves as ‘having’ one of the three abovementioned ethnicities. Different methods were applied to recruit these respondents: Public family assistance services and specific ‘ethnic’ organizations were contacted and, in combination with the method of snowball sampling, this allowed for the recruitment of a diverse research sample. In total, 42 respondents were interviewed: 24 women and 18 men. With respect to their ethnic background, 14 Belgian-Flemish individuals (five men and nine women), 13 Italians (seven men and six women) and 15 Moroccans (six men and nine women) participated in the research. Of these respondents, 15 were in a partner relationship while the others were married. Concerning family composition (older and/or younger children), educational background, and employment status, variation was pursued and obtained. The Italian and Moroccan parents were mostly born in Belgium, or migrated to Belgium before their children were born. In each of the categories – be it gender and/or ethnicity – there were respondents who were higher or lower educated; some worked as labourers, others as clerks, teachers, researchers or social workers. The unemployed women were housewives and the two unemployed men were looking for a job. Yet, in this sample, Belgian-Flemish respondents were generally higher educated than Italian and Moroccan respondents. This sample configuration made it possible to study the research topics from a diversity of perspectives. All interviews were executed and transcribed by the researcher, and afterwards coded via a code scheme in Atlas-ti. In the following sections we discuss the results of this study more in-depth via illustrative key narratives that reveal important underlying processes.

Ethnic ways of seeing the world: Pasta in the blood

The context dependency of ethnic self-perception and awareness is apparent in the comparison between the narratives of, on the one hand, parents who claim they are of Italian or Moroccan origin and, on the other hand, Belgian-Flemish-origin parents. The latter never explicitly referred to their ethnic identity in the interviews. With respect to the former, what is most striking is the explicit way Italian respondents – although for some decades now an ‘established community’ in Flanders – discuss their ethnic identity. Their ‘ethnic ways of seeing the world’ are salient throughout the narratives, and this concerns feelings of strong intra-group bonds – ‘Italians, we are the same pasta’ (45-year-old Italian man) – and undeniable blood ties, as the narrative of the 29-year-old Italian woman in the beginning of this article illustrates. Italian respondents construct clear differences between themselves and the significant other – in these cases, ‘Belgians’, the dominant group: ‘Belgians will always be Belgians. That’s the way it is. No, I don’t have anything against Belgians, but they are put together completely differently [than Italians]’ (35-year-old Italian man). These Italian respondents have the feeling it is easier to understand each other when they are among co-ethnics, and this emotional bond – this ‘fictive kinship’ – is absent when they are among Belgian friends (for similar findings see also Bentley, 1987; Haynes, 2003; Roosens, 1998).

While these ethnic representations surfaced rather spontaneously in the interviews with Italian respondents, this differs strongly from the way Belgian-Flemish but also Moroccan respondents discuss their ethnicity. In the context of Flanders, Belgian-Flemish respondents appear to ‘forget’ that they themselves have a particular ethnic background, a common phenomenon among dominant groups (Lewis, 2004). For them, their identification processes needed to be elicited in the interviews. Moroccan respondents, on the other hand, are less explicit than Italian respondents when it comes to identifying themselves as clearly different from Belgians. They focus on emphasizing feelings of ‘ethnic pride’ towards their children and themselves, primarily in reaction to the experience of racism and discrimination: ‘We are Moroccans and we are proud of it’ (33-year-old Moroccan woman), ‘we try to pass along the love for Morocco … so they [his children] take some pride in their country’ (35-year-old Moroccan man) and ‘by making their [his children] own values as strong as possible … You are not worse than any other. Even if someone treats you badly. … These things have to be clear for them’ (40-year-old Moroccan man).

These differences are probably due to the general assumption that Italian migrants in Belgium are perceived as a very well-integrated or even assimilated – and therefore invisible – minority in broader society. As an illustration to this, the current prime minister of Belgium, Elio Di Rupo, is of Italian origin but is seldom identified as being primarily of Italian descent. Thus, this conscious explication of a particular ‘Italian identity’ could be a strategy to differentiate themselves from a Belgian majority that seems to ‘overlook’ or not to notice the Italian presence anymore. This resembles the idea of symbolic ethnicity as discussed by Gans (1979) which also incorporates elements of the increased ‘ethnic choice’ theory of Waters (1990). However, as will be discussed later, parents’ choices are limited as they are confronted with social institutions (and other individuals) that still stress their otherness to some extent. Nevertheless, there also seems to be a need to feel recognized in their ethnic otherness; the arrival of Moroccan (and Turkish) migrants reduced the ‘strangeness’ of Italian migrants in Belgium, who then underwent some sort of ‘whitening process’ (Sanders, 2002: 334). Moroccans are commonly perceived in Flanders as the most prominent ‘outsiders’, and their ‘otherness’ is accentuated by their Islamic background, which is perceived as clearly different from the dominant Christian background of the broader society. This will be elaborated on more in-depth in the next section.

However, besides (the idea of) ethnicity being in your blood, supplementary elements are necessary for constructing an ethnic identity: the cultural stuff (Barth, 1969; Roosens, 1994). The narratives show how cultural emblems such as language and religion are put forward as being essential for your ethnic identity and thus to pass along by parents. With respect to language – especially ‘the mother tongue’ – Italian and Moroccan parents share similar perceptions and strategies related to their similar position as linguistic minorities. These parents perceive the Italian or Moroccan/Berber language as their ‘mother tongue’, something you need to be able to speak if you want to ‘claim’ your ethnic identity. You are your language, as it were, as these narratives illustrate: ‘When she [her daughter] was little, I always spoke Italian to her. Just so she could speak our language’ (42-year-old Italian woman) and ‘So he has to know three languages fluently: so the mother tongue, that’s Berber, and French, and Dutch’ (28-year-old Moroccan woman).

Similar to these narratives concerning language, parents also put their religious background on front stage when it comes to the issue of in-group continuity. Yet, in this case, Belgian-Flemish and Italian parents share a similar dominant position – Christian-Catholicism is by far the most prevalent religion in Belgium/Flanders – and the idea that it is a part of their ethnicity needed to be triggered (similar to the position of Belgian-Flemish ethnicity, above). For Moroccan respondents there was no doubt that their Islamic background is central to their identity (little or no reference was made to specific Islamic traditions) as this key narrative illustrates ‘We, of course, we are Muslims’ (29-year-old Moroccan woman).

Nevertheless, when ‘interreligious relations’ are discussed, even self-identified non-Catholic ‘majority group members’ seem to use religion as the all-encompassing and explanatory difference between the two religious groups, as this narrative shows:

The difference between the religions. They will never accept us as Catholics, and we will never accept them as Muslims. That’s just the way it is. And you can say as much as you want: I respect that, if they want to pray seven times a day. Then they do that, but do not involve in that, you know? And with involving I mean: do not tell me how bad I am as a Catholic. I don’t need to know. I don’t believe anyway. (33-year-old Italian man)

As a conclusion to this section one can say that ethnic membership and its related cultural emblems are perceived as almost innate and essential elements for ‘the self’ but also to pass on to the next generation. In the next section we discuss how these dispositions can be related to specific actions and strategies.

Dispositions generating praxis: Strategies to ensure ethno-cultural continuity

As language and religion are put forward as the two most important cultural emblems constituting ethnic identity, it is important to study how these dispositions generate strategies and actions to ensure continuity. The analysis shows that, concerning language, only Italian and Moroccan parents develop specific strategies, whereas concerning religion, also Belgian-Flemish parents feel the need to design specific strategies.

The role of the cultural apparatus: Reflecting dominant ethnicity

While the previous sections discussed the processes of ethnic self-perception and the strategies to ensure ethno-cultural continuity, the interaction between these processes and the cultural apparatus was left out of the analysis. Nevertheless, individuals are always constrained by social structures and processes of domination. When this perspective is applied to the perceptions and strategies discussed above, it becomes clear to what extent parents feel their ideas about the upbringing of their children are considered deviant, particularly in the field of education (Reay, 2010). This is predominantly true with respect to the language dilemma that parents face. As said, in Flanders’ official (civic) integration policy the mastery of Dutch is put forward as the single most important cultural emblem for newcomers to acquire. Dutch proficiency is a prerequisite to becoming a successful citizen. This ideology is also omnipresent in the Flemish educational system (Clycq et al., 2013). It is to be expected that parental strategies focusing on mastering the ‘mother tongue’ conflict with such a policy, as the following narrative shows:

Oh no, I don’t listen to them [official pupil guidance centers]. I think that’s bullshit. [They said] we [herself and her sisters] weren’t capable of doing general education. I reminded them of that, when they said your child [her daughter] will never become an Einstein. [ … ] I said, now I have a degree and a good job. [They said] your daughter doesn’t understand many Dutch words, because she’s raised Italian. But I said, she’ll learn it with time. (35-year-old Italian woman)

In a society where the mastery of Dutch is put forward as the superhighway to a successful life, the linguistic strategies of these Italian and Moroccan parents are pathologized and their ideas of upbringing are considered as hindering instead of helping their own children (Clycq et al., 2013). One of the consequences is these parents feel they are marginalized and have to rely upon themselves to ensure familial continuity, since the mainstream social institutions do not support their practices and strategies (Reay, 2010).

The narratives of Moroccan parents demonstrate that they feel not only their mother tongue but also their religious background is under incessant attack by processes of domination. The Christian tradition is omnipresent in Belgium and Flanders and especially in the educational system where Catholic schools enroll the vast majority of all pupils. For Belgian and Italian parents who are and/or feel affiliated with this religion, this presence feels self-evident and fairly unproblematic, similar to the feelings of Belgian parents in this research who do not feel their mother tongue – being Dutch – is problematized by the dominant group. For Moroccan parents with a Muslim background this is perceived very differently, as the following key narrative illustrates:

With Christmas, they say in school ‘a Christmas tree, and Christmas this and Christmas that’, and then at our home [his children ask] ‘Why don’t we have a Christmas tree?’ And then you have to explain it again: ‘Soon we’ll have our Festival of Sacrifice, and then we’ll have our celebrations and new clothes and candy … ’ We try to emphasize these things … I would be the first to send my children to an Islamic school because, as a parent, you are relieved from that overall responsibility to transmit all these cultural things. And then you can just concentrate on raising your children, playing with them and doing all sorts of things because the rest is taken care of [in an Islamic school]. But not in our case. You always have to react to these things they take with them from school, like now with Sinterklaas [a traditional Belgian and Dutch festival for children] and Christmas. … And then you always have to explain yourself. And again, you get frustrated – what should I do with this? Whereas in an Islamic school this would all be self-evident. (35-year-old Moroccan man)

This narrative poignantly demonstrates how social reality can be perceived very differently when taking into account your specific social position. Never in their narratives did Belgian origin parents discuss or allude to similar feelings and dilemmas with respect to their ethnic identity. The next section takes a deeper look into this.

The invisibility of dominant ethnicity

Whereas in the previous sections, different features of Moroccan and Italian ethnicity are being used by respondents to explain certain social phenomena, features of the dominant ethnic group are never discussed and remain invisible. Yet the meanings attributed to features such as ‘Flemishness’ (or ‘Belgianness’) and Christianity reveal how dominant groups can influence the representation of social reality (Bourdieu, 1990).

The differential positions of dominant emblems can best be illustrated by using the concept of gender, namely how it intersects with ethnicity and religion. When Moroccan or Italian respondents are talking about gender differences in their families of origin – in particular the different restrictions for daughters and sons – these respondents relate the existence of these differences to the ‘Italian’ or ‘Moroccan’ identity of the people involved, stating things like ‘it is because people are Italian/Moroccan’ or ‘it is typically Italian/Moroccan that sons get more freedom than daughters’ (Clycq, 2012). However, when Belgian respondents talk about very similar gender differences they never relate this to their ethnic background: these differences have nothing to do with ‘Flemishness or Belgianness’. They refer to ‘those days when people did not know better’ or to ‘men’ who always want to restrict daughters more than sons. In all these narratives Belgian ethnicity – in contradiction to Italian and Moroccan ethnicity – remains invisible and unproblematized. It does not seem a useful nor relevant element to explain certain phenomena.

A more in-depth analysis even shows that, when discussing the gender inequalities between different ethnic groups, respondents make another interesting variation. Belgian and interestingly also Italian respondents relate the perceived gender differences between these two groups to the ‘Italianness’ of Italians and for example not to their (shared) Christian background. When Italian and Belgian respondents are discussing perceived gender inequalities in Moroccan families, they explicitly state it is not the ‘Moroccanness’ of these individuals (because that would be racist), but their Islamic background that is the ‘cause’ (because it is ‘common knowledge’ that Islam is oppressive, see Clycq, 2012). On the other hand, Moroccan respondents who talk about gender inequalities in ‘their community’ relate them to Moroccan ethnicity and explicitly not to their Islamic background. Similar to the way Italian respondents talk about this topic, it is not because of their Christian but because of their Italian background.

Several mechanisms seem to be influencing the interpretations and perceptions of the respondents. On the one hand, like Flemishness, Christianity appears to be invisible. For example, at no point in their conversations do respondents, regardless of their own religious background or ideology (Christian, Muslim or atheist), link gender inequalities to Christianity. Moreover, the shared religious background between Belgians and Italians seems to have as a consequence that another central aspect than religion has to be brought into the picture to create a boundary, namely ‘Italianness’, which can account for the perceived (gender) differences. As a consequence, like Belgian-Flemish ethnicity, Christianity is never problematized when talking about (gender) differences in the family and it is never used as an explanation for problematic situations.

This analysis shows that dominant representations appear to be deep-rooted, self-evident and unproblematic. At no moment is the dominant ethnicity problematized in the same way as is the case for dominated ethnicities and it seems to be the privilege of the dominant group to remain invisible and, in this case, the neutral viewpoint when it comes to their ethnic or religious perspective. Moreover, this representation of the causes of gender inequality and the self-evident nature of gender equality in Flemish-origin families is also legitimized by the discourses of individuals with a dominated ethnicity.

Ethnicity and intergenerational continuity in the upbringing of children

The context of family socialization processes – and in particular the upbringing of children – strongly encourages parents to make explicit (i) what they feel is crucial to pass on to their children, and (ii) what they deem necessary to ensure the essential amount of continuity and family social cohesion (Bourdieu, 1990). In this article the focus lay on these elements related to one’s ethnic identity and we tried to illuminate the complex ways parents discuss this intergenerational transmission of ethnicity and culture. To answer this question the narratives and strategies developed by minority and majority parents were studied in-depth. An important limitation of this study is the fact that it is based on discourses without linking these discourses to praxis. That makes it difficult to be (more) certain about specific actions individuals claim to undertake. However, since in most cases both parents were interviewed separately on similar topics and courses of action, we were able to see if both parents discussed the same strategies, which could mean that these actions are or were indeed undertaken.

To start with, parents’ perception and representation of the deep-rootedness of their ethnic identity resembles the idea of cognitive schemes or mental structures developed by Brubaker, et al. (2004). Parents feel ethnicity is in your blood; it is something you cannot deny (Gil-White, 1999). Taking this reasoning a step further, if it is in your blood, it will also be – to some extent – in the blood of your offspring. Furthermore, you share with your co-ethnics some sort of blood tie (Roosens, 1994). Current ethnicity theories mainly focus on the malleability and context dependency of identity. Some, however, argue that primordialist arguments should be integrated in conceptualizations of ethnicity (Brubaker et al., 2004, Roosens, 2005), especially since the strategies and practices that parents develop strongly relate to these mental structures, being logical consequences of these cognitive schemes. This seems clear when one studies that which gives content to one’s ethnicity: the cultural emblems that are felt to be less stable and can be lost – especially with respect to the transposition of these emblems from one generation to the other (Epstein, 1978; Roosens, 1994). As a consequence, parents, feeling the burden of continuity resting upon their shoulders, are forced by specific social processes to carefully consider how to deal with these issues.

In these parental discourses clear dispositions to perceive in a certain ‘ethnic way’ were present and they seem to generate actions that are logical in their context, even though they might have negative consequences for their children (e.g. prohibiting interreligious relations). Moreover, some strategies might even be socially sanctioned when they are perceived as inadequate or deviant by the dominant group (e.g. mother tongue transmission in the educational field), but parents feel it is their duty to transmit this to their children. Consequently, some of these parental perceptions of the in-group and the out-group are explainable from their perspective and social location, yet they can hold quite negative perceptions about ‘the other’.

A similar reasoning can be found with respect to the discourses on dominant (white) ethnicity. Whereas this dominant perspective often strongly corresponds with the ‘norm’ in a specific field, the dominated groups frequently challenge these normalized patterns of perceiving and acting, and thus make explicit which homogenizing processes are at work. With respect to the current article, the narratives of the dominated perspectives show how the ‘feel for the game’ is a socially constructed and a majority dominated one. Yet this does not necessarily imply that each ethnic minority is by definition dominated, because the specific cultural emblems used – sometimes instrumentally – allow for shifts in power relations.

This is illustrated most explicitly when focusing on the discourses and dispositions with respect to language and religion. Concerning language, particularly the proficiency in one’s ‘mother tongue’, Belgian-Flemish parents do not feel they need to design careful strategies to ensure linguistic continuity, for the Belgian Constitution as well as the cultural apparatus ensure this for them. For ethno-linguistic minorities such as the Italian and Moroccan parents, the situation is completely different. In their perspective the Italian or Moroccan/Berber language has an important role as a cultural emblem in and of itself, but also as a means for enabling communication between generations. Yet this does not correspond with the broader social field in the way it does for Belgian-Flemish parents. Given this specific configuration, their desire for linguistic and intergenerational continuity, they feel their family cohesion is threatened. Developing counterstrategies, such as not sending one’s children to mainstream day-care centres but keeping their upbringing (at least the first three years of their lives) within the family context, becomes a logical and effective strategy to ensure continuity and cohesion. Moreover, this differs from the perspective of Belgian-Flemish lower SES families who do not send their children to day-care centres; rather than relating this decision to their ethno-cultural identity, they relate it, for example, to their financial situation.

The story becomes a bit more complex when one focuses on that other cultural emblem put forward by parents as essential: religion. The dominant religion in Flanders is Christian-Catholic and is as such reflected in its privileged position in social institution such as the education and the political system. Thus, the narratives make clear that Belgian-Flemish and Italian parents feel confident that their religion as such is not threatened. This is different for the Moroccan parents in this article, who all claim to be Muslim and who explicitly stressed the importance of religion in their narratives. However, when parents were confronted with the idea of interethnic marriages, it became poignantly clear that they saw clear boundaries between Christians and Muslims and employed strategies to discourage or prohibit interreligious relations. The threat of losing religious continuity is expressed by various parents, irrespective of their ethnic or religious background, and mostly linked with the idea of interreligious marriages. Parents are afraid their children will marry outside of the religious group and family, and group cohesion will be lost due to the perceived fundamental incompatibility of the different religions. Also in this case, these strategies seem logical, given the fact that religion is presented as crucial to one’s ethnic identity. Therefore, informing your children of the dangers of, and in some cases even prohibiting them to engage in, interreligious relationships is a way to reduce the possibility of these events taking place (Clycq, 2012). At the same time, almost none of these parents do seem to hold explicit racist feelings towards the religious other, for they simultaneously emphasize the need for interethnic and even interreligious friendship relations. Yet, given their perspectives on social reality and on intergenerational continuity, this is no contradictory concourse of discourse and praxis; the exclusion of ‘the other’ seems indeed a logical course of action with the goals that parents have in mind.

In this paper we studied the complex ways in which discourse and praxis, or the ‘ethnic ways of seeing the world’ and the ‘doing being ethnic’ strategies, are intertwined. Certain dispositions generate actions that are self-evident and normal within the context individuals are living in, but sometimes they do not seem clear to the outside world. Often a specific logic is related to these mental structures, which needs to be uncovered to reveal the underlying processes. More in particular, the narratives show that the construction of boundaries, or in some cases we might even say fault lines, is intentional but at the same time sheds light on the underlying processes related to ethnicity, culture and religion. There are specific and socially accepted dispositions and strategies for categorizing individuals into different ethno-cultural and status groups, and while specific boundaries seem open for bridging relations, others seem more stringent and crossing them is felt to be trespassing. Nevertheless, some fault lines can be deconstructed and transformed while others are reproduced and remain strong over the years. This article discusses the work of parents as active agents in reproducing or transforming certain boundaries and social categories. Power relations play a crucial role in these processes and help explain why certain strategies are developed given the social location of these parents. Nevertheless, studying how children react to these strategies and how they are influenced by their parents is necessary to shed more light on the impact of these processes and strategies.